Category: What We’re Reading

We’ve constantly been sharing a list of our recent reads in our weekly emails for The Good Investors.

Do subscribe for our weekly updates through the orange box in the blog (it’s on the side if you’re using a computer, and all the way at the bottom if you’re using mobile) – it’s free!

But since our readership-audience for The Good Investors is wider than our subscriber base, we think sharing the reading list regularly on the blog itself can benefit even more people. The articles we share touch on a wide range of topics, including investing, business, and the world in general. 

Here are the articles for the week ending 02 March 2025:

1. Satya Nadella – Microsoft’s AGI Plan & Quantum Breakthrough – Dwarkesh Patel and Satya Nadella

Dwarkesh Patel

Where is the value going to be created in AI?

Satya Nadella

That’s a great one. So I think there are two places where I can say with some confidence. One is the hyperscalers that do well, because the fundamental thing is if you sort of go back to even how Sam and others describe it, if intelligence is log of compute, whoever can do lots of compute is a big winner.

The other interesting thing is, if you look at underneath even any AI workload, like take ChatGPT, it’s not like everybody’s excited about what’s happening on the GPU side, it’s great. In fact, I think of my fleet even as a ratio of the AI accelerator to storage, to compute. And at scale, you’ve got to grow it…

…Satya Nadella

So in fact it’s manna from heaven to have these AI workloads because guess what? They’re more hungry for more compute, not just for training, but we now know, for test time. When you think of an AI agent, it turns out the AI agent is going to exponentially increase compute usage because you’re not even bound by just one human invoking a program. It’s one human invoking programs that invoke lots more programs. That’s going to create massive, massive demand and scale for compute infrastructure. So our hyperscale business, Azure business, and other hyperscalers, I think that’s a big thing.

Then after that, it becomes a little fuzzy. You could say, hey, there is a winner-take-all model- I just don’t see it. This, by the way, is the other thing I’ve learned: being very good at understanding what are winner-take-all markets and what are not winner-take-all markets is, in some sense, everything. I remember even in the early days when I was getting into Azure, Amazon had a very significant lead and people would come to me, and investors would come to me, and say, “Oh, it’s game over. You’ll never make it. Amazon, it’s winner-take-all.”

Having competed against Oracle and IBM in client-server, I knew that the buyers will not tolerate winner-take-all. Structurally, hyperscale will never be a winner-take-all because buyers are smart.

Consumer markets sometimes can be winner-take-all, but anything where the buyer is a corporation, an enterprise, an IT department, they will want multiple suppliers. And so you got to be one of the multiple suppliers.

That, I think, is what will happen even on the model side. There will be open-source. There will be a governor. Just like on Windows, one of the big lessons learned for me was, if you have a closed-source operating system, there will be a complement to it, which will be open source.

And so to some degree that’s a real check on what happens. I think in models there is one dimension of, maybe there will be a few closed source, but there will definitely be an open source alternative, and the open-source alternative will actually make sure that the closed-source, winner-take-all is mitigated.

That’s my feeling on the model side. And by the way, let’s not discount if this thing is really as powerful as people make it out to be, the state is not going to sit around and wait for private companies to go around and… all over the world. So, I don’t see it as a winner-take-all.

Then above that, I think it’s going to be the same old stuff, which is in consumer, in some categories, there may be some winner-take-all network effect. After all, ChatGPT is a great example.

It’s an at-scale consumer property that has already got real escape velocity. I go to the App Store, and I see it’s always there in the top five, and I say “wow, that’s pretty unbelievable”.

So they were able to use that early advantage and parlay that into an app advantage. In consumer, that could happen. In the enterprise again, I think there will be, by category, different winners. That’s sort of at least how I analyze it…

…Satya Nadella

The way I come at it, Dwarkesh, it’s a great question because at some level, if you’re going to have this explosion, abundance, whatever, commodity of intelligence available, the first thing we have to observe is GDP growth.

Before I get to what Microsoft’s revenue will look like, there’s only one governor in all of this. This is where we get a little bit ahead of ourselves with all this AGI hype. Remember the developed world, which is what? 2% growth and if you adjust for inflation it’s zero?

So in 2025, as we sit here, I’m not an economist, at least I look at it and say we have a real growth challenge. So, the first thing that we all have to do is, when we say this is like the Industrial Revolution, let’s have that Industrial Revolution type of growth.

That means to me, 10%, 7%, developed world, inflation-adjusted, growing at 5%. That’s the real marker. It can’t just be supply-side.

In fact that’s the thing, a lot of people are writing about it, and I’m glad they are, which is the big winners here are not going to be tech companies. The winners are going to be the broader industry that uses this commodity that, by the way, is abundant. Suddenly productivity goes up and the economy is growing at a faster rate. When that happens, we’ll be fine as an industry.

But that’s to me the moment. Us self-claiming some AGI milestone, that’s just nonsensical benchmark hacking to me. The real benchmark is: the world growing at 10%.

Dwarkesh Patel

Okay, so if the world grew at 10%, the world economy is $100 trillion or something, if the world grew at 10%, that’s like an extra $10 trillion in value produced every single year. If that is the case, you as a hyperscaler… It seems like $80 billion is a lot of money. Shouldn’t you be doing like $800 billion?

If you really think in a couple of years, we could be really growing the world economy at this rate, and the key bottleneck would be: do you have the compute necessary to deploy these AIs to do all this work?

Satya Nadella

That is correct. But by the way, the classic supply side is, “Hey, let me build it and they’ll come.” That’s an argument, and after all we’ve done that, we’ve taken enough risk to go do it.

But at some point, the supply and demand have to map. That’s why I’m tracking both sides of it. You can go off the rails completely when you are hyping yourself with the supply-side, versus really understanding how to translate that into real value to customers.

That’s why I look at my inference revenue. That’s one of the reasons why even the disclosure on the inference revenue… It’s interesting that not many people are talking about their real revenue, but to me, that is important as a governor for how you think about it.

You’re not going to say they have to symmetrically meet at any given point in time, but you need to have existence proof that you are able to parlay yesterday’s, let’s call it capital, into today’s demand, so that then you can again invest, maybe exponentially even, knowing that you’re not going to be completely rate mismatched.

Dwarkesh Patel

I wonder if there’s a contradiction in these two different viewpoints, because one of the things you’ve done wonderfully is make these early bets. You invested in OpenAI in 2019, even before there was Copilot and any applications.

If you look at the Industrial Revolution, these 6%, 10% build-outs of railways and whatever things, many of those were not like, “We’ve got revenue from the tickets, and now we’re going to…”

Satya Nadella

There was a lot of money lost.

Dwarkesh Patel

That’s true. So, if you really think there’s some potential here to 10x or 5x the growth rate of the world, and then you’re like, “Well, what is the revenue from GPT-4?”

If you really think that’s the possibility from the next level up, shouldn’t you just, “Let’s go crazy, let’s do the hundreds of billions of dollars of compute?” I mean, there’s some chance, right?

Satya Nadella

Here’s the interesting thing, right? That’s why even that balanced approach to the fleet, at least, is very important to me. It’s not about building compute. It’s about building compute that can actually help me not only train the next big model but also serve the next big model. Until you do those two things, you’re not going to be able to really be in a position to take advantage of even your investment.

So, that’s kind of where it’s not a race to just building a model, it’s a race to creating a commodity that is getting used in the world to drive… You have to have a complete thought, not just one thing that you’re thinking about.

And by the way, one of the things is that there will be overbuild. To your point about what happened in the dotcom era, the memo has gone out that, hey, you know, you need more energy, and you need more compute. Thank God for it. So, everybody’s going to race.

In fact, it’s not just companies deploying, countries are going to deploy capital, and there will be clearly… I’m so excited to be a leaser, because, by the way; I build a lot, I lease a lot. I am thrilled that I’m going to be leasing a lot of capacity in ’27, ’28 because I look at the builds, and I’m saying, “This is fantastic.” The only thing that’s going to happen with all the compute builds is the prices are going to come down…

…Satya Nadella

This has been another 30-year journey for us. It’s unbelievable. I’m the third CEO of Microsoft who’s been excited about quantum.

The fundamental breakthrough here, or the vision that we’ve always had is, you need a physics breakthrough in order to build a utility-scale quantum computer that works. We took the path of saying, the one way for having a less noisy or more reliable qubit is to bet on a physical property that by definition is more reliable and that’s what led us to the Majorana zero modes, which was theorized in the 1930s. The question was, can we actually physically fabricate these things? Can we actually build them?

So the big breakthrough effectively, and I know you talked to Chetan, was that we now finally have existence proof and a physics breakthrough of Majorana zero modes in a new phase of matter effectively. This is why we like the analogy of thinking of this as the transistor moment of quantum computing, where we effectively have a new phase, which is the topological phase, which means we can even now reliably hide the quantum information, measure it, and we can fabricate it. And so now that we have it, we feel like with that core foundational fabrication technique out of the way, we can start building a Majorana chip.

That Majorana One which I think is going to basically be the first chip that will be capable of a million qubits, physical. And then on that, thousands of logical qubits, error-corrected. And then it’s game on. You suddenly have the ability to build a real utility-scale quantum computer, and that to me is now so much more feasible. Without something like this, you will still be able to achieve milestones, but you’ll never be able to build a utility-scale computer. That’s why we’re excited about it…

…Satya Nadella

It’s a great question. One thing that I’ve been excited about is, even in today’s world… we had this quantum program, and we added some APIs to it. The breakthrough we had maybe two years ago was to think of this HPC stack, and AI stack, and quantum together.

In fact, if you think about it, AI is like an emulator of the simulator. Quantum is like a simulator of nature. What is quantum going to do? By the way, quantum is not going to replace classical. Quantum is great at what quantum can do, and classical will also…

Quantum is going to be fantastic for anything that is not data-heavy but is exploration-heavy in terms of the state space. It should be data-light but exponential states that you want to explore. Simulation is a great one: chemical physics, what have you, biology.

One of the things that we’ve started doing is really using AI as the emulation engine. But you can then train. So the way I think of it is, if you have AI plus quantum, maybe you’ll use quantum to generate synthetic data that then gets used by AI to train better models that know how to model something like chemistry or physics or what have you. These two things will get used together.

So even today, that’s effectively what we’re doing with the combination of HPC and AI. I hope to replace some of the HPC pieces with quantum computers.

2. Microsoft’s Majorana 1 chip carves new path for quantum computing – Catherine Bolgar

Microsoft today introduced Majorana 1, the world’s first quantum chip powered by a new Topological Core architecture that it expects will realize quantum computers capable of solving meaningful, industrial-scale problems in years, not decades.

It leverages the world’s first topoconductor, a breakthrough type of material which can observe and control Majorana particles to produce more reliable and scalable qubits, which are the building blocks for quantum computers.

In the same way that the invention of semiconductors made today’s smartphones, computers and electronics possible, topoconductors and the new type of chip they enable offer a path to developing quantum systems that can scale to a million qubits and are capable of tackling the most complex industrial and societal problems, Microsoft said…

…This new architecture used to develop the Majorana 1 processor offers a clear path to fit a million qubits on a single chip that can fit in the palm of one’s hand, Microsoft said. This is a needed threshold for quantum computers to deliver transformative, real-world solutions – such as breaking down microplastics into harmless byproducts or inventing self-healing materials for construction, manufacturing or healthcare. All the world’s current computers operating together can’t do what a one-million-qubit quantum computer will be able to do…

…The topoconductor, or topological superconductor, is a special category of material that can create an entirely new state of matter – not a solid, liquid or gas but a topological state. This is harnessed to produce a more stable qubit that is fast, small and can be digitally controlled, without the tradeoffs required by current alternatives…

…This breakthrough required developing an entirely new materials stack made of indium arsenide and aluminum, much of which Microsoft designed and fabricated atom by atom…

…Commercially important applications will also require trillions of operations on a million qubits, which would be prohibitive with current approaches that rely on fine-tuned analog control of each qubit. The Microsoft team’s new measurement approach enables qubits to be controlled digitally, redefining and vastly simplifying how quantum computing works.

This progress validates Microsoft’s choice years ago to pursue a topological qubit design – a high risk, high reward scientific and engineering challenge that is now paying off. Today, the company has placed eight topological qubits on a chip designed to scale to one million…

…But reaching the next horizon of quantum computing will require a quantum architecture that can provide a million qubits or more and reach trillions of fast and reliable operations. Today’s announcement puts that horizon within years, not decades, Microsoft said.

Because they can use quantum mechanics to mathematically map how nature behaves with incredible precision – from chemical reactions to molecular interactions and enzyme energies – million-qubit machines should be able to solve certain types of problems in chemistry, materials science and other industries that are impossible for today’s classical computers to accurately calculate…

…Most of all, quantum computing could allow engineers, scientists, companies and others to simply design things right the first time – which would be transformative for everything from healthcare to product development. The power of quantum computing, combined with AI tools, would allow someone to describe what kind of new material or molecule they want to create in plain language and get an answer that works straightaway – no guesswork or years of trial and error.

“Any company that makes anything could just design it perfectly the first time out. It would just give you the answer,” Troyer said. “The quantum computer teaches the AI the language of nature so the AI can just tell you the recipe for what you want to make.”…

…Qubits can be created in different ways, each with advantages and disadvantages. Nearly 20 years ago, Microsoft decided to pursue a unique approach: developing topological qubits, which it believed would offer more stable qubits requiring less error correction, thereby unlocking speed, size and controllability advantages. The approach posed a steep learning curve, requiring uncharted scientific and engineering breakthroughs, but also the most promising path to creating scalable and controllable qubits capable of doing commercially valuable work.

The disadvantage is – or was – that until recently the exotic particles Microsoft sought to use, called Majoranas, had never been seen or made. They don’t exist in nature and can only be coaxed into existence with magnetic fields and superconductors. The difficulty of developing the right materials to create the exotic particles and their associated topological state of matter is why most quantum efforts have focused on other kinds of qubits…

…Majoranas hide quantum information, making it more robust, but also harder to measure. The Microsoft team’s new measurement approach is so precise it can detect the difference between one billion and one billion and one electrons in a superconducting wire – which tells the computer what state the qubit is in and forms the basis for quantum computation.

The measurements can be turned on and off with voltage pulses, like flicking a light switch, rather than finetuning dials for each individual qubit. This simpler measurement approach that enables digital control simplifies the quantum computing process and the physical requirements to build a scalable machine…

…Majorana 1, Microsoft’s quantum chip that contains both qubits as well as surrounding control electronics, can be held in the palm of one’s hand and fits neatly into a quantum computer that can be easily deployed inside Azure datacenters.

3. The most underreported and important story in AI right now is that pure scaling has failed to produce AGI – Gary Marcus

On the order of half a trillion dollars has been invested on a premise that I have long argued was unlikely to succeed: the idea (sometimes informally referred to as the scaling hypothesis) that we could get to “artificial general intelligence” simply by adding more and more data and GPUs…

…Virtually all of the generative AI industry has been built on this presumption; projects like the OpenAI/Oracle/Softbank joint venture Stargate, allegedly another half trillion dollars, are also largely based on this premise…

…But I always knew it couldn’t last forever. When I said as much, the field was absolutely furious at me…

…The first signs that the pure scaling of data and compute might in fact be hitting a wall came from industry leaks from people like famed investor Marc Andreessen, who said in early November 2024 that current models are “sort of hitting the same ceiling on capabilities.” Then, in December, Microsoft CEO Satya Nadella echoed many of my 2022 themes, saying at a Microsoft Ignite event, “in the last multiple weeks there is a lot of debate on have we hit the wall with scaling laws. Is it gonna continue? Again, the thing to remember at the end of the day these are not physical laws. There are just empirical observations that hold true just like Moore’s Law did for a long period of time and so therefore it’s actually good to have some skepticism some debate.”…

…Finally, and perhaps most significantly: Elon Musk said over that weekend that Grok 3, with 15x the compute of Grok 2, and immense energy (and construction and chop) bills, would be “the smartest AI on the earth.” Yet the world quickly saw that Grok 3 is still afflicted by the kind of unreliability that has hobbled earlier models. The famous ML expert Andrej Karpathy reported that Grok 3 occasionally stumbles on basics like math and spelling. In my own experiments, I quickly found a wide array of errors, such as hallucinations (e.g, it told me with certainty that there was a significant 5.6-sized earthquake on Feb. 10 in Billings, Montana, when no such thing had happened) and extremely poor visual comprehension (e.g. it could not properly label the basic parts of a bicycle).

Nadella, in his December speech, pointed to test-time compute, in which systems are allowed extra time for “reasoning” as the next big thing, and to some degree he is right; it is the next big thing, a new thing to try to scale, since merely scaling compute and data is no longer bringing the massive returns it once did. At least for a while, adding more and more computing time will help, at least on some kinds of problems…

…although DeepSeek lowered the costs of training these new systems, they are still expensive to operate, which is why companies like OpenAI are limiting their usage. When customers begin to realize that even with the greater expenses, errors still seep in, they are likely to be disappointed. One irate customer cc:d me yesterday on a several page demand for a refund, writing in part that “GPT-4o Pro [which includes access to test time compute] has consistently underperformed,” and enumerated problems such as “Severely degraded memory” and “Hallucinations and Unreliable Answers.”…

…the illustrious Stanford Natural Language Processing group reached a similar conclusion, reading between the lines of OpenAI’s recent announcement in the same way I did. In their words, Altman’s recent OpenAI roadmap was “the final admission that the 2023 strategy of OpenAI, Anthropic, etc. ‘“simply scaling up model size, data, compute, and dollars spent will get us to AGI/ASI’) is no longer working!”

In short, half a trillion dollars have been invested in a misguided premise; a great deal more funding seems to be headed in the same direction for now.

4. Is Microsoft’s Copilot push the biggest bait-and-switch in AI – Tien Tzuo

Over a year ago, Microsoft launched Copilot Pro, an AI assistant embedded in its Office suite, with a $20/month price. The uptake apparently was abysmal. By October, they admitted that the way they were selling Copilot was not working out.

So what did they do? They forced it on Microsoft users by including Copilot in Office, and hiking up subscription fees. Microsoft first made this change in Asia, then fully pulled the plug across the globe last month, impacting 84 million subscribers. To add insult to injury, Microsoft renamed the product to Microsoft 365 Copilot. You didn’t want to pay for Copilot? Well, now you are…

…Not only is Microsoft’s Copilot rollout deceptive, it’s also embarrassingly disastrous.

This goes to show that even tech giants, including a major backer of AI pioneer OpenAI, can suffer the hype and competitive pressure surrounding AI. And it’s a stark reminder that what businesses should really be focused on instead is value — communicating it clearly and delivering it tangibly to customers…

…Well, there’s a good contrast to Microsoft’s approach — from Adobe.

Adobe took a different approach with its AI rollout last fall, resisting the temptation to immediately monetize its new video generator, instead using it to boost adoption and engagement. By positioning AI as a value-add rather than a paid extra, Adobe was playing the long game, building a loyal user base that would be ripe for future upselling once they experienced AI’s benefits for themselves.

5. Broken Markets!? – The Brooklyn Investor

So, I keep hearing that markets are broken, or that the market is as expensive as ever. I know I keep saying this and I am sounding like a broken record, but I am not so sure…

…But if you look at individual stocks, markets are clearly differentiating between individual stocks. Look at Nvidia vs. Intel. If nobody was really evaluating them and the market was ignoring fundamentals, you would think both stocks would be performing similarly. But they are clearly not. People are clearly differentiating between winners and losers. It’s a separate question whether they are over-discounting their views. That’s a different discussion, and contrary to the view that passive investing is killing fundamental analysis.

Another example: JPM, the better bank, is trading at 2.4x book, versus C, which is a crappier one, selling at 0.8x book. You can’t complain that the market is broken just because you don’t agree with it. On the other hand, Buffett in the 50s loved the fact that institutional investors of the time completely ignored company analysis / valuation…

…Look at all the rich folks at the Berkshire Hathaway annual meeting. Look at BRK stock since 1970. How often did it look ‘overpriced’? What about the market? What if people sold out when they thought BRK was overpriced? Or the market? Would they be as rich as they are now? Probably not. Maybe there are some smart folks that got in and out of BRK over the years, but I would bet that the richest of them are the ones that just held it and did nothing.

I keep telling people this, but if you look at all the richest people in the world, a lot of them are people who held a single asset, and held it for decades. Look at Bill Gates. What if he was hip to value investing and knew more about stocks and values. He may have sold out of MSFT when it looked really overvalued. What about Buffett? What about Bezos?

A lot of the rich used to be real estate moguls, and I thought a lot of them were wealthy because real estate was not very liquid, so they had no choice but to hold on even in bad times. Stocks have what you may call the “curse of liquidity”. It’s so easy to say, holy sh*t, something bad is going to happen, and *click*, you can get out of the market. Back in the 90s, we used to fear a 1-800 crash; people will call their brokers’ automated trade execution lines, 1-800-Sell-Everything, and go “Get me out of the market!!! See everything NOW!!!”, and the market would not be able to open the next morning. Some of us truly feared that, and hedge funds talked about that sort of thing all the time. But you can’t do that with your house.

---

Disclaimer: The Good Investors is the personal investing blog of two simple guys who are passionate about educating Singaporeans about stock market investing. By using this Site, you specifically agree that none of the information provided constitutes financial, investment, or other professional advice. It is only intended to provide education. Speak with a professional before making important decisions about your money, your professional life, or even your personal life. We currently have a vested interest in Amazon and Microsoft. Holdings are subject to change at any time.

We’ve constantly been sharing a list of our recent reads in our weekly emails for The Good Investors.

Do subscribe for our weekly updates through the orange box in the blog (it’s on the side if you’re using a computer, and all the way at the bottom if you’re using mobile) – it’s free!

But since our readership-audience for The Good Investors is wider than our subscriber base, we think sharing the reading list regularly on the blog itself can benefit even more people. The articles we share touch on a wide range of topics, including investing, business, and the world in general. 

Here are the articles for the week ending 23 February 2025:

1. Weekend Thoughts: Demand Shocks and How Utilities Went From Sleepy to AI Darlings – Andrew Walker

That history has completely changed with AI over the past 18 months. Suddenly, the U.S. is experiencing real electricity demand growth for the first time in decades, and the rush for AI players to secure an enormous amount of electricity for their big AI datacenters has left the U.S. short electricity and scrambling to reopen old plants. The combination has created a bonanza for utilities; the whole sector screamed higher in 2024, and many of the utilities proved to be better AI plays than MSFT or even NVDA in 2024!…

…I mention it because I think that’s a really interesting set up to pattern to match / look for in the future. Find a staid, boring industry that most investors won’t look at that trades for a reasonable multiple (as utilities did pre-AI boom), and buy them just before a demand shock. You could argue if you do / time it right your downside is protected (it’s not like the stocks are going to go down when a demand shock doesn’t come; they’re not pricing one in!), and if you’re right you can make multiples of your money.

It worked for utilities in 2024, and it worked for all sorts of COVID beneficiaries in 2020/2021 (cleaning products, home gym equipment, etc.).

Now, there is one interesting wrinkle to this thesis: industries with long lead times benefit the most from a demand shock. Consider power: if you have a demand shock on the power side, then you eventually need to build more power plants to handle that shock. Building power plants takes a long time; I’d guess it takes 3-5 years from start to finish to build a baseload natural gas plant, and probably 20 years to build a nuke (if you can even get one built!). So a demand shock in power can create super normal pricing for a very long time. Most other demand shocks can be responded to much faster. A simple example: a demand shock in oil can be met relatively quickly; it only takes a few months to start spinning up new Permian wells!

2. How I avoided China’s biggest bond default – Willie Keng

In 2016, I told myself:

“Stay away from this Chinese company at ALL cost.”

It was a Thursday. I was in a room filled with high-level executives dressed in black suits – the company’s chief financial officer, finance managers, bankers, analysts and fund managers were there…

…Back then, even Chinese property developers from blue-chips to junk-rated companies all wanted to raise debt. China’s skyline was dominated by the grey, skeletal skyscrapers wrapped in scaffoldings. Tower cranes with long, mechanical arms swung on these buildings — there were noise of rapid urban growth.

And China’s property bond market was hot like an iron.

It was common to have 3-4 such company meetings in a single day. These were eye-opening — and very exciting times, as developers rushed to raise as much “cheap debt” as possible…

…In 2016, I saw the early warning signs of China’s bigger property developer, Evergrande…

…During that management [sic: meeting] held on Thursday, I concluded its leverage was too high…

…In 2021, Evergrande defaulted.

I’ve dug out some short-cut questions from my old notebooks:

1. How do you make money? (where are your sources of revenues from?)…
…11. How much debt are you refinancing over the next few years?
12. Who are your main bankers?
13. How do you hedge your interest rates? At what cost?
14. How much committed credit facilities you have with banks?
15. Why are you borrowing X debt, what are you using for?…
…23. How are you planning to refinance your short-term loans and who are the lenders? At what interest rate? Is it possible to get “secured” and “unsecured” funding?

3. The Magnificent Seven, MKL – The Brooklyn Investor

The question is, basically, what’s up with the Mag 7? How crazy is it? Is it Nifty-fifty all over again?…

…The Mag 7 is trading at 41x 2024 earnings, which is close to ttm earnings. and 33.5x 2025 expected earnings, and 28x 2026 expected earnings. I know, expected earnings is sort of rubbish. Nobody gets that right. But we gotta start somewhere, right?

By the way, if you exclude NVDA and TSLA, the above would be 32.6x, 29.6x and 25.8x P/Es, respectively. In this case, this is valid to do, because you can actually create a portfolio of high ROE, decent growth stocks at that valuation.

And then look at the ROE of each of these companies. And then look at the historic growth rates, 5 and 10 year of these companies…

… Let’s say the Mag 7 was a modern-day, techful version of a conglomerate like BRK. Its subsidiaries have grown tremendously in the past 5 and 10 years. Earnings will collectively grow 23% in 2025 and 19% in 2026 (willful suspension of disbelief may be key here), and look at the high ROE of each division (OK, I was too lazy to calculate a weighted average).

And this conglomerate is trading for 34x earnings! Or less than 30x if you don’t want NVDA and TSLA. Think about that for a second. How many ideas with those metrics can you find?…

… It’s easy to call AMZN a retailer, for example. YouTube is a big part of Google, and the rest of Google is advertising. So is Facebook. They compete with linear television, radio and other forms of entertainment in the past, and they make money from advertising, just like old media (including magazines too…). So we can call it media and advertising, not even “new” media. Just media. Tesla is an automaker. AAPL is more like the old Sony; consumer electronics. Basically every single consumer electronic product ever invented rolled into one product. They do media too; music, streaming etc. Gaming too. Only NVDA and MSFT sort of feel like the conventional ‘tech’.

My point was going to be, the Mag 7 domination may or may not be a problem, but it is quite diversified as a virtual conglomerate.

4. The Great EBITDA Illusion –  Stephen Clapham

KPMG examined 1800 transactions between 2013 and 2018 and found that both the number of adjustments to EBITDA increased, as did the value. The number increased from 5.8 to 7.9 per transaction and the value increased…

…Pro-forma adjustments have risen by 10% and were in over 60% of deals. These include cost-related adjustments, to reflect the future value of cost reductions, and revenue run-rate adjustments, including planned price increases, or the expected impact of a new product line. In my experience, cost savings tend to have a higher deliverability as they are more within management’s control; but it’s a rare business which can increase price without affecting volumes, while capacity increases are often tricky to model…

…S&P’s Leveraged Finance group have also looked at the topic of EBITDA adjustments, but through a different lens.

“Our six-year study on EBITDA addbacks appears to shows a positive correlation between the magnitude of addbacks at deal inception and the severity of management projection misses.”

They highlight that addbacks represent a median 30% of management adjusted EBITDA at deal inception. They consider management projections to be aggressive and U.S. speculative-grade corporate issuers generally “present earnings, debt, and leverage projections in their marketing materials at deal inception that they cannot realize”…

…This is of real significance, especially to lenders…

…Forecasts made in M&A deals turn out badly with leverage nearly twice the projection in year 1 and worse by end year 2. Most of the miss is down to over-estimating adjusted EBITDA. The median miss in year one was 34%, rising to 35% in year two…

…Leverage forecasts made in leveraged buyout transactions are much worse with actual leverage of 8.2x vs a 3.9x forecast…

…The S&P report concludes:

“Our six-year study continues to underscore that addbacks and company-adjusted EBITDA are a poor predictor of profitability. Our substantial dataset makes it clear that management teams and equity sponsors regularly miss their projections by a large margin, and that the magnitude of the misses is positively correlated with addbacks and firms that we rate lower. This suggests that inflated addbacks may help companies with higher financial risk get deals done.”

The data is clear and there is no reason to doubt it. What surprises me is that private equity and credit funds continue to engage in such practices and that allocators and credit investors appear relaxed. That may be justified given past performance, but as I have written here several times, I don’t believe that the historical record is anywhere near sustainable.

5. There Goes My Hero – Ben Carlson

My family took its first and only Disney trip in the summer of 1990.

We rode some rollercoasters. Went to one of the waterparks. Decently fun trip from what I can remember as a 4th grader.

The strange part was that my older brother Jon was lethargic the whole trip. I still remember a picture of him taking a nap on a bench in the middle of the day. Something was off.

I was nine, so I didn’t think anything of it. My mother, a registered nurse, knew something was wrong so when we got home, they took Jon to the doctor.

He was diagnosed with a rare form of leukemia just before heading into the 7th grade…

…Jon endured months of chemotherapy and radiation, after which the only solution was a bone marrow transplant. My parents weren’t a match. Luckily, my sister and I were both were.

I was the bone marrow donor. There was no guarantee it would work, but miraculously, it did. Jon’s cancer went into remission.

It was a terrible year for our family but Jon was a trooper. He never once complained. Even though he lived in the hospital on and off for months at a time and lost all of his hair he never felt sorry for himself…

…Last year, he was diagnosed with stage 4 pancreatic cancer. Last week he passed away just shy of his 46th birthday.

Jon was a tough son of a bitch and went out swinging.

The original plan was to manage the pancreatic cancer with chemo until Jon died but he didn’t want to just wither away. He called specialists all over the country, finally finding a doctor who would give him an experimental drug that allowed him to stop receiving chemo.

And it actually worked for a while. The cancer spread slowed. Eventually it would stop working but it gave us an extra six months or so…

…Grief is strange. Although you know millions and millions of other people have felt it, it still feels like the most personal of all emotions. I guess it is in some ways depending on the person and how they were lost.

At times, I’ve felt like there’s a black cloud hanging over my head. Other times, it’s as if there is a dull knife stuck in the back of my head. Sometimes it crashes into you all at once like a wave.

But it also forces you to reminisce about the good times. These past few months, it’s almost felt like my life has slowly flashed before my eyes through the lens of all the memories of my brother…

…After his bone marrow transplant, Jon was approached by the Make a Wish Foundation — anything he wanted, within reason.

He could have asked to meet his favorite celebrity or athlete. He could have asked for a room full of video games. He could have asked for a four-wheeler or a jetski or some other fun toy like that.

Instead, Jon requested a two-week all-expenses-paid vacation to Hawaii for our entire family. We got to swim with dolphins, fly in a helicopter, see some volcanoes, play on the beach, and more. They even sent a limo to our house to drive us to the airport.

I didn’t realize it at the time, but it was like Jon instinctively knew our family needed that after what we all went through. I still can’t believe a 12-year-old had the foresight to be so selfless, especially when no one would have blamed him for being as selfish as he wanted.

Jon was wise beyond his years and valued experiences with loved ones more than material possessions…

…As we worked through his financial situation it became abundantly clear he was more than prepared for something like this than I ever could have imagined. There was a large life insurance policy. He was holding far too much cash for a person his age.

Jon why do you have so much cash?

Ben, I knew something like this was going to happen. I’ve known it since I was 12 years old.

That bout with cancer changed his entire perception of risk. He’s been working and saving since age 19 because there was always a voice in the back of his head telling him something like this could happen again…

…He also left behind some life advice for his kids that helps explain the kind of guy he was:

Be happy with what you have, you don’t need as much as you think.

Never leave anyone behind.

Life is way better than a screen, go live it.

Our mantra is to go live like Jon. I’m so lucky to have him as part of my life while he was here.

---

Disclaimer: The Good Investors is the personal investing blog of two simple guys who are passionate about educating Singaporeans about stock market investing. By using this Site, you specifically agree that none of the information provided constitutes financial, investment, or other professional advice. It is only intended to provide education. Speak with a professional before making important decisions about your money, your professional life, or even your personal life. We currently have a vested interest in Alphabet (parent of Google), Amazon, Apple, Markel, Meta Platforms (parent of Facebook), Microsoft, and Tesla. Holdings are subject to change at any time.

We’ve constantly been sharing a list of our recent reads in our weekly emails for The Good Investors.

Do subscribe for our weekly updates through the orange box in the blog (it’s on the side if you’re using a computer, and all the way at the bottom if you’re using mobile) – it’s free!

But since our readership-audience for The Good Investors is wider than our subscriber base, we think sharing the reading list regularly on the blog itself can benefit even more people. The articles we share touch on a wide range of topics, including investing, business, and the world in general. 

Here are the articles for the week ending 16 February 2025:

1. The real threat to American prosperity – Daron Acemoglu

American economic success in the era after the second world war depended on innovation, which in turn relied on strong institutions that encouraged people to invest in new technologies, trusting that their inventiveness would be rewarded. This meant a court system that functioned, so that the fruits of their investments could not be taken away from them by expropriation, corruption or chicanery; a financial system that would enable them to scale up their new technologies; and a competitive environment to ensure that incumbents or rivals couldn’t block their superior offerings. These kinds of institutions matter under all circumstances, but they are especially critical for economies that rely heavily on innovation.

Stability requires that people trust institutions, and institutions become more likely to fail when people think they are failing. This is what explained the sudden meltdown of US economic dynamism…

…Economic growth in the US was rapid for most of the post-1980 era, but about half of the country didn’t benefit much from this. In a pattern unparalleled in the industrialised world, Americans with less than a college degree experienced a real (inflation-adjusted) decline in their wages between 1980 and 2013, while those with postgraduate degrees experienced robust growth…

…Many Americans felt that they no longer had much of a political voice. In surveys, more than 80 per cent started saying that politicians did not care about what people like them thought…

…But perhaps the most important determinant of this dwindling trust in institutions was that the US had become much more polarised, making it increasingly difficult to satisfy the majority of the voters. The flames of grievance were powerfully fanned by social media, which deepened polarisation. This then further reduced trust in democracy and in public institutions. Worse, with intensifying distrust, something essential to democracy — compromise — became more and more challenging.

By the 2010s something unprecedented was happening. Ever since data on this had been collected, an overwhelming majority of Americans saw democracy as the “only game in town” and gave it strong support relative to alternatives such as monarchy, military dictatorship or rule by unelected experts. That began changing, especially among young people, who reported growing scepticism about democracy and much more lukewarm support for these institutions.

The cracks were visible long before Trump was first elected in November 2016. He was in many ways a symptom of those troubled times…

…Turning points are useful to locate because they are symbolic of deeper causes of social change. In hindsight, an obvious turning point came just before Trump’s second inauguration. Biden, who had four years ago made defence of democracy a main agenda item, pre-emptively pardoned his family and a number of politicians and public servants, including former Republican Congresswoman Liz Cheney and the former medical adviser to the president, Anthony Fauci. The optics were clear and ugly: Biden and his camp by this point had so little trust in US institutions that they thought only such pre-emptive pardons could stop Trump’s retribution (and making the reality worse than the optics, it was only the enemies of Trump who were close to Biden that counted)…

…While Trump’s domestic agenda intensified the loss of trust in US institutions and expertise in government, his relations with foreign allies did the same for the so-called rules-based order. Of course, there was some truth to critics’ contention that these rules were designed for America’s benefit and that when they didn’t serve it well, they were bent or broken by US politicians, diplomats and companies. But the world was not ready for Trump’s tariffs, threats and military expansionist rhetoric towards Panama, Greenland and even Canada.

This set the scene for a series of catastrophic governmental failures. With morale gone and key personnel fired, the US state was ill-equipped to deal with emergencies. When new pandemics arrived, the response was haphazard, and unpreparedness cost tens of thousands of lives. The few remaining independent media sources uncovered a glaring and dangerous lack of oversight of critical infrastructure, including nuclear reactors and cyber security.

But the real extent of the damage became clear only with the tech meltdown of 2030. Economists and historians have now shown that a lot of this was the outcome of institutional failures and growing concentration in the industry. After Trump lifted all roadblocks ahead of AI acceleration and cryptocurrency speculation, there was initially a boom in the tech sector. But within a few years the industry had become even more consolidated than before, and both insiders and outsiders came to realise that only companies favoured by the administration could survive…

…By late 2029, many commentators were questioning what was going on in the tech industry, which had invested heavily in AI but had little to show for this in terms of innovation or productivity growth. There was huge enthusiasm and investment in cryptoassets, which were one by one revealed to be scams costing regular Americans billions of dollars. The AI empire had no clothes by this point, because the competitive energy had been sucked out of it. It took a while longer for the market to realise that, but when it did, a massive stock market crash followed.

This is the kind of shock that a dynamic economy can recover from, with new innovators coming in, government experts using fiscal policy and other interventions to prevent the crash from translating into a deep recession, and all sorts of people still believing in their ability to make a difference. But once malaise about US institutions had sunk in and experts were no longer around in the government, the crash became a recession and then a depression.

The depression continued and intensified. Many now understood that institutions needed to be fixed, but after the damage that Biden and Trump had done and the polarisation that had reached even higher peaks, rebuilding them proved difficult. American innovators and scientists started emigrating to Canada and the European Union. Some even went to China.

America’s collapse thus followed Hemingway’s famous line on bankruptcy. It happened gradually, as shared prosperity, high-quality public services and the operation of democratic institutions weakened, and then suddenly, as Americans stopped believing in those institutions.

2. The Drug Industry Is Having Its Own DeepSeek Moment – David Wainer

In 2020, less than 5% of large pharmaceutical transactions worth $50 million or more upfront involved China. By 2024, that number had surged to nearly 30%, according to DealForma. A decade from now, many drugs hitting the U.S. market will have originated in Chinese labs…

…China’s biotech boom mirrors its rise in tech. In both cases, China has moved up the value chain, from manufacturing goods to becoming a more sophisticated hub for innovation, competing in industries once dominated by the U.S. There are several reasons for the industry’s growth. For one, many top scientists trained in the U.S. have returned to China over the past decade, fueling the emergence of biotech hubs around Shanghai. And just as DeepSeek built a formidable chatbot—allegedly on a lean budget with limited access to semiconductors—Chinese biotech companies are also scrappier, capitalizing on a highly skilled, lower-cost workforce that can move faster.

Additionally, companies can conduct clinical trials at a fraction of what they would cost in the U.S., while recent changes in the Chinese regulatory system have streamlined and accelerated the approval process to get a study started. 

For now, much of China’s biotech innovation is incremental rather than groundbreaking. Many companies focus on improving existing drugs—tweaking the chemistry, enhancing efficacy or differentiating them in key ways.

But Chinese innovation is steadily improving and is already starting to disrupt the U.S. drug-development ecosystem…

…Chief executives of large pharmaceutical companies are broadening their horizons. Why spend $10 billion acquiring a U.S. biotech with a mid-stage drug when a similar molecule can be licensed from China for a fraction of the price?…

…In late 2024, after scouring the market for obesity assets—presumably eyeing U.S. companies like Viking Therapeutics, which trades at a market value of around $3.7 billion—Merck chose to license an oral GLP-1 drug from China’s Hansoh Pharma. The deal: $112 million upfront, with potential milestone payments of up to $1.9 billion…

…These “bargain” deals are great for Big Pharma. But for U.S. biotech companies—and their venture-capital backers—they are creating real challenges. Investors increasingly struggle to value early-stage biotechs because it is difficult to predict what competition might emerge from China.

3. All of us could be wrong about DeepSeek and OpenAI – Chin Hui Leong

China’s DeepSeek has unleashed a new wave of AI hype.

But amid the noise, one thing is clear: everyone has an opinion, and no one has the answers….

…When Apple (NASDAQ: AAPL) unveiled its iPhone in 2007, many analysts dismissed its hardware-focused strategy.

Their argument hinged on a familiar pattern: over time, consumer hardware tends to become commoditised. If the iPhone becomes popular, they reasoned, its unique appeal would fade as competitors come in with cheaper imitations.

This wasn’t a baseless concern.

The personal computer (PC) era, the previous dominant computing platform, was marked by fierce price competition among hardware manufacturers. Even Apple’s Macintosh PC had fallen victim to the cutthroat competition in the 1980s and 1990s.

In short, the precedent was clear: hardware eventually becomes a commodity.

However, this time, things would be different.

Today, nearly 18 years later, Apple boasts over 2.35 billion devices in circulation, generating upwards of US$200 billion in annual iPhone revenue. Clearly, the popular smartphone has defied the conventional wisdom of hardware commoditisation.

Therein lies a lesson.

When considering the future of AI, the iPhone’s success serves as a crucial reminder: be wary of preconceived notions…

…Too often, we fall prey to the “Highlander” fallacy, assuming that one side can only win if the other loses.

This zero-sum mindset blinds us from a range of possible future scenarios.

Think about the mobile operating system (OS) market.

On one side, you’ve got Apple’s closed iOS, with 2.35 billion devices, and on the other, Google’s open-source Android, with a massive three billion devices.

Crucially, they’ve each found their own area to thrive in.

Apple continues to dominate in the premium smartphone market, while Android is all about getting Google services out there.

Going back to AI models: can OpenAI replicate this coexistence, thriving alongside open-source models?

Could we see large, proprietary models handling general use cases while smaller, specialised models address niche needs? Could there be a main AI model, featuring a supporting cast of smaller models?

Your guess is as good as mine…

…Do you know who were among the biggest “losers” in the shift from desktop to mobile?

In my book, it may be Microsoft and Nvidia.

Nvidia tried to break into the smartphone market but threw in the towel when it failed to get a foothold in the market. Microsoft, on the other hand, had long held a monopoly in the desktop OS market but failed to extend its dominance to mobile devices.

But are we really going to brand Microsoft and Nvidia as losers, even though they got the short end of the stick in the smartphone arena?

Today, both are at the forefront of the AI revolution, proving that setbacks don’t preclude future triumphs…

…Amid the noise, it’s important to remember that ChatGPT is barely two years old, a stark reminder of the industry’s infancy.

If history teaches us anything, we may want to put our egos aside and accept that there are developments that cannot be known ahead of time.

The AI landscape is still being written.

4. Deep Research and Knowledge Value – Ben Thompson

I found a much more beneficial use case the next day. Before I conduct a Stratechery Interview I do several hours of research on the person I am interviewing, their professional background, the company they work for, etc.; in this case I was talking to Bill McDermott, the Chairman and CEO of ServiceNow, a company I am somewhat familiar with but not intimately so. So, I asked Deep Research for help…

…I found the results eminently useful, although the questions were pretty mid; I did spend some time doing some additional reading of things like earnings reports before conducting the Interview with my own questions. In short, it saved me a fair bit of time and gave me a place to start from, and that alone more than paid for my monthly subscription.

Another compelling example came in researching a friend’s complicated medical issue; I’m not going to share my prompt and results for obvious reasons. What I will note is that this friend has been struggling with this issue for over a year, and has seen multiple doctors and tried several different remedies. Deep Research identified a possible issue in ten minutes that my friend has only just learned about from a specialist last week; while it is still to be determined if this is the answer he is looking for, it is notable that Deep Research may have accomplished in ten minutes what has taken my friend many hours over many months with many medical professionals.

It is the final example, however, that is the most interesting, precisely because it is the question on which Deep Research most egregiously failed. I generated a report about another friend’s industry, asking for the major players, supply chain analysis, customer segments, etc. It was by far my most comprehensive and detailed prompt. And, sure enough, Deep Research came back with a fully fleshed out report answering all of my questions.

It was also completely wrong, but in a really surprising way. The best way to characterize the issue is to go back to that famous Donald Rumsfeld quote:

There are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns — the ones we don’t know we don’t know.

The issue with the report I generated — and once again, I’m not going to share the results, but this time for reasons that are non-obvious — is that it completely missed a major entity in the industry in question. This particular entity is not a well-known brand, but is a major player in the supply chain. It is a significant enough entity that any report about the industry that did not include them is, if you want to be generous, incomplete.

It is, in fact, the fourth categorization that Rumsfeld didn’t mention: “the unknown known.” Anyone who read the report that Deep Research generated would be given the illusion of knowledge, but would not know what they think they know…

…What Deep Research reveals is how much more could be known. I read a lot of things on the Internet, but it’s not as if I will ever come close to reading everything. Moreover, as the amount of slop increases — whether human or AI generated — the difficulty in finding the right stuff to read is only increasing. This is also one problem with Deep Research that is worth pointing out: the worst results are often, paradoxically, for the most popular topics, precisely because those are the topics that are the most likely to be contaminated by slop. The more precise and obscure the topic, the more likely it is that Deep Research will have to find papers and articles that actually cover the topic well…

…There is a good chance that Deep Research, particularly as it evolves, will become the most effective search engine there has ever been; it will find whatever information there is to find about a particular topic and present it in a relevant way. It is the death, in other words, of security through obscurity. Previously we shifted from a world where you had to pay for the news to the news being fed to you; now we will shift from a world where you had to spend hours researching a topic to having a topic reported to you on command.

Unless, of course, the information that matters is not on the Internet. This is why I am not sharing the Deep Research report that provoked this insight: I happen to know some things about the industry in question — which is not related to tech, to be clear — because I have a friend who works in it, and it is suddenly clear to me how much future economic value is wrapped up in information not being public. In this case the entity in question is privately held, so there aren’t stock market filings, public reports, barely even a webpage! And so AI is blind…

…That, by extension, is why AI’s like Deep Research are one of the most powerful arguments yet for prediction markets. Prediction markets had their moment in the sun last fall during the U.S. presidential election, when they were far more optimistic about a Trump victory than polls. However, the potential — in fact, the necessity — of prediction markets is only going to increase with AI. AI’s capability of knowing everything that is public is going to increase the incentive to keep things secret; prediction markets in everything will provide a profit incentive for knowledge to be disseminated, by price if nothing else.

It is also interesting that prediction markets have become associated with crypto, another technology that is poised to come into its own in an AI-dominated world; infinite content generation increases the value of digital scarcity and verification, just as infinite transparency increases the value of secrecy. AI is likely to be the key to tying all of this together: a combination of verifiable information and understandable price movements may the only way to derive any meaning from the slop that is slowly drowning the Internet.

This is the other reality of AI, and why it is inescapable. Just as the Internet’s transparency and freedom to publish has devolved into torrents of information of questionable veracity, requiring ever more heroic efforts to parse, and undeniable opportunities to thrive by building independent brands — like this site — AI will both be the cause of further pollution of the information ecosystem and, simultaneously, the only way out…

…Secrecy is its own form of friction, the purposeful imposition of scarcity on valuable knowledge. It speaks to what will be valuable in an AI-denominated future: yes, the real world and human-denominated industries will rise in economic value, but so will the tools and infrastructure that both drive original research and discoveries, and the mechanisms to price it. The power of AI, at least on our current trajectory, comes from knowing everything; the (perhaps doomed) response of many will be to build walls, toll gates, and marketplaces to protect and harvest the fruits of their human expeditions.

5. AI and the Mag 7 – Daniel Rasmussen

Last summer, Goldman Sachs was estimating a $1T spend on AI capex in the coming years, and the numbers have only gone up since then, with most of it concentrated in the Mag 7 that dominate the public markets…

…It’s necessary as an investor to at least consider how these bets might go awry…

…The skeptic’s case starts with the possibility that the Mag 7 is suffering from a classic case of “competition neglect,” where “subjects in competitive settings overestimate their own skill and speed in responding to common observable shocks and underestimate the skill and responsiveness of their competitors,” as Robin Greenwood and Samuel Hanson put it in their paper, “Waves in Ship Prices and Investment.” When shipping prices increase, shipping companies all decide to invest in ships—after all, their models are all saying these investments will be profitable at current rates. That investment not only drives up the price of building new ships, it causes a glut of supply once they are built, resulting in poor returns on these pro-cyclical investments, as low as -36%, according to Greenwood and Hanson. Meanwhile, those who invest at the bottom of that cycle—when current shipping prices are low and there’s no one else building at the shipyards—earn returns as high as 24%.

Rather than ships, today’s AI capex “is a euphemism for building physical data centers with land, power, steel and industrial capacity,” as Sequoia Capital’s David Cahn puts it…

…OpenAI, SoftBank, and the federal government’s $500 billion Project Stargate is the culmination of this race to convert tech companies into industrial manufacturers. But even winning this race could be a Pyrrhic victory. Capex at these levels is an asset-heavy business model. Asset-heavy business models historically have lower returns on capital, especially when sunk costs meet increased competition.

In this scenario, perhaps Stargate is the AI equivalent of overinvesting in new ships at the same moment that everyone else is overinvesting in ships, leading to a supply glut, price drops, and poor investment returns…

…We still don’t have many economical use cases for AI. Even in low-compute mode, a single prompt on ChatGPT’s o3 model costs $20 to perform. High-compute mode can cost much more….

…While Anthropic CEO Dario Amodei is confident AI can beat humans at most things in 2-3 years, that doesn’t mean we will all be using AI that way. There’s a difference between what can be automated and what is cost-effective to automate. Daron Acemoglu, Institute Professor at MIT, estimates that only a quarter of AI-exposed tasks will be cost-effective to automate within the next 10 years. An MIT research paper looked at jobs in non-farm businesses and found 36% of tasks in jobs they studied could be automated by AI vision models, but only 8% were economically worth automating.

Scaling laws are an assumption that brute force will get us more and more powerful AI. For AI investors, it’s a playbook to outspend the competition, win the market, and trust that, eventually, more infrastructure and better chips will bring costs down and make more tasks economical to automate. But shooting for scale and achieving high ROI are not usually achieved at the same time.

Shortly after Stargate was announced, it was soon overshadowed by bigger news about China’s DeepSeek model. While the exact specs are a subject of debate, DeepSeek shattered the cost-to-performance expectations that investors and the Mag 7 have been working from…

…We’ve only just entered the true product-building era for AI. How many people today think of the internet as a product? The internet is not a single thing but a collection of services and products on common digital infrastructure (e.g., TCP/IP protocol, which was built by DARPA with US taxpayer money and isn’t a business anyone is making money on). Similarly, AI models could, like other commodities, utilities, and infrastructure projects, become a part of everything we use rather than a distinct product. Usage patterns are starting to reflect this: we are using these models less directly and more through other services built on top of them.

---

Disclaimer: The Good Investors is the personal investing blog of two simple guys who are passionate about educating Singaporeans about stock market investing. By using this Site, you specifically agree that none of the information provided constitutes financial, investment, or other professional advice. It is only intended to provide education. Speak with a professional before making important decisions about your money, your professional life, or even your personal life. We currently have a vested interest in Alphabet (parent of Google), Apple, and Microsoft. Holdings are subject to change at any time.

We’ve constantly been sharing a list of our recent reads in our weekly emails for The Good Investors.

Do subscribe for our weekly updates through the orange box in the blog (it’s on the side if you’re using a computer, and all the way at the bottom if you’re using mobile) – it’s free!

But since our readership-audience for The Good Investors is wider than our subscriber base, we think sharing the reading list regularly on the blog itself can benefit even more people. The articles we share touch on a wide range of topics, including investing, business, and the world in general. 

Here are the articles for the week ending 09 February 2025:

1. Robert Litan: An Economist Walks Into a Bar at TEDxKC (Transcript) – Robert Litan

First guy, he approaches the first woman that he sees, offers her a drink. She turns him down. He, then, decides to walk his way down the bar. And, of course, all the women watching this, they see what he’s up to. And they all turn him down…

…He hasn’t learned from this experience, in the real world. So he decides to go to the virtual world. He goes to the Internet and joins Cupid.com and he tries the same technique, and sure enough, with the same result. They all turn him down…

…Cupid.com is in trouble too. And the reason they are, is that the women who have joined Cupid.com are being inundated with offers for men for dates. They get turned off, they quit. And if they quit, men quit. Cupid is in trouble. Who are you going call, to solve this problem. Know the answer is more obvious than ghost busters. You call an economist. Don’t laugh, you call economists. In fact, you call two of them.

This is Muriel Niederle of Stanford, and Dan Ariely of Duke. And they spend a lot of time, studying the problem of artificial scarcity and abundance, in the online dating context, which is a reason Cupid call them up. And they wanted to know how to fix their problem and the two economists said they had an idea, that was as simple as it was profound. Just put a sharp limit on the number of date offers that men could make to women each month. This is the notion of artificial scarcity. Taking what looks like an abundant resource, which is date offers, and artificially constraining them.

And the economists said to Cupid that if you do this, the men will take their offer seriously. They’ll look at more than just the women’s pictures and they’ll actually look at their profiles. And the women will know this, and they’ll be more likely to accept date-proposals. Artificial scarcity helped save Cupid.com, and other dating sites that copied the technique…

…Google collects about $50 billion a year, from advertisers, large and small, seeking placement on that right hand side. They auction off the site. But that’s not how the system started, because when Google was launched, online advertising was in its infancy, and Google, believe it or not, went door to door, advertiser to advertiser, trying to get them to place for an ad next to a search term. Highly laborious, you quickly can see that this is not going to scale, as the number of searches explodes on Google.

And so the founder of Google asked two young engineers, Eric Veach and Salar Kamangar, to come up with an automatic system, that would solve this problem. Well, they were instinctively attracted to auctions. But they were thinking about another problem. That is if they auction off the sites, they fear that the advertisers would bid a very low price, and then incrementally raise their prices just a little bit, and keep the auctions going on forever. And if this happened, and a lot of searches were also going on at the same time, the whole site would crash.

So, as an engineering solution, they came up with this idea. That the winning auction, or the winning placement will be the price, the second highest price that was bid plus one penny. This will cut off the auctions, greatly simplify the process, and in the process also solve another problem called “the winner’s curse“. I’m sure that many of you that have participated in auctions may have regretted winning because you felt like you’ve paid too much. Pretty obvious point…

…“You know, those two engineers, they have reinvented what this guy came up with.” This is William Vickrey, he was an economist at Colombia, who proved mathematically, that the second price auction was the ideal solution to the winner’s curse. And you know what, that won him the Nobel Prize in Economics in 1996.

2. Emergent Layers, Chapter 2: Overserved and Underserved Customers – Alex Danco

Returning to disruption theory, the critical element we’re going to use from that framework is the idea of the overserved customer: the customer who is being served too much by incumbents. In mature industries, where everybody agrees what the scarce resource is and the core constraints are well understood and organized around, we see this happen a lot. As incumbent companies compete with each other for business, and customers are all being served adequately (for the understood job at hand), competition becomes a feature race where products improve or expand at a greater rate than customers’ capacity to use them. There’s a misalignment between what the customer needs and is getting, with that misalignment falling onto the side of “I’m spending way too much of my money or time for this.” Crucially, when customers are overserved for a particular job, it introduces the critical space and oxygen required for a new competitor with some sort of scalable, technological advantage to enter the market at the low end. The nature of over-service creates powerful incentives for incumbents to not engage with disruptive entrants, but rather to retreat upmarket towards higher profit margins…

…For a more recent but still “classic” example, let’s look at Airbnb. Airbnb was able to get off the ground because there was a critical subset of customers in the hospitality industry — initially young people, although not exclusively so — who were overserved by many aspects of the hotel industry. Hotels were serving customers along many axes of performance — comfort, privacy, loyalty reward programs, and so forth — that just weren’t very important to a specific subset of customers who didn’t care too much about all that stuff; they just want a place to stay. This gave Airbnb the critical oxygen necessary to get a foot in the door, and then expand upwards from a dramatically cheaper cost structure than Marriott can possibly compete with. The overserved customer is a very potent and dangerous one: they know what they’re looking for, and they don’t need to be educated when a new entrant comes along with the right proposition. If that new entrant gets a few critical things right, they’re looking at a large group of early adopters that need little prodding, little education and little advance notice. That’s a great basis to start a company.

Let’s now consider another kind of pain: underserved customers. Their pain appears to be more straightforward: they have some fundamental need that isn’t being met. But this situation is trickier than it seems: if a group of customers have a genuine need, then why aren’t companies stepping in to offer solutions? What’s the catch? It could be because the solutions are genuinely too hard, or face technical or feasibility obstacles. It could also be that customers aren’t aware they have the problem. Either way, that’s tough…

…Now let’s put these two types of customer pain together. What would happen if a customer were both overserved and underserved at the same time? Is this possible?

As it turns out, this situation is not only possible, but occurs regularly. And it’s highly volatile. The trick to figuring out how this works requires venturing one step beyond disruption theory, and recasting the job-to-be-done as a stack itself with a hierarchy of low-level to high-level needs…

…We can characterize the initial job where customers are being served as being at level j, where incumbents vie for customer dollars and products will inevitably trend towards over-service. Meanwhile, we can characterize the higher-order job as being at level j+1, which encompass the customer’s higher level objectives, and where companies are not, for whatever reason, currently serving anyone…

…Consider Uber: you have a large group of customers (myself included) who are overserved by owning their own vehicle. If your car sits idle & parked more than 95% of the time (which is about average in North America), you are clearly overserved by owning this car! Yet at the same time, that same set of customers is underserved at level j+1, or the reason why they own a car in the first place — “I need to get to specific places at specific times”. You have a schedule to keep, and it’s hard.

Notice that both of these conditions must hold true in order for Uber to work. If customers were not overserved, it would be difficult for them to abandon their current solution. (Consider someone who drives their vehicle for a living, many hours per day. They are significantly less overserved by their vehicle, and quite unlikely to switch to using Uber for the equivalent job.) At the same time, if they weren’t underserved for a higher-level job (get me places at a certain time), then the only way for a new solution to be truly compelling would be dramatically lower price — which makes for a tough business model. This is another thing outside observers get wrong about Uber when they exclaim, “I don’t see how this is cheaper than owning a car!” Well, here’s the thing — Uber doesn’t have to be cheaper than driving, because it’s superior to driving your own vehicle in many ways! You don’t have to worry about parking, insurance, drinking, maintenance, gas, or anything else. The simultaneous condition of being overserved and underserved by existing solutions is what made Uber so compelling, in a way that other ride-sharing services or carpooling didn’t quite get right. Uber works because it’s cheap, but its appeal is because it’s better…

…If customers only check off the “underserved” box, then it seems likely you’re dealing with a problem that’s a. very hard, or b. the customer isn’t aware they have. This isn’t a great position to be in — it’ll be very hard to build an initial solution and attract early adopters.

If they only check off the “overserved” box, then customers know what they want — but it may be that they’re only motivated by price. And that’s also not a great position to be in: you may get lots of adopters really quickly, but find it very difficult to extract any profit from them…

…The particular combination of customers overserved at level j while being underserved at level j+1, when it happens, explains how from time to time we see markets where the demand is zero and then all of a sudden a vertical line straight up.

3. Why Housing May Be In for Another Cost Shock Next Year – Tracy Alloway, Joe Weisenthal, and Lee Everett

Lee (04:44):

It’s interesting. I think stress is hitting sort of all sides of the market. You have your bigger, more well established shops that have been managing through this, able to handle the higher rate environment, but have obviously taken a very real valuation hit on their existing portfolios. Like 20% to 30% depending upon the portfolio composition. At the same time you’ve had record demand hitting the sector because cost to buy housing is exceptionally unattainable today. And then on the other side you’re having a very material impact on the supply side and I think that’s what’s really unique. If you think back to September, the 10-year was around a 3.6%, I think, the day Chair Powell cut us by 50 basis points. Well, we’re at almost a 4.6% today and I remember that night you heard reports about developers out at local dinners and they were calling it Fed Day and getting ready to put shovels in the ground.

Joe (05:37):

Drinking champagne and stuff like that.

Lee (05:38):

Exactly. And what you’ve seen instead is increased stress on both the short end and the long end of the curve. That’s given you trouble on the short end, to start new housing, and trouble on the long end to afford longer term for ownership housing…

…Lee (11:29):

Yes, I think frankly we’re about to transition from what has been a very renter friendly market to again a landlord friendly market over the course of the next two to three years. And that’s going to be particularly driven by what we’re seeing on the supply side. We’re going to have over a million units come to market over a two-year period here in ’24 and ’25, but peak supply is hitting in the next six months and if you look at relative time from a) peak supply and then b) to getting to a level of lower supply than you saw last cycle, every major market in the country will be there by the end of 2026.

Joe (12:13):

Be where?

Lee (12:15):

Delivering less housing units than they did on average from ’17 to ’19 in apartment buildings. So you’re going to go below prior cycle supply very quickly. At the same time, we do have exceptionally strong labor markets here and the demand story has been outstanding. So 2024 is going to end the year, depending upon the data provider you use, as the first or third highest year for rental demand ever. 2021 was the prior record. So we’re seeing people form rental households at unprecedented rate in the US and as that supply comes down, you’re going to see that demand struggle to frankly find high quality, well-located assets to move in, and you’re likely to see that relationship flip at that point.

Tracy (13:08):

So the other thing that affects multifamily housing construction other than interest rates has to be just general confidence, I guess, in the direction of the economy, the direction of the world and certainly there’s a lot going on right now. We’re recording this on January 28th and there’s news that the Trump administration is freezing a whole bunch of federal spending. I think it’s something like 20% of federal spending. That includes presumably stuff like Section 8 and other affordable housing measures. Would that be expected to hit multifamily as well?

Lee (13:46):

Yeah, and I think it’s probably easiest to sort of start at the top, right? When you’re building multifamily, you’re generally trying to build to an acceptable return on cost, but frankly what we’re doing is putting an investor’s money together and generating returns for them. Multifamily isn’t built for free and it can’t be in this sort of economic world and a general rule of thumb is a 6+% return on cost. So cost to build, you want to yield over 6% of that to get a building to pencil. That tracks up closer to 7% depending upon the institution, because you need to build to that yield on cost, you have to have rents that are high enough to generate enough rental revenue to drive that return. So in order to build today, you have to build it exceptionally high rent levels, because of the cost to build, because of the cost of interest rates.

The only way to drop that is to drop the cost and that cost drop typically comes for affordable housing from the federal government, be it HUD grants that are then deployed through the local housing agency, be it LIHTC, be it any sort of an ensemble of ways to cut costs. That’s how you can get to affordable rents on the supply side. And then on the demand side, you can cut rents by literally giving people a rent check, which is what Section 8 is. And that again comes from the federal government via grants given to the local housing agencies to deploy. And if that money dries up, you have immense problems in terms of a) fueling the demand for these people, because you’re cutting rent on the Section 8 side and b) encouraging future construction of affordable apartment buildings…

…Joe (17:47):

Let’s talk about deportation impacts on labor. What are the estimates for what percentage of the multifamily workforce, whether it’s construction or maintenance, whatever else, is undocumented labor?

Lee (18:01):

It’s estimated 20% of construction workers in this country are undocumented labor. I’d venture to guess it’s similar for the whole multifamily industry when you look at staffing and things along those lines, and I think when you look at a combination of deportation of construction workers as well as the sheer amount of labor it’s going to require to rebuild huge swaths of California, I think you could be looking at a massive deficit in labor within the construction space. And when you think about that, that’s going to be your strongest lever that’s going to hit your cost to build and that’s what’s going to drive up those rents that are necessary. Is all of this immense pressure you’re going to see in the labor costs.

4. Test-Time Search: A Path To AGI – Akash Bajwa

The GPT family of models performed poorly relative to o3 on the ARC benchmark because large models memorise knowledge rather than reasoning processes…

…As an example, Meta intentionally overtrained Llama 3 on 15 trillion tokens to lower inference costs (as they served their billions of users). The model weights become more optimised for common patterns and in-distribution tasks, trading off generalisability to novel tasks.

This architecture combined with ‘internet scale’ data has produced incredible recent advances, but the next leap will come from a new paradigm – instead of outputs, models will be trained on reasoning steps…

…This new vector of scaling will rely on a combination of synthetic and human generated reasoning data. As we’ll see, both will be expensive forms of reinforcement learning (o3’s performance of 87.5% on ARC AGI in high-compute mode cost thousands of $ per task)…

…Synthetic data will be most useful for domains where functional verification is possible, e.g. code, maths and engineering…

…Scaling inference time compute is in line with the Bitter Lesson – there are only 2 techniques that scale indefinitely with compute: learning & search.

DeepMind’s AlphaGo used Monte Carlo Tree Search during test time to attain superhuman status – if stripped of this capabilities, it drops in Elo from ~5,200 to 3,000 (top humans are around ~3,800)…

…The exorbitant costs stem from the many, many Chains Of Thought generated as the model searches for the chains that lead to the right answer – all of the other tokens are useless, but cost a lot to generate…

…Functionally verifiable domains are the most amenable to synthetic CoTs because engineering the reward is much easier than in domains where subjectivity is involved…

…Code execution provides an unambiguous, binary reward signal – either the code executes successfully or it fails, creating clearly defined success criteria for training.

In functionally verifiable domains, the correct CoT tokens become training data…

…Over time, this should have a deflationary effect on the inference cost of reasoning models, as we’ve seen with frontier models in the pre-training paradigm…

…As pre-training gains plateau (or become too expensive), we’ve found a new vector of scaling (test time search) that is demonstrating a path to truly general intelligence.

Data acquisition/generation remains the bottleneck on progress, not compute. Microsoft’s announcement of $80bn in capex for 2025 underscores the Street’s underestimation of hyperscaler capex and compute buildout.

The implications of inference scaling run up and down the stack. Instead of the densely interconnected supercomputers of the pre-training paradigm, we will see more distribution of workloads, perhaps some even running locally. How will market share evolve as companies look to optimise test time search workloads – will AI ASICs eat into Nvidia market share?

Instead of prohibitively expensive pre-training runs, enterprises developing their own models may opt to train smaller models with reasoning cores and decide when to scale up test time search for certain economically valuable tasks. The result is the alchemy of capex to opex and fixed costs to variable costs. CIOs will decide which tasks merit more investment and test time search – inevitably, this will still be cheaper than human labour.

5. Don’t Freak Out – Ben Carlson

The common theme across the Apollo missions was the sheer amount of planning involved.  There were months and months of simulations and training exercises to review every possible scenario. They wanted every process to be automatic.

But there was always the risk of an unplanned error, considering they were propelling these giant hunks of metal through space using rocket fuel that would allow them to reach speeds of more than 24,000 miles per hour…

…When Apollo 13 had an explosion mid-flight, it wasn’t something anyone thought could have been even a remote possibility. Astronaut Jack Swigert explained it after the fact like this:

Nobody thought the spacecraft would lose two fuel cells and two oxygen tanks. It couldn’t happen. If somebody had thrown that at us in the simulator, we’d have said, ‘Come on, you’re not being realistic.’

This is why NASA trained the astronauts in one skill more than any other leading up to their space flights — the art of not panicking. The only reason they could turn the Apollo 13 spacecraft around 200,000 miles from earth following an explosion onboard is because the astronauts and everyone on the ground remained levelheaded. No one freaked out.

Or if they were freaking out internally, they didn’t act on those emotions.

In a nutshell, that is successful investing.

---

Disclaimer: The Good Investors is the personal investing blog of two simple guys who are passionate about educating Singaporeans about stock market investing. By using this Site, you specifically agree that none of the information provided constitutes financial, investment, or other professional advice. It is only intended to provide education. Speak with a professional before making important decisions about your money, your professional life, or even your personal life. We currently have a vested interest in Alphabet (parent of Google), Meta Platforms, and Microsoft. Holdings are subject to change at any time.

We’ve constantly been sharing a list of our recent reads in our weekly emails for The Good Investors.

Do subscribe for our weekly updates through the orange box in the blog (it’s on the side if you’re using a computer, and all the way at the bottom if you’re using mobile) – it’s free!

But since our readership-audience for The Good Investors is wider than our subscriber base, we think sharing the reading list regularly on the blog itself can benefit even more people. The articles we share touch on a wide range of topics, including investing, business, and the world in general. 

Here are the articles for the week ending 02 February 2025:

1. DeepSeek: The View from China – Jordan Schneider, Irene Zhang, Angela Shen, and Yiwen

In this newsletter, we share a translation of insights from a January 26 closed-door session hosted by Shixiang 拾象, a VC spun out from Sequoia China. Attended by dozens of AI researchers, investors, and industry insiders, the event captures how the Chinese AI community is processing the DeepSeek shock…

…The CEO of Scale.ai said that DeepSeek has 50,000 chips, but that is definitely not reality. According to public information, DeepSeek had 10,000 old A100 chips and possibly 3,000 H800 cards before the ban. DeepSeek pays great attention to compliance and has not purchased any non-compliant GPUs, so it should have few chips. The way the United States uses GPUs is too extravagant…

…In the short-term, everyone will be driven to think about how to make AI more efficient. In the long-run, questions about computing power will remain. Demand for compute remains strong and no company has enough…

…Why did DeepSeek catch up so fast?

Reasoning models require high-quality data and training. For LLMs or multimodal AI, it’s difficult to catch up with a closed source model from scratch. The architecture of pure reasoning models hasn’t changed much, so it’s easier to catch up in reasoning.

One reason R1 caught up quickly was that the task was not particularly difficult. Reinforcement learning only made the model choices more accurate. R1 did not break through the efficiency of Consensus 32, spending 32 times the efficiency, which is equivalent to moving from deep processing to parallelization, which is not pushing the boundaries of intelligence, just making it easier….

…AI is similar to a step function, where the compute requirements for followers have decreased by a factor of 10. Followers have historically had lower compute costs, but explorers still need to train many models. The exploration of new algorithms and architectures will not stop. Behind the step function, there are significant investments by many people, meaning compute investments will continue to advance. Many resources will also be allocated to products. Apart from reasoning, there are other directions that are compute-intensive. While the vast amount of compute resources spent by explorers may not be visible, without such investment, the next “step” might not occur. Additionally, many are dissatisfied with current architectures and RL methods, and progress will continue.

When exploring directions, performance achieved with 10,000 GPUs may not always be significantly better than that of 1,000 GPUs, but there is a threshold somewhere. It’s unlikely that meaningful results can be achieved with only 100 GPUs because the iteration time for each solution would be too long…

…The question of why OpenAI and Anthropic did not do work in DeepSeek’s direction is a question of company-specific focus. OpenAI and Anthropic might have felt that investing their compute towards other areas was more valuable.

One hypothesis for why DeepSeek was successful is that unlike Big Tech firms, DeepSeek did not work on multi-modality and focused exclusively on language. Big Tech firms’ model capabilities aren’t weak, but they have to maintain a low profile and cannot release too often. Currently, multimodality is not very critical, as intelligence primarily comes from language, and multimodality does not contribute significantly to improving intelligence…

…2025 will, first and foremost, see interest in new architectures beyond Transformers. Some initial exploration is already underway, aiming to reduce costs while pushing the boundaries of intelligence. Secondly, the potential of reinforcement learning (RL) has yet to be tapped into completely. On the product side, there is significant interest in agents, though they have yet to see widespread application…

…It is reported that Meta is still in the process of reproducing DeepSeek, but so far, this has not significantly impacted their infrastructure or long-term roadmap. In the long run, beyond exploring the boundaries of the technology, cost efficiency must also be considered. Lowering costs will let us have more fun…

…From the developer’s perspective, models like Claude-3.5-Sonnet have been specifically trained for tool use, making them highly suitable for agent development. In contrast, models like DeepSeek have not yet focused on this area, but the potential for growth with DeepSeek is immense…

…Currently, reinforcement learning (RL) solves problems with standard answers but has not achieved breakthroughs beyond what AlphaZero accomplished. In fact, it is often simpler. Distillation addresses problems with standard answers, and RL methods work effectively when training with such answers. This explains why distillation and RL have made rapid progress in recent years.

Humanity’s demand for intelligence is vastly underestimated. Many critical problems, such as cancer and SpaceX’s heat shield materials, remain unsolved. Existing AI primarily automates tasks, but there are numerous unsolved challenges ahead. Looking forward, the potential for explosive growth is immense, and the advancement of intelligence cannot stop…

…Domestic Chinese companies were previously constrained by computing power, but now it’s proven that the potential technical space is vast. For more efficient models, we might not need especially large cards — we can provide relatively customized chips that can be adapted for compatibility with AMD and ASIC. From an investment perspective, Nvidia’s moat is very high, but ASIC will have yet greater opportunities.

The DeepSeek situation isn’t really about compute — it’s about America realizing China’s capabilities and efficiency. DeepSeek isn’t Nvidia’s vulnerability; Nvidia will grow as long as AI grows. Nvidia’s strength is its ecosystem, which has been built up over a long time. Indeed, when technology develops rapidly, the ecosystem is crucial. The real crisis comes, though, when technology matures like electricity: it becomes commoditized; then, everyone will focus on products, and many ASIC chips will emerge for specific scenario optimization…

…Open source controls the margins of the whole market. If open source can do 95% of what closed source can do and closed source is too expensive, then open source can be used completely. If the capabilities of open source and closed source do not differ greatly, then this presents a big challenge for closed source…

…AI explorers definitely need more computing power; China, as a follower, can leverage its engineering advantages. How Chinese large-model teams use less computing power to produce results, thereby having some definite resilience — or even doing better — might end up being how the US-China AI landscape plays out in the future.

2. Explaining International Valuations –  Daniel Rasmussen

Perhaps the single greatest divergence in equity markets has been the continued outperformance of US versus international equities—and thus the widening of the valuation gap between the US and the rest of the world…

…By far the most significant difference, explaining about half the valuation gap, is the domicile of listing. US-listed stocks are substantially more expensive than internationally listed stocks for no reason other than the place of listing.

It’s particularly interesting that the regression shows having a higher percentage of sales in the US results in cheaper valuations. A key driver of this is that several of the US tech giants most responsible for high US equity valuations having a relatively low percentage of sales in the US (Alphabet, Microsoft, and Tesla at around 50%; Apple, Netflix, Meta, and NVIDIA at around 40%). The big question, then, is why half the valuation gap is explained simply by being listed on US exchanges. Even large internationally listed companies with >40% of their revenue coming from the US, like Toyota, Mitsubishi, Roche or Deutsche Telekom (which owns T-Mobile), trade at steep value multiples relative to US peers.

Were a larger percentage of the valuation gap explained by fundamentals, we’d expect such a gap to persist. But given that the valuation gap is primarily explained simply by the location of listing, we think there’s a strong reason to expect a convergence—and therefore to favor international over US-listed stocks, despite their terrible relative performance over the past decade.

3. The Most Impressive Prediction of All Time – Jeffrey Emanuel

My candidate for the most impressive prediction of all time came from a person who is practically unknown in the West except for a relatively small group of historians and people interested in niche subjects. The person I’m thinking of is named Pyotr Durnovo, and he was an Imperial Russian government official who lived from 1842 to 1915.

We will discuss more about him later and how his life experience may have prepared him to be able to make such an impressive prediction, but the short version of it is that he initially studied to be in the Navy and served there for around a decade, and then became the Director of Police for the Ministry of Internal Affairs for the entire Russian Empire under Tsar Alexander III. Later, he served as the Minister of the Interior under Tsar Nicholas II (the one who was ultimately executed with his family by the Bolsheviks in 1917 during the Russian Revolution).

So what is this prediction he made, anyway, and why is it so impressive? Well, in 1914, six months prior to the outbreak of World War 1, Durnovo wrote a truly remarkable ~7,600-word memorandum for Tsar Nicholas II and his top 2 or 3 ministers, which we know was given to them, since it was found in Nicholas’ papers and later published in 1922 by communist historians after the revolution. If they had only read it carefully and took its warnings more seriously, the world we live in today might look very different!…

…For one, it predicted an imminent war on the horizon, which he ultimately blamed on the collision course between England and Germany, which were the two greatest industrial powers at the time. This was certainly not some earth shattering or special prediction; a lot of people predicted some kind of big conflict, and it was often said that “war was in the air” at the time…

…It’s how he analyzed the situation, and then used that reasoning to predict the exact groupings of countries that would participate in the conflict and on which side, and how the situation would evolve from there, that is so impressive…

…His predictions about alliances and national behaviors were almost unbelievably specific and ran counter to the conventional wisdom of the time:

• He predicted that Italy would not side with Germany despite being part of the Triple Alliance, and would instead join the opposing side if victory seemed likely, seeking territory from both Austria and Turkey. This is exactly what happened; Italy joined the Allies in 1915 after negotiating for territorial concessions.
• He predicted that Romania would remain neutral until it was clear which side would win, then join the victorious side to claim territory. This also came true— Romania entered the war in 1916 on the Allied side after significant Russian successes.
• Most surprsingly, he predicted that Bulgaria would side against Serbia and by extension against Russia, despite Russia being Bulgaria’s historic liberator from Ottoman rule— a prediction that seemed almost unthinkable to most observers at the time. This came true exactly as he foresaw, with Bulgaria joining the Central Powers in 1915.
• He correctly predicted that Serbia and Montenegro would side against Austria, while Greece would likely remain neutral until the outcome was more or less predetermined.
• He predicted unrest among Muslims in the Caucasus and Turkestan (which occurred).
• He predicted the possibility of Afghanistan moving against Russia (which happened in 1919).
• He predicted serious complications in Poland (the Polish-Soviet War of 1919-1921).
• He predicted an uprising in Finland if Sweden joined Germany (Finland did declare independence in 1917)…

…If all of that weren’t already so ridiculous to get right, he went way beyond all that to realize that, regardless of who won, the war would lead to “social revolution” in both the defeated AND victorious countries, starting with the losing side and then spreading to the winners. This was perhaps his most extraordinary prediction, as it came true in spectacular fashion:

• Russia, despite being on the winning side, experienced the Bolshevik Revolution in 1917; we will go into much more detail about these predictions below.
• Germany, after losing the war, experienced the German Revolution of 1918-1919; Durnovo predicted that unrest and revolution would be specifically tied to economic factors and class interests rather than purely political ones: he outlined how German workers would turn against the agricultural interests that had dominated pre-war German policy once defeat cut off their export markets and industrial employment, and this exact dynamic played out in the German Revolution of 1918-1919.

Now, you might object here that “Well, it’s not that crazy to believe there might be a revolution in a country which suffered massive losses in a catastrophic war; lots of people might have predicted that.” But the thing is, Durnovo went so far beyond merely predicting that there would be a Russian Revolution. He basically predicted every contour of the Revolution, the driving forces behind it, how it impacted different segments of Russian society, and how it would all unfold, step by step!…

…So how was Durnovo able to accomplish this incredible feat of prediction? Obviously, he was a genius of the first order, which is perhaps not so surprising given that he was a close relative of the famous Tolstoy family. But raw IQ is certainly not enough, nor is being well informed and knowledgeable. What kind of man could see so clearly what virtually everyone else missed? He was a complex character whose very contradictions likely enabled his extraordinary insights; he was, at the same time:

• A conservative police chief who often expressed liberal thoughts in private
• A supposed reactionary who opposed anti-Semitic measures and defended Jews
• A cynical operator who nevertheless would help others when he could
• A man capable of both strict officialdom and surprising gentleness
• A high official who preferred informal interactions (his subordinates would warn visitors not to address him as “Your Excellency”)

These contradictions suggest someone who wasn’t bound by conventional ideological frameworks or social expectations— a crucial trait for seeing beyond accepted wisdom. He also had a wide range of professional experience that prepared him to see things in a multi-faceted, sophisticated way, as by 1915, he had done the following:

• Naval officer (9 years of far-sea cruises)
• Military legal training
• Assistant Prosecutor in various parts of Russia
• Director of Police Department for 10 years
• Assistant Minister of Interior under multiple ministers
• Minister of Interior
• Member of State Council

This combination of experiences was extraordinary and atypical to say the least:

• His naval and legal background gave him insight into the military, maritime trade, and the Russian legal system.
• His prosecutorial work exposed him to conditions across Russia, not just in the big cities.
• His police work gave him unparalleled insight into social discontent and the strategies and thinking of professional revolutionaries like Lenin, Stalin, and Trotsky.
• His ministerial positions showed him the workings (and limitations) of state power.

He also occupied a unique position as both an insider and an outsider: 

• He was from old nobility but not wealthy or particularly influential
• He reached high office but was temporarily dismissed in disgrace (a sordid story in which Durnovo had his secret police officers search the private letters of a foreign ambassador— inside an embassy building no less— so they could steal love letters sent by Durnovo’s mistress to the ambassador; when the ambassador complained to Tsar Alexander III, he was furious, ordering his minister to “remove this swine within twenty-four hours.”)
• He was a conservative who often disagreed with other conservatives
• He understood both state power and its limitations

This dual perspective may have freed him from the groupthink that afflicted both conservative and liberal circles.

4. USA, Inc – Michael Batnick

Consider this face blower of a stat from Goldman: “Since 1992, earnings growth in the US has outpaced earnings in non-US developed economies by an annual average of 2.4 percentage points.”

Most of the world is barely earning more than they were prior to the pandemic. The U.S. looks like an unstoppable freight train…

…The one sided performance has driven valuations between us and the rest of the world to record levels. We’ve all seen a version of these charts before…

…BUT! These charts aren’t comparing apples with apples. Goldman notes that only 1% of the U.K. market is in technology companies. Another example they cite is that energy is 5% of S&P 500 earnings, 19% of UK, and just 1% of Japan. We’re not comparing apples with apples.

They did a great job adjusting for differences in sector weights…

…The U.S. still trades at a premium to the rest of the world ex-India, but not as much as the prior chart would have you believe. Before any adjustments, the Eurozone trades at a 39% discount to the U.S. And after the adjustments, that falls to 23%.

5. DeepSeek FAQ – Ben Thompson

Let’s work backwards: what was the V2 model, and why was it important?

The DeepSeek-V2 model introduced two important breakthroughs: DeepSeekMoE and DeepSeekMLA. The “MoE” in DeepSeekMoE refers to “mixture of experts”. Some models, like GPT-3.5, activate the entire model during both training and inference; it turns out, however, that not every part of the model is necessary for the topic at hand. MoE splits the model into multiple “experts” and only activates the ones that are necessary; GPT-4 was a MoE model that was believed to have 16 experts with approximately 110 billion parameters each.

DeepSeekMoE, as implemented in V2, introduced important innovations on this concept, including differentiating between more finely-grained specialized experts, and shared experts with more generalized capabilities. Critically, DeepSeekMoE also introduced new approaches to load-balancing and routing during training; traditionally MoE increased communications overhead in training in exchange for efficient inference, but DeepSeek’s approach made training more efficient as well.

DeepSeekMLA was an even bigger breakthrough. One of the biggest limitations on inference is the sheer amount of memory required: you both need to load the model into memory and also load the entire context window. Context windows are particularly expensive in terms of memory, as every token requires both a key and corresponding value; DeepSeekMLA, or multi-head latent attention, makes it possible to compress the key-value store, dramatically decreasing memory usage during inference.

I’m not sure I understood any of that.

The key implications of these breakthroughs — and the part you need to understand — only became apparent with V3, which added a new approach to load balancing (further reducing communications overhead) and multi-token prediction in training (further densifying each training step, again reducing overhead): V3 was shockingly cheap to train. DeepSeek claimed the model training took 2,788 thousand H800 GPU hours, which, at a cost of $2/GPU hour, comes out to a mere $5.576 million.

That seems impossibly low.

DeepSeek is clear that these costs are only for the final training run, and exclude all other expenses; from the V3 paper:

Lastly, we emphasize again the economical training costs of DeepSeek-V3, summarized in Table 1, achieved through our optimized co-design of algorithms, frameworks, and hardware. During the pre-training stage, training DeepSeek-V3 on each trillion tokens requires only 180K H800 GPU hours, i.e., 3.7 days on our cluster with 2048 H800 GPUs. Consequently, our pre- training stage is completed in less than two months and costs 2664K GPU hours. Combined with 119K GPU hours for the context length extension and 5K GPU hours for post-training, DeepSeek-V3 costs only 2.788M GPU hours for its full training. Assuming the rental price of the H800 GPU is $2 per GPU hour, our total training costs amount to only $5.576M. Note that the aforementioned costs include only the official training of DeepSeek-V3, excluding the costs associated with prior research and ablation experiments on architectures, algorithms, or data.

So no, you can’t replicate DeepSeek the company for $5.576 million.

I still don’t believe that number.

Actually, the burden of proof is on the doubters, at least once you understand the V3 architecture. Remember that bit about DeepSeekMoE: V3 has 671 billion parameters, but only 37 billion parameters in the active expert are computed per token; this equates to 333.3 billion FLOPs of compute per token. Here I should mention another DeepSeek innovation: while parameters were stored with BF16 or FP32 precision, they were reduced to FP8 precision for calculations; 2048 H800 GPUs have a capacity of 3.97 exoflops, i.e. 3.97 billion billion FLOPS. The training set, meanwhile, consisted of 14.8 trillion tokens; once you do all of the math it becomes apparent that 2.8 million H800 hours is sufficient for training V3. Again, this was just the final run, not the total cost, but it’s a plausible number.

Scale AI CEO Alexandr Wang said they have 50,000 H100s.

I don’t know where Wang got his information; I’m guessing he’s referring to this November 2024 tweet from Dylan Patel, which says that DeepSeek had “over 50k Hopper GPUs”. H800s, however, are Hopper GPUs, they just have much more constrained memory bandwidth than H100s because of U.S. sanctions.

Here’s the thing: a huge number of the innovations I explained above are about overcoming the lack of memory bandwidth implied in using H800s instead of H100s. Moreover, if you actually did the math on the previous question, you would realize that DeepSeek actually had an excess of computing; that’s because DeepSeek actually programmed 20 of the 132 processing units on each H800 specifically to manage cross-chip communications. This is actually impossible to do in CUDA. DeepSeek engineers had to drop down to PTX, a low-level instruction set for Nvidia GPUs that is basically like assembly language. This is an insane level of optimization that only makes sense if you are using H800s.

Meanwhile, DeepSeek also makes their models available for inference: that requires a whole bunch of GPUs above-and-beyond whatever was used for training…

…Is this why all of the Big Tech stock prices are down?

In the long run, model commoditization and cheaper inference — which DeepSeek has also demonstrated — is great for Big Tech. A world where Microsoft gets to provide inference to its customers for a fraction of the cost means that Microsoft has to spend less on data centers and GPUs, or, just as likely, sees dramatically higher usage given that inference is so much cheaper. Another big winner is Amazon: AWS has by-and-large failed to make their own quality model, but that doesn’t matter if there are very high quality open source models that they can serve at far lower costs than expected.

Apple is also a big winner. Dramatically decreased memory requirements for inference make edge inference much more viable, and Apple has the best hardware for exactly that. Apple Silicon uses unified memory, which means that the CPU, GPU, and NPU (neural processing unit) have access to a shared pool of memory; this means that Apple’s high-end hardware actually has the best consumer chip for inference (Nvidia gaming GPUs max out at 32GB of VRAM, while Apple’s chips go up to 192 GB of RAM).

Meta, meanwhile, is the biggest winner of all. I already laid out last fall how every aspect of Meta’s business benefits from AI; a big barrier to realizing that vision is the cost of inference, which means that dramatically cheaper inference — and dramatically cheaper training, given the need for Meta to stay on the cutting edge — makes that vision much more achievable.

Google, meanwhile, is probably in worse shape: a world of decreased hardware requirements lessens the relative advantage they have from TPUs. More importantly, a world of zero-cost inference increases the viability and likelihood of products that displace search; granted, Google gets lower costs as well, but any change from the status quo is probably a net negative…

...How did DeepSeek make R1?

DeepSeek actually made two models: R1 and R1-Zero. I actually think that R1-Zero is the bigger deal…

…R1-Zero, however, drops the HF part — it’s just reinforcement learning. DeepSeek gave the model a set of math, code, and logic questions, and set two reward functions: one for the right answer, and one for the right format that utilized a thinking process. Moreover, the technique was a simple one: instead of trying to evaluate step-by-step (process supervision), or doing a search of all possible answers (a la AlphaGo), DeepSeek encouraged the model to try several different answers at a time and then graded them according to the two reward functions.

What emerged is a model that developed reasoning and chains-of-thought on its own…

…Here again it seems plausible that DeepSeek benefited from distillation, particularly in terms of training R1. That, though, is itself an important takeaway: we have a situation where AI models are teaching AI models, and where AI models are teaching themselves.

---

Disclaimer: The Good Investors is the personal investing blog of two simple guys who are passionate about educating Singaporeans about stock market investing. By using this Site, you specifically agree that none of the information provided constitutes financial, investment, or other professional advice. It is only intended to provide education. Speak with a professional before making important decisions about your money, your professional life, or even your personal life. We currently have a vested interest in Alphabet (parent of Google), Apple, Meta Platforms, Microsoft, Netflix, and Tesla. Holdings are subject to change at any time.

We’ve constantly been sharing a list of our recent reads in our weekly emails for The Good Investors.

Do subscribe for our weekly updates through the orange box in the blog (it’s on the side if you’re using a computer, and all the way at the bottom if you’re using mobile) – it’s free!

But since our readership-audience for The Good Investors is wider than our subscriber base, we think sharing the reading list regularly on the blog itself can benefit even more people. The articles we share touch on a wide range of topics, including investing, business, and the world in general. 

Here are the articles for the week ending 26 January 2025:

1. Thoughts On A Month With Devin – Hamel Husain, Isaac Flath, and Johno Whitaker

Unlike typical AI assistants, Devin operates through Slack and spins up its own computing environment. When you chat with Devin, you’re talking to an AI that has access to a full computing environment – complete with a web browser, code editor, and shell. It can install dependencies, read documentation, and even preview web applications it creates…

…The experience is designed to feel like chatting with a colleague. You describe what you want, and Devin starts working. Through Slack, you can watch it think through problems, ask for credentials when needed, and share links to completed work. Behind the scenes, it’s running in a Docker container, which gives it the isolation it needs to safely experiment while protecting your systems. Devin also provides a web interface, which also allows you to gain access to its envirnoment and watch it work with IDEs, Web Browsers and more in real time…

…Our first task was straightforward but real: pull data from a Notion database into Google Sheets. Devin tackled this with surprising competence. It navigated to the Notion API documentation, understood what it needed, and guided me through setting up the necessary credentials in Google Cloud Console. Rather than just dumping API instructions, it walked me through each menu and button click needed – saving what would typically be tedious documentation sleuthing. The whole process took about an hour (but only a few minutes of human interaction). At the end, Devin shared a link to a perfectly formatted Google Sheet containing our data.

The code it produced was a bit verbose, but it worked. This felt like a glimpse into the future – an AI that could handle the “glue code” tasks that consume so much developer time. Johno had similar success using Devin to create a planet tracker for debunking claims about historical positions of Jupiter and Saturn. What made this particularly impressive was that he managed this entirely through his phone, with Devin handling all the heavy lifting of setting up the environment and writing the code…

…Over the course of a month, we systematically documented our attempts across these categories:

1. Creating new projects from scratch
2. Performing research tasks
3. Analyzing & Modifying existing projects

The results were sobering. Out of 20 tasks, we had 14 failures, 3 successes (including our 2 initial ones), and 3 inconclusive results. Even more telling was that we couldn’t discern any pattern to predict which tasks would work. Tasks that seemed similar to our early successes would fail in unexpected ways…

…Working with Devin showed what autonomous AI development aspires to be. The UX is polished – chatting through Slack, watching it work asynchronously, seeing it set up environments and handle dependencies. When it worked, it was impressive.

But that’s the problem – it rarely worked. Out of 20 tasks we attempted, we saw 14 failures, 3 inconclusive results, and just 3 successes. More concerning was our inability to predict which tasks would succeed. Even tasks similar to our early wins would fail in complex, time-consuming ways…

…This reflects a pattern we’ve observed repeatedly in AI tooling. Social media excitement and company valuations have minimal relationship to real-world utility. We’ve found the most reliable signal comes from detailed stories of users shipping products and services. For now, we’re sticking with tools that let us drive the development process while providing AI assistance along the way.

2. Transcript: The Hidden History of Eurodollars, Part 1: Cold War Origins – Joe Weisenthal, Tracy Alloway, Lev Menand, and Josh Younger

Tracy (01:30):
It can be admittedly confusing. So why don’t we just define it right away. So eurodollars are dollar-denominated bank deposits held at foreign banks or overseas branches of US banks. And you can think of them as basically offshore dollars that sit outside the US banking system and kind of away from the Federal Reserve. They’re basically a very special form of money. You could call them shadow money.

Joe (01:57):
And it’s totally gigantic. So it’s almost $10 trillion. And I just find it so interesting, right? Because when I think of dollars, they’re either coming from, you know, the government spends dollars into existence or US bank credit. US banks [have a] license to de facto create dollars or deposits at will. And yet, eurodollars are kind of this weird thing, I guess because they’re not that.

Tracy (02:21):
Yeah, they’re not either of those. And eurodollars didn’t just spring up fully formed out of thin air. They were the result of a series of decisions all aimed at solving particular problems…

…Josh (04:27):
So eurodollars are among the most important financial instruments in the world and they are really the backbone of the global dollar system. But they come from very humble beginnings, very idiosyncratic start. And really it all started in Yugoslavia…

…So in 1945 in November, there’s a communist revolution and the US is miffed in a bunch of ways, but one of them is that the old government owes them money. And so the question is, how are they going to get it? And a few months later, Tito asked for his gold back because the Yugoslavia government had $70 million worth of gold in New York. And the Secretary of State, who was George Marshall of the Marshall Plan, he realizes he’s got a bargaining chip, which is the gold. It’s in New York and they don’t get it back until they settle their claims.

Now, even people within the State Department were kind of skeptical of this, the Yugoslavian government is obviously furious. And so are the Russians who, at this point, you know, Tito and Stalin have a falling out eventually a few years later. But at this point, they’re quite closely aligned..

…The Russians get the sense that the US is willing to use gold as a bargaining chip. They’d previously actually been building up dollar balances in New York. This is this kind of a misnomer about the post-war period. There’s this sense that that the Russians are extracting all their resources from the US, but they’re actually building up reserves of dollars because the thought is ‘We’re probably going to need to trade with these people. We have a trading company based in the US and they need resources.’ And so they’re building up foreign currency deposits and gold, but in 1947, they realize it’s not going to go well, potentially. And they pull all the gold out. They actually just called banks in New York and they say ‘We want our gold back.’ A massive reversal of the policy.

And the question is, where’s it going to go? And so they need dollars because the US dollar is the currency of foreign exchange. If they want to trade with the West, they have to trade in dollars. They need gold because gold is the basis for the monetary system. And so the question is, where can they put gold and dollars in a safe place that’s still on the right side of what was then already known as the iron curtain?

And so it turns out Paris is the ticket. They’ve actually been secretly stockpiling cash in gold in Paris. They put it in briefcases. They would fly people to Paris and put it in the consulate offices. They would just build up piles of cash and gold. And in particular, there’s a bank — BCEN — I won’t try to do it in French. And BCEN is owned by, or run by, a notorious communist sympathizer, who has a very good relationship with the Politburo. And so this is a friendly bank. And so they take on deposit the Soviet money and BCEN’s moniker in the Telex system they used to communicate was “Eurobank.”

And so, eurodollars were initially, in the late forties, just deposits issued by Eurobank, BCEN, generally for the Soviets, although also for the Chinese. And slowly this starts to percolate. There’s another communist-owned bank in London. There’s one in Brussels, which DCIA just describes as run by ‘someone with few scruples, I think is the way they put it. And so there’s some friendlies across Europe who are willing to take their money and the eurodollar market begins this way, which is preemptive sanctions evasion, basically…

…And so the first use case of eurodollars is sanctions evasion. The second use is to facilitate cross-Iron Curtain trade, although that’s a pretty small business. And so the third, and much larger business, is cross-border interest rate arbitrage. And that sounds really technical, but what it’s really doing is using foreign exchange markets and derivative markets to source dollars that the UK in particular needs in this post-war environment.

So imagine a eurodollar bank, a euro bank, takes in a eurodollar deposit, which means it gets a dollar in cash — let’s think of a physical bill, that’s an asset. It issues a eurodollar liability. And then, what is it going to do next? Because it needs to do some sort of investing. And what it does is it exchanges that dollar asset for a sterling cash, and it invests that sterling cash in some short term sterling investment — short bills or something like that. And after it does that, it says ‘I want to hedge my foreign exchange risk, because now I have a dollar liability and a sterling asset. So I’m going to use the foreign exchange forward market to agree to sell that sterling back for dollars at some point in the future at a fixed price that we agree on today.’

So that’s the bank’s position. Who’s on the other side of that trade? Let’s say a corporation, a manufacturing entity, they make radios, and that radio production process requires inputs. Those inputs are imported. And so that radio production company needs dollars with which to buy the raw materials that it uses to make the radio that it then sells for dollars in foreign markets. And so, they get those dollars from the eurobank, in exchange for the sterling they have on hand, they go buy all the parts, but they want to make sure that they know how much they’re going to receive in local currency at the end of the production process. When they sell that radio abroad, they don’t want the value of the dollar to go down. So they sell those dollars forward in exchange for sterling. And so they’ve entered into a derivative agreement, which is the opposite of the one that the euro bank has or the euro banking system.

And so then they put together the radio, they sell it abroad, they receive dollar proceeds, they turn those into sterling, which is what they pay their employees in, that’s what they pay for their land and equipment in. And that exchange rate was the one they agreed upon in advance through the foreign exchange forward contract. And so, basically what’s happening is the euro banks are pulling in dollars from abroad, distributing them through the foreign exchange market that’s trading onshore to those that need dollars today, and then providing hedges to those that will receive dollars in the future. And in the case of the euro bank, the dollars they’ll owe in the future, potentially, to their eurodollar deposit holder.

Lev (18:32):
Think about this from the perspective of the City of London coming out of the war and those bankers and the world that they grew up in, which is a world that we’ve completely forgotten, but was the world of sterling dominance before the First World War and the role that the empire played in financing global trade.

What we’re looking at in the 1950s is a group of London-based financial institutions trying to figure out a way to continue their dominance in a global economy that runs on dollars now and not on sterling. And so, the eurodollars are sort of worth the risk to the City of London, and to some extent to UK financial regulators like the Bank of England, because they need to fix their business model for a dollar world, and they want to get in on the dollar world…

…Josh (20:43):
And so this cross-border interest rate arbitrage is really just the way markets distribute the currency according to who needs it and provide the hedges that facilitate the functioning of British corporations as well. It’s what we’d call now like a use case, right? This is like a real underlying use case that doesn’t involve the Soviet Union for dollar deposits issued by non-US banks, which is, you can’t emphasize enough how fundamentally strange that is because if I tried to make dollars by writing it on piece of paper, I don’t think I’d get very far. But at the time, that’s essentially what these banks are doing.

And in particular London is a more, let’s say, reputable locale, particularly banks that are not known to be communist sympathizers. There’s a little bit of a funny thing about being a communist bank, but we won’t get into that specifically, but these are blue chip banks in London issuing dollar deposits. And that means you can use them for things and you can feel more comfortable…

…Lev (26:54):
Although, just let’s size this a little bit, right? It was a billion dollars in, say, 1960, which is maybe the equivalent of $50 billion today…

…So we have way more to go in terms of the growth of this market subsequent to 1960. It’s still pretty nascent in 1960…

…Josh (31:08):
So the question at this point is, it’s a nascent market, it’s half a Tether, and it’s unclear whether or not it’s become a big major global actor. We know it eventually becomes that, but at the time, that’s super unclear, but it becomes eventually and soon the solution to a big problem. So eurodollars are the solution to big problem because, in the background of all of this buildup, there’s massive trouble brewing and the whole global edifice of the dollar system is starting to crack.

And the question is, you know, how are we going to save it? Or should we?

3. Emergent Layers, Chapter 1: Scarcity, Abstraction & Abundance – Alex Danco

One foundational principle of the tech world is that as it builds upwards and outwards into the rest of the world, it’s doing so by building on top of these abundant resources and progressively leveraging them. We can think about the world that we know and understand today — with its constraints, and business models and maturing industries that are generally understood by all — as forming a layer, which we’ll call layer i. In time, as certain elements become abstracted and subsequently abundant, others emerge as newly scarce, or in play for new reasons and in new business models. The critical skill for understanding how this works (which is worth practicing!) is being able to work one’s way up and down between stack layers so as to understand when an abundant and scalable element has blossomed at layer i of a stack, and its scarce, non-scalable counterpart has emerged at a new layer — which we’ll call layer i+1…

…Microsoft

The original scarce resource at layer i = PC hardware. In the early days of PCs, manufacturers could compete along many axes of performance — memory, speed, functionality, and so forth — while being sufficiently differentiated from one another. But it was very hard to standardize common functions and applications that people could run across any computer, making it difficult for these use cases to grow rapidly — until Bill Gates and Paul Allen realized, Hey, there isn’t a software industry yet but there’s gonna be, so we should start it. Microsoft abstracted away the capabilities of a computer into software, so now anyone else could write their own software on top of Microsoft’s software without having to worry about the underlying machinery. PCs became an abundantly available commodity, and Microsoft became dominant and mega-profitable. A new scarce resource emerged at layer i+1: the ability to connect these PCs and get them to talk to one another…

…Facebook

Scarce resource at layer i = connections between humans using the internet. The internet was awash in people and content, but authentic human interaction was still relatively scarce and difficult. As such, all of the attempts at connecting people to content and advertising and services were feature-stuffed, spammy, bloated and bad. The critical step forward that Facebook accomplished was abstracting away the “reciprocal friendship” into a functioning social graph. And we’ve seen what’s happened since: Facebook, and social connectivity in general, has exploded and become a newly abundant resource. Facebook became dominant and mega-profitable…

…One critical aspect of this layering is that at each higher level of abstraction, the lever with which one can create value and extract profit becomes successively longer. You can see this by looking at market cap per employee of these dominant companies:

Intel: 106k employees, 55B revenue, 149B mkt cap

Microsoft: 120k employees, 93B revenue, 429B mkt cap

Google / Alphabet: 60k employees 75B revenue, 510B mkt cap

Facebook: 13k employees, 6B revenue, 320B mkt cap…

…A non-obvious but critical point to appreciate here is that for of the first n movers mobilizing around a scarce element, the arrival and eventual dominance of the last mover will be seen as a Black Swan event of sorts. By abstracting away the scarce resource instead of organizing around its scarcity, these companies become the first to be fully playing in the sandbox at level i+1, as opposed to the non-scalable scarcity-governed sandbox at level i…

…The last decade saw plenty of startups go after the transportation market, and I’m sure all of them described themselves as “scalable” in their investor decks. Meanwhile, the whole valley was busy passing on Uber because it was initially just a better way to do a black car service, and few people understood the true scalable potential in abstracting away the driver-rider trust required for UberX. The take home lesson here should be taken to heart: when the first n companies go after an issue, no matter what language they use in their pitch, their business models typically don’t truly venture beyond the constraints at layer i that anybody can see and understand. They’re easier to work through, make more sense to “rational investors”, and require fewer non-linear leaps of thinking to understand. As such, when the last mover emerges at level i+1, they’re a Black Swan event: few people foresaw their opportunity, their impact is enormous, and everybody rationalizes what happened after the fact…

…At level i+1 of the stack, the newly valuable resource is that which emerges as scarce out of the transition from scarcity to abstraction to abundance at layer i.

4. The Default Position: LevFin’s Latest Game Just Got Shut Down…Sort Of – JunkBondInvestor

Serta was no small player. We’re talking about the company behind Serta and Beautyrest—the beds you see in every department store in America. But by 2020, they were in serious trouble. Drowning in debt and sales were tanking.

That’s when a group of savvy lenders saw their opportunity. Already holding a chunk of Serta’s debt, they approached with what would become lawyers’ new favorite playbook.

The deal? A group holding 51% of their term loans would provide new money, but only if they got to exchange their old loans for new “super-senior” debt that jumps to the front of the line. The other 49%? They didn’t even get a phone call.

Here’s a sobering fact: non-participating lenders saw their position so deeply subordinated that their recovery prospects plummeted. The new super-senior debt was worth nearly full value, while the excluded lenders saw their position crater.

But here’s where they screwed up.

Their loan agreement only allowed “open market purchases.” Serta’s lawyers tried arguing that their private backroom deal counted as “open market” because… well, just because.

The Fifth Circuit wasn’t having any of it. They said what everyone was thinking: A private deal with hand-picked lenders isn’t an “open market” any more than a private club is a public park…

…On the exact same day—I’m not making this up—a New York court looked at pretty much the identical deal from Mitel Networks and said “Sure, go right ahead.”…

…Mitel pulled the exact same move as Serta. They were drowning in debt, so they cut a deal with friendly lenders to jump them to the front of the line. New super-priority debt paper. Everyone else got pushed to the back.

So what made this different from Serta?

Three words. That’s it. Instead of requiring “open market purchases,” Mitel’s agreement just said they could “purchase by way of assignment.” No mention of open markets anywhere.

The New York court basically said: “Look, if you didn’t want the company doing private deals, you should have said so in the contract.” Those excluded lenders who were screaming about their “sacred rights”? The court told them their rights weren’t so sacred after all.

Here’s the brutal truth—the same transaction either flies or dies based entirely on a few words in your documents. If that doesn’t scare the hell out of every lender out there, it should.

5. Tyler Cowen – The #1 Bottleneck to AI progress Is Humans – Dwarkesh Patel and Tyler Cowen

Dwarkesh Patel 00:00:11
Why won’t we have explosive economic growth, 20% plus, because of AI?

Tyler Cowen 00:00:17
It’s very hard to get explosive economic growth for any reason, AI or not. One problem is that some parts of your economy grow very rapidly, and then you get a cost disease in the other parts of your economy that, for instance, can’t use AI very well.

Look at the US economy. These numbers are guesses, but government consumption is what, 18%? Healthcare is almost 20%. I’m guessing education is 6 to 7%. The nonprofit sector, I’m not sure the number, but you add it all up, that’s half of the economy right there.

How well are they going to use AI? Is failure to use AI going to cause them to just immediately disappear and be replaced? No, that will take, say, 30 years. So you’ll have some sectors of the economy, less regulated, where it happens very quickly. But that only gets you a modest boost in growth rates, not anything like the whole economy grows 40% a year.

Dwarkesh Patel 00:01:04
The mechanism behind cost disease is that there’s a limited amount of laborers, and if there’s one high productivity sector, then wages everywhere have to go up. So your barber also has to earn twice the wages or something. With AI, you can just have every barbershop with 1,000 times the workers, every restaurant with 1,000 times the workers, not just Google. So why would the cost disease mechanism still work here?

Tyler Cowen 00:01:25
Cost disease is more general than that. Let’s say you have a bunch of factors of production, say five of them. Now, all of a sudden, we get a lot more intelligence, which has already been happening, to be clear.

Well, that just means the other constraints in your system become a lot more binding, that the marginal importance of those goes up, and the marginal value of more and more IQ or intelligence goes down. So that also is self-limiting on growth, and the cost disease is just one particular instantiation of that more general problem that we illustrate with talk about barbers and string quartets.

Dwarkesh Patel 00:01:57
If you were talking to a farmer in 2000 BC, and you told them that growth rates would 10x, 100x, you’d have 2% economic growth after the Industrial Revolution, and then he started talking about bottlenecks, what do you say to him in retrospect?

Tyler Cowen 00:02:11
He and I would agree, I hope. I think I would tell him, “Hey, it’s going to take a long time.” And he’d say, “Hmm, I don’t see it happening yet. I think it’s going to take a long time.” And we’d shake hands and walk off into the sunset. And then I’d eat some of his rice or wheat or whatever, and that would be awesome.

Dwarkesh Patel 00:02:29
But the idea that you can have a rapid acceleration in growth rates and that bottlenecks don’t just eat it away, you could agree with that, right?

Tyler Cowen 00:02:38
I don’t know what the word “could” means. So I would say this: You look at market data, say real interest rates, stock prices, right now everything looks so normal, startlingly normal, even apart from AI. So what you’d call prediction markets are not forecasting super rapid growth anytime soon…

…Dwarkesh Patel 00:03:13
In his talk yesterday, Chad Jones said that the main variable, the main input into his model for growth, is just population. If you have a doubling, an order of magnitude increase in the population, you plug that number in in his model, you get explosive economic growth.

Tyler Cowen 00:03:26
I don’t agree.

Dwarkesh Patel 00:03:27
Why not buy the models?

Tyler Cowen 00:03:28
His model is far too much a one-factor model, right? Population. I don’t think it’s very predictive. We’ve had big increases in effective world population in terms of purchasing power. A lot of different areas have not become more innovative. Until the last, say, four years, most of them became less innovative.

So it’s really about the quality of your best people or institutions, as you and Patrick were discussing last night. And there it’s unclear what’s happened, but it’s also fragile. There’s the perspective of the economist, but also that of the anthropologist, the sociologist.

They all matter. But I think the more you stack different pluralistic perspectives, the harder it is to see that there’s any simple lever you can push on, intelligence or not, that’s going to give you breakaway economic growth.

Dwarkesh Patel 00:04:11
What you just said, where you’re bottlenecked by your best people, seems to contradict what you were saying in your initial answer, that even if you boost the best parts, you’re going to be bottlenecked by the restaurants…

…Here’s a simple way to put it. Most of sub-Saharan Africa still does not have reliable clean water. The intelligence required for that is not scarce. We cannot so readily do it.

We are more in that position than we might like to think, but along other variables. And taking advantage of the intelligence from strong AI is one of those.

Dwarkesh Patel 00:04:53
So about a year ago, your co-writer on Martial Revolution, Alex Tabarrok, had a post about the extreme scarcity of high-IQ workers. And so if the labor force in the United States is 164 million people, if one in a thousand of them are geniuses, you have 164,000 geniuses. That’s why you have to do semiconductors in Taiwan, because that’s where they’re putting their nominal amount of geniuses. We’re putting ours in finance and tech.

If you look at that framework, we have a thousand times more of those kinds of people. The bottlenecks are going to eat all that away? If you ask any one of these people, if you had a thousand times more of your best colleague, your best coworker, your best co-founder, the bottlenecks are going to eat all that away? Your organization isn’t going to grow any faster?

Tyler Cowen 00:05:32
I didn’t agree with that post. If you look at labor market data, the returns to IQ as it translates into wages, they’re amazingly low. They’re pretty insignificant.

People who are very successful, they’re very smart, but they’re people who have say eight or nine areas where they’re like, on a scale of 1 to 10, there are nine. Like they have one area where they’re just like an 11 and a half on a scale of 1 to 10. And then on everything else, they’re an eight to a nine and have a lot of determination.

And that’s what leads to incredible success. And IQ is one of those things, but it’s not actually that important. It’s the bundle, and the bundles are scarce. And then the bundles interacting with the rest of the world.

---

Disclaimer: The Good Investors is the personal investing blog of two simple guys who are passionate about educating Singaporeans about stock market investing. By using this Site, you specifically agree that none of the information provided constitutes financial, investment, or other professional advice. It is only intended to provide education. Speak with a professional before making important decisions about your money, your professional life, or even your personal life. We currently have a vested interest in Alphabet (parent of Google), Meta Platforms (parent of Facebook), and Microsoft. Holdings are subject to change at any time.

We’ve constantly been sharing a list of our recent reads in our weekly emails for The Good Investors.

Do subscribe for our weekly updates through the orange box in the blog (it’s on the side if you’re using a computer, and all the way at the bottom if you’re using mobile) – it’s free!

But since our readership-audience for The Good Investors is wider than our subscriber base, we think sharing the reading list regularly on the blog itself can benefit even more people. The articles we share touch on a wide range of topics, including investing, business, and the world in general. 

Here are the articles for the week ending 19 January 2025:

1. OpenAI o3 Breakthrough High Score on ARC-AGI-Pub – François Chollet

OpenAI’s new o3 system – trained on the ARC-AGI-1 Public Training set – has scored a breakthrough 75.7% on the Semi-Private Evaluation set at our stated public leaderboard $10k compute limit. A high-compute (172x) o3 configuration scored 87.5%.

This is a surprising and important step-function increase in AI capabilities, showing novel task adaptation ability never seen before in the GPT-family models. For context, ARC-AGI-1 took 4 years to go from 0% with GPT-3 in 2020 to 5% in 2024 with GPT-4o. All intuition about AI capabilities will need to get updated for o3…

…The high-efficiency score of 75.7% is within the budget rules of ARC-AGI-Pub (costs <$10k) and therefore qualifies as 1st place on the public leaderboard!

The low-efficiency score of 87.5% is quite expensive, but still shows that performance on novel tasks does improve with increased compute (at least up to this level.)

Despite the significant cost per task, these numbers aren’t just the result of applying brute force compute to the benchmark. OpenAI’s new o3 model represents a significant leap forward in AI’s ability to adapt to novel tasks. This is not merely incremental improvement, but a genuine breakthrough, marking a qualitative shift in AI capabilities compared to the prior limitations of LLMs. o3 is a system capable of adapting to tasks it has never encountered before, arguably approaching human-level performance in the ARC-AGI domain.

Of course, such generality comes at a steep cost, and wouldn’t quite be economical yet: you could pay a human to solve ARC-AGI tasks for roughly $5 per task (we know, we did that), while consuming mere cents in energy. Meanwhile o3 requires $17-20 per task in the low-compute mode. But cost-performance will likely improve quite dramatically over the next few months and years, so you should plan for these capabilities to become competitive with human work within a fairly short timeline.

o3’s improvement over the GPT series proves that architecture is everything. You couldn’t throw more compute at GPT-4 and get these results. Simply scaling up the things we were doing from 2019 to 2023 – take the same architecture, train a bigger version on more data – is not enough. Further progress is about new ideas…

…Passing ARC-AGI does not equate to achieving AGI, and, as a matter of fact, I don’t think o3 is AGI yet. o3 still fails on some very easy tasks, indicating fundamental differences with human intelligence.

Furthermore, early data points suggest that the upcoming ARC-AGI-2 benchmark will still pose a significant challenge to o3, potentially reducing its score to under 30% even at high compute (while a smart human would still be able to score over 95% with no training). This demonstrates the continued possibility of creating challenging, unsaturated benchmarks without having to rely on expert domain knowledge. You’ll know AGI is here when the exercise of creating tasks that are easy for regular humans but hard for AI becomes simply impossible…

…To adapt to novelty, you need two things. First, you need knowledge – a set of reusable functions or programs to draw upon. LLMs have more than enough of that. Second, you need the ability to recombine these functions into a brand new program when facing a new task – a program that models the task at hand. Program synthesis. LLMs have long lacked this feature. The o series of models fixes that.

For now, we can only speculate about the exact specifics of how o3 works. But o3’s core mechanism appears to be natural language program search and execution within token space – at test time, the model searches over the space of possible Chains of Thought (CoTs) describing the steps required to solve the task, in a fashion perhaps not too dissimilar to AlphaZero-style Monte-Carlo tree search. In the case of o3, the search is presumably guided by some kind of evaluator model. To note, Demis Hassabis hinted back in a June 2023 interview that DeepMind had been researching this very idea – this line of work has been a long time coming.

So while single-generation LLMs struggle with novelty, o3 overcomes this by generating and executing its own programs, where the program itself (the CoT) becomes the artifact of knowledge recombination. Although this is not the only viable approach to test-time knowledge recombination (you could also do test-time training, or search in latent space), it represents the current state-of-the-art as per these new ARC-AGI numbers.

Effectively, o3 represents a form of deep learning-guided program search. The model does test-time search over a space of “programs” (in this case, natural language programs – the space of CoTs that describe the steps to solve the task at hand), guided by a deep learning prior (the base LLM). The reason why solving a single ARC-AGI task can end up taking up tens of millions of tokens and cost thousands of dollars is because this search process has to explore an enormous number of paths through program space – including backtracking.

2. Energy Cheat Sheet – Brian Potter

Most energy we consume gets wasted. Of the 93.6 quads (~27,400 TWh) the US consumed in 2023, only around 1/3rd of that went towards producing useful work. The rest was lost due to various inefficiencies, such as heat engine and transmission losses…

…Another obvious fact is that despite the burgeoning construction of renewable energy infrastructure, the majority of our energy still comes from burning hydrocarbons. Petroleum, coal, and natural gas combined are responsible for roughly 82% of total energy consumption in the US.

Related to this fact is that electricity generation is a relatively small fraction of our energy system: roughly ⅓ of energy inputs go towards generating electricity. For residential and commercial consumption, only around half of energy use comes from electricity. For industrial and transportation energy (the two largest sources of consumption), electricity is around 13% and less than 0.1%.

What this chart makes clear, but also sort of abstracts away, is the enormous amount of infrastructure we’ve built for moving around hydrocarbons. The US has close to 1 million oil and natural gas wells, 3 million miles of natural gas pipeline, 145,000 gas stations, and capacity to refine 18.4 million barrels of oil a day.

This is why environmental advocates often focus on electrifying everything: decarbonizing energy infrastructure requires much more than just building low-carbon sources of energy like solar panels and wind turbines — it requires fundamentally reworking how our society moves energy around. It’s also why eliminating roadblocks and bottlenecks to energy infrastructure construction is so important.

We can also dive deeper and look at a sector-by-sector breakdown of energy use. The residential sector uses around 11.5 quads (3370 TWh) of energy, a little over 12% of total US energy consumption…

…One major takeaway here is that most residential energy consumption goes into heating things up: Space heating (5.74 quads), water heating (1.69 quads), and clothes dryers (0.26 quads) together account for ⅔rds of residential energy consumption.4 You sometimes see air conditioners decried as wasteful by energy-minded environmentalists, but air conditioning is a much smaller share of energy consumption than heating…

…Most transportation energy in the US is consumed in the form of gasoline and diesel fuel, with a relatively small amount of jet fuel. If we look at it by transportation mode, most energy (~78%) is consumed by cars, trucks, and motorcycles…

…The huge amount of energy used by transportation also means that households are using a lot of energy that isn’t captured by the residential energy consumption statistics above. In fact, in a year, the average US household consumes more energy from burning gasoline (~24,000 kilowatt-hours) than what’s used by the entire rest of the house (~22,500 kilowatt-hours).

The commercial sector is not that different from the residential sector, with heating air and water using the largest fraction, with cooling and ventilation (ie: moving air around) also using large fractions.5 As with residential, its energy consumption is roughly split between electricity and natural gas…

…With industrial energy use, we see a lot of the same patterns that we see in other sectors. One is that utility electricity is a relatively small amount of industrial energy consumption (less than 20%). Most industrial energy comes from burning fuel (mostly natural gas) directly. Once again, we see that heating things up accounts for a huge fraction of energy consumption: roughly half of all manufacturing energy goes into process heating: If we add process heat to residential and commercial air and water heating, we find that roughly 20% of total US energy consumption goes towards heating things up…

…It’s clear that most energy used in the US is ultimately wasted, with only a small fraction being used to perform useful work (moving cars, heating homes, operating electronics, and so on). Moving energy around and changing its form can’t be done perfectly efficiently (thanks in part to the 2nd law of thermodynamics), and all those conversions we require to get energy where it needs to be and in the form we need it whittle away the energy available to get things done…

…The biggest source of losses is probably heat engine inefficiencies. In our hydrocarbon-based energy economy, we often need to transform energy by burning fuel and converting the heat into useful work. There are limits to how efficiently we can transform heat into mechanical work (for more about how heat engines work, see my essay about gas turbines).

The thermal efficiency of an engine is the fraction of heat energy it can transform into useful work. Coal power plant typically operates at around 30 to 40% thermal efficiency. A combined cycle gas turbine will hit closer to 60% thermal efficiency. A gas-powered car, on the other hand, operates at around 25% thermal efficiency. The large fraction of energy lost by heat engines is why some thermal electricity generation plants list their capacity in MWe, the power output in megawatts of electricity…

…The low thermal efficiency of ICE cars and heat engines in general and the high efficiency of electrical equipment (especially things like heat pumps) are the biggest counterweight to the high energy capacity of hydrocarbons. The gas tank on an ICE car technically stores much more energy than a Tesla battery pack but only a small fraction of that gasoline energy can be converted into useful motion. Switching to EVs, even if that electricity is still provided by burning fossil fuels, could save large amounts of energy (and thus carbon emissions), as it could mean switching from a 25% efficient gasoline engine to a 60% efficient combined cycle gas turbine. And of course, with electric vehicles, there’s the possibility of powering them by non-carbon emitting sources of electricity like solar or wind. 

3. Stocks Are More Expensive Than They Used to Be – Michael Batnick

In January 2018, they wrote an article, CAPE Fear: Why CAPE Naysayers Are Wrong. The article featured yours truly…

…It’s hard to believe seven years have passed since this article. It’s harder to believe that the S&P 500 is up almost 100% since their article came out, and delivered the highest 7-year performance for any CAPE starting at 33x. I did not see this coming. At all.

My whole thing was, yes, valuations are high. But companies are better today and deserve the premium multiple. I was not saying that a high CAPE is bullish. In fact, I ended most of my posts on this topic with the message of, “Expect lower returns.” I’ve never been happier to be wrong.

I want to return to some of the arguments I made, and what the CAPE zealots missed.

To use a long-term average that goes back to the late 1800s is foolish for three reasons. First, we didn’t have CAPE data back in 1929. It was first “discovered” in the late 90s. The discovery of data in financial markets changes the very essence of it. Markets are not governed by the laws of physics. They’re alive. They adapt and evolve and adjust, like an micro organism.

Second, the CAPE ratio has been rising over time since the 1980s. We’ve only visited the long-term average once in the last 25 years, and that was at the bottom of the GFC. If that’s what it takes to return to the long-term average, maybe you should reconsider what an appropriate comp level really is.

Third, and most important, the companies are far better today than they were in the past.

4. AI’s Uneven Arrival – Ben Thompson

What o3 and inference-time scaling point to is something different: AI’s that can actually be given tasks and trusted to complete them. This, by extension, looks a lot more like an independent worker than an assistant — ammunition, rather than a rifle sight. That may seem an odd analogy, but it comes from a talk Keith Rabois gave at Stanford:

So I like this idea of barrels and ammunition. Most companies, once they get into hiring mode…just hire a lot of people, you expect that when you add more people your horsepower or your velocity of shipping things is going to increase. Turns out it doesn’t work that way. When you hire more engineers you don’t get that much more done. You actually sometimes get less done. You hire more designers, you definitely don’t get more done, you get less done in a day.

The reason why is because most great people actually are ammunition. But what you need in your company are barrels. And you can only shoot through the number of unique barrels that you have. That’s how the velocity of your company improves is adding barrels. Then you stock them with ammunition, then you can do a lot. You go from one barrel company, which is mostly how you start, to a two barrel company, suddenly you get twice as many things done in a day, per week, per quarter. If you go to three barrels, great. If you go to four barrels, awesome. Barrels are very difficult to find. But when you have them, give them lots of equity. Promote them, take them to dinner every week, because they are virtually irreplaceable. They are also very culturally specific. So a barrel at one company may not be a barrel at another company because one of the ways, the definition of a barrel is, they can take an idea from conception and take it all the way to shipping and bring people with them. And that’s a very cultural skill set.

The promise of AI generally, and inference-time scaling models in particular, is that they can be ammunition; in this context, the costs — even marginal ones — will in the long run be immaterial compared to the costs of people, particularly once you factor in non-salary costs like coordination and motivation…

…What will become clear once AI ammunition becomes available is just how unsuited most companies are for high precision agents, just as P&G was unsuited for highly-targeted advertising. No matter how well-documented a company’s processes might be, it will become clear that there are massive gaps that were filled through experience and tacit knowledge by the human ammunition.

SaaS companies, meanwhile, are the ad agencies. The ad agencies had value by providing a means for advertisers to scale to all sorts of media across geographies; SaaS companies have value by giving human ammunition software to do their job. Ad agencies, meanwhile, made money by charging a commission on the advertising they bought; SaaS companies make money by charging a per-seat licensing fee. Look again at that S-1 excerpt I opened with:

Our business model focuses on maximizing the lifetime value of a customer relationship. We make significant investments in acquiring new customers and believe that we will be able to achieve a positive return on these investments by retaining customers and expanding the size of our deployments within our customer base over time…

The positive return on investment comes from retaining and increasing seat licenses; those seats, however, are proxies for actually getting work done, just as advertising was just a proxy for actually selling something. Part of what made direct response digital advertising fundamentally different is that it was tied to actually making a sale, as opposed to lifting brand awareness, which is a proxy for the ultimate goal of increasing revenue. To that end, AI — particularly AI’s like o3 that scale with compute — will be priced according to the value of the task they complete; the amount that companies will pay for inference time compute will be a function of how much the task is worth. This is analogous to digital ads that are priced by conversion, not CPM.

The companies that actually leveraged that capability, however, were not, at least for a good long while, the companies that dominated the old advertising paradigm. Facebook became a juggernaut by creating its own customer base, not by being the advertising platform of choice for companies like P&G; meanwhile, TV and the economy built on it stayed relevant far longer than anyone expected. And, by the time TV truly collapsed, both the old guard and digital advertising had evolved to the point that they could work together.

If something similar plays out with AI agents, then the most important AI customers will primarily be new companies, and probably a lot of them will be long tail type entities that take the barrel and ammunition analogy to its logical extreme. Traditional companies, meanwhile, will struggle to incorporate AI (outside of whole-scale job replacement a la the mainframe); the true AI takeover of enterprises that retain real world differentiation will likely take years.

None of this is to diminish what is coming with AI; rather, as the saying goes, the future may arrive but be unevenly distributed, and, contrary to what you might think, the larger and more successful a company is the less they may benefit in the short term. Everything that makes a company work today is about harnessing people — and the entire SaaS ecosystem is predicated on monetizing this reality; the entities that will truly leverage AI, however, will not be the ones that replace them, but start without them.

5. Don’t let interest-rate predictions dictate your investment decisions – Chin Hui Leong

A little over a year ago, the US Federal Reserve signalled its intention to cut interest rates three times in 2024. This commentary sparked a flurry of predictions, with market watchers vying to outguess the Fed on the number, timing, and size of these cuts. Goldman Sachs, for instance, boldly predicted five cuts.

We ended up with just three interest-rate cuts in 2024 – a significant miss, to say the least…

…According to Visual Capitalist, four firms – Morgan Stanley, Bank of America, Citigroup and Nomura – pencilled in a one-percentage-point cut for 2024. Credit should be given where it’s due: their forecasts were right.

However, did getting these predictions right matter in the end? As it turns out, not so much.

Morgan Stanley, Bank of America and Citi set 2024’s S&P 500 price targets at 4,500, 5,000 and 5,100 respectively… 

…The S&P 500, of course, closed the year at 5,881…

…Forecasts and expectations may look similar, but they are different. My friend Eugene Ng puts it best: Forecasts rely on knowing when something will occur. Expectations, on the other hand, are the acknowledgement of what’s likely to occur without professing insight into when it will happen.

For example, it’s reasonable to expect the stock market to fall by 10 per cent or more sometime in the future. After all, history has shown that corrections are a common occurrence…

…In my eyes, calmness can be achieved by having the right expectations, and preparing well for any market turbulence even when we don’t know when the market will fall.

If you are prepared, you will have fewer worries. If you worry less, you will stand a better chance of doing better than average. And that’s more than any investor can hope for, whether the forecasts are right or wrong.

---

Disclaimer: The Good Investors is the personal investing blog of two simple guys who are passionate about educating Singaporeans about stock market investing. By using this Site, you specifically agree that none of the information provided constitutes financial, investment, or other professional advice. It is only intended to provide education. Speak with a professional before making important decisions about your money, your professional life, or even your personal life. We currently have a vested interest in Alphabet (parent of Deepmind), Meta Platforms (parent of Facebook), and Tesla. Holdings are subject to change at any time.

We’ve constantly been sharing a list of our recent reads in our weekly emails for The Good Investors.

Do subscribe for our weekly updates through the orange box in the blog (it’s on the side if you’re using a computer, and all the way at the bottom if you’re using mobile) – it’s free!

But since our readership-audience for The Good Investors is wider than our subscriber base, we think sharing the reading list regularly on the blog itself can benefit even more people. The articles we share touch on a wide range of topics, including investing, business, and the world in general. 

Here are the articles for the week ending 05 January 2025:

1. Mike Alkin – Talking Uranium (Transcript here) – Bill Brewster and Mike Alkin

Alkin: So coming to this market, I did that. I spent a good almost couple of years doing supply/demand on my own. There’s 430 reactors around the world. And understanding the country where they operate, the attitude towards nuclear, understanding the math involved. Often as investors, you look for heuristics. How many reactors are there? How many pounds per reactor would there be? You’re looking for rules of thumb. As you start peeling the onion back, I realize that rules of thumb don’t apply here because the amount of uranium needed for the reactor fleet around the world is not always the same. It depends upon enrichment capacity. We won’t go down that rabbit hole, but there’s a whole other segment you need to learn.

As I was doing that, I would go to these conferences and I would talk to nuclear fuel buyers, people who buy this stuff. It was hard for me at first to really understand what I was dealing with because as somebody at that time having well over 20 years of experience as a hedge fund investor, I talked to people in all industries that were on all sides of the equation. But the people buying it typically were curious as to what we were thinking when we were questioning them. If we were talking to a buyer at a company that was buying a product, they would say “What are you as an investor hearing? What are you hearing from the other side? What are my competitors saying? What are you hearing about inventories?” They were inquisitive. That was not this cohort. As I started speaking to nuclear fuel buyers, I was met with an enormous wall put in front of me telling me, “I’m an outsider, I’m not a nuclear engineer, I don’t know what I’m doing, I should basically stay away and they’ve got it.”

I thought it was that attitude that just said to me, “Something’s not right here because the numbers I’m coming up with, whether I’m looking at inventories or the amount of the cost of the supply, or the actual demand” – for context, at the time the price of uranium was $17, $18, $19 a pound. It would say what it was trading for in the market. As I did the analysis, I realized that the average cost was somewhere in the mid-$50s. I’m not that sharpest tool in the shed but I know that if something costs you mid-$50s to make, you can’t sell it for $17 for very long. So it was then that I had to peel back the onion saying, “Why are they producing it at that price?” Then you start to understand that the uranium market is one driven mostly by long term contracts. Well north of 80% on average will trade in a long-term window with contracts that cover 5, 7, 10, 12, 15 years depending on the contract. But that’s where most of the pounds trade. After the Fukushima event, a lot of these uranium producers, when the spot market had declined precipitously, were still selling into much higher prices. My understanding of that when I was talking to fuel buyers at these nuclear conferences, they were telling me that the price of uranium was $17 and $18, it was going to $10, it was going to $5. There was all this uranium out there.

That’s not what my math was showing me. What my math was showing me was that the model was that the long term contracts that had been signed before Fukushima melted down in 2011 were going to start to expire and rather rapidly. Uranium producers could not sell $17, $18, $20 uranium when it cost him 2.5 times that. At some point, production would have to start to shut down.

So you ask, “Do you think you’re crazy?” Yes, because as I’m talking to people who are obviously very sharp – they’re nuclear engineers – but it’s understanding, as you realize, as an investor, you have to understand incentives and you have to understand market structure. Charlie Munger would always say, “Show me the incentive, I’ll show you the outcome.” It was as I was starting to go and talk to these folks and realizing a couple of things. Number one is, they had no interest in what I was learning on my journey. Even though I’m not a nuclear engineer, I’m still somebody who’s a market participant. I’m still somebody that while I don’t speak their language, sitting at a dinner table or a lunch table or at a bar having a beer with them, I certainly could hold my own in supply/demand conversation. And as I would talk about what I was learning and uncovering, I was shot down at every step. I thought, “Wow, that’s interesting because I’m seeing a recency bias. What is now will always be.” So they were kind of latched onto that.

Then as I started peeling that, I’m thinking, “Why is this?” I’ve been doing this a very long time. Over the years, I’ve been wrong many times. I’ve been right more often than not. But you’re wrong and you try and understand where you’ve been wrong. I was thinking, “What is it? Why are they so uninterested in hearing what an outsider’s view is?” As I started to explore that more, you start to understand the makeup and the cost structure of a nuclear reactor, which I have known, but it really started to come into clear vision for me was the fuel. Uranium is just one part of the fuel cycle that goes in. You have uranium, they convert uranium from a powder into a gas. It then gets enriched, it then gets fabricated into pellets. That takes 18 to 24 months to do this stuff. There’s many different stages of the fuel cycle. As I was starting to think about what are the costs of that, all those stages are probably around 20% to 25%. What’s the cost of the uranium? That depends on the price. But it could be mid-single digits, high-single digits, somewhere around that. As you start talking to them about that, you realize it’s not a meaningful cost.

For comparative purposes, if I’m running a natural gas power plant or a coal power plant, my feedstock, the natural gas and the coal are 80% to 90% of the cost of operating it. Here, the uranium is single digits cost of operating it. The vision that started to come to me was uninterested market participants. They’re in the market very infrequently. Why are they uninterested? Because the cost is de minimis. Not to say it’s meaningless, but it’s de minimis. Then as I started to explore and ask questions, “Why are you not as concerned about this?” I was obviously met with a wall.

But what started to come to me was – and I asked flat out at a particular dinner at a World Nuclear Conference – I asked one, actually there were four fuel buyers at a dinner, I said, “If you all had a really enterprising fuel buyer that did the supply/demand work and said, “I think consensus is wrong. Here we are, $17, $18, $20 a pound. We should be buying uranium because the forecasts going out of the future are for deficits to be forming.” Let me ask you a question. Do you all, if the price were to go parabolic and you had all these great cost savings for your plant, do you participate that in any way, shape or form? Are you rewarded financially? Are you rewarded with a promotion?” The answer was I got laughed at. “What are you talking about? We’re paid to secure fuel.” These were buyers. As you come to a market as an investor, you think buyers are traders – they’re commercial creatures. These aren’t. These are really smart nuclear engineers that happen to buy a product that happens to not be a major cost component. There’s infrequent price discovery on their part and so it’s a lesson in understanding incentives and market structure…

…Alkin: One of the things you see now is you have expert networks who provide hedge funds and mutual funds experts to speak to in any industry. If you’re a hedge fund wanting to get up to speed right now on the nuclear power industry, you’re going to say, “Get me three nuclear fuel buyers. I’d like to speak to them about uranium.” They’re going to get on the phone and they’re going to speak to them. For years – though I’m sure they’ve been doing this – they can get on the phone and speak to three fuel buyers and they say, “Yeah, there’s plenty of uranium out there.” Those are the same folks, when the price was $17 was telling me that, versus here you’re seeing floors and ceilings at $125 and $135. They are the gift that keep on giving. Yet the way the structure of the research process is, they’re going to expert networks. They find these people, and if you don’t understand how the sausage is made, you’re going to be misled. They’re not purposely misleading you. It’s just what their own beliefs are. For me, that’s a beautiful thing. I’ve been doing this a long time now, almost 30 years as a professional investor, and I’ve never seen a cohort of people who are so uninterested in hearing the other side of the story. So far I’ve seen them prices move up 4x in there against them and they still have the same attitude.

Brewster: To your point, it doesn’t sound like they’re very incentivized to care.

Alkin: There’s very little to no incentive to care, other than maybe you would think pride? I don’t know. But it doesn’t matter. It’s just not a thing. We actually chuckle because when we go to these conferences, you talk to them in a hallway or in a bar, it’s as though you’re an adversary. It’s very bizarre. They don’t have an incentive. It doesn’t matter what they pay. So that’s the bizarre thing.

2. Chip Cities Rise in Japan’s Fields of Dreams – Gearoid Reidy

In Chitose, a city of 100,000 in the northernmost main island of Hokkaido, billboards seek recruits for the Self-Defense Forces, which saw a 50% shortfall last year. When I arrived on a fully booked plane from Tokyo packed with salarymen in cheap suits and expensive watches, it was easy to see where the competition was coming from: a half-dozen towering cranes jutting into the sky, a jarring contrast against the surrounding countryside…

…Those cranes are building the first fab for Rapidus Corp., a public-private venture that aims to skip Japan to the head of the chip production queue. Founded just two years ago, it hopes to produce cutting-edge, 2-nanometer chips by 2027, in cooperation with IBM Corp. It’s fraught with risks, and the government’s record in promoting industry is spotty. But this is just the latest and most ambitious example of a series of bets on chips, with Prime Minister Shigeru Ishiba recently pledging an extra ¥10 trillion ($66 billion) on top of ¥3.9 trillion invested since 2021. Near the other end of the Japanese archipelago, 1,500 kilometers (930 miles) to the southwest, is another. In Kumamoto, on the island of Kyushu, mass production is soon set to begin at a $7 billion semiconductor plant.

Here, Taiwan Semiconductor Manufacturing Co., drawn by government subsidies and the region’s supply chain, opened its first Japanese plant in February. A second is in the works, with authorities lobbying for a third. It’s triggered an influx of Taiwanese workers into a city where until recently almost everyone was Japanese…

…As many as 6,000 laborers are employed to build Rapidus. But talk is of the arrival of permanent workers once test production begins. That’ll bring at least 1,000 high-earning jobs, along with their supply chains. On my visit, ASML Holding NV, the Dutch maker of chip-testing tools, had just opened offices, with 50 staff expected. Every second building seems to be being torn down and rebuilt…

…The scale of the ambition creates the risk of spectacular failure, one many in Japan’s media fully expect. Skepticism is warranted, considering previous government-led efforts, from DRAM maker Elpida Memory Inc., sold to Micron Technology Inc. after its 2012 bankruptcy, to troubled Japan Display Inc.

The economy was already doing well even before talk of Rapidus, Mayor Ryuichi Yokota told me, describing the fab as a “Big Bang” that has the city scrambling. Yet at night, when the construction crews leave, the silence is deafening. I couldn’t feel the billions I expected to find flowing, just a cold wind that would soon begin to turn to snow…

…The risk from disaster is unpredictable; but what if these experiments simply don’t work out? Japan has spent billions on subsidies to bring a foreign company in Kumamoto. And when it comes to Rapidus, the risks are immense. Even if the company can find the talent it needs (the country is expected to have a shortfall of 40,000 engineers), the technology succeeds and yields are acceptable, it still has to outcompete rivals — including TSMC — to attract customers with an unproven product.

Chitose mayor Yokota shrugged off these concerns. “I’m convinced it will succeed,” he said, resolute that researchers currently studying with IBM in the US will return, like Meiji-era scholars, with secrets Japan can use to rebuild.

3. Before Berkshire: Warren Buffett’s Tab Card Triumph – Kingswell and Alice Schroeder

He decided that he would come in and invest in this company — Mid-Continent Tab Card Co. — but, interestingly, he did not take Wayne and John’s word for it. The numbers they gave him were really enticing, but again he went through and he acted like a horse handicapper.

Here’s another point of departure from what almost anybody else would do. Everybody that I know — or knew as an analyst — would have created a model for this company and would have projected out its earnings and would have looked at its return on investment in the future. Warren didn’t do that. In fact, in going through hundreds of his files, I’ve never seen anything that resembled a model.

What he did is he did what you would do with a horse. He figured out the one or two factors that could make the horse succeed or fail — and, in this case, it was sales growth and making the cost advantage continue to work. Then, he took all of the historical data, quarter by quarter for every single plant, he got the similar information as best he could from every competitor they had, and he filled pages with little hen scratches of all this information and he studied that information.

And, then, he made a yes/no decision. He looked at it: They were getting 36% margins [and] they were growing over 70% a year on a million of sales. Those were the historic numbers. He looked at them in great detail — just like a horse handicapper studying the tip sheet — and then he said to himself, “I want a 15% return on $2 million of sales.” And then he said, “Yeah, I can get that.” And he came in as an investor.

So what he did is he incorporated his whole earnings model and compounding discounted cash flow into that one sentence. “I want 15% on $2 million of sales.”

Why 15%? Because Warren is not greedy. He always wants a mere 15% day one return on an investment and then it compounds from there. That’s all he has ever wanted. He’s happy with that. It’s a very simple thing. There’s nothing fancy about it…

…The $2 million of sales was pretty simple, too. It had $1 million [and] it was growing 70%. There was a big margin of safety built into these numbers. It had a 36% profit margin and he said, “I’ll take half that.”

He ended up putting $60,000 of his personal non-partnership money into this company, which was about 20% of his net worth at the time. He got 16% of the company’s stock, plus some subordinated notes.

4. China’s Bond Yields Scream the ‘D’ Word – Lingling Wei

Over the past week, just as Chinese leaders tried to get the public—and markets—excited with another round of stimulus talk, China’s 10-year sovereign yield kept falling to fresh lows. Now, the yield is around 1.7%, a full percentage-point plunge from a little over a year ago. The return on the 30-year government bond has also dropped below 2%.

The sovereign-debt yield still has a ways to go before falling to zero, but the speed of the drop is astonishing. The lower the yield falls, the deeper the market is signaling economic stress.…

…In reality, Beijing is sticking to the formula of boosting demand through investment. The official thinking is, investment creates jobs, which would in turn create demand. That means more roads will be built, factories will be expanded and debts will continue to rise. Already, residents in some cities are complaining about the inconvenience from old roads being dredged up as authorities search for ways to invest.

One big irony is the source of bond buying—the force pushing down the yields.

State-owned banks, insurance firms and funds, the very institutions Beijing is counting on to support the economy, are the major purchasers of government bonds. These institutions would rather park their money in the safety of bonds than financing business projects or otherwise putting it to work.

“What’s good to invest in these days when demand is so low?” a Chinese banker told me, referring to weak business and consumer spending.

5. An Interview with Gregory Allen About the State of China Chip Export Controls – Ben Thompson and Gregory Allen

Here’s the question though. China doesn’t generally seem to be operating, and for good reason under the circumstances, under a real stringent return on invested capital calculation. I mean the 7nm chips that are being produced, we know with I think a pretty high degree of certainty, the yields are terrible.

GA: The yields are dreadful.

But they’re doing it anyway just because it needs to be done and this sort of ties into another thing. You referenced Dylan Patel and SemiAnalysis, who have been pretty strident critics of the enforcement of chip controls. But I think a good point he has made is that China, unlike the US, is not necessarily constrained in power or in the ability to build a ton of data centers, and so there’s a bit where they could just sort of — it’s not great, but they could just be way less efficient and accomplish similar things. Is there a bit where these expert controls are fashioned with Western/US constraints and concerns about how you go about building this stuff that might make them less impactful in the long run?

GA: Yeah, the export controls have not achieved their wildest dreams. There was a faction in the Biden administration that says, “Bwahaha, we found the secret weapon, and China’s AI dreams are gone” — that theory is just dead. Where we are now is at more of a cost imposition strategy. “We are going to make this as expensive and complicated as possible for you to do it, we’re going to try and slow you down, we’re going to try and increase your costs, and that is the race that we’re going to run”.

I mean, if you think about it, we’re switching from a mode in which the US AI ecosystem and the Chinese AI ecosystem were largely fused such that if we’re running a race, you can imagine there’s US people giving China Gatorade and those new Nike shoes that make you run faster. Now we’re moving to a moment where we’re trying to trip them in the race, that’s the change in mindset that we’ve experienced, and it’s not working to its most extreme form, but there is real cost imposition takes the form of the fact that SMIC has to operate at these dreadful yields. The economics are terrible, the fact that when they’re building all of these data centers, they’re having to use lousy chips, they’re having to buy more of them, and they’re having to deal with the higher energy costs of all of that.

It’s true that China does have just this extraordinary willingness to spend, but the point is we’re in this race, we’re in this competition, and it gives us an edge, not an infinite edge, but a meaningful edge.

This is a field, maybe you don’t have an answer to this, but there are some that argue that actually the better approach to some of these chips is a much more expensive, a much more high speed memory approach that has much lower latency using SRAM instead of High Bandwidth Memory. Is there a possibility that we actually pushed China down a different route towards developing these chips that maybe ends up being better because we thought HBM was the right way?

GA: I think that’s probably not what’s going to happen. It’s definitely worth saying that that could happen, a version of that kind of happened with YMTC and their NAND memory. There were multiple different approaches they could have taken technologically. All the Western and US allied Asian firms picked one way because it was obviously the best economics, and they held all the intellectual property, they held all the patents and so YMTC basically said, “Okay, we’re going to go down this other road and because we’re so heavily subsidized, it doesn’t really matter that it’s going to be more expensive”, and they did ultimately figure out how to get it work.

I think what you’re describing, the SRAM in massive quantities thing verges on the neuromorphic architecture, and it’s not that that’s impossible, and it’s not that that’s never going to happen, but it’s clearly not the right step for China right now. I think they have a path to domestic HBM production and that’s so much easier for them to chase than a SRAM revolution. I think traditionally they would just wait for somebody else to try and figure out and demonstrate that it’s possible and then they would throw infinite resources at it…

...For all of these chip controls, all this stuff that you’ve covered and written about, does any of it matter, if you add it all up, in comparison to that point that they don’t have EUV?

GA: EUV is the highest return on investment export control that we have had and are likely to have. It’s definitely the case that some of the other stuff hurts. If you talk about SMIC, for example, increasing their yields on their 7nm line and expanding the capacity of their 7nm line, they actually are bottlenecked by US equipment, a lot of US metrology equipment, etc. But if you want to talk about why they can’t—

But they do have the equipment, they just need to figure out how to duplicate it. The challenge with EUV is they don’t even have one, so duplicating it is that much harder.

GA: Yes exactly, it’s a lot harder to reverse engineer something that you don’t have a copy of, it really helps to have a copy of it. So I would say the EUV thing really matters, but there’s areas where China is facing headwinds that aren’t part of the EUV story.

So just to take one example, in DRAM, Micron still doesn’t use EUV in their production of DRAM, and they’re a globally competitive firm. So CXMT, the Chinese domestic champion of DRAM, the reason why they’re not currently globally competitive is not the absence of EUV, but I do think you could make a story that it is the absence of all this other stuff that we’ve been refusing to sell…

…You’re not necessarily like a geopolitical analyst, but the thing that scares me about all this, I think I’ve asked you this every time, it still scares me, is we’re talking and saying the administration needs to do better at enforcing these laws that guarantee a power imbalance in the long run, that is usually very destabilizing. China might think, if we’re going to have a fundamental power imbalance, then how about we take Taiwan off the board because that will screw everyone? Now we’re equal again. Do you worry about this? You’re a strong advocate for doing this better.

GA: So. Number one is, I don’t know that I ever agree that the balance of power is the stable universe. In 1994, the Taiwanese defense budget was half of that of the Chinese defense budget, now the Chinese defense budget is infinity times that of the Taiwanese defense budget. And by contrast, in 1997, I think there was a single U.S aircraft carrier battle group that was more than capable of defeating the entire Chinese Navy and the entire Chinese Air Force, that was a massive power imbalance and it was a very stable relationship. And by the way, it was a relationship in which a lot of people got rich and had productive free trade and all these kinds of happy relationships. So the idea that power parity is the path to peace here, don’t know that I necessarily agree with that, I don’t think the historical record really bears that out.

Now, you could argue if we’re going to make bold moves and try and seize a decisive advantage, could those bold moves be destabilizing? Yeah, I think definitely think so.

---

Disclaimer: The Good Investors is the personal investing blog of two simple guys who are passionate about educating Singaporeans about stock market investing. By using this Site, you specifically agree that none of the information provided constitutes financial, investment, or other professional advice. It is only intended to provide education. Speak with a professional before making important decisions about your money, your professional life, or even your personal life. We currently have a vested interest in ASML and TSMC. Holdings are subject to change at any time.

We’ve constantly been sharing a list of our recent reads in our weekly emails for The Good Investors.

Do subscribe for our weekly updates through the orange box in the blog (it’s on the side if you’re using a computer, and all the way at the bottom if you’re using mobile) – it’s free!

But since our readership-audience for The Good Investors is wider than our subscriber base, we think sharing the reading list regularly on the blog itself can benefit even more people. The articles we share touch on a wide range of topics, including investing, business, and the world in general. 

Here are the articles for the week ending 22 December 2024:

1. Meet Willow, our state-of-the-art quantum chip – Hartmut Neven

Errors are one of the greatest challenges in quantum computing, since qubits, the units of computation in quantum computers, have a tendency to rapidly exchange information with their environment, making it difficult to protect the information needed to complete a computation. Typically the more qubits you use, the more errors will occur, and the system becomes classical.

Today in Nature, we published results showing that the more qubits we use in Willow, the more we reduce errors, and the more quantum the system becomes…

…This historic accomplishment is known in the field as “below threshold” — being able to drive errors down while scaling up the number of qubits…

…There are other scientific “firsts” involved in this result as well. For example, it’s also one of the first compelling examples of real-time error correction on a superconducting quantum system — crucial for any useful computation, because if you can’t correct errors fast enough, they ruin your computation before it’s done. And it’s a “beyond breakeven” demonstration, where our arrays of qubits have longer lifetimes than the individual physical qubits do, an unfakable sign that error correction is improving the system overall.

As the first system below threshold, this is the most convincing prototype for a scalable logical qubit built to date. It’s a strong sign that useful, very large quantum computers can indeed be built…

…As a measure of Willow’s performance, we used the random circuit sampling (RCS) benchmark. Pioneered by our team and now widely used as a standard in the field, RCS is the classically hardest benchmark that can be done on a quantum computer today…

…Willow’s performance on this benchmark is astonishing: It performed a computation in under five minutes that would take one of today’s fastest supercomputers 1025 or 10 septillion years. If you want to write it out, it’s 10,000,000,000,000,000,000,000,000 years. This mind-boggling number exceeds known timescales in physics and vastly exceeds the age of the universe. It lends credence to the notion that quantum computation occurs in many parallel universes, in line with the idea that we live in a multiverse, a prediction first made by David Deutsch…

…Willow was fabricated in our new, state-of-the-art fabrication facility in Santa Barbara — one of only a few facilities in the world built from the ground up for this purpose. System engineering is key when designing and fabricating quantum chips: All components of a chip, such as single and two-qubit gates, qubit reset, and readout, have to be simultaneously well engineered and integrated. If any component lags or if two components don’t function well together, it drags down system performance…

…The next challenge for the field is to demonstrate a first “useful, beyond-classical” computation on today’s quantum chips that is relevant to a real-world application. We’re optimistic that the Willow generation of chips can help us achieve this goal. So far, there have been two separate types of experiments. On the one hand, we’ve run the RCS benchmark, which measures performance against classical computers but has no known real-world applications. On the other hand, we’ve done scientifically interesting simulations of quantum systems, which have led to new scientific discoveries but are still within the reach of classical computers. Our goal is to do both at the same time — to step into the realm of algorithms that are beyond the reach of classical computers and that are useful for real-world, commercially relevant problems.

2. X (previously Twitter) thread on quantum computing and Google’s Willow – Jeffrey Scholz

Like a regular computer, a quantum computer keeps bits in groups. So a 64 bit quantum computer would have a vector of 64 2d vectors serving as it’s “word.”

Here is where the speedup happens: in a regular computer, each of the 64 bits don’t know anything about the value of any of the other 64 bits.

If we want one bit to affect another bit, we have to explicilty combine them with a logic gate.

However, in a quantum computer, each of the 64 qbits can “talk to each other” via “quantum entanglement.”

Running a quantum circuit means you plug in a quantum vector, run it through a bunch of matrix multiplications, then collapse the output.

The final vector will be the correct answer. Technically, quantum computers can give wrong answers, but if you run the computation multiple times, then you will get the correct answer on average…

…The current problem with quantum computers is that as the circuit gets bigger, they become less correct on average. All of the “talking to each other” creates so much noise the system stops working.

Once your probability of being correct drops below a certain threshold your quantum computer becomes useless. This is a major blocker for current quantum compute.

Let’s look at a specific (oversimplified but helpful) example. Suppose you shine a laser beam into an ice cube.

Actually simulating what the laser will do when it exits the ice cube is very hard to predict because some quantum phenomena is involved.

To actually compute what the laser will do means you have to explicilty compute quantum entanglement, which is slow for classical computers but “built in” to a quantum computer.

However, you can *estimate* the distribution of how the laser will scatter without a quantum computer, so you can have at least a rough idea if your answer might be correct…

…By analogy, this is what Google was doing. The computation Google was doing was a “pseudo-random quantum circuit” (think pseudoranom ice cube) but we know a quantum circuit is just matrix multiplications (on crack). Therefore, it is a bunch of random matrix multiplications with an output that looks right.

Google’s actual breakthrough was that the output of the circuit “looks correct” — which sounds underwhealming — and compared to the headlines, it definitely is. The academic breakthrough is that Google was able to use a larger circuit and notice an apparent *increase* in accuracy when modeling how a laser shines through an ice cube. That is noteworthy.

You can definitely tell if a computation has failed, and it seemed to be failing less as the circuit got bigger…

…However, note that the problem is “rigged” in favor of quantum computers. The benchmark is explicitly modeling a quantum phenomenon, so *of course* we get a speedup.

In other words, Google created a random distribution on the output that “seems correct.” Why does it “seem correct?” well because by design, the computation cannot be run on a classical computer. But if we can’t run it on a classical computer, how do we know the quantum computer is actually giving the right answer? The answer is we don’t, and this is a serious gap…

…Quantum computing is kind of at the stage right now where some smart teenager wired a few logic gates together in a random fashion and said “hey look, my circuit made a random output and didn’t explode!” Compared to previous attempts, it is an improvement. But he is still a long way from training an LLM.

3. Volatility: A Double-Edged Sword for Long-Term Equity Investors – Daniel Crowley

The ability to measure risk in a portfolio has long been a puzzle for the financial world. When Harry Markowitz introduced Modern Portfolio Theory in 1952, he revolutionized how institutions approached risk and return. His use of standard deviation as a proxy for volatility offered a clean, mathematical way to quantify the unpredictability of markets. It gave investors a seemingly precise tool to compare assets and assess portfolio risk. Over time, this approach became gospel, with concepts like beta and the Sharpe ratio reinforcing volatility as the core measure of risk.

But here’s the problem: volatility tells only part of the story. Financial markets don’t follow the neat patterns of a normal distribution, which is what these models assume. Extreme events occur far more often than traditional models predict. We’ve seen this play out time and again—from the collapse of Long-Term Capital Management to the Great Financial Crisis. The models couldn’t account for the market’s tendency to behave irrationally and with far greater extremes than the math suggested. That’s why I’ve come to view volatility not as risk itself but as a signal, an invitation to investigate further…

…Volatility is often misunderstood because it treats upward and downward price movements as equal. A stock with erratic upward swings may have high volatility but poses little risk if the business fundamentals are sound. Conversely, a stock that steadily declines might appear “safe” on paper but can quietly destroy wealth.

The market’s reliance on volatility as a measure of risk often misses these nuances.

This misunderstanding creates a divide among investors. On one side are those who cling to volatility as the ultimate arbiter of risk, building models that rely on neat equations and assumptions about market behavior. On the other are those who dismiss it entirely, treating volatility as irrelevant noise.

My view lies somewhere in the middle. Volatility is neither good nor bad—it’s just a clue. It’s a signal to dig deeper and assess whether the market’s movements are justified by changes in a business’s intrinsic value.

What I’ve come to appreciate about volatility is its ability to surface opportunity. Markets are emotional, driven by fear, greed, and short-term thinking. Prices frequently diverge from reality, creating moments where high-quality businesses are available at steep discounts. When markets panic, as they did during the COVID-19 pandemic or the Great Financial Crisis, those who can stay calm and look beyond the noise can identify extraordinary opportunities.

Volatility, far from being a risk, is often the price of admission for outsized returns.

4. The AI nuclear renaissance – SMRs role – Rihard Jarc

The global nuclear power market is about 10% of global electricity (about $350-$400B annually) and around 32% of zero-carbon electricity generation.

As of 2023, nuclear energy accounted for about 18.6% of total electricity generation in the United States. The International Energy Agency (IEA) highlights that global nuclear power output must more than double by 2050 to meet net-zero emission targets. Most of the U.S.’s nuclear power plants are over 50 years old and nearing the end of their operational lives. While their lifespans have been extended to support the grid, they will need to be replaced in the coming decades…

…The introduction of ChatGPT and the AI boom that we have experienced in the last 2 years have only accelerated as AI workloads and AI chips consume much more energy than traditional data center workloads. This Nuclear Energy expert gives a good example:

» If you provide a simple search in Google, you consume 0.3 W per hour of electricity. If you do the same with ChatGPT or Alexa or Gemini, any AI that we can imagine, this 0.3 W transforms into 2.9 W, so it means 10X the consumption.«…

…Driven by artificial intelligence (AI), cloud computing, and digital transformation, U.S. data centers consumed an estimated 150 TWh of electricity in 2023, equivalent to around 3% of the nation’s power demand. According to Goldman Sachs estimates, data center demand hovered at 340 TWh in 2023 globally, which is about 1.3% of worldwide electricity use. U.S. data center power use is expected to triple between 2023 and 2030 roughly and will require about 47 gigawatts of new generation capacity…

…Nuclear energy has become very attractive because companies want to be carbon-neutral and have stable power. An additional benefit of nuclear power is that it can provide more stable long-term contracts that are less sensitive to inflation and supply chain problems…

…Interest in nuclear energy, particularly Small Modular Reactors (SMRs), is growing as they have been heralded as a solution to streamline nuclear power production, offering flexibility, lower upfront costs, and modular deployment. The simplest way to imagine SMR is that it is a smaller version of the traditional nuclear reactor. One of their most significant benefits is that they are modular. They are designed to be built in factories, not on-site. Because they are built in factories, they are easier to assemble and control. From quality checks to a more predictable supply chain and quality of workers. When assembled, they are then shipped to the site of the nuclear plant, where they are stacked together to form the whole plant. In terms of energy output, traditional nuclear plants have outputs between 1,000-1,600 megawatts of electric (MWe) per reactor, while SMRs are around 50-300 MWe per module. Some SMRs are also said to be safer due to passive safety features, which rely on natural processes like convection to prevent meltdowns in emergencies. But they also come with cons. The primary one is that they are much smaller than traditional nuclear plants, so they do not have the cost benefits of economy of scale. Because of that, producing the same amount of energy is more expensive than on a traditional nuclear plant…

…Over 25 countries, according to the International Atomic Energy Agency (IAEA), are investing in SMRs. In March, Wood Mackenzie estimated the pipeline of SMR projects was worth more than $176 billion and that SMRs could account for as much as 30% of the global nuclear fleet by 2050…

…We can look at the example of NuScale, which has its Pressurised Water Reactor design. Their levelized cost of electricity ranges from $89-135/MWh, while traditional nuclear plants are in the $110-160/MWh. However, looking at the most traditional alternative in data centers, which is combined solar and gas, gas costs $45-70/MWh, and solar plus storage costs $30-60/MWh…

…State-backed projects in countries like China and Russia have made more progress, leveraging integrated supply chains, controlled costs, and assured revenue streams. But even for them, the costs to build these reactors compared to first estimates are still much bigger…

…We must also face reality, which says that only 2 SMRs are operational right now, one of which is in Russia and the other one in China.

Another important topic when assessing nuclear energy is the problem of nuclear waste and its storage. Most SMR designs produce a similar amount of nuclear waste on a unit production basis than traditional nuclear plants, so the problem of storing nuclear waste stays.

5. How to invest without relying on target prices – Chin Hui Leong

The US stock market is soaring to new heights. But what does that mean for your stock returns in 2025? I would like to give you a definite answer but if I did so, I would be lying to you. In fact, you should view anyone who gives you target prices with suspicion.

Here’s the hard truth: No one can control where the market is headed in the short term. Yet, the allure of target prices persists…

…The answer lies in the inherent difficulty in predicting the future of rapidly evolving technologies.

The best example is Amazon.com. In mid-2010, when I first invested in the company, it had just reported US$24.5 billion in annual revenue, primarily from its online retail business. Here is the twist: it was impossible to know what the business would look like a decade later…

…Fast forward to 2023, and AWS had become a financial cash cow with nearly US$90 billion in annual revenue and an impressive US$24.6 billion in operating income. In other words, AWS, an insignificant division back in 2009, had generated more operating income in 2023 than the entire company’s revenue in 2009…

…I like to go back to the reason why valuation is used in the first place: to reduce your investment risk. The way I see it, valuation is one of the many ways you can employ to manage risk. But valuation is not the only risk in investing.

A weak, shrinking business can pose risks that no amount of stock valuation can solve. Hence, starting with high-quality businesses is my preferred approach.

---

Disclaimer: The Good Investors is the personal investing blog of two simple guys who are passionate about educating Singaporeans about stock market investing. By using this Site, you specifically agree that none of the information provided constitutes financial, investment, or other professional advice. It is only intended to provide education. Speak with a professional before making important decisions about your money, your professional life, or even your personal life. We currently have a vested interest in Alphabet (parent of Google) and Amazon. Holdings are subject to change at any time.

We’ve constantly been sharing a list of our recent reads in our weekly emails for The Good Investors.

Do subscribe for our weekly updates through the orange box in the blog (it’s on the side if you’re using a computer, and all the way at the bottom if you’re using mobile) – it’s free!

But since our readership-audience for The Good Investors is wider than our subscriber base, we think sharing the reading list regularly on the blog itself can benefit even more people. The articles we share touch on a wide range of topics, including investing, business, and the world in general. 

Here are the articles for the week ending 08 December 2024:

1. Why China’s Economy Opened Up in the 1970s – Joe Weisenthal, Tracy Alloway, and Odd Arne Westad

Joe (13:32):

What does it mean when you talk about history being “contingent?” You used that word a couple of times and I actually don’t know if I fully understand what that means, but when you’re telling these stories, or this story, and you’re keeping in mind the contingency in history, can you talk a little bit more about this idea?

Odd (13:48):

So you’ll see from the book that we go in and out from the sort of micro to the macro level of telling history. And if you look at the night when the coup against the radicals — the so-called Gang of Four within the party — took place, which we describe in some detail, you know, what happens from hour to hour…

Joe (14:10):

Right, this was the moment in which the left faction, after Mao dies, was arrested, and allowed for a sort of more moderate path to emerge.

Odd (14:21):

That’s right. And it was in effect a military coup. I mean, it was undertaken by the military and the security forces against the people who Mao himself had put in charge of the party, including his widow who was most prominent of all, Jiang Qing. Now that night, and the following few days, things could have ended up very differently. I mean, Shanghai, the biggest city in China by far, was still under control of the radicals. There were military units that supported the radical approach to politics. This could have ended up very differently from what it did.

And as we describe in the book, some of the plotters, some of the coup-makers themselves, in those days that followed the coup itself, were completely surprised by how little resistance there had been from the left. And how little chaos there had been on the streets. So that’s what I mean with it being contingent. I mean, this is something that obviously connects to the larger picture that we see today — going back to your sort of three level version of what happened in China. But it didn’t seem that obvious at the time. And it could have gone in very different directions from what we’re seeing today.

Tracy (15:30):

How important was the fraying of the relationship between China and the Soviet Union in the 1960s, early 1970s to spurring or catalyzing that opening up? Because it does feel like the sudden emergence of the Soviet Union as an external enemy, it feels like that led China in some respects to open up to the US and some other countries.

Odd (15:56):

This is a sort of trajectory that I think it’s really important to get right, because what Mao and his group of leaders did in the late 1960s was to turn to the United States as an ally — a pseudo ally, security ally — against the Soviet Union because they were so deadly afraid that there would be a war with the Soviets — a war that China certainly would have lost, given the state that Chinese communists themselves had pulled China into during the Cultural Revolution. So what Mao did was to turn to the enemy far away, the United States, to help back him against an enemy much closer to home, the Soviet Union, which they had this falling out with mainly for ideological reasons.

From Mao’s perspective, this was always intended to be a strictly security oriented pseudo alliance. It was directed against the Soviet Union. Mao to the end of his days was puzzled that United States would support the real communists, meaning him, against the fake communists, meaning the Soviet Union. But as long as they were willing to do that, he was certainly willing to reap the benefits. But he never intended that this would have any effect in terms of the increasingly radical communist direction that he was taking for China internally, domestically.

So that’s when what happens in 1976, after Mao’s death, becomes so significant, because the people who then took over, they thought, ‘Aha! We have this relationship between United States. They are supporting us for their own reasons in the Cold War against the Soviet Union. We can now also make use of this to supercharge Chinese reform.’ If it hadn’t been for that relationship, strictly security oriented, that already existed between China and the United States, I doubt that that would be possible. So it’s very important when about the longer term US-China relationship to think about that origin and how this actually got started. Very different from the way most people think about it, where the security element and the reform element are sort of conflated into one…

…Odd (36:05):

I think it was both. I mean in the Xi Jinping case, I think he was picked by the party as the, what Chinese would call, the core leader, back in the early twenty-teens, in response to what was seen as a bunch of real problems, from a Chinese Communist Party perspective, over liberalization, decentralization, corruption, strength of private companies that meddled in a lot of things that the communists didn’t want them to meddle in. They wanted to get a strong leader in who could deal with those issues, in a way that his predecessors, Jiang Zemin [and] Hu Jintao, had not been able to do it. So they wanted a strong leader. It’s just that, I think even for many communist leaders of that generation, they got more than they bargained for. So that’s where the personality aspect comes in. They got a leader who really wanted to return, at least on some issues, to the Maoist or even the sort of pre-Mao period, in terms of the CCP’s history and emphasizes the party’s position over what even many party leaders back 10 [or] 15 years ago thought would be good for China.

And it’s a classic example of responding to real world problems — not unknown in this country, right? — by going very far in one direction, hoping that that would resolve the problem that is there, and then getting stuck in a way with the kind of leader that you have in this case, in Xi Jinping. So I think that’s the story, the way we can tell it now. I hope at some point to be able to tell that story based on archives and primary documents, as an historian, we can’t do that yet. But I think at some point, we’ll be able to do that, and then it’ll be fascinating to test that hypothesis about how this happened.

Tracy (37:54):

So just on the revolution from below point, one of the things that you emphasize in the book is a lot of the stuff that happens in this time period is a result of people feeling that they are heading somewhere, that there’s a grander Chinese vision that can be achieved. And so that motivates people to actually do something. I’m curious, just going up to the present day, do you get a sense that people feel that? That there’s like a direction that China is heading in that it’s clear to people what they are trying to do?

Odd (38:33):

At the moment, absolutely not. I think it’s very, very clear that a lot of people in China do not understand where the country is heading and what the reasons are. And you know, you don’t spend much time in Beijing before you realize that these days. I think it was very different in the time period that we are talking about, which was generally a time of uplift, at least in economic and and social terms. And it’s right to say, I mean as many historians have said, that there was an element of a bargain in this. That, at least for some Chinese, not not everyone, but for some Chinese, maybe particularly in business, that would accept a dictatorship for what it was and then went on getting rich and and establishing some of these great or middling fortunes that you find so many of in China today. And that is good. I mean that was positive. It was much, much better than the dark past that we described at the beginning of the book.

It was just that, China wasn’t able to take what, in our view, is a necessary step to improve its political system, its overall attempt at trying to become a more open, more pluralistic country in the period when the going was good, when there was a general sense that China was making advances, domestically and internationally. Now, I think even if people from within the Chinese Communist Party after Xi Jinping would try to move in a direction of increased liberalization — which I think they will have to do at some point because people are just very unhappy with the kind of system that is there at the moment — it would be much more difficult, because the going is not that good. And probably it’s never going to be that good again. I mean, it was a remarkable period of economic transformation, 10% per year growth rates. It would’ve been possible to carry out necessary reform. But these people didn’t want to do it because they had become so preoccupied with holding onto power themselves. And I think, historically, that that might turn out to be the biggest mistake that the Chinese Communist Party has made.

2. Tim Cook Wants Apple to Literally Save Your Life – Steven Levy and Tim Cook

Some companies charge for AI-enhanced services. Did you consider that?

We never talked about charging for it. We view it sort of like multitouch, which enabled the smartphone revolution and the modern tablet.

You’ve personally been using Apple Intelligence for a while. What has been most useful for you?

We’re an email-based company, and I get enormous numbers from users, employees, partners, and so forth. Having it summarize author responses is a game changer, and having it prioritize things for you so you’re not doing your usual triage. Then, of course, there are fun things like the Image Playground.

I’ve heard you say that Apple Intelligence could make you funnier, which seems strange.

I think it can make you friendlier, which, in many ways, can be funnier as well.

Having AI speak for people makes me wonder whether the nature of communication will degrade. If Apple Intelligence writes something funny, who’s being funny, the sender or the AI?

It’s still coming from you. It’s your thoughts and your perspective. You and I both remember the productivity that came from the advent of the personal computer. It was no longer you punching your calculator, you were doing something on a spreadsheet. It was no longer you at the typewriter, you were using a word processor. Logic Pro helps musicians create music, but they’re still the author.

One of your demos involves a fictional recent graduate applying for a job. The cover letter is colloquial and somewhat sophomoric, but with Apple Intelligence a single click changes it to look like a savvy, smart person wrote it. If I’m a recruiter who hired that person, maybe I will feel tricked if they don’t live up to the professionalism of that letter.

I don’t think so. By using the tool, it comes across as more polished. It’s still your decision to use the tool. It’s like you and I collaborating on something—one plus one can equal more than two, right?…

…When you’re thinking about things late at night, don’t you sometimes ask what it would mean if computers had superhuman intelligence?

Oh, of course. Not just for Apple, but for the world. There’s so much extraordinary benefit for humanity. Are there some things you have to have guardrails on? Of course. We’re very deeply considerate about things that we do and don’t do. I hope that others are as well. AGI itself is a ways away, at a minimum. We’ll sort out along the way what the guardrails need to be in such an environment…

…Meta and Snap are leading us to mixed-reality glasses that we’d wear continually. Is the bigger, heavier Vision Pro ultimately headed that way?

Yes, it’s a progression over time in terms of what happens with form factors. AR is a huge deal. With Vision Pro, we’ve progressed to what is clearly the most advanced technology we’ve ever done, and I think the most advanced technology in the world in terms of electronics problems. We’ll see where it goes.

Apple has created a lot of consumer tools for medical technology. What’s the strategy for biological metrics and prosthetics?

It’s clear to me that if you zoom out way into the future, and you look back and ask what Apple’s biggest contribution was, it will be in the health area. That’s what I really believe. When we started pulling that string with the Apple Watch, it was a cascade of events. We started with something simple, like monitoring your heart rate, and then figured out we could pick up heart signals to get to an EKG and an AFib determination. Now we are monitoring sleep apnea. I’ve gotten so many notes over time from people who would have not survived had it not been for the alert on their wrist.

Apple plans to give AirPods the ability to correct for hearing loss. I bet the makers of expensive hearing aids are freaking out.

It’s not about competing against hearing aids on the market. It’s about trying to convince people who have hearing loss to use their AirPods. The vast majority of people with hearing issues have not been diagnosed. For some people, hearing aids have a stigma, and we can counter that with AirPods. And we can have people diagnose themselves. It’s the democratization of health…

…We’re doing this interview at Apple Park, which is now seven years old. Have you been surprised by anything that couldn’t have been anticipated when it was just blueprints?

It’s promoted collaboration even more than I thought. That was a key component of the design, but there are so many places here where you just unexpectedly run into people. In the cafeteria, at the coffee bar, outside when you’re going across the pathway. Also, there’s a connection here to Steve that is incredible and very deep. We have the theater named after him and think about him all the time, but I can feel him in other spaces too.

3. 2024: The State of Generative AI in the Enterprise – Tim Tully, Joff Redfern, Derek Xiao, with Claude Sonnet 3.5

AI spending surged to $13.8 billion this year, more than 6x the $2.3 billion spent in 2023—a clear signal that enterprises are shifting from experimentation to execution, embedding AI at the core of their business strategies…

…Today, 60% of enterprise generative AI investments come from innovation budgets, reflecting the early stages of generative AI adoption. However, with 40% of generative AI spending sourced from more permanent budgets—58% of which is redirected from existing allocations—businesses are demonstrating a growing commitment to AI transformation…

…While foundation model investments still dominate enterprise generative AI spend, the application layer is now growing faster, benefiting from coalescing design patterns at the infrastructure level. Companies are creating substantial value by using these tools to optimize workflows across sectors, paving the way for broader innovation…

…In 2024, much of the action happened at the application layer. With many architectural design patterns established, app layer companies are leveraging LLMs’ capabilities across domains to unlock new efficiencies and capabilities. Enterprise buyers are seizing the moment, pouring $4.6 billion into generative AI applications in 2024, an almost 8x increase from the $600 million reported last year…

…Code copilots lead the charge with 51% adoption, making developers AI’s earliest power users…

…Support chatbots have captured significant usage, with 31% enterprise adoption…

…Enterprise search + retrieval and data extraction + transformation (28% and 27%, respectively) reflect a strong drive to unlock and harness the valuable knowledge hidden within data silos scattered across organizations…

…Meeting summarization ranks fifth in use cases (24% adoption), saving time and boosting productivity by automating note-taking and takeaways…

…When selecting generative AI applications, enterprises have clear priorities: Return on investment and industry-specific customization matter most when selecting new tools. Surprisingly, price isn’t a major issue; just 1% of the enterprise leaders we surveyed mentioned price as a selection concern. Buyers are playing the long game: They are far more focused on tools that can deliver measurable value (30%) and that understand the unique context of their work (26%) over those offering the lowest price tag (1%)…

…When AI pilots stutter or stall, it’s often due to challenges not adequately considered during the selection process. Although buyers aren’t checking price tags, implementation costs, cited in 26% of failed pilots, frequently catch them off guard. Data privacy hurdles (21%) and disappointing return on investment (ROI) (18%) also throw pilots off course. Technical issues, especially around hallucinations (15%), round out the top reasons for failure…

…Traditionally slow to adopt tech, healthcare is now leading generative AI adoption with $500 million in enterprise spend…

…Historically resistant to tech, the legal industry ($350 million in enterprise AI spend) is now embracing generative AI to manage massive amounts of unstructured data and automate complex, pattern-based workflows…

…With its complex data, strict regulations, and critical workflows, financial services ($100 million in enterprise AI spend) are primed for AI transformation…

…From Hollywood screens to creators’ smartphones, generative AI is reshaping media and entertainment ($100 million in enterprise AI spend)…

…Foundation models still dominate. The LLM layer commands $6.5 billion of enterprise investment…

…Rather than relying on a single provider, enterprises have adopted a pragmatic, multi-model approach. Our research shows organizations typically deploy three or more foundation models in their AI stacks, routing to different models depending on the use case or results…

…Among closed-source models, OpenAI’s early mover advantage has eroded somewhat, with enterprise market share dropping from 50% to 34%. The primary beneficiary has been Anthropic,* which doubled its enterprise presence from 12% to 24% as some enterprises switched from GPT-4 to Claude 3.5 Sonnet when the new model became state-of-the-art. When moving to a new LLM, organizations most commonly cite security and safety considerations (46%), price (44%), performance (42%), and expanded capabilities (41%) as motivations…

…To power RAG, enterprises must store and access relevant query knowledge efficiently. While traditional databases like Postgres (15%) and MongoDB (14%) remain common, AI-first solutions continue to gain ground. Pinecone,* an AI-native vector database, has already captured 18% of the market.

4. An Interview with Understanding AI Author Timothy B. Lee – Ben Thompson and Timothy B. Lee

As a side note, just as you sort of referenced it in passing, there is always the question of where are the productivity gains, when it came to, first the PC, and then the Internet? Is your sense that those just take a while to show up? Is there just a massive amount of consumer surplus that is not measured? What’s your big picture take on that question?

TL: There’s a couple of things. One is it takes a while to show up because to really get the big gains from a new general purpose technology, often you need to reorganize a lot of other business processes. There’s a famous analogy economists like to use for when they originally electrified the economy. The first thing they try to do is they tried to take the old steam-powered factories that just had one big crank shaft and put an electric motor in and that didn’t get you much improvement because the electricity was not cheap.

It was arguably worse.

TL: But then ten to twenty years later, people figured out, “Oh, we can have a bunch of small electric motors, one at each workstation, and now factories can be a lot more efficient”, but you had to build new factories and new businesses to do that…

…Believe me, I think we’re around the same age, I know exactly what you mean and feel. That said, I feel like the big company — Wikipedia came out back when I was in college, or around that time and of course everyone, professors or teachers, banned the use of it. But what you quickly realized is that the key way to use Wikipedia is the sources. You go to Wikipedia, and then it has links to all the sources, then you have your original source documentation. I do feel like ChatGPT is just such a better version of that, particularly with the search version, and when it does sources, it’s just like, “What if we make a Wikipedia that just fills all sort of weight and space about knowledge”, and it’s pretty tough to beat in that regard.

TL: Yeah, absolutely. And as with Wikipedia, you have to be smart about it. You can’t assume that everything is accurate, you have to check your work. But I definitely find, anytime I have, if I’m trying to make a list of things and I want to know all the companies in a particular category, it’s a pain in the ass to find that on Google. Whereas if you ask ChatGPT, “Here’s like three companies in this category, give me more on the list”, it’ll know a bunch more of them. There’s so many things like that. So yeah, definitely, I don’t want to say never use it or it’s not useful. It’s definitely useful, but it’s 1% to 2% more productive over the course of a week rather than really transformational…

...Again, to go back to your perspective of looking at it over the last 18, 20 months since you started, do you think we’ve hit a wall with AI? You started wondering this publicly actually last December when Gemini came out and you felt a little underwhelmed, particularly given Google’s advantages. You weren’t sure at the time, was Google underperforming for Google specific reasons, maybe have we gotten as far as we can with GPT-4? What’s your evaluation 11 months on from that article?

TL: The thing I’ve noticed is that we keep hearing about there’s going to be a GPT-5—

It’s not here.

TL: There’s going to be a new big model and it hasn’t been released and I don’t have enough sources in the inside to those companies to know why that’s happening. But it could be they’re just still working on it and it’s going to come out next month and blow my mind, but every month that ticks by makes me a little more skeptical. Especially because the other thing trend we’ve seen is these companies are releasing these smaller models that are almost as good as the big models.

And then even to some extent, I was pretty impressed by o1, but what o1 did is kind of different. It wasn’t like scaling up the model, it’s like we’re going to do more inference time compute. In certain ways, it was much better, but it wasn’t better overall.

So my still pretty rough hypothesis, but my hypothesis is that there’s kind of a limit to what the current LLM architectures can do and we’re sort bumping up against that in various — I mean, another thing, we’ve had multimodal models that are much better, so we can do real-time voice and we can do images, so there’s new things it can do. But in terms of just the increase of overall reasoning capability, it doesn’t seem like we’ve had a big jump, really since March of 2023 when GPT-4 came out, and so I’m not going to make a strong prediction because again, it could come out next month and amaze me, but every month that ticks by I get a little bit more wondering what’s going on.

What do you think is the limitation? Is it data, compute or is it just a fundamental limitation of the transformer architecture?

TL: My guess is it’s a fundamental limitation of the transformer architecture, and I think the main issue is that the transformer architecture requires all of the model state to be in these vectors for individual words, and then it keeps a record of that forever — the whole context, there’s no process where you summarize and abstract a way. If you think about your life, you think about something that happened ten years ago, you don’t remember every single thing you said, everything that others said, you have a abstract memory that, “Oh, in 2014 I remember I lived in this place and I had this job”, and things you learn kind work their way into the brain, but it’s organized in a good way. LLMs just don’t have a way to do that.

So if I think about how people expect that at some point you’re going to have an LLM who’s like a personal assistant who maybe will work with you over your career and know all your habits and make all your appointments stuff and to do that, I just think this architecture where you remember every token exactly and do attention over that whole corpus, I don’t have any way of synthesizing and abstracting and forgetting unimportant things, just as a computer scientist, that doesn’t seem viable to me…

…Do you think there’s a bubble now then?

TL: That’s always a hard question to say. Part of what’s hard about bubbles is that often people start calling a bubble pretty early and then the bubble keeps growing and people keep saying there’s a bubble.

Right. If people think there’s a bubble, there is not a bubble, that’s my heuristic.

TL: Well, there’s that, but also, at some point, the stock or the house price or whatever will peak and then go down, and the people who said it was a bubble right at the top will be right, but some people who called it way at the beginning were probably wrong.

I do expect a period where AI gets overly frothy and then crashes. Whether we’re currently there or just headed for that, is a little hard to say. I do not expect a dot-com bust level expansion, because as you were saying, I do think that this technology has clear benefits, it’s mostly big technology companies, it’s not as venture-funded. In fact, some of the early really crazy-funded companies have already been acquired.

So, yeah, I think the level of hype right now is a little too high and there’ll be some pullback, but I don’t think you’ll see a big crash and I don’t think you’ll see much of a pullback from deployment, because I think there really is enough value here that there’s going to be a big market for a lot of people working on it, and a lot of valuable stuff will come out of it in a pretty direct way.

I saw a new theory this week that actually really resonated with me. So this might be new to you, so I’m going to drop it to you on the spot. I think the big question on if you’re thinking about bubbles, you go back to a Carlota Perez model of the importance of bubbles and driving, you go back to the dot-com era, the really important part was the telecoms build out, which was, at the time, some people called it, and in retrospect, clearly insane. If you’re rolling out all this fiber and everyone’s doing it, the costs are going to go to zero, you’re all going to go bankrupt because it’s all financed by debt, as large infrastructure usually is. But the long-term payoff from that was massive, right? That, basically, booted off the whole Web 2.0 era where now everyone, suddenly, had broadband. Recessions suck, but there was a huge societal benefit that did come from that build out.

You go back to previous ones, whether it be electricity or steam, you had these similar cycles and the big question was, “What’s the societal beneficial output of an AI bubble if there is a bubble?” and chips never quite fit, because chips wear out and chips get better. So, if you buy a bunch of chips, but they’re five-year-old chips, what’s the benefit there? Doug O’Laughlin put this tweet out here, that has been really striking to me. He said, “Internet Bubble:Telecom::AI:Power/DCs”, and to me, that makes sense. If you’re going to actually build more nuclear power, or you’re going to do massive investments in solar and batteries, or whatever it might be to fuel these sorts of things, those are investments that, 1) can definitely make you go bankrupt because you’re taking out a bunch of debt to fund it, but 2) will retain value for many, many, many years to come. What do you think of that analogy? To me, it seems pretty compelling.

TL: Yeah, I one hundred percent agree with that. I mean, I was actually going to say the part of it that seems most bubbly is this stuff about Microsoft leasing out Three Mile Island for 20 years. Again, we were talking before is, “Do I think scaling law thing is going to run out of steam?”, my guess is it probably will. I don’t know if we’re on the verge of that, but, anyway, so I would not be surprised if people look back ten years from now, and say, “Oh, man, all that money companies spent on data centers and power is, that was kind of a waste of money”. But then, like you said, the country needs more power, and at some point, probably, we’ll want to be training really big models and so, if we have a bunch of huge data centers that we can use to train models, probably, we’ll get some value out of that. It’s tech companies spending the money so the social cost is not probably that high.

5. 7% of Book Value; 1x EBITDA; Cash is 2.5x Larger than Market Cap – Dirtcheapstocks

Highlands REIT, Inc. (Ticker HHDS) was created in 2016 when it was spun out of InvenTrust Properties Corp.

HHDS was formed to hold non-core assets of InvenTrust.

Today, HHDS owns 13 apartment houses, 3 retail properties, 1 office property and 1 correctional facility…

…HHDS has:

• $205MM of book value.
• $16.7MM of net operating income (NOI) in 2023.
• $17MM of NOI in 2022.
• $85MM of net debt.
• 57% of NOI generated from multifamily assets

What do you think? Is Highlands worth book value? Is it worth half of book value?

If we want to value the business at an 8 cap, the equity must be worth $124MM.

Within the last two weeks, HHDS has been valued as low as $14.4MM.

That’s less than 1x NOI, and 7% of book value…

…Most companies valued at $14MM might have a few hundred shareholders of record. Apple is valued at $3.5 Trillion, and it has 23,000 record holders.

Highlands has 143,000 record holders…

…Here’s my theory: When Highlands was spun out of InvenTrust, every shareholder was given ownership individually. There are 143,000 separate people/entities that own this stock. And this stock was an afterthought. It was just a few noncore assets being spun out of a $2 billion REIT…

…HHDS, perhaps wanting to ward off future material purchases by Mackenzie, announced a tender offer in October 2023. While Mackenzie was tendering at $0.04/share earlier that summer, HHDS was willing to pay $0.12 – $0.17/share. What’s more, HHDS was committing $20MM to the share buyback.

HHDS would repurchase 13-19% of its shares if fully subscribed.

A few weeks later, HHDS increased the buyback to $25MM!

In the end, $23.7MM was spent to buy in 169MM shares – nearly 20% of the outstanding share count…

…HHDS showed up as an expert market security, even though it’s SEC registered.

But I found that the traditional expert market brokers couldn’t buy shares.

Then I went to alternative market brokers. They’d be happy to take my money, and told me I could get as much volume at $0.10 as my heart desired.

---

Disclaimer: The Good Investors is the personal investing blog of two simple guys who are passionate about educating Singaporeans about stock market investing. By using this Site, you specifically agree that none of the information provided constitutes financial, investment, or other professional advice. It is only intended to provide education. Speak with a professional before making important decisions about your money, your professional life, or even your personal life. We currently have a vested interest in Alphabet (parent of Google), Apple, Meta, Microsoft, and MongoDB. Holdings are subject to change at any time.