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 May 2024:

1. Why Xi Jinping is afraid to unleash China’s consumers – Joe Leahy

Both inside and outside China, there is a strongly held view among many economists that the country could secure a further period of robust growth if it were able to boost consumption by its own citizens. Indeed, faced with a property crisis, President Xi Jinping has taken some one-off measures to stimulate consumption to offset a fall in domestic demand.

But Xi has eschewed more radical medicine, such as cash transfers to consumers or deeper economic reforms. His latest campaign is instead to unleash “new quality productive forces” — more investment in high-end manufacturing, such as EVs, green energy industries and AI.

According to analysts, the reasons for the lack of more radical action on consumption range from a need to generate growth quickly by pumping in state funds — this time into manufacturing — to the more deep-seated difficulties of reforming an economy that has become addicted to state-led investment.

Ideology and geopolitics also play roles. For Xi, China’s most powerful leader since Mao Zedong, the greater the control his country exerts over global supply chains, the more secure he feels, particularly as tensions rise with the US, analysts argue. This leads to an emphasis on investment, particularly in technology, rather than consumption.

Under Xi, security has also increasingly taken precedence over growth. Self-reliance in manufacturing under extreme circumstances, even armed conflict, is an important part of this, academics in Beijing say…

…The pressure on Beijing to find a new growth model is becoming acute, analysts say. China has become too big to rely on its trading partners to absorb its excess production.

“The exit strategy has to be, at the end of the day, consumption — there’s no point producing all this stuff if no one’s going to buy it,” says Michael Pettis, a senior fellow at the Carnegie Endowment in Beijing.

Few projects capture Xi’s vision for 21st-century Chinese development as well as Xiongan, a new city being built on marshlands about 100km from Beijing…

…Xiongan unites many of Xi’s favourite development themes. Through vast investment in mega-infrastructure projects such as a high-speed rail hub, Xiongan aims to bring state-owned enterprises, universities and entrepreneurs together to concentrate on high-technology innovation, from autonomous vehicles and life sciences to biomanufacturing and new materials. As of last year, about 1mn people were living there, $74bn had been invested and 140 companies had set up there, Beijing says.

Conspicuously absent from the city plans are strategies to encourage the thing China’s economy lacks most — domestic consumption. In guidelines released in 2019 for Xiongan by Xi’s cabinet, the State Council, there was no mention of the term “consumption”, except for “water consumption”…

…China’s investment to gross domestic product ratio, at more than 40 per cent last year, is one of the highest in the world, according to the IMF, while private consumption to GDP was about 39 per cent in 2023 compared to about 68 per cent in the US. With the property slowdown, more of this investment is pouring into manufacturing rather than household consumption, stimulating oversupply, western critics say…

…Economists suspect that behind the rhetoric, the investment in manufacturing is partly pragmatic. With the property market still falling three years after the crisis began, and many indebted provinces ordered to suspend large infrastructure projects, Xi needs to find growth somewhere to meet his 5 per cent target for this year.

“The bottom line is they want growth in output and they want the jobs associated with that growth,” says Stephen Roach, a faculty member at Yale and former chair of Morgan Stanley Asia. He says when “they’re clamping down on property, it doesn’t leave them with much choice but to go for a production-oriented growth stimulus”…

…In areas vital to China’s national security, the country needed supply chains that “are self-sufficient at critical moments”, he said. “This will ensure the economy functions normally in extreme circumstances.”

HKU’s Chen says China no longer measures its “national power” in purely economic terms “but more importantly, in terms of military . . . capacity. And this is why manufacturing is very important”.

He says in this vision of the world, consumption is a lower priority…

…The Rhodium Group argues that some of the loans that flowed into the industrial sector last year went to local government finance vehicles, the heavily indebted off-balance sheet investment holding companies of provinces and municipalities.

While large sums still went to manufacturers, they “do not have a strong appetite to expand capacity given falling prices”, Rhodium said in a report.

Economists say that for consumers to feel comfortable to spend more, particularly after the property slump, China needs to step up its development of social welfare programmes and healthcare. While China has made strides in building out its public pension and healthcare systems, they are still lacking.

But such solutions would take a long time to boost consumer confidence and would require massive new funding from government coffers that are running dry.

Greater consumption would also necessarily mean reducing the role of manufacturing or investment in the economy. This could be done by unwinding China’s intricate system of subsidies to producers, which includes government infrastructure investment, access to cheap labour, land and other credit, says Pettis.

But if that was done in a big bang fashion, the share of household consumption to GDP would increase while overall GDP would contract as manufacturers suffered. This was obviously not a politically preferable option for Xi.

2. Strategy Reviews – John H. Cochrane

After an extensive extended and collective deliberation, the Fed adopted a new strategy framework known as Flexible Average Inflation Targeting. This framework was explicitly designed by a worldview that “the federal funds rate is likely to be constrained by its effective lower bound more frequently than in the past,”  and a consequent judgement that “downward risks to employment and inflation have increased.” A shift to “inclusive” employment, a return to the old idea that economic “shortfalls” can be filled, and a promise not to preempt future inflation but rather let inflation run hot above 2% to make up past shortfalls followed. These promise of future dovishness were hoped to stimulate demand in the short run.

In short, the Fed adopted an elaborately-constructed new-Keynesian forward-guidance defense against the perceived danger of deflation and stagnation at the zero bound.

No sooner was the ink dry on this grand effort, however, than inflation shot up to 8%, and the zero bound seemed like a quaint worry. Something clearly went drastically wrong. Naturally, the first question for a strategy review is, how can we avoid having that happen again?

Inflation eased without interest rates substantially higher than inflation or a large recession. I think I have a (and the only) clear and simple explanation for that, but I promised not to digress into a fiscal theory today. Still inflation is persistently high, raising the obvious worry that it’s 1978 again. Obviously, central banks have a range of worries on which to focus a new strategy, not just a return to a long-lasting zero bound. (Though that could happen too.)…

…React or guide? It seems clear to me that policy will have to be described more in terms of how the Fed will react to events, rather than in standard forward guidance terms, unconditional promises of how the funds rate will evolve. It will involve more “data-dependent” rather than “time-dependent” policy.

In part, that must come, I think, as a result of the stunning failure of all inflation forecasts, including the Fed’s. Forecasts did not see inflation coming, did not see that it would surge up once it started, and basically always saw a swift AR(1) response from whatever it was at any moment back to 2%. Either the strategy review needs to dramatically improve forecasts, or the strategy needs to abandon dependence on forecasts to prescribe a future policy path, and thus just state how policy will react to events and very short-term forecasts. I state that as a question for debate, however…

…Fiscal limitations loom. Debt to GDP was 25% in 1980, and still constrained monetary policy. It’s 100% now, and only not 115% because we inflated away a bunch of it.  Each percentage point of real interest rate rise is now quickly (thanks to the Treasury’s decision to issue short, and the Fed’s QE which shortened even that maturity structure)  a percentage point extra interest cost on the debt, requiring a percent of GDP more primary surplus (taxes less spending). If that fiscal response is not forthcoming, higher interest rates just raise debt even more, and will have a hard time lowering inflation. In Europe, the problem is more acute, as higher interest costs could cause sovereign defaults. Many central banks have been told to hold down interest rates to make debt more sustainable. Those days can return…

…Ignorance. Finally, we should admit that neither we, nor central banks, really understand how the economy works and how monetary policy affects the economy. There is a complex verbal doctrine that bounces around central banks, policy institutions, and private analysts, asserting that interest rates have a relatively mechanical, reliable, and understood effect on “spending” through a “transmission mechanism” that though operating through “long and variable lags” gives the Fed essentially complete control over inflation in a few years. The one thing I know from 40 years of study, and all of you know as well, is that there is no respectable well-tested economic model that produces anything like that verbal doctrine. (More here.)  Knowing what you don’t know, and that nobody else does either, is knowledge. Our empirical knowledge is also skimpy, and the historical episodes underlying that experience come with quite different fiscal and financial-structure preconditions. 1980 was a different world in many ways, and also combined fiscal and microeconomic reform with high interest rates.

3. Big Tech Capex and Earnings Quality – John Huber

Capex is not only growing larger, but the rate of growth is set to accelerate this year as they invest in the AI boom. Combined capex at MSFT, GOOG and META is set to grow around 70% in 2024. As a percentage of sales, capex will grow from 13% of sales in 2023 to around 20% in 2024…

…Bottom line: the other Big Techs are getting far more capital intensive than they have in the past. Their FCF is currently lagging net income because of the large capex, and this will eventually flow through to much higher depreciation charges in the coming years.

This is not necessarily worrying — if the returns on these investments are good, then sales growth will be able to absorb these much higher expenses. But this is not a sure thing, so I like to use P/FCF metrics as I think a large majority of the assets they’re investing in will need to be replaced. This means the capex levels we see currently could be recurring. So, while the P/E ratios range from 25 to 35, the P/FCF ranges from 40-50.

Again, if the investments are able to earn good returns, then profit margins will remain intact, but one thing to notice is FCF margins (while very strong) have not kept up with GAAP profit margins: e.g. at MSFT, FCF margins have declined slightly from 28% to 26% over the last decade while net margins have expanded from 25% to 36%, leaving GAAP profit margins far in excess of FCF margins. Eventually, as growth slows these margins will tend to converge as depreciation “catches up” to cash capex spend. Whether net margins come down or FCF margins move up simply depends on the returns on capital earned and the growth it produces.

I’m not predicting a poor result, but I’m mindful of how difficult it will be given how different the companies are today. They used to grow with very little capital invested, but now they have a mountain of capital to deploy, which is obviously much harder at 7 times the size:…

…I don’t think anyone (including management) yet knows what the returns on the $150 billion of investments that these three companies will spend in 2024. They are optimistic, but it’s not clear cut to me.

Think about how much profit needs to be generated annually to earn acceptable returns on this capex: a 10% return would require $15 billion of additional after tax profits in year 1. As Buffett points out, if you require a 10% return on a $150 billion investment but get nothing in year 1, then you’d need $32 billion in year 2, and just one more year of deferred returns would require a massive $50 billion profit in year 3.

What’s staggering is that the above is the return needed to earn 10% on just one year’s worth of capex. Even if we assume that capex growth slows from 70% this year down to 0% in 2025 and stays there, MSFT, GOOG and META will invest an additional $750 billion of capital over the next 5 years!

What’s staggering is that the above is the return needed to earn 10% on just one year’s worth of capex. Even if we assume that capex growth slows from 70% this year down to 0% in 2025 and stays there, MSFT, GOOG and META will invest an additional $750 billion of capital over the next 5 years!

4. A Few Short Stories – Morgan Housel

Thirty-seven thousand Americans died in car accidents in 1955, six times today’s rate adjusted for miles driven.

Ford began offering seat belts in every model that year. It was a $27 upgrade, equivalent to about $190 today. Research showed they reduced traffic fatalities by nearly 70%.

But only 2% of customers opted for the upgrade. Ninety-eight percent of buyers preferred to remain at the mercy of inertia.

Things eventually changed, but it took decades. Seatbelt usage was still under 15% in the early 1980s. It didn’t exceed 80% until the early 2000s – almost half a century after Ford offered them in all cars.

It’s easy to underestimate how social norms stall change, even when the change is an obvious improvement. One of the strongest forces in the world is the urge to keep doing things as you’ve always done them, because people don’t like to be told they’ve been doing things wrong. Change eventually comes, but agonizingly slower than you might assume…

…When Barack Obama discussed running for president in 2005, his friend George Haywood – an accomplished investor – gave him a warning: the housing market was about to collapse, and would take the economy down with it.

George told Obama how mortgage-backed securities worked, how they were being rated all wrong, how much risk was piling up, and how inevitable its collapse was. And it wasn’t just talk: George was short the mortgage market.

Home prices kept rising for two years. By 2007, when cracks began showing, Obama checked in with George. Surely his bet was now paying off?

Obama wrote in his memoir:

George told me that he had been forced to abandon his short position after taking heavy losses.

“I just don’t have enough cash to stay with the bet,” he said calmly enough, adding, “Apparently I’ve underestimated how willing people are to maintain a charade.”

Irrational trends rarely follow rational timelines. Unsustainable things can last longer than you think…

…John Nash is one of the smartest mathematicians to ever live, winning the Nobel Prize. He was also schizophrenic, and spent most of his life convinced that aliens were sending him coded messages.

In her book A Beautiful Mind, Silvia Nasar recounts a conversation between Nash and Harvard professor George Mackey:

“How could you, a mathematician, a man devoted to reason and logical proof, how could you believe that extraterrestrials are sending you messages? How could you believe that you are being recruited by aliens from outer space to save the world?” Mackey asked.

“Because,” Nash said slowly in his soft, reasonable southern drawl, “the ideas I had about supernatural beings came to me the same way that my mathematical ideas did. So I took them seriously.”

This is a good example of a theory I have about very talented people: No one should be shocked when people who think about the world in unique ways you like also think about the world in unique ways you don’t like. Unique minds have to be accepted as a full package.

5. An Interview with Databricks CEO Ali Ghodsi About Building Enterprise AI – Ben Thompson and Ali Ghodsi

So you said you came over to the U.S. in 2009. Did you go straight to UC Berkeley? There’s some great videos of you giving lectures on YouTube. You’re still an adjunct professor there. Do you ever teach anymore or is this a, “Homeboy made good, we’ll give him the title forever”, sort of situation?

AG: No, I teach about a class a year and I still enjoy really doing that. I imagine if I had nothing to do, that’s a job I would actually enjoy doing.

So yeah, I came to the United States just to stay here one year and do research at UC Berkeley and just ended up staying another year, another year, another year. And the timing was — we didn’t know it at the time, but Dave Patterson, who was a professor at UC Berkeley, and now Turing Award winner, which is the Nobel Prize in computer science essentially, said at the time, “We’ve had Moore’s Law, but we no longer know how to make the computers faster and cramming more transistors. That era is over, so computers are not going to get any faster”, and we know he was right, they’re all between two and four gigahertz since then.

So we need the new computer, and the new computer is the cloud, and it also needs new software, so we built all this software stack — the era of data and AI. So it was the perfect time. I always regretted, “Why was I not born in the ’50s or ’60s when computers happened?” — well, actually it kind of happened again in ’08, ’09, ’10, and Berkeley was at the forefront of that. So we were super lucky to see that kind of revolution and being part of that…

…The general idea is you mentioned you started out with Mesos where you needed to compute in parallel instead of serially so you have to have a cluster of computers, not just one. Spark lets you basically do the same thing with data, spread it out over a huge number of computers. You can end up with massive amounts of data, structured, unstructured, people will call it like a “data lake”. There’s a data lake, there’s a data warehouse, there’s a Data Lakehouse. Walk me through the distinction and where that applies to Databricks and its offering.

AG: At the time, the world was kind of split. Those that have structured data, structured data are things that you can represent in tables with rows and columns, those were in data warehouses and you could connect your BI tools, business intelligence tools, that lets you ask questions about the past from those rows and columns. “What was my revenue last week in different regions, by different products, by different SKUs?”, but you couldn’t ask questions about the future.

Then at the same time, we had these future looking workloads, which were, “Okay, we have all kinds of text, images, and unstructured data that’s coming into the enterprise,” and that you couldn’t store in these structured tables, they cannot be represented as tables of rows and columns, those you stored in what’s called data lakes. But then the good news was if you knew what you were doing, you could ask questions about the future, “What’s my revenue going to be next week? Which customer is going to churn next?”. But these worlds were living completely separately and securing them was very hard and there was a lot of redundant stacks that were being built up at the same time.

Our idea was how do we, 1) unify this and 2) how do we disrupt the existing ecosystem? How do we create the company that’s disruptive? And our idea was what if we have open source technology, everybody stores all their data, both the structured and unstructured data in the lake, which is just basically almost free storage by the cloud vendors, but we standardize the format in an open source format, so it almost becomes like USB — you can plug anything in there. Then we build an engine that can do both the BI stuff, backwards looking questions, and the futuristic AI stuff, and that’s what we call the Lakehouse, which is a portmanteau of data lakes and their warehouses. The marketing firms we talked to and anyone we’d ask said, “This is a terrible idea”…

…So you’ve been using the word AI a lot. Did you use the word AI a lot five years ago?

AG: I think we used the word ML quite a bit.

Yeah, machine learning. That’s right, there’s a big word branding moment. I mean, there was the ChatGPT moment, so I guess there’s two questions. Number one, did that shift how you were thinking about this space, or was this already super clear to you? But then number two, I have to imagine it fundamentally changed the way your customers were thinking about things and asking about things.

AG: Yeah, from day one we were doing machine learning. We actually built MLlib as part of Spark already before we actually started Databricks. Actually the first use case of Spark in 2009 was to participate in the Netflix competition of recommending the best movie, and we got the second prize actually, we didn’t get the first prize.

The whole point about being able to distribute broadly and do things in a highly parallel manner, I mean we’re basically in that world.

AG: Exactly. Well, a lot of people also use that parallel worlds to just do backwards processing, like a data warehouse that tells you, “Tell me everything about the past”, and it’s great to see trend lines about the past, but to use this kind of more advanced statistical approach, that’s when you venture into machine learning. We were doing it already in 2012, ’13, I tried to push the company to use the phrase AI instead of ML, the most hardcore academics in the company were against it. They said that AI was a buzzword but I said, “No, I think that’s actually what resonates with people”. But at the same time we were seeing more and more deep neural networks so these neural networks are getting stacked do better and better.

Around 2018 is when we started seeing especially language processing, natural language processing, getting more and more applications on the platform. We saw insurance companies using them to analyze huge amounts of texts to assess risks, we saw translation starting to happen, we saw pharma companies analyzing big amounts of electronic medical records that were written, unstructured text. So it was pretty clear that something is going on with NLP [Natural Language Processing] and that just accelerated during the pandemic. So we saw it, we already had over a thousand customers using these kind of transformer models. So when ChatGPT came out, we kind of thought it’s a nothing burger, but of course we were wrong in that it was an absolute awareness revolution.

Yes, exactly.

AG: What we took for granted was not what the rest of the world was taking for granted. So we feel like the world woke up to AI in November 2022 with ChatGPT, though the truth is it had been going on for 20 years.

That’s what strikes me. That’s the biggest impact is number one, you had the total rebranding of everything to AI, my washing machine now has AI, what a miracle. But just the fact that you went through this when you started with Spark, you thought this is a great idea, no one knows what it is. Now suddenly people are asking you, knocking on your door, “We have data on your thing, can we run ChatGPT on it?” — is that how those conversations went?

AG: Yeah, I mean literally before ChatGPT, I would tell the marketing department to tone down the AI language because customers would say, “Hey, this AI stuff is futuristic, we have concrete problems right now with data that we need to solve”, so I actually shot down a marketing campaign and marketing was really upset about it, which said, “Customer X is a data and AI company, Customer Y is a data and AI company”. They had it ready to go and I shot it down and I said, “We don’t want to push so hard on AI because people don’t really want AI”, and then literally after ChatGPT happened, I told them, “Hey, that campaign from a couple of years ago, maybe we should run it now” — which we did actually and people loved it. So yeah, it’s just the market was just not ready…

…All right, number three, Databricks solves Mosaic ML’s need to build a sales force and Mosaic ML solves Databricks need to build a sustainable differentiated business around an open source project.

AG: Yes, I think you are 99% right? I would modify that last sentence to say —

I didn’t give you enough credit for how much you had differentiated to date?

AG: No, I actually think that you kind of were spot on, but I would say with open source, I would say that it was Mosaic ML having a research team that really was deep in LLM research and AI, it was hard to come by at the time and it was very, very hard actually to hire those researchers that really gave us that. And then the know-how to customize LLMs on your data in a secure way.

How does that work? How do you do that?

AG: So this is what their specialty was. When everybody else was building one giant model or a few giant models that are supposed to be very smart, these guys, their business model was, “We build it again and again and again, custom either from scratch or from an existing checkpoint, you tell us or we can fine tune it, but we can help you build an LLM for you and we will give you the intellectual property of that LLM and its weights”. That way you as a customer can compete with your competitors and in the long run you become a data and AI leader just like our billboards that I had banned a few years earlier say. You’re going to be a data and AI company. It doesn’t matter if you’re a pharma company or a finance company or a retail company, you’re actually going to be a data and AI company, but for that you need intellectual property. Elon Musk is not just going to call OpenAI for his self-driving capabilities, he needs to have his own. Same thing is going to be true for you in finance, retail, media. So that was their specialty, but we had the data.

Is that actually true though? Do they actually need to have their own intellectual property or is there a sense — my perception, and I picked up on this, I was at some sort of conference with a bunch of CEOs, it struck me how they had this perception of, “We’ve had this data for years, we were right to hold onto it, this is so valuable!”, and I’m almost wondering, are you now so excited about your own data that you’re going to be over protective of it? You’re not going to want to do anything, you’re actually going to sort of paralyzed by, “We have so much value here, we have to do it ourselves”, and miss out on leveraging it sooner rather than later because you’re like, “It has to be just us”.

AG: No, I do think that people have now realized how valuable their data is, there’s no doubt about that and it is also true, I believe in it. The way I think of it is that you can think of the world as two kind of parallel universes that coexist these days with LLMs. We’re super focused on one, which is the kind of open Internet and the whole crawl of everything that’s in it and all of the history of mankind that has been stored there. Then you’re building LLMs that’s trained on that and they become intelligent and they can reason and understand language, that’s what we’re focused on.

But we’re ignoring this other parallel universe, which is every company on the planet that you join have you sign an NDA, an employee agreement, and then that gives you access to all this proprietary data that they have on the customers and everything else, and they have always been protective of that. The LLMs today that we are training and we’re talking about, they don’t understand that data, they do not understand the three letter acronyms in any organization on the planet.

So we do the boring LLMs and the boring AI for those enterprises. We didn’t have quite the muscle to do it without Mosaic, they really understood how to build those LLMs, we had the data already. So we had the data and we had the sales force, Mosaic did not have the data, they did not have the sales force, they did have the know-how of how to build those custom models.

I don’t think that the companies are hamstrung and they’re not going to do anything with it, they want to do things with it. I mean, people are ready to spend money to do this. It’s just that I feel like it’s a little bit of a 2007 iPhone moment. iPhone comes out, every company on the planet says, “We have to build lots of iPhone apps, we have to”. Then later it turns out, “Well, okay, every company building a flashlight app is maybe not the best use of resources, in fact, maybe your iPhone will just have a flashlight in it”. So then it comes back to what special data do you have that no one else has, and how can we actually monetize that?

How does it actually work to help companies leverage that? So you released a state-of-the-art open LLM, DBRX, pretty well regarded. Do you do a core set of training on open data on whatever might exist and then you’d retrain it with a few extra layers of the company’s proprietary data and you have to do that every time? How modular is that? How does that actually work in practice?

AG: Yeah, there’s a whole slew of different techniques ranging from very, very lightweight fine tuning techniques. The most popular one is called LoRA, low rank adaptation, to actually training a chunk of the model. So you take an existing model that’s already trained and it already works and you customize a bunch of the layers to what’s called CPT, continuous pre-training, in which case you actually train all of the layers of the model, an existing model that’s already baked and ready, but you train all of the layers. It costs more to do that to all the way if you’re doing something really different. So if the domain that you’re using for the data set is significantly different, then you want actually what’s called pre-train, which is train the model from scratch. If you’re a SaaS application and LLMs is the core of the offering, you probably want to have a pre-trained model from scratch, so we can do all of those.

I would say the industry is not actually a hundred percent, the research is not a hundred percent clear today of when should you use which, where. We have a loose idea that if you don’t have huge amounts of data and it’s kind of similar in domain to what the LLM already can do, then you can probably use the more lightweight ones, and if your data is very different and it’s significant, then probably the lightweight mechanisms are not good for you, and so on. So we have a research team that really can do this really, really well for enterprises. But I think a lot of progress is going to happen in the next few years to determine how can we do this automatically? How do we know when to use them? And there might be new techniques also that are developed.

What’s the trade-off? I imagine you talk to a company, we absolutely want the most accurate model for sure, we want it totally customized to us. And then you’re like, “Okay, that’s going to cost XYZ, but then also to serve it is going to cost ABC”. The larger a model is the more expensive it is to serve and so your inference costs are just going to overwhelm even the upfront costs. What’s that discussion like and trade-off like that you’re having with your customers?

AG: Well, the vast majority have lots of specific tasks that they want to do. So again, a lot of people are thinking of things like ChatGPT, which are sort of completely general purpose open-ended, ask me anything. But enterprise typically have, “Okay, I want to extract labels from all this core piece of data, I want to do it every day like ten times a day”, or, “I want to tag all of these articles with the right tags and I want to do that very accurately”. So then actually for those specific tasks, it turns out you can have small models. The size of the model helps you actually be much cheaper and that matters at scale and then they are really, really concerned about quality and accuracy of that, but for that specific task, it doesn’t need to nail a super balanced answer to the question of whether there was election fraud or not in 2020.

(laughing) Right.

AG: It just needs to really extract those tags really, really well, so then there are techniques you can use to that. There is a way where you can actually have your cake and eat it too, assuming that the task you want to do is somewhat narrow.

But we also have customers that are, “No, I’m building a complete interactive general-purpose application in say, many of the Indian dialects in India, and I want to do that, and existing models are not very good at that, help me do that”. Then you have to go for a bigger model but bigger is usually more expensive. Of course, we are using the mixture of experts architecture, which we think is where the world is headed and which is what people also think what GPT-4 was based on, but we’ve also seen with Llama 3 from Meta that dense models, that are not mixture of experts, are also excellent and they’re doing really, really well…

…Is there a difference between domestic and international in terms of the aggressiveness with which they’re approaching AI?

AG: Yeah, I would say that China is moving very, very fast on AI. Some Asian countries, there’s less regulation. Europe, I feel is lagging always, has been lagging a few years behind the United States, and they’re concerned about — there’s also competitive concerns with so many American companies, cloud companies and so on from Europe. So Europe is a little bit more regulated and a few years usually lagging United States.

That’s what we’re seeing, but there’s regional differences. Like India is very interesting because it’s moving so fast, there’s no signs of anything that’s recession-like over there. There are markets like Brazil and so on that are doing really well. So really, you have to go case-by-case, country-by-country. We have significant portion now of our business in Europe as well, and also now a growing business in Asia and also Latin America.

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. 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 12 May 2024:

1. From Blueprint to Therapy: The Evolution and Challenges of Nucleic Acid Interventions – Biocompounding

Nucleic Acid Therapies (NATs) offer a targeted approach to rectify these underlying genetic issues. By employing strategies like antisense oligonucleotides, mRNA therapy, RNA interference, or CRISPR-based gene editing, NATs can directly modify or regulate the expression of genes responsible for the disease. These therapies can repair or silence defective genes, replace missing genes, or modulate gene expression, thereby addressing the root cause of the disease at the molecular level. This precision in targeting genetic defects makes NATs a promising and revolutionary approach in modern medicine, potentially offering cures or significant treatment improvements for numerous genetic and acquired diseases.

The modalities of NATs vary based on their mechanism of action, type of nucleic acid used, and therapeutic goals. Here’s an introduction to the different modalities of NATs:

1. Antisense Oligonucleotides (ASOs): These are short, synthetic strands of DNA or RNA that are designed to bind to specific RNA molecules within a cell. By binding to their target RNA, ASOs can interfere with the process of protein production. They can inhibit the expression of a gene, modify RNA splicing, or promote the degradation of the RNA molecule. ASOs are used in conditions like Duchenne Muscular Dystrophy and Spinal Muscular Atrophy. Example: Sarepta Therapeutics
2. RNA Interference (RNAi): This modality uses small interfering RNA (siRNA) or microRNA (miRNA) to silence specific genes. RNAi works by degrading the mRNA of a target gene, preventing it from being translated into a protein. This approach is particularly useful in diseases where inhibiting the expression of a certain gene can be therapeutic. RNAi has been explored for various applications including cancer therapy and viral infections. Currently FDA-approved siRNAs are used in conditions such as Hereditary transthyretin-mediated amyloidosis. Example: Alnylam Pharmaceuticals
3. AAV Gene Therapy: Adeno-associated virus (AAV) vectors are commonly used in gene therapy. AAVs are small viruses that can deliver genetic material into cells without causing disease. In AAV gene therapy, the therapeutic gene is packaged into an AAV vector, which then delivers the gene into patient’s cells. This modality is useful for treating genetic disorders, such as Hemophilia A, by providing a functional copy of a defective or missing gene. Example: Spark Therapeutics
4. mRNA Therapy: mRNA therapies involve the use of messenger RNA to produce therapeutic proteins inside the body. Unlike traditional gene therapy that alters the DNA within cells, mRNA therapy delivers mRNA that is translated into the desired protein, offering a temporary but effective treatment. This approach has gained significant attention, especially in the development of COVID-19 vaccines. Currently there are several attempts to develop cancer vaccines by the key players in this space. Example: Moderna, BioNTech, Pfizer
5. CRISPR/Cas9 and Genome Editing: This revolutionary technology enables precise editing of the genome. CRISPR/Cas9 can be used to add, delete, or alter specific DNA sequences in the genome, offering the potential to correct genetic defects at their source. While still in the experimental stages for many applications, it holds promise for treating a range of genetic disorders. In Dec 2023, the first ever FDA-approved CRISPR-based gene therapy was used to treat sickle cell disease. Example: CRISPR Therapeutics, Vertex

2. China Is Still Rising – Nicholas Lardy

Those who doubt that China’s rise will continue point to the country’s weak household spending, its declining private investment, and its entrenched deflation. Sooner than overtake the United States, they argue, China would likely enter a long recession, perhaps even a lost decade.

But this dismissive view of the country underestimates the resilience of its economy. Yes, China faces several well documented headwinds, including a housing market slump, restrictions imposed by the United States on access to some advanced technologies, and a shrinking working-age population. But China overcame even greater challenges when it started on the path of economic reform in the late 1970s. While its growth has slowed in recent years, China is likely to expand at twice the rate of the United States in the years ahead.

Several misconceptions undergird the pessimism about China’s economic potential…

…A second misconception is that household income, spending, and consumer confidence in China is weak. The data do not support this view. Last year, real per capita income rose by 6 percent, more than double the growth rate in 2022, when the country was in lockdown, and per capita consumption climbed by nine percent. If consumer confidence were weak, households would curtail consumption, building up their savings instead. But Chinese households did just the opposite last year: consumption grew more than income, which is possible only if households reduced the share of their income going to savings…

…Another misconception concerns the potential for a collapse in property investment. These fears are not entirely misplaced; they are supported by data on housing starts, the number of new buildings on which construction has begun, which in 2023 was half what it was in 2021. But one has to look at the context. In that same two-year period, real estate investment fell by only 20 percent, as developers allocated a greater share of such outlays to completing housing projects they had started in earlier years. Completions expanded to 7.8 billion square feet in 2023, eclipsing housing starts for the first time. It helped that government policy encouraged banks to lend specifically to housing projects that were almost finished; a general easing of such constraints on bank loans to property developers would have compounded the property glut…

…By 2014, private investment composed almost 60 percent of all investment—up from virtually zero percent in 1978. As private investment is generally more productive than that of state companies, its expanding share of total investment was critical to China’s rapid growth over this period. This trend went into reverse after 2014 when Xi Jinping, having just assumed the top leadership position, aggressively redirected resources to the state sector. The slowdown was modest at first, but by 2023, private investment accounted for only 50 percent of total investment…

…But here again, the pessimism is not supported by the data. First, almost all the decline in the private share of total investment after 2014 resulted from a correction in the property market, which is dominated by private companies. When real estate is excluded, private investment rose by almost ten percent in 2023. Although some prominent Chinese entrepreneurs have left the country, more than 30 million private companies remain and continue to invest.

3. The Cloud Under The Sea – Josh Dzieza

In the family tree of professions, submarine cable work occupies a lonely branch somewhere between heavy construction and neurosurgery. It’s precision engineering on a shifting sea using heavy metal hooks and high-tension lines that, if they snap, can cut a person in half. In Hirai’s three decades with Kokusai Cable Ship Company (KCS), he had learned that every step must be followed, no matter how chaotic the situation. Above all else, he often said, “you must always be cool.”…

…The world’s emails, TikToks, classified memos, bank transfers, satellite surveillance, and FaceTime calls travel on cables that are about as thin as a garden hose. There are about 800,000 miles of these skinny tubes crisscrossing the Earth’s oceans, representing nearly 600 different systems, according to the industry tracking organization TeleGeography. The cables are buried near shore, but for the vast majority of their length, they just sit amid the gray ooze and alien creatures of the ocean floor, the hair-thin strands of glass at their center glowing with lasers encoding the world’s data.

If, hypothetically, all these cables were to simultaneously break, modern civilization would cease to function. The financial system would immediately freeze. Currency trading would stop; stock exchanges would close. Banks and governments would be unable to move funds between countries because the Swift and US interbank systems both rely on submarine cables to settle over $10 trillion in transactions each day. In large swaths of the world, people would discover their credit cards no longer worked and ATMs would dispense no cash. As US Federal Reserve staff director Steve Malphrus said at a 2009 cable security conference, “When communications networks go down, the financial services sector does not grind to a halt. It snaps to a halt.”…

…Fortunately, there is enough redundancy in the world’s cables to make it nearly impossible for a well-connected country to be cut off, but cable breaks do happen. On average, they happen every other day, about 200 times a year. The reason websites continue to load, bank transfers go through, and civilization persists is because of the thousand or so people living aboard 20-some ships stationed around the world, who race to fix each cable as soon as it breaks.

The industry responsible for this crucial work traces its origins back far beyond the internet, past even the telephone, to the early days of telegraphy. It’s invisible, underappreciated, analog. Few people set out to join the profession, mostly because few people know it exists…

…Once people are in, they tend to stay. For some, it’s the adventure — repairing cables in the churning currents of the Congo Canyon, enduring hull-denting North Atlantic storms. Others find a sense of purpose in maintaining the infrastructure on which society depends, even if most people’s response when they hear about their job is, But isn’t the internet all satellites by now? The sheer scale of the work can be thrilling, too. People will sometimes note that these are the largest construction projects humanity has ever built or sum up a decades-long resume by saying they’ve laid enough cable to circle the planet six times…

…The world is in the midst of a cable boom, with multiple new transoceanic lines announced every year. But there is growing concern that the industry responsible for maintaining these cables is running perilously lean. There are 77 cable ships in the world, according to data supplied by SubTel Forum, but most are focused on the more profitable work of laying new systems. Only 22 are designated for repair, and it’s an aging and eclectic fleet. Often, maintenance is their second act. Some, like Alcatel’s Ile de Molene, are converted tugs. Others, like Global Marine’s Wave Sentinel, were once ferries. Global Marine recently told Data Centre Dynamics that it’s trying to extend the life of its ships to 40 years, citing a lack of money. One out of 4 repair ships have already passed that milestone. The design life for bulk carriers and oil tankers, by contrast, is 20 years.

“We’re all happy to spend billions to build new cables, but we’re not really thinking about how we’re going to look after them,” said Mike Constable, the former CEO of Huawei Marine Networks, who gave a presentation on the state of the maintenance fleet at an industry event in Singapore last year. “If you talk to the ship operators, they say it’s not sustainable anymore.”

He pointed to a case last year when four of Vietnam’s five subsea cables went down, slowing the internet to a crawl. The cables hadn’t fallen victim to some catastrophic event. It was just the usual entropy of fishing, shipping, and technical failure. But with nearby ships already busy on other repairs, the cables didn’t get fixed for six months. (One promptly broke again.)

But perhaps a greater threat to the industry’s long-term survival is that the people, like the ships, are getting old. In a profession learned almost entirely on the job, people take longer to train than ships to build.

“One of the biggest problems we have in this industry is attracting new people to it,” said Constable. He recalled another panel he was on in Singapore meant to introduce university students to the industry. “The audience was probably about 10 university kids and 60 old gray people from the industry just filling out their day,” he said. When he speaks with students looking to get into tech, he tries to convince them that subsea cables are also part — a foundational part — of the tech industry…

…To the extent he is remembered, Cyrus Field is known to history as the person responsible for running a telegraph cable across the Atlantic Ocean, but he also conducted what at the time was considered an equally great technical feat: the first deep-sea cable repair.

Field, a 35-year-old self-made paper tycoon, had no experience in telegraphy — which helps explain why, in 1854, he embarked on such a quixotic mission…

…“When it was first proposed to drag the bottom of the Atlantic for a cable lost in waters two and a half miles deep, the project was so daring that it seemed to be almost a war of the Titans upon the gods,” wrote Cyrus’ brother Henry. “Yet never was anything undertaken less in the spirit of reckless desperation. The cable was recovered as a city is taken by siege — by slow approaches, and the sure and inevitable result of mathematical calculation.”

Field’s crew caught the cable on the first try and nearly had it aboard when the rope snapped and slipped back into the sea. After 28 more failed attempts, they caught it again. When they brought it aboard and found it still worked, the crew fired rockets in celebration. Field withdrew to his cabin, locked the door, and wept.

Cable repair today works more or less the same as in Field’s day. There have been some refinements: ships now hold steady using automated dynamic positioning systems rather than churning paddle wheels in opposite directions, and Field’s pronged anchor has spawned a medieval-looking arsenal of grapnels — long chains called “rennies,” diamond-shaped “flat fish,” spring-loaded six-blade “son of sammys,” three-ton detrenchers with seven-foot blades for digging through marine muck — but at its core, cable repair is still a matter of a ship dragging a big hook along the ocean floor. Newfangled technologies like remotely operated submersibles can be useful in shallow water, but beyond 8,000 feet or so, conditions are so punishing that simple is best…

…Debates about the future of cable repair have become a staple of industry events. They typically begin with a few key facts: the ships are aging; the people are aging; and it’s unclear where the money will come from to turn things around.

For much of the 20th century, cable maintenance wasn’t a distinct business; it was just something giant, vertically integrated telecom monopolies had to do in order to function. As they started laying coaxial cables in the 1950s, they decided to pool resources. Rather than each company having its own repair vessel mostly sitting idle, they divided the oceans into zones, each with a few designated repair ships.

When the telcos were split up at the turn of the century, their marine divisions were sold off. Cable & Wireless Marine became Global Marine. AT&T’s division is now the New Jersey-based SubCom. (Both are now owned by private equity companies; KCS remains a subsidiary of KDDI.) The zone system continued, now governed by contracts between cable owners and ship operators. Cable owners can sign up with a nonprofit cooperative, like the Atlantic Cable Maintenance & Repair Agreement, and pay an annual fee plus a day rate for repairs. In exchange, the zone’s three ships — a Global Marine vessel in Portland, UK, another in Curaçao, and an Orange Marine vessel in Brest, France — will stand ready to sail out within 24 hours of being notified of a fault.

This system has been able to cope with the day-to-day cadence of cable breaks, but margins are thin and contracts are short-term, making it difficult to convince investors to spend $100 million on a new vessel.

“The main issue for me in the industry has to do with hyperscalers coming in and saying we need to reduce costs every year,” said Wilkie, the chair of the ACMA, using the industry term for tech giants like Google and Meta. “We’d all like to have maintenance cheaper, but the cost of running a ship doesn’t actually change much from year to year. It goes up, actually. So there has been a severe lack of investment in new ships.”

At the same time, there are more cables to repair than ever, also partly a result of the tech giants entering the industry. Starting around 2016, tech companies that previously purchased bandwidth from telcos began pouring billions of dollars into cable systems of their own, seeking to ensure their cloud services were always available and content libraries synced. The result has been not just a boom in new cables but a change in the topology of the internet. “In the old days we connected population centers,” said Constable, the former Huawei Marine executive. “Now we connect data centers. Eighty percent of traffic crossing the Atlantic is probably machines talking to machines.”…

…In 2022, the industry organization SubOptic gathered six cable employees in their 20s and 30s for a panel on the future of the industry. Most of them had stumbled into their jobs inadvertently after college, and the consensus was that the industry needed to be much better about raising public awareness, especially among the young.

“I don’t know if anyone saw, but during the pandemic, submarine cables actually went viral on TikTok,” said one panelist, a young cable engineer from Vodafone. “People didn’t know they existed, and then suddenly, out of nowhere, they were viral. I think it’s engaging with youth and children through their own avenues — yes, you can have science museums and things like that, but they are online, they are on their iPads, they’re on their phones.”

“We’ve got some pretty senior decision-makers and influencers in the subsea cable industry here,” said one audience member. “Did any of us know that we went viral on TikTok?” he asked, to laughter.

“As this panel rightfully said upfront, it’s not that we have a brand problem,” said another audience member, “we just don’t have a brand at all.”

4. Looking for AI use-cases – Benedict Evans

I’ve been thinking about this problem a lot in the last 18 months, as I’ve experimented with ChatGPT, Gemini, Claude and all the other chatbots that have sprouted up: ‘this is amazing, but I don’t have that use-case’.

The one really big use-case that took off in 2023 was writing code, but I don’t write code. People use it for brainstorming, and making lists and sorting ideas, but again, I don’t do that. I don’t have homework anymore. I see people using it to get a generic first draft, and designers making concept roughs with MidJourney, but, again, these are not my use-cases. I have not, yet, found anything that matches with a use-case that I have. I don’t think I’m the only one, either, as is suggested by some of the survey data – a lot of people have tried this, especially since you don’t need to spend $12,000 on a new Apple II, and it’s very cool, but how much do we use it, and what for?…

…Suppose you want to analyse this month’s customer cancellations, or dispute a parking ticket, or file your taxes – you can ask an LLM, and it will work out what data you need, find the right websites, ask you the right questions, parse a photo of your mortgage statement, fill in the forms and give you the answers. We could move orders of magnitude more manual tasks into software, because you don’t need to write software to do each of those tasks one at a time. This, I think, is why Bill Gates said that this is the biggest thing since the GUI. That’s a lot more than a writing assistant.

It seems to me, though, that there are two kinds of problem with this thesis.

The narrow problem, and perhaps the ‘weak’ problem, is that these models aren’t quite good enough, yet. They will get stuck, quite a lot, in the scenarios I suggested above. Meanwhile, these are probabilistic rather than deterministic systems, so they’re much better for some kinds of task than others. They’re now very good at making things that look right, and for some use-cases this is what you want, but for others, ‘looks right’ is different to ‘right’…

…The deeper problem, I think, is that no matter how good the tech is, you have to think of the use-case. You have to see it. You have to notice something you spend a lot of time doing and realise that it could be automated with a tool like this…

…The cognitive dissonance of generative AI is that OpenAI or Anthropic say that we are very close to general-purpose autonomous agents that could handle many different complex multi-stage tasks, while at the same time there’s a ‘Cambrian Explosion’ of startups using OpenAI or Anthropic APIs to build single-purpose dedicated apps that aim at one problem and wrap it in hand-built UI, tooling and enterprise sales, much as a previous generation did with SQL. Back in 1982, my father had one (1) electric drill, but since then tool companies have turned that into a whole constellation of battery-powered electric hole-makers. One upon a time every startup had SQL inside, but that wasn’t the product, and now every startup will have LLMs inside.

I often compared the last wave of machine learning to automated interns. You want to listen to every call coming into the call centre and recognise which customers sound angry or suspicious: doing that didn’t need an expert, just a human (or indeed maybe even a dog), and now you could automate that entire class of problem. Spotting those problems and building that software takes time: machine learning’s breakthrough was over a decade ago now, and yet we are still inventing new use-cases for it – people are still creating companies based on realising that X or Y is a problem, realising that it can be turned into pattern recognition, and then going out and selling that problem.

You could propose the current wave of generative AI as giving us another set of interns, that can make things as well as recognise them, and, again, we need to work out what. Meanwhile, the AGI argument comes down to whether this could be far, far more than interns, and if we had that, then it wouldn’t be a tool anymore.

5. TIP622: Finding Certainty In An Uncertain World w/ Joseph Shaposhnik – Clay Finck and Joseph Shaposhnik

[00:29:29] Joseph Shaposhnik: I think of the credit bureaus and I think of a partner of theirs, which we’ll spend a minute talking about in a second, but. As you may know, the credit bureaus, there’s three of them in the United States. And they run an incredible oligopoly. If you want to secure a mortgage, get a car loan, rent a home, they’re involved in all of those decision making situations by the owners of those assets.

[00:29:55] Joseph Shaposhnik: As an example, if you go for a mortgage, all three credit bureaus will be pinned to get a score on you. All of them will be paid a couple of dollars for that score and all of that information that they’re pulling is contributory data. So there’s a relatively insignificant amount of incremental cost to generate that score and deliver it to the customer.

[00:30:21] Joseph Shaposhnik: You know, it’s a 95% incremental margin business. I mean, this is an incredible business. It’s basically an override on all economic activity in the United States and outside the United States where they play. And they’re just incredible businesses. But surprisingly not incredible stocks. You know, how could that be?

[00:30:40] Joseph Shaposhnik: It’s shocking to give you a sense organic growth if you look back, the last 5 years for the businesses have been approximately 7% a year. So, 3 or 4 times. Global GDP or U.S. GDP. They’ve outgrown the S&P the average S&P business over that period of time. They started with 30% EBITDA margins at the beginning of the 5 year period, so very profitable businesses.

[00:31:09] Joseph Shaposhnik: Yet over the last five years, two out of the three credit bureaus have underperformed the S&P, and over a 10 year period, they’ve been just in line performers with the S&P and so, I mean, they run an oligopoly. How could that possibly be? I used to be the credit bureau analyst at TCW, so I’m very familiar with these businesses, and they’re just incredible companies.

[00:31:33] Joseph Shaposhnik: And what happened is all three of these businesses spent more money on M&A than they generated in free cash flow over that five year period of time. They spent more money on M&A than all of the free cash flow they generated over the last five years. And they generate a lot of free cash flow. And let me say, let me just tell you, this is not on synergistic M&A.

[00:31:57] Joseph Shaposhnik: This was, I mean, they would call it synergistic, but it’s very difficult to synergize a near utility that they operate. And instead of just sticking to their knitting, they decided to acquire a lot of different data assets. that were incredibly expensive, generally from private equity, which doesn’t give assets away.

[00:32:18] Joseph Shaposhnik: And those returns are always, the returns on those businesses are always going to be lower than the returns on this incredible oligopoly that they run. And so, as interestingly as that, so of course, margins have been under pressure, returns have gone way down for these businesses because of all the acquisitions, these poor acquisitions at high multiples.

[00:32:42] Joseph Shaposhnik: And one of the most surprising things is we looked at the data on this, two out of the three businesses engaged in near zero share of purchases over that five year period of time. So you have this incredible business, you know, these three businesses that run an oligopoly, basically just an override on all economic activity.

[00:33:03] Joseph Shaposhnik: And they find all of these other businesses more attractive to allocate capital to than their own business, which is a 95% incremental margin business. Incredible. No wonder the stocks have not performed well, even though those businesses and those stocks should be like shooting fish in a barrel.

[00:33:20] Joseph Shaposhnik: So it’s incredible they bought back no, no stock, two out of the three businesses bought back no meaningful amount of stock. And not surprisingly, those businesses underperformed. In contrast to that, they have a partner, which is Fair Isaacs. And so Fair Isaacs, which is, the ticker is FICO provides the formula to the credit bureaus, which generates the score.

[00:33:44] Joseph Shaposhnik: The credit bureaus contribute the data. and the data with the formula creates a score that they can then sell to their end customers. So the bureaus pay FICO a fee for the formula, and they take the formula, and they generate a score, and they sell it to their customer. So you would think that FICO is basically in this ecosystem, has similar growth dynamics, has similar returns going into that 5 year period of time, similar EBITDA margins, tied to the same end markets, relatively similar company.

[00:34:17] Joseph Shaposhnik: Yet, over that five year period of time, FICO took all of its free cash flow, all of it, and used it to repurchase its shares. And so over the last five years, FICO has reduced share count by 20%, has engaged in no meaningful acquisitions to dilute its incredible franchise, and has generated a five bagger over the last five years.

[00:34:44] Joseph Shaposhnik: compared to the bureaus that have generated 15 to 100% return, total return over that 5 year period of time. So, a 5 bagger, which has outperformed the market by a ton, compared to an underperforming or an inline performance for the bureaus, I think just tells the tale of how important great capital allocation decision making is, how important it is to be aligned with a management team that understands how to generate value for shareholders.

[00:35:13] Joseph Shaposhnik: And I think for us and for everybody, serves as a warning when we think about investing with teams that are acquiring businesses in general and certainly acquiring businesses That are not as attractive as the core business. So capital allocation makes or breaks stories all the time, and incentives generally drive these decisions, but often times it just takes and an investor oriented CEO to see the big opportunity, which is usually in its core and not far afield.

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 Meta Platforms. 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 May 2024:

1. Karen Karniol-Tambour on investment anomalies at Sohn 2024 (transcript here) – Karen Karniol-Tambour and Jawad Mian

Jawad Mian (00:30): So 6 months ago equities were rallying in anticipation of lower interest rates but now we’ve seen year-to-date equities are rallying despite higher bond yields. So with a strong economy and inflation less of an issue, are you reverting to the typical inverse relationship between equities and bonds.

Karen Karniol-Tambour (00:49): The relationship between equities and bonds – it’s not an immutable fact of life. It’s not just a thing that occurs. It’s a function of the fundamental building blocks in stocks and bonds. When you look at stocks and bonds, they have a lot of things in common. They’re all future cash flows you’re discounting to today. So if you raise that, it’s bad for both and they both don’t do great when inflation is strong. The real inverse comes from their reaction to growth for the reason you’re saying. If growth is strong, then you can get equities rising and at the same time you can actually get the central bank tightening in response to that growth, which is bad for the bonds. And actually, the anomaly has been the years leading up to 2022, where inflation was just just a non-factor and the only dominant macro issue was growth. And so we’ve gotten really used to the idea that stocks and bonds have this inverse relationship. But that’s actually the anomaly. It’s not that normal to have a world where inflation just doesn’t matter. And finally, we live through this period where it’s like, “Wait a minute, inflation, its gravitational pull was at such a low level it was irrelevant – it’s becoming relevant again.” And we got this positive correlation where they both did badly, because you need to tighten in response to that inflation rearing its head.

Today – knock on wood – we look like we’re back to a world where inflation is not a non-issue, but it’s not a dominant issue, where we can have the kind of market action we’ve enjoyed so far in 2024, where we find out growth’s pretty damn resilient, growth’s doing great, companies can do well, earnings can do well, and at the same time the FED can ease less-than-expected or tighten relative to expectations at the same time. If they were tightening to stop very bad inflation, that would be a very different outcome. So the fundamental question as an investor is sort of where is the gravitational pull of inflation going to be? Is this going to be a major topic that then lead stocks and bonds sometimes to act the same way? Or is it going to go back to being kind of a non-issue?…

…Mian (02:53): A second anomaly. For the last 50 years, we’ve seen the US budget deficit average around 3% and it’s projected to be 6% over the next decade. So far we have seen markets being willing to finance these record deficits, in contrast to the UK for example. How come?

Karniol-Tambour (03:11): I think the best answer to this starts with actually the current account deficits, because obviously that’s part of who’s buying all the bonds we’re issuing. And it is a really weird anomaly because the United States is buying way more foreign goods than they’re buying ours. And typically if countries do that, their currency is weak because they have to convince someone to hold all the currency on the other side of that, so they have to attract all this financing. That the United States is running a massive current account deficit and yet the dollar is strong because what’s happening on the other end is people are just so enthusiastic about buying dollar financial assets. It’s so extreme that I think the United States has kind of a version of a Dutch disease.

So the classic Dutch disease is, you’re Saudi Arabia, you have oil. No one’s buying oil because you’re Saudi Arabia. No one’s thinking, “I really want Saudi oil.” They just need to fill up their car. So whatever the gas, is the gas is. But as Saudi Arabia, you get uncompetitive outside of it because money’s flooding in just for your oil, for nothing else. The United States has kind of become that on financial assets, which is people aren’t really thinking “I just want US financial assets.” It’s just that United States financial assets have done so well, they’re the dominant part of the index in stocks and in bonds. So anyone that needs to save any money around the world just ends up in US assets. As long as you care at all about market cap – which anyone reasonable would – and you’re going to the big market around the world, if you’re saving, you’re giving the United States money. And so we’re ending up with this flood of money that is a huge anomaly where we actually have a rising currency making everything else kind of uncompetitive, because people just want to buy stocks and bonds and no one else enjoys that. So we can run these huge deficits and sort of not worry about it.

2. Remembering Daniel Kahneman: A Mosaic of Memories and Lessons – Evan Nesterak and many others

To be continued …

By Richard Thaler, Professor of Behavioral Science and Economics, University of Chicago

My fondest memories of working with Danny come from 1984 to ’85 when I spent a year visiting him in Vancouver at The University of British Columbia. Danny had just begun a new project with Jack Knetsch on what people think is fair in market transactions and they invited me to join them. We had the then-rare ability to ask survey questions to a few hundred randomly selected Canadians each week. We would draft three versions of five questions, fax them to Ottawa Monday morning, get the results faxed back to us Thursday afternoon. Who needs Mturk! We then spent the weekend digesting the results and writing new questions.

We learned that raising the price of snow shovels the morning after a blizzard might make sense to an economist, but would make customers angry. Danny displayed two of his most prominent traits. He was always a skeptic, even (especially?) about his own ideas, so we stress-tested everything. And he was infinitely patient in that pursuit. Was our finding just true for snow shovels? What about water after a hurricane? Flu medicine? How about late-season discounts (which of course are fine). It was total immersion; meeting in person several times a week and talking constantly. We were in the zone.

Although we spent another year together in New York seven years later, we were unable to recreate that intensity. We had too many other balls in the air. But we continued our conversations and friendship until the end. Every conversation ended the same way: “To be continued.”…

...I’m more like a spiral than a circle

By Dan Lovallo, Professor of Strategy, Innovation and Decision Sciences, University of Sydney

Many people have heard that Danny changes his mind—a lot. This is certainly true. I have never written even a 5,000-word essay with him that didn’t take a year. Let me add another dimension to the discussion. During our last working dinner at a bistro in New York, and possibly out of mild frustration, I said, “Danny, you know you change your mind a lot.” It wasn’t a question. He continued chewing. I continued my line of non-question questioning: “And often you change it back to what it was at the beginning.”

Danny, having finished his bite and without missing a beat, looked up and in his characteristic lilt said, “Dan, that’s when I learn the most.” Then using his finger he drew a circle in space. “I don’t go around and around a problem. It might seem like it, but I am getting deeper and deeper.” The circle morphed into a three-dimensional spiral. “So, you’re missing all the learning,” he explained, as he displayed the invisible sculpture. “I’m more like a spiral than a circle.” Happy with this new idea, Danny grinned as only Danny could…

…A case in character

By Angela Duckworth, Professor of Psychology, University of Pennsylvania

One evening, more than twenty years ago, I was the last one in the lab when the phone rang. “Hello?” I said, I hope not brusquely. I was a Ph.D. student at the time and eager to get back to my work. “Hello?” came the reply of an uncommonly polite older gentleman, whose accent I couldn’t quite place. “I’m so sorry to trouble you,” he continued. “I believe I’ve just now left my suitcase there.” Ah, this made sense. We’d hosted an academic conference that day. “It’s a terrible inconvenience, I know, but might you keep it somewhere until I can return to pick it up?” “Sure,” I said, cradling the receiver and grabbing a notepad. “How do you spell your name?” “Thank you so very much. It’s K-A-H-N-E-M-A-N.” I just about fainted. “Yes, Dr. Kahneman,” I said, coming to my senses, likely more deferentially than when I’d first picked up.

When I hung up, I thought to myself, Oh, it’s possible to be a world-famous genius—the most recently anointed Nobel laureate in economics, among other honors—and interact with anybody and everybody with utmost respect and dignity, no matter who they are. In the years that followed, I got to know Danny Kahneman much better, and when I did, that view was only confirmed. Confirmation bias? Halo effect? No and no. What then? Character. The world is mourning the loss of Danny Kahneman the genius, as we should, but I am missing Danny Kahneman the person…

…Anxious and unsure

By Eric Johnson, Professor of Business, Columbia University

A few months before the publication of Thinking, Fast and Slow in 2011, the Center for Decision Sciences had scheduled Danny to present in our seminar series. We were excited because he had decided to present his first “book talk” with us. Expecting a healthy crowd, we scheduled the talk in Uris 301, the biggest classroom in Columbia Business School.

I arrived in the room a half hour early to find Danny, sitting alone in the large room, obsessing over his laptop. He confided that he had just changed two-thirds of the slides for the talk and was quite anxious and unsure about how to present the material. Of course, after the introduction, Danny presented in his usual charming, erudite style, communicating the distinction between System 1 and System 2 with clarity to an engaged audience. Afterwards, I asked him how he thought it went, and he said, “It was awful, but at least now I know how to make it better.” Needless to say, the book went on to become an international bestseller.

This was not false modesty. Having studied overconfidence throughout his career, Danny seemed immune to its effects. While surely maddening to some coauthors, this resulted in work that was more insightful and, most importantly to Danny and to us, correct. He was not always right, but always responsive to evidence, supportive or contradictory. For example, when some of the evidence cited in the book was questioned as a result of the replication crisis in psychology, Danny revised his opinion, writing in the comments of a critical blog: “I placed too much faith in underpowered studies.”

The best tribute to Danny, I believe, is adopting this idea, that science and particularly the social sciences, is not about seeming right, but instead, being truthful…

…Practical problem solving

By Todd Rogers, Professor of Public Policy, Harvard University

I was part of a group helping some political candidates think about how to respond to untrue attacks by their political rivals. We focused on what cognitive and social psychology said about persuasive messaging. Danny suggested a different emphasis I hadn’t considered.

He directed us to a literature in cognitive psychology on cognitive associations. Once established, associations cannot simply be severed; attempting to directly refute them often reinforces them, and logical arguments alone can’t undo them. But these associations can be weakened when other competing associations are created.

For instance, if falsely accused of enjoying watching baseball, I’d be better off highlighting genuine interests—like my enjoyment of watching American football or reality TV—to dilute the false association with baseball. This anecdote is one small example of the many ways Danny’s profound intellect has influenced practical problem-solving. He’ll be missed and remembered.

Premortems

By Michael Mauboussin, Head of Consilient Research, Morgan Stanley

The opportunity to spend time with Danny and the chance to interview him were professional delights. One of my favorite lessons was about premortems, a technique developed by Gary Klein that Danny called one of his favorite debiasing techniques. In a premortem, a group assumes that they have made a decision (which they have yet to do), places themselves in the future (generally a year from now), and pretends that it worked out poorly. Each member independently writes down the reasons for the failure.

Klein suggested that one of the keys to premortems was the idea of prospective hindsight, that putting yourself into the future and thinking about the present opens up the mind to unconsidered yet relevant potential outcomes. I then learned that the findings of the research on prospective hindsight had failed to replicate—which made me question the value of the technique.

Danny explained that my concern was misplaced and that prospective hindsight was not central to the premortem. Rather, it was that the technique legitimizes dissent and allows organizations the opportunities to consider and close potential loopholes in their plans. That I had missed the real power of the premortem was a revelation and a relief, providing me with a cherished lesson…

…Eradicating unhappiness

By George Loewenstein, Professor of Economics and Psychology, Carnegie Mellon University

For Danny, research was intensely personal. He got into intellectual disputes with a wide range of people, and these would hurt him viscerally, in part because it pained him that people he respected could come to different conclusions from those he held so strongly. He came up with, or at least embraced, the concept of “adversarial collaboration” in which researchers who disagreed on key issues would, however, agree upon a definitive test to determine where reality lay. A few of these were successful, but others (I would say most) ended with both parties unmoved, perhaps reflecting Robert Abelson’s insight that “beliefs are like possessions,” and, hence subject to the endowment effect.

I was spending time with Danny when he first got interested in hedonics—happiness—and that was a personal matter as well. His mother was declining mentally in France, and he agonized about whether to visit her; the issue was that she had anterograde amnesia, so he knew that she would forget his visit as soon as it ended. The criterion for quality of life, he had decided, should be the integral of happiness over time; so that—although she would miss out on the pleasure of remembering it—his visit would have value if she enjoyed it while it was happening.

Showing the flexibility of his thinking, and his all-too-rare willingness to learn from the data, his perspective changed as he studied happiness. He became more concerned about the story a life tells, including, notably, its peak and end; he concluded that eradicating unhappiness was a more important goal than fostering happiness, and began to draw a sharp distinction between happiness and life satisfaction, perhaps drawing, again, on his own experience. He always seemed to me to be extremely high in life satisfaction, but considerably less so in happiness.

3. Paradox of China’s stock market and economic growth – Glenn Luk

Joe Weisenthal of Bloomberg and the Odd Lots posed this question on Twitter/X:

“Given that the stock market hasn’t been especially rewarding to the volume-over-profits strategy undertaken by big Chinese manufacturers, what policy levers does Beijing have to sustain and encourage the existing approach?”

Many people may have noticed that despite the impressive growth of Chinese manufacturers in sectors like electric vehicles, the market capitalizations of these companies are dwarfed by Tesla. This seeming paradox lies at the heart of the the question posed by Joe.

In 2020, I shared an observation that China cares a lot more about GDP than market capitalization. I was making this observation in the context of Alibaba1 but would soon broaden the observation to encapsulate many more situations. In sharp contrast to Americans, Beijing just does not seem to care that much about equity market valuations but do seem to very much care about domestic growth and economic development…

…With respect to private sector market forces, Chinese policymakers tend to see its role as coordinators of an elaborate “game” that is meant to create an industry dynamic that drives desired market behaviors. The metaphor I sometimes use is as the Dungeon Master role in Dungeons & Dragons.

These “desired market behaviors” tend to overwhelmingly revolve around this multi-decade effort to maximize economic development and growth. Beijing has been very consistent about the goal to become “fully developed” by the middle of the 21st century.

To date, I would say that Chinese policymakers have been relatively successful using the approaches and principles described above to drive economic growth:

• Priority on labor over capital / wage growth over capital income growth. Prioritizing labor is a key pillar of China’s demand-side support strategy. Growth in household income drives growth in domestic demand (whether in the form of household gross capital formation or expenditures).
• Setting up rules to foster the create competitive industry dynamics and motivate economic actors to reinvest earnings back into growth.
• Periodic crackdowns to disrupt what is perceived to be rent-seeking behavior, particularly from private sector players that have accumulated large amounts of equity capital (vs. small family businesses):
  • Anti-competitive behavior (e.g. Alibaba e-commerce dominance in the late 2010s)
  • Regulatory arbitrage (moral hazards inherent in Ant Financial’s risk-sharing arrangement with SOE banks)
• Societal effects (for-profit education driving “standing on tiptoes” approach to childhood education)
• Supply-side support to encourage dynamic, entrepreneurial participation from private sector players like in the clean energy transition to drive rapid industry through scale and scale-related production efficiencies. China has relied on supply-side strategies to support economic for decades despite repeated exhortations by outsiders to implement OECD-style income transfers.
• Encouraging industry consolidation (vs. long drawn-out bankruptcies) once sectors have reached maturity although there are often conflicting motivations between Beijing and local governments.

A consistent theme is Beijing’s paranoia to rent-seeking behavior by capitalists (especially those who have accumulated large amounts of capital). It is sensitive to the potential stakeholder misalignment when capitalists — who are primarily aligned with one stakeholder class (fiduciary duty to equity owners).

It would prefer that rent-seeking behavior be handled by the party instead, whose objective (at least in theory) is to distribute these rents back to “The People” — although naturally in practice it never turns out this way; Yuen Yuen Ang has written multiple volumes about the prevalence of Chinese-style corruption and its corrosive economic effects.

So to bring it back to Joe’s question, the answer on whether Chinese policymakers can continue these policies going forward very much revolves around this question of rent-seeking: is it better to be done by the government or by private sector capitalists? What should be abundantly clear is that Beijing is definitive on this question: the party will maintain a monopoly on rent-seeking.

4. What Surging AI Demand Means for Electricity Markets – Tracy Alloway, Joe Weisenthal, and Brian Janous

Brian (09:58):

Yeah, and you’re right, I mean it’s not like we didn’t know that Microsoft had a partnership with OpenAI and that AI was going to consume energy. I think everyone though was a bit surprised at just how quickly what ChatGPT could do just captured the collective consciousness.

You probably remember when that was released. I mean it really sort surprised everyone and it became this thing where suddenly, even though we sort of knew what we were working on, it wasn’t until you put it out into the world that you realize maybe what you’ve created. That’s where we realized we are running up this curve of capability a lot faster than we thought. A number of applications that are getting built on this and the number of different ways that it’s being used and how it’s just become sort of common parlance. I mean, everyone knows what Chat GPT-3 is, and no one knew what it was the month before that.

So there was a bit, I think of a surprise in terms of just how quickly it was going to capture the collective consciousness and then obviously lead to everything that’s being created as a result. And so we just moved up that curve so quickly and I think that’s where the industry maybe got, certainly the utilities were behind because as you may have seen there, a lot of them are starting to restate their low-growth expectations.

And that was something that was not happening right before that. And so we’ve had massive changes just in the last two years of how utilities are starting to forecast what forecast. So if you take a look at a utility like Dominion in Virginia, so that’s the largest concentration of data centers in the United States. So they’re pretty good representative of what’s happening. If you go back to 2021, they were forecasting load growth over a period of 15 years of just a few percent.

I mean it was single-digit growth over that entire period. So not yearly growth, but over 15 years, single-digit growth. By 2023, they were forecasting to grow 2X over 15 years. Now keep in mind this is an electric utility. They do 10-year planning cycles. So because they have very long lead times for equipment for getting rights of away for transmission lines, they aren’t companies that easily respond to a 2X order of magnitude growth changed over a period of 15 years.

I mean, that is a massive change for electric utility, particularly given the fact that the growth rate over the last 15 to 20 years has been close to zero. So there’s been relatively no load growth in 15 to 20 years. Now suddenly you have utilities having to pivot to doubling the size of their system in that same horizon.

Tracy (13:10):

I want to ask a very basic question, but I think it will probably inform the rest of this conversation, but when we say that AI consumes a lot of energy, where is that consumption actually coming from? And Joe touched on this in the intro, but is it the sheer scale of users on these platforms? Is it, I imagine the training that you need in order to develop these models. and then does that energy usage differ in any way from more traditional technologies?

Brian (13:43):

Yeah, so whenever I think about the consumption of electricity for AI or really any other application, I think you have to start at sort of the core of what we’re talking about, which is really the human capacity for data, like whether it’s AI or cloud, humans have a massive capacity to consume data.

And if you think about where we are in this curve, I mean we’re on some form of S-curve of human data consumption, which then directly ties to data centers, devices, energy consumption ultimately, because what we’re doing is we’re turning energy into data. We take electrons, we convert them to light, we move them around to your TV screens and your phones and your laptops, etc. So that’s the uber trend that we’re riding up right now. And so we’re climbing this S-curve. I don’t know that anyone has a good sense of how steep or how long this curve will go.

If you go back to look at something like electricity, it was roughly about a hundred year. S-curve started in the beginning of last century. And it really started to flat line, as I mentioned before, towards the beginning of this century. Now we have this new trajectory that we’re entering, this new S-curve that we’re entering that’s going to change that narrative. But that S-curve for electricity took about a hundred years.

No one knows where we are on that data curve today. So when you inject something like AI, you create a whole new opportunity for humans to consume data, to do new things with data that we couldn’t do before. And so you accelerate us up this curve. So we were sitting somewhere along this curve, AI comes along and now we’re just moving up even further. And of course that means more energy consumption because the energy intensity of running an AI query versus a traditional search is much higher.

Now, what you can do with AI obviously is also much greater than what you can do with a traditional search. So there is a positive return on that invested energy. Oftentimes when this conversation comes up, there’s a lot of consternation and panic over ‘Well, what are we going to do? We’re going to run out of energy.’

The nice thing about electricity is we can always make more. We’re never going to run out of electricity. Not to say that there’s not times where the grid is under constraint and you have risks of brownouts and blackouts. That’s the reality. But we can invest more in transmission lines, we can invest more in power plants and we can create enough electricity to match that demand.

Joe (16:26):

Just to sort of clarify a point and adding on to Tracy’s question, you mentioned that doing an AI query is more energy intensive than, say, if I had just done a Google search or if I had done a Bing search or something like that. What is it about the process of delivering these capabilities that makes it more computationally intensive or energy intensive than the previous generation of data usage or data querying online?

Brian (16:57):

There’s two aspects to it, and I think we sort of alluded to it earlier, but the first is the training. So the first is the building of the large language model. That itself is very energy intensive. These are extraordinarily large machines, collections of machines that use very dense chips to create these language models that ultimately then get queried when you do an inference.

So then you go to ChatGPT and you ask it to give you a menu for a dinner party you want to have this weekend, it’s then referencing that large language model and creating this response. And of course that process is more computationally intensive because it’s doing a lot more things than a traditional search does. A traditional search just matched the words you put into a database of knowledge that it had put together, but these large language models are much more complex and then therefore the things you’re asking it to do is more complex.

So it will almost by definition be a more energy intensive process. Now, that’s not to say that it can’t get more efficient and it will, and Nvidia just last week was releasing some data on some of its next generation chips that are going to be significantly more efficient than the prior generation.

But one of the things that we need to be careful of is to think that because something becomes more efficient, then therefore we’re going to use less of the input resource. In this case, electricity. That’s not how it works, because going back to the concept of human capacity for consuming data, all we do is we find more things to compute. And this is, you’ve probably heard of Jevons paradox, and this is the idea that, well, if we make more efficient steam engines, he was an economist in the 1800s and he said ‘Well, if make more efficient steam engines, then we’ll use less coal.’

And he is like ‘No, that’s not what’s going to happen. We’re going to use more coal because we’re going to mechanize more things.’ And that’s exactly what we do with data just because we’ve had Moore’s Law for years, and so chips has become incredibly more efficient than they were decades ago, but we didn’t use less energy. We used much more energy because we could put chips in everything.

So that’s the trend line that we’re on. It’s still climbing that curve of consumption. And so no amount of efficiency is going to take us at this point, at least because I don’t believe we’re anywhere close to the bend in that S-curve. No amount of efficiency is going to take us off of continuing to consume more electricity, at least in the near term…

…Brian (22:35):

Well, this is where it gets a little concerning is that you have these tech companies that have these really ambitious commitments to being carbon neutral, carbon negative, having a hundred percent zero carbon energy a hundred percent of the time, and you have to give them credit for the work they’ve done.

I mean, that industry has done amazing work over the last decade to build absolutely just gigawatts upon gigawatts of new renewable energy projects in the United States all over the world. They’ve been some of the biggest drivers in the corporate focus on decarbonization. And so you really have to give that industry credit for all it’s done and all the big tech companies have done some amazing work there.

The challenge though that we have is the environment that they did that in was that no growth environment we were talking about. They were all growing, but they were starting from a relatively small denominator 10 or 15 years ago. And so there was a lot of overhang in the utility system at that time because the utilities had overbuilt ahead of that sort of flatlining. So there was excess capacity on the system.

They were growing inside of a system that wasn’t itself growing on a net basis. So everything they did, every new wind project you brought on, every new solar project you bought on, those were all incrementally reducing the amount of carbon in the system. It was all net positive.

Now we get into this new world where their growth rates are exceeding what the utilities had ever imagined in terms of the absolute impact on the system. The utilities’ response is ‘The only thing we can do in the time horizon that we have is basically build more gas plants or keep online gas plants or coal plants that we were planning on shuttering.’

And so now that the commitments that they have to zero carbon energy to be carbon negative, etc., are coming into contrast with the response that the utilities are laying out in their what’s called integrated resource plans or IRPs.

And we’ve seen this recently just last week in Georgia. We’ve seen it in Duke and North Carolina, Dominion and Virginia. Every single one of those utilities is saying ‘With all the demand that we’re seeing coming into our system, we have to put more fossil fuel resources on the grid. It’s the only way that we can manage it in a time horizon we have.’ Now, there’s a lot of debate about whether that is true, but it is what’s happening…

…Brian (30:29):

That’s right. And that’s the big challenge that good planners have today is what loads do you say yes to and what are the long-term implications of that? And we’ve seen this play out over the rest of the globe where you’ve had these concentrations of data centers. This is a story that we saw in Dublin, we’ve seen it in Singapore, we’ve seen it in Amsterdam.

And these governments start to get really worried of ‘Wait a minute, we have too many data centers as a percentage of overall energy consumption.’ And what inevitably happens is a move towards putting either moratoriums on data center build out or putting very tight restrictions on what they can do and the scale at which they can do it. And so we haven’t yet seen that to any material degree in the United States, but I do think that’s a real risk and it’s a risk that the data center industry faces.

I think somewhat uniquely in that if you’re the governor of a state and you have a choice to give power to a say new EV car factory that’s going to produce 1,500, 2,000 jobs versus a data center that’s going to produce significantly less than that, you’re going to give it to the factory. The data centers are actually the ones that are going to face likely the most constraints as governments, utilities, regulators start wrestling with this trade-off of ‘Ooh, we’re going to have to say no to somebody.’…

…Tracy (36:36):

What are the levers specifically on the tech company or the data center side? Because again, so much of the focus of this conversation is on what can the utilities do, what can we do in terms of enhancing the grid managing supply more efficiently? But are there novel or interesting things that the data centers themselves can do here in terms of managing their own energy usage?

Brian (37:02):

Yes. There’s a few things. I mean, one is data centers have substantial ability to be more flexible in terms of the power that they’re taking from the grid at any given time. As I mentioned before, every data center or nearly every data center has some form of backup generation. They have some form of energy storage built into this.

So the way a data center is designed, it’s designed like a power plant with an energy storage plant that just happens to be sitting next to a room full of servers. And so when you break it down into those components, you say, okay, well how can we better optimize this power plant to be more of a grid resource? How can we optimize the storage plant to be more of a grid resource? And then in terms of even the servers themselves, how can we optimize the way the software actually operates and is architected to be more of a grid resource?

And that sort of thinking is what is being forced on the industry. Frankly, we’ve always had this capability. I mean, we were doing, I mean we did a project like 2016 with a utility where we put in flexible gas generators behind our meter because the utility was going to have to build a new power plant if we didn’t have a way to be more flexible.

So we’ve always known that we can do this, but the industry has never been pressurized to really think innovatively about how can we utilize all these assets that we have inside of the data center plant itself to be more part of the grid. So I think the most important thing is really thinking about how data centers become more flexible. There’s a whole ‘nother line of thinking, which is this idea of, well, utilities aren’t going to move fast enough, so data centers just need to build all their own power plants.

And this is where you start hearing about nuclear and SMRs and infusion, which is interesting, except it doesn’t solve the problem this decade. It doesn’t solve the problem that we’re facing right now because none of that stuff is actually ready for prime time. We don’t have an SMR that we can build today predictably on time, on budget.

So we are dependent on the tools that we have today, which are things like batteries, grid enhancing technologies, flexible load, reconductoring transmission lines to get more power over existing rights of ways. So there’s a number of things we can do with technologies we have today that are going to be very meaningful this decade and we should keep investing in things that are going to be really meaningful next decade. I’m very bullish on what we can do with new forms of nuclear technology. They’re just not relevant in the time horizon. The problem we’re talking about [now].

Joe (39:52):

At some point, we’re going to do an Odd Lots episode specifically on the promise of small modular reactors and why we still don’t have them despite the seeming benefits. But do you have a sort of succinct answer for why this sort of seeming solution of manufacturing them faster, etc., has not translated into anything in production?

Brian (40:14)

Well, quite simply, we just forgot how to do it. We used to be able to build nuclear in this country. We did it in the seventies, we did it in the eighties, but every person that was involved in any one of those projects is either not alive or certainly not still a project manager at a company that would be building nuclear plants, right?

I think we underestimate human capacity to forget things. Just because we’ve done something in the past doesn’t mean that we necessarily can do it. Again, we have to relearn these things, and as a country, we do not have a supply chain. We don’t have a labor force. We don’t have people that manage construction projects that know how to do any of these things.

And so when you look at what South Korea is doing, you look at what China’s doing, they’re building nuclear plants with regularity. They’re doing it at a very attractive cost. They’re doing it on a predictable time horizon, but they have actually built all of those resources that we just simply don’t have in this country that we need and we need to rebuild that capability. It just doesn’t exist today…

…Brian (41:50):

Absolutely. And so if you go back to the era that we’ve been in of relative no load growth, if you’re a utility regulator and utility comes and asks you for a billion dollars for new investment and you’re used to saying ‘no,’ you’re used to saying ‘Well, wait a minute. Why do you need this? What is this for? How is this going to help manage again, reliability, cost, predictability, etc.?’

Now you’re in this whole new world and going back to this concept of we easily forget things — no one who’s a regulator today or the head of utility today has ever lived through an environment where we’ve had this massive expansion of the demand for electricity. So everyone now, including the regulators are having to relearn, okay, how do we enable utility investment in a growth environment? It’s not something they’ve ever done before. And so they’re having to figure out, okay, how do we create the bandwidth for utilities to make these investments?

Because one of the fundamental challenges that utilities have is that they struggle to invest if there’s no customer sitting there asking for the request, so they can’t sort of invest. I mean, if I’m Nvidia and I’m thinking about the world five years from now and think ‘Wow, how many chips do I want to sell in 2030?’ I can go out and build a new factory. I can go out and invest capital and I can go do all, I mean, I don’t need to have an order from a Microsoft or an Amazon or a Meta to go do that. I can build speculatively.

Utilities can’t really do that. They’re basically waiting for the customer to come ask for it. But when you have all this demand show up at the same time, well, what happens? The lead time start to extend. And so instead of saying ‘Yeah, I’ll give you that power in a year or two years,’ it’s now like, ‘Well, I’ll give it to you in five to seven years.’ And so that’s an unsustainable way to run the electric utility grid. So we do need regulators to adapt and evolve to this new era of growth.

5. Reflections from the heart of Japan’s ancient cedar forest – Thomas Chua

Yakushima was particularly memorable, an island near Kagoshima famous for its wildlife and ancient cedar forests. These majestic cedars, some of the oldest trees in the world, grow steadily through centuries, unaffected by the transient storms and seasonal fluctuations.

This is Sennensugi, which means a thousand-year-old cedar tree even though it’s still young. Yakushima’s oldest tree (and the oldest tree in Japan) is Jōmon Sugi, which is estimated to be between 2,170 and 7,200 years old.

This resonates deeply with my investment strategy. Just as these enduring cedars are not swayed by the fleeting changes in their environment, I focus on “Steady Compounders”—companies with significant economic moats and consistent intrinsic value growth.

When friends learn about my extensive travels, they often ask, “What about your investments? Don’t you need to monitor them constantly?” What they usually mean about ”monitoring” isn’t analyzing quarterly business results, but rather obsessively tracking stock prices and consuming every tidbit of news to stay perpetually informed.

However, I liken such constant vigilance to setting up a camera in a forest to watch the trees grow, this approach isn’t just tedious—it’s unnecessary and potentially harmful, often prompting rash decisions.

Everyone invests to grow wealth, but understanding why you invest is crucial. For me, it serves to enrich my curiosity and intellect, rewards my eagerness to learn, and more importantly, grants me the freedom to live life on my terms and cherish moments with my loved ones.

Therefore, I don’t pursue obscure, unproven companies which require intensive monitoring. Instead, I look for Steady Compounders — firms with a significant economic moat that are growing their intrinsic value steadily.

Like the steady growth of Yakushima’s cedars, these firms don’t need constant oversight; they thrive over long periods through economic cycles, much as the cedars endure through seasonal changes. Investing in such companies gives me the freedom to explore the world, knowing my investments are growing steadily, mirroring the quiet, powerful ascent of those ancient trees.

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, Meta Platforms, 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 28 April 2024:

1. 10 Questions with Chris Beselin – Michael Fritzell and Chris Beselin

Today, I’ll be interviewing Chris Beselin, who runs a Vietnam-focused activist fund and two tech businesses from his base in Ho Chi Minh City…

…3. Why do you Vietnam has been so successful as an economy – why has it developed faster than almost any other nation on earth?

There are a range of factors, of course, but just to outline a few:

It’s a balanced economy and growth model – it’s not your typical emerging market, where the economy is overly dependent on one or a handful of commodities.

Rather, the Vietnamese growth model has multiple core engines: its one of the most trade-focused economies in the world (measured as (export+import)/GDP) with free trade agreements signed with countries representing 60% of global GDP, it has a young and well-educated population where English proficiency is on par with e.g. India and South Korea, it has a sizeable and confident middle class that is rapidly growing and it has a stable government that has been focused on pro-market deregulations for the past 35 years.

And in contrast to what many people think from the outset, Vietnamese exports are primarily engineering-driven (as opposed to lower value-add textiles and similar). Around 45% of the exports are electronics, smartphones, laptops and machinery components. In this sense, my conviction is that Vietnam is much more the next South Korea or Japan than the next China.

To me, this all boils down to the fact that the number one asset of the country is its young, savvy and hungry engineering population (ca. 100,000 engineers are educated per year in Vietnam, of which around 50% are within software). The attractiveness of the Vietnamese engineering talent pulls foreign capital to invest in onshore engineering-centered manufacturing, which in turn has vast ripple effects on the employment of thousands of additional factory workers around the engineers…

…5. What misconceptions do you think foreigners typically have about the country?

I think there are many. Just to name a few:

The first one is perhaps “Vietnam is almost like China, but smaller and less developed”. I went through a bit of the difference in the fabric of the economies and demographics previously, but then there is also the very important difference in politics. Geopolitically, Vietnam is not and will never be or perceive itself to be a global superpower like China – it doesn’t have any geopolitical ambitions outside its own borders like China has.

Vietnam is primarily interested in developing its economy through trade and FDI, this in turn means that Vietnam in practice benefits from being geopolitically neutral between East and West and by trading/being friends with “everyone”. So far the country has managed this balance very astutely for decades.

Another common misconception (particularly for Westerners growing up during the Vietnam War) is that “Vietnam is just getting back on its feet after the recent war”. Obviously, this perspective is wildly outdated, but it’s still surprisingly common among foreign visitors. To put it into perspective, perhaps a suitable analogy is if you would have been saying/thinking similar things about France or the UK in the mid 90s… (also then ca. 50 years from the end of the Second World War, just like Vietnam is today 50 years away from its war ending in 1975).

2. Private Credit Offers No Extra Gains After Fees, New Study Finds – Justina Lee

A trio of academics has a bold take on the booming $1.7 trillion private credit market: after accounting for additional risks and fees, the asset class delivers virtually no extra return to investors…

…“It’s not a panacea for investors where they can earn 15% risk-free,” said Michael Weisbach, a finance professor at Ohio State University who co-wrote the research with Isil Erel and Thomas Flanagan. “Once you adjust for the risk, they basically are getting the amount they deserve, but no more.”

Behind the research is complex math to try to untangle the alpha part of a return that’s down to skill, and the beta part that might just come from stumbling into a bull market. While comparing stock pickers to a market benchmark like the S&P 500 is standard by now, it’s not obvious what the right yardstick is for private-credit funds, which make idiosyncratic and opaque loans to a wide array of companies…

..The three economists dissected MSCI data on 532 funds’ cash flows, covering their incoming capital and distributions to investors. They compare the industry’s performance to stock and credit portfolios with similar characteristics, whose fluctuations end up explaining the majority of private-credit returns. The study makes the case that these private credit funds also carry some equity risk, since around 20% of their investments contain equity-like features such as warrants.

After accounting for those risks, they find that there’s still alpha left on the table — which only vanishes once fees paid to these managers are deducted…

…As private markets boom, some quants — most notably Cliff Asness of AQR Capital Management — have suggested that investors are being misguided by returns that mask volatility and may be less impressive than they appear.

True at Adams Street Partners, who co-wrote one of the first papers on private-credit performance, cautions that until the industry faces its first downturn it may be hard to determine real alpha. But he says the NBER study is a good step toward digging beneath the surface of private-credit returns.

3. Americans are still not worried enough about the risk of world war – Noah Smith

When Americans think of World War 2, we usually think of the roughly four years of the war that we participated in, from late 1941 through 1945. Those years were indeed the most climactic and destructive of the war, by far, but the war actually began earlier. In fact, although the official start date is September 1, 1939, it’s easy to make an argument that the war began long before that…

…But throughout the 1930s, there were a number of conflicts that led into World War 2, and eventually merged with that overall conflict, like tributaries emptying into a great river. Let’s do a quick timeline of these.

In 1931-32, Japan seized Manchuria from China, an act that led inexorably to a wider war between the two powers. The Manchurian war and occupation also set Japan on a path toward militarism, bringing to power the regime that would ultimately prosecute WW2 itself.

In 1935-36, fascist Italy invaded and conquered Ethiopia. The League of Nations halfheartedly tried to stop the war and failed, leading to the League being discredited and the post-WW1 order being greatly weakened. That emboldened the fascist powers. Ethiopian resistance to Italian rule would eventually become a minor theater of WW2.

From 1935 through 1939, Japan and the Soviet Union fought an undeclared border war, ultimately culminating in major battles in 1939, which the USSR won. That led to Japan seeking an alliance with Nazi Germany, and eventually led to the Soviets’ entry into the war against Japan at the very end of WW2. (The realization that Japan couldn’t defeat the Soviets and conquer Siberian oil fields also prompted Japan to try to take Southeast Asian oil instead, when it needed oil to prosecute its war against China; this led to Pearl Harbor and the Pacific War.)

From 1936 through 1939, Nazi Germany, fascist Italy, and the Soviet Union fought each other in a proxy war: the Spanish Civil War. Units from all three powers officially or unofficially engaged in the fighting. When the Nationalists won, it emboldened the fascist powers even further.

In 1937, Japan invaded the remainder of China, in what’s called the Second Sino-Japanese War. This became a major theater of World War 2, accounting for almost as many deaths as the Nazi invasion of the USSR. It also prompted Japan to go to war with Britain and the U.S., in order to seize the oil fields of Indonesia to support the invasion of China. (The fact that we don’t count this as the start of WW2 seems like pure eurocentrism to me.)

In 1939, before the Soviet Union joined World War 2, it invaded Finland in what’s known as the Winter War, seizing some territory at great cost. This war would continue all the way through WW2 itself.

So there were no fewer than six wars in the 1930s that ended up feeding into World War 2 itself!..

…So if you were living at any point in 1931 through 1940, you would already be witnessing conflicts that would eventually turn into the bloodiest, most cataclysmic war that humanity has yet known — but you might not realize it. You would be standing in the foothills of the Second World War, but unless you were able to make far-sighted predictions, you wouldn’t know what horrors lurked in the near future.

In case the parallel isn’t blindingly obvious, we might be standing in the foothills of World War 3 right now. If WW3 happens, future bloggers might list the wars in Ukraine and Gaza in a timeline like the one I just gave.

Or we might not be in the foothills of WW3. I think there’s still a good chance that we can avert a wider, more cataclysmic war, and instead have a protracted standoff — Cold War 2 — instead. But I’m not going to lie — the outlook seems to be deteriorating. One big reason is that China appears to be ramping up its support for Russia…

…So it makes sense to view the Ukraine War as a European proxy conflict against Russia. But what’s more ominous is that it also makes an increasing amount of sense to view it as a Chinese proxy conflict against Europe.

A little over a year ago, there were reports that China was sending lethal aid to Russia. Believing these reports, I concluded — perhaps prematurely — that China had gone “all in” on Russia’s military effort. Some of the reports were later walked back, but the fact is, it’s very hard to know how much China is doing to provide Russia with drones and such under the table. But now, a year later, there are multiple credible reports that China is ramping up aid in various forms.

For example, the U.S. is now claiming that China is providing Russia with both material aid and geospatial intelligence (i.e. telling Russia where Ukrainian units are so Russia can hit them):

The US is warning allies that China has stepped up its support for Russia, including by providing geospatial intelligence, to help Moscow in its war against Ukraine.

Amid signs of continued military integration between the two nations, China has provided Russia with satellite imagery for military purposes, as well as microelectronics and machine tools for tanks, according to people familiar with the matter.

China’s support also includes optics, propellants to be used in missiles and increased space cooperation, one of the people said.

President Joe Biden raised concerns with Xi Jinping during their call this week about China’s support for the Russian defense industrial base, including machine tools, optics, nitrocellulose, microelectronics, and turbojet engines, White House National Security Council spokesperson Adrienne Watson said.

This is very similar to the aid that Europe and the U.S. are providing Ukraine. We also provide geospatial intelligence, as well as materiel and production assistance. If reports are correct — and this time, they come from the U.S. government as well as from major news organizations — then China is now playing the same role for Russia that Europe and the U.S. have been playing for Ukraine.

In other words, the Ukraine War now looks like a proxy war between China and Europe…

…Of course, World War 3 will actually begin if and when the U.S. and China go to war. Almost everyone thinks this would happen if and when China attacks Taiwan, but in fact there are several other flashpoints that are just as scary and which many people seem to be overlooking.

First, there’s the South China Sea, where China has been pressing the Philippines to surrender its maritime territory with various “gray zone” bullying tactics…

…The U.S. is a formal treaty ally of the Philippines, and has vowed to honor its commitments and defend its ally against potential threats.

And then there’s the ever-present background threat of North Korea, which some experts believe is seriously considering re-starting the Korean War. That would trigger a U.S. defense of South Korea, which in turn might bring in China, as it did in the 1950s.

It’s also worth mentioning the China-India border. China has recently reiterated its claim to the Indian state of Arunachal Pradesh, calling it “South Tibet” and declaring that the area was part of China since ancient times. India has vigorously rejected this notion, of course. An India-China border war might not start World War 3, but the U.S. would definitely try to help India out against China, as we did in 2022…

…America hasn’t mustered the urgency necessary to resuscitate its desiccated defense-industrial base. China is engaging in a massive military buildup while the U.S. is lagging behind. This is from a report from the Center for Strategic and International Studies:

The U.S. defense industrial base…lacks the capacity, responsiveness, flexibility, and surge capability to meet the U.S. military’s production needs as China ramps up defense industrial production. Unless there are urgent changes, the United States risks weakening deterrence and undermining its warfighting capabilities against China and other competitors. A significant part of the problem is that the U.S. defense ecosystem remains on a peacetime footing, despite a protracted war in Ukraine, an active war in the Middle East, and growing tensions in the Indo-Pacific in such areas as the Taiwan Strait and Korean Peninsula.

The United States faces several acute challenges.

First, the Chinese defense industrial base is increasingly on a wartime footing and, in some areas, outpacing the U.S. defense industrial base…Chinese defense companies…are producing a growing quantity and quality of land, maritime, air, space, and other capabilities. China…is heavily investing in munitions and acquiring high-end weapons systems and equipment five to six times faster than the United States…China is now the world’s largest shipbuilder and has a shipbuilding capacity that is roughly 230 times larger than the United States. One of China’s large shipyards, such as Jiangnan Shipyard, has more capacity than all U.S. shipyards combined, according to U.S. Navy estimates.

Second, the U.S. defense industrial base continues to face a range of production challenges, including a lack of urgency in revitalizing the defense industrial ecosystem…[T]here is still a shortfall of munitions and other weapons systems for a protracted war in such areas as the Indo-Pacific. Supply chain challenges also remain serious, and today’s workforce is inadequate to meet the demands of the defense industrial base.

“The Chinese defense industrial base is increasingly on a wartime footing.” If that isn’t a clear enough warning, I don’t know what would be. You have now been warned!

4. Memory Shortage and ASML – Nomad Semi

Although we are down to 3 major DRAM manufacturers, memory has always been a commodity since the 1970s. The wild swing in gross margin for SK Hynix in a year proves the point. Supply is the underlying driver of the memory cycle. Memory prices should always trend down over time with cost as memory manufacturers migrate to the next process node that allows for higher bit per wafer. There is always a duration mismatch in demand and supply due to the inelasticity of supply. When there is supernormal profit, capex will go up and supply will follow in 2 to 3 years. Supply will exceed demand and DRAM prices should fall to cost theoretically. This post will instead focus on the current upcycle and how it could actually be stronger than the 2016 cycle.

How we get here today is a result of the proliferation of AI and the worst downcycle since the GFC. To be fair, the 2022 downcycle was driven by a broad-based inventory correction across all the verticals rather than very aggressive capex hike from the memory producer. The downcycle led to negative gross margin for SK Hynix, Micron and Kioxia. Negative gross margin led to going concern risk, which led to Hynix and Micron cutting their capex to the lowest level in the last 7 years. This is despite the fact that we have moved from 1x nm to 1b nm which will require higher capex per wafer.

HBM (High Bandwidth Memory) has become very important in AI training, and you can’t run away from talking about HBM if you are looking at DRAM…

…HBM affects current and future DRAM supply in 2 different ways. The 1st is cannibalization of capex from DRAM and NAND where Fabricated Knowledge gave a very good analogy. The 2nd is as Micron mentioned in the last call that “HBM3E consumes approximately three times the wafer supply as DDR5 to produce a given number of bits in the same technology node”. In fact, this ratio will only get worse in 2026 as HBM4 can consume up to 5x the wafer supply as DDR5. The way it works is a HBM die size is double that of DDR5 (which already suffers from single digit die size penalty vs DDR4). HBM die size will only get bigger as more TSV is needed. Yield rate of HBM3e is below 70% and will only get harder as more dies are stacked beyond 8-hi. Logic base die of the HBM module is currently produced in-house by Micron and Hynix although this could be outsourced to TSMC for HBM4. In summary, not only is HBM consuming more of current wafer supply, but it is also cannibalizing the capex for future DRAM and NAND capacity expansion

In past upcycles, capex will often go up as memory producers gain confidence. Nobody wants to lose market share as current capex = future capacity → future market share. However, SK Hynix and Micron will be unable to expand their DRAM wafer capacity meaningfully in 2024 and 2025.

SK Hynix has limited cleanroom space available for DRAM expansion (~45k wpm) at M16 and this will be fully utilized by 2025. The company will have to wait till 2027 before Yong-In fab can be completed. Even when the balance sheet situation for SK Hynix improves in 2025, it will be limited by its cleanroom space.

For Micron, the situation is slightly better. Taichung fab also has limited space available for capacity expansion, but this will likely be earmarked for HBM production. Micron will have to wait until the new Boise fab is ready in 2026. Both Micron and Hynix will be limited in capacity expansion in 2025 against their will.

5. Artificial intelligence is taking over drug development – The Economist

The most striking evidence that artificial intelligence can provide profound scientific breakthroughs came with the unveiling of a program called AlphaFold by Google DeepMind. In 2016 researchers at the company had scored a big success with AlphaGo, an AI system which, having essentially taught itself the rules of Go, went on to beat the most highly rated human players of the game, sometimes by using tactics no one had ever foreseen. This emboldened the company to build a system that would work out a far more complex set of rules: those through which the sequence of amino acids which defines a particular protein leads to the shape that sequence folds into when that protein is actually made. AlphaFold found those rules and applied them with astonishing success.

The achievement was both remarkable and useful. Remarkable because a lot of clever humans had been trying hard to create computer models of the processes which fold chains of amino acids into proteins for decades. AlphaFold bested their best efforts almost as thoroughly as the system that inspired it trounces human Go players. Useful because the shape of a protein is of immense practical importance: it determines what the protein does and what other molecules can do to it. All the basic processes of life depend on what specific proteins do. Finding molecules that do desirable things to proteins (sometimes blocking their action, sometimes encouraging it) is the aim of the vast majority of the world’s drug development programmes.

Because of the importance of proteins’ three-dimensional structure there is an entire sub-discipline largely devoted to it: structural biology. It makes use of all sorts of technology to look at proteins through nuclear-magnetic-resonance techniques or by getting them to crystallise (which can be very hard) and blasting them with x-rays. Before AlphaFold over half a century of structural biology had produced a couple of hundred thousand reliable protein structures through these means. AlphaFold and its rivals (most notably a program made by Meta) have now provided detailed predictions of the shapes of more than 600m.

As a way of leaving scientists gobsmacked it is a hard act to follow. But if AlphaFold’s products have wowed the world, the basics of how it made them are fairly typical of the sort of things deep learning and generative AI can offer biology. Trained on two different types of data (amino-acid sequences and three-dimensional descriptions of the shapes they fold into) AlphaFold found patterns that allowed it to use the first sort of data to predict the second. The predictions are not all perfect. Chris Gibson, the boss of Recursion Pharmaceuticals, an AI-intensive drug-discovery startup based in Utah, says that his company treats AlphaFold’s outputs as hypotheses to be tested and validated experimentally. Not all of them pan out. But Dr Gibson also says the model is quickly getting better…

…A lot of pharma firms have made significant investments in the development of foundation models in recent years. Alongside this has been a rise in AI-centred startups such as Recursion, Genesis Therapeutics, based in Silicon Valley, Insilico, based in Hong Kong and New York and Relay Therapeutics, in Cambridge, Massachusetts. Daphne Koller, the boss of Insitro, an AI-heavy biotech in South San Francisco, says one sign of the times is that she no longer needs to explain large language models and self-supervised learning. And Nvidia—which makes the graphics-processing units that are essential for powering foundation models—has shown a keen interest. In the past year, it has invested or made partnership deals with at least six different AI-focused biotech firms including Schrodinger, another New York based firm, Genesis, Recursion and Genentech, an independent subsidiary of Roche, a big Swiss pharmaceutical company.

The drug-discovery models many of the companies are working with can learn from a wide variety of biological data including gene sequences, pictures of cells and tissues, the structures of relevant proteins, biomarkers in the blood, the proteins being made in specific cells and clinical data on the course of disease and effect of treatments in patients. Once trained, the AIs can be fine tuned with labelled data to enhance their capabilities.

The use of patient data is particularly interesting. For fairly obvious reasons it is often not possible to discover the exact workings of a disease in humans through experiment. So drug development typically relies a lot on animal models, even though they can be misleading. AIs that are trained on, and better attuned to, human biology may help avoid some of the blind alleys that stymie drug development.

Insitro, for example, trains its models on pathology slides, gene sequences, MRI data and blood proteins. One of its models is able to connect changes in what cells look like under the microscope with underlying mutations in the genome and with clinical outcomes across various different diseases. The company hopes to use these and similar techniques to find ways to identify sub-groups of cancer patients that will do particularly well on specific courses of treatment.

Sometimes finding out what aspect of the data an AI is responding to is useful in and of itself. In 2019 Owkin, a Paris based “AI biotech”, published details of a deep neural network trained to predict survival in patients with malignant mesothelioma, a cancer of the tissue surrounding the lung, on the basis of tissue samples mounted on slides. It found that the cells most germane to the AI’s predictions were not the cancer cells themselves but non-cancerous cells nearby. The Owkin team brought extra cellular and molecular data into the picture and discovered a new drug target. In August last year a team of scientists from Indiana University Bloomington trained a model on data about how cancer cells respond to drugs (including genetic information) and the chemical structures of drugs, allowing it to predict how effective a drug would be in treating a specific cancer.

Many of the companies using AI need such great volumes of high quality data they are generating it themselves as part of their drug development programmes rather than waiting for it to be published elsewhere. One variation on this theme comes from a new computational sciences unit at Genentech which uses a “lab in the loop” approach to train their AI. The system’s predictions are tested at a large scale by means of experiments run with automated lab systems. The results of those experiments are then used to retrain the AI and enhance its accuracy. Recursion, which is using a similar strategy, says it can use automated laboratory robotics to conduct 2.2m experiments each week…

…The world has seen a number of ground breaking new drugs and treatments in the past decade: the drugs targeting GLP-1 that are transforming the treatment of diabetes and obesity; the CAR-T therapies enlisting the immune system against cancer; the first clinical applications of genome editing. But the long haul of drug development, from discerning the biological processes that matter to identifying druggable targets to developing candidate molecules to putting them through preclinical tests and then clinical trials, remains generally slow and frustrating work. Approximately 86% of all drug candidates developed between 2000 and 2015 failed to meet their primary endpoints in clinical trials. Some argue that drug development has picked off most of biology’s low-hanging fruit, leaving diseases which are intractable and drug targets that are “undruggable”.

The next few years will demonstrate conclusively if AI is able to materially shift that picture. If it offers merely incremental improvements that could still be a real boon. If it allows biology to be deciphered in a whole new way, as the most boosterish suggest, it could make the whole process far more successful and efficient—and drug the undruggable very rapidly indeed. The analysts at BCG see signs of a fast-approaching AI-enabled wave of new drugs. Dr Pande warns that drug regulators will need to up their game to meet the challenge. It would be a good problem for the world to have. 

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 DeepMind), 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 21 April 2024:

1. The Anguish of Central Banking – Arthur F. Burns

Why, in particular, have central bankers, whose main business one might suppose is to fight inflation, been so ineffective in dealing with this worldwide problem?

To me, as a former central banker, the last of these questions is especially intriguing. One of the time-honored functions of a central bank is to protect the integrity of its nation’s currency, both domestically and internationally. In monetary policy central bankers have a potent means for fostering stability of the general price level. By training, if not also by temperament, they are inclined to lay great stress on price stability, and their abhorrence of inflation is continually reinforced by contacts with one another and with like-minded members of the private financial community. And yet, despite their antipathy to inflation and the powerful weapons they could wield against it, central bankers have failed so utterly in this mission in recent years. In this paradox lies the anguish of central banking…

…Analyses of the inflation that the United States has experienced over the past fifteen years frequently proceed in three stages. First are considered the factors that launched inflation in the mid-1960s, particularly the governmental fine tuning inspired by the New Economics and the loose financing of the war in Vietnam. Next are considered the factors that led to subsequent strengthening of inflationary forces, including further policy errors, the devaluations of the dollar in 1971 and 1973, the worldwide economic boom of 1972-73, the crop failures and resulting surge in world food prices in 1973-74, the extraordinary increases in oil prices that became effective in 1974, and the sharp deceleration of productivity growth from the late 1960s onward. Finally, attention is turned to the process whereby protracted experience with inflation has led to widespread expectations that it will continue in the future, so that inflation has acquired a momentum of its own.

I have no quarrel with analyses of this type. They are distinctly helpful in explaining the American inflation and, with changes here and there, that in other nations also. At the same time, I believe that such analyses overlook a more fundamental factor: the persistent inflationary bias that has emerged from the philosophic and political currents that have been transforming economic life in the United States and elsewhere since the 1930s. The essence of the unique inflation of our times and the reason central bankers have been ineffective in dealing with it can be understood only in terms of those currents of thought and the political environment they have created…

…the period from World War II to the mid-1960s was marked not only by a dampening of the business cycle but also by persistent increases in the prosperity of American families…

…This experience of economic progress strengthened the public’s expectations of progress. What had once been a quiet personal feeling that the long future would be better than the past, particularly for one’s children, was transformed during the postwar years into an articulate and widespread expectation of steady improvement in living standards—indeed, into a feeling of entitlement to annual increases in real income.

But the rapid rise in national affluence did not create a mood of contentment. On the contrary, the 1960s were years of social turmoil in the United States, as they were in other industrial democracies…

…In the innocence of the day, many Americans came to believe that all of the new or newly discovered ills of society should be addressed promptly by the federal government. And in the innocence of the day, the administration in office attempted to respond to the growing demands for social and economic reform while waging war in Vietnam on a rising scale. Under the rubric of the New Economics, a more activist policy was adopted for the purpose of increasing the rate of economic growth and reducing the level of unemployment…

…The interplay of governmental action and private demands had an internal dynamic that led to their concurrent escalation. When the government undertook in the mid-1960s to address such “unfinished tasks” as reducing frictional unemployment, eliminating poverty, widening the benefits of prosperity, and improving the quality of life, it awakened new ranges of expectation and demand. Once it was established that the key function of government was to solve problems and relieve hardships—not only for society at large but also for troubled industries, regions, occupations, or social groups—a great and growing body of problems and hardships became candidates for governmental solution…

…Many results of this interaction of government and citizen activism proved wholesome. Their cumulative effect, however, was to impart a strong inflationary bias to the American economy. The proliferation of government programs led to progressively higher tax burdens on both individuals and corporations. Even so, the willingness of government to levy taxes fell distinctly short of its propensity to spend. Since 1950, the federal budget has been in balance in only five years. Since 1970, a deficit has occurred in every year. Not only that, but the deficits have been mounting in size. Budget deficits have thus become a chronic condition of federal finance; they have been incurred when business conditions were poor and also when business was booming. But when the government runs a budget deficit, it pumps more money into the pocketbooks of people than it withdraws from their pocketbooks; the demand for goods and services therefore tends to increase all around. That is the way the inflation that has been raging since the mid-1960s first got started and later kept being nourished.

The pursuit of costly social reforms often went hand in hand with the pursuit of full employment. In fact, much of the expanding range of government spending was prompted by the commitment to full employment. Inflation came to be widely viewed as a temporary phenomenon—or, provided it remained mild, as an acceptable condition. “Maximum” or “full” employment, after all, had become the nation’s major economic goal— not stability of the price level. That inflation ultimately brings on recession and otherwise nullifies many of the benefits sought through social legislation was largely ignored…

…And so I finally come to the role of central bankers in the inflationary process. The worldwide philosophic and political trends on which I have been dwelling inevitably affected their attitudes and actions. In most countries, the central bank is an instrumentality of the executive branch of government—carrying out monetary policy according to the wishes of the head of government or the ministry of finance. Some industrial democracies, to be sure, have substantially independent central banks, and that is certainly the case in the United States. Viewed in the abstract, the Federal Reserve System had the power to abort the inflation at its incipient stage fifteen years ago or at any later point, and it has the power to end it today. At any time within that period, it could have restricted the money supply and created sufficient strains in financial and industrial markets to terminate inflation with little delay. It did not do so because the Federal Reserve was itself caught up in the philosophic and political currents that were transforming American life and culture…

…Facing these political realities, the Federal Reserve was still willing to step hard on the monetary brake at times—as in 1966, 1969, and 1974—but its restrictive stance was not maintained long enough to end inflation. By and large, monetary policy came to be governed by the principle of undernourishing the inflationary process while still accommodating a good part of the pressures in the marketplace. The central banks of other industrial countries, functioning as they did in a basically similar political environment, appear to have behaved in much the same fashion.

In describing as I just have the anguish of central banking in a modern democracy, I do not mean to suggest that central bankers are free from responsibility for the inflation that is our common inheritance. After all, every central bank has some room for discretion, and the range is considerable in the more independent central banks. As the Federal Reserve, for example, kept testing and probing the limits of its freedom to undernourish the inflation, it repeatedly evoked violent criticism from both the Executive Branch and the Congress and therefore had to devote much of its energy to warding off legislation that could destroy any hope of ending inflation. This testing process necessarily involved political judgments, and the Federal Reserve may at times have overestimated the risks attaching to additional monetary restraint…

…Monetary theory is a controversial area. It does not provide central bankers with decision rules that are at once firm and dependable. To be sure, every central banker has learned from the world’s experience that an expanding economy requires expanding supplies of money and credit, that excessive creation of money will over the longer run cause or validate inflation, and that declining interest rates will tend to stimulate economic expansion while rising interest rates will tend to restrict it; but this knowledge stops short of mathematical precision…

…It is clear, therefore, that central bankers can make errors—or encounter surprises—at practically every stage of the process of making monetary policy. In some respects, their capacity to err has become larger in our age of inflation. They are accustomed, as are students of finance generally, to think of high and rising market interest rates as a restraining force on economic expansion. That rule of experience, however, tends to break down once expectations of inflation become widespread in a country. At such a time, lenders expect to be paid back in cheaper currency, and they are therefore apt to demand higher interest rates. Since borrowers have similar expectations, they are willing to comply. An “inflation premium” thus gets built into nominal interest rates. In principle, no matter how high the nominal interest rate may be, as long as it stays below or only slightly above the inflation rate, it very likely will have perverse effects on the economy; that is, it will run up costs of doing business but do little or nothing to restrain overall spending. In practice, since inflationary expectations, and therefore the real interest rates implied by any given nominal rate, vary among individuals, central bankers cannot be sure of the magnitude of the inflation premium that is built into nominal rates. In many countries, however, these rates have at times in recent years been so clearly below the ongoing inflation rate that one can hardly escape the impression that, however high or outrageous the nominal rates may appear to observers accustomed to judging them by a historical yardstick, they have utterly failed to accomplish the restraint that central bankers sought to achieve. In other words, inflation has often taken the sting out of interest rates— especially, as in the United States, where interest payments can be deducted for income tax purposes…

…There is a profound difference between the effects of mistaken judgments by a central bank in our age of inflation and the effects of such judgments a generation or two ago. In earlier times, when a central bank permitted excessive creation of money and credit in times of prosperity, the price level would indeed tend to rise. But the resulting inflation was confined to the expansion phase of the business cycle; it did not persist or gather force beyond that phase. Therefore, people generally took it for granted that the advance of prices would be followed by a decline once a business recession got under way. That is no longer the case.

Nowadays, businessmen, farmers, bankers, trade union leaders, factory workers, and housewives generally proceed on the expectation that inflation will continue in the future, whether economic activity is booming or receding. Once such a psychology has become dominant in a country, the influence of a central bank error that intensified inflation may stretch out over years, even after a business recession has set in. For in our modern environment, any rise in the general price level tends to develop a momentum of its own. It stimulates higher wage demands, which are accommodated by employers who feel they can recover the additional costs through higher prices; it results in labor agreements in key industries that call for substantial wage increases in later years without regard to the state of business then; and through the use of indexing formulas, it leads to automatic increases in other wages as well as in social security payments, various other pensions, and welfare benefits, in rents on many properties, and in the prices of many commodities acquired under long-term contracts…

…If my analysis of central banking in the modern environment is anywhere near the mark, two conclusions immediately follow. First, central banks have indeed been participants in the inflationary process in which the industrial countries have been enmeshed, but their role has been subsidiary. Second, while the making of monetary policy requires continuing scrutiny and can stand considerable improvement, we would look in vain to technical reforms as a way of eliminating the inflationary bias of industrial countries. What is unique about our inflation is its stubborn persistence, not the behavior of central bankers. This persistence reflects the fundamental forces on which I dwelt earlier in this address—namely, the philosophic and political currents of thought that have impinged on economic life since the Great Depression and particularly since the mid-1960s…

…The precise therapy that can serve a nation best is not easy to identify, and what will work well in one country may work poorly in another. In the case of the American inflation, which has become a major threat to the well-being of much of the world as well as of the American people, it would seem wise to me at this juncture of history for the government to adopt a basic program consisting of four parts. The first of these would be a legislative revision of the federal budgetary process that would make it more difficult to run budget deficits and that would serve as the initial step toward a constitutional amendment directed to the same end. The second part would be a commitment to a comprehensive plan for dismantling regulations that have been impeding the competitive process and for modifying others that have been running up costs and prices unnecessarily. The third part would be a binding endorsement of restrictive monetary policies until the rate of inflation has become substantially lower. And the fourth part would consist of legislation scheduling reductions of business taxes in each of the next five years—the reduction to be quite small in the first two years but to become substantial in later years. This sort of tax legislation would release powerful forces to improve the nation’s productivity and thereby exert downward pressure on prices; and it would also help in the more immediate future to ease the difficult adjustments forced on many businesses and their employees by the adoption of the first three parts of the suggested program.

2. Two Things I’m Not Worried About – Ben Carlson

Here are two things a lot of other people are worried about but not me:

Stock market concentration. Here’s a chart from Goldman Sachs that shows by one measure, the U.S. stock market is as concentrated as it has ever been:

To which my reply is: So what?

Yes, the top 10 stocks make up more than one-third of the S&P 500. All this tells me is that the biggest and best companies are doing really well. Is that a bad thing?

Stock markets around the globe are far more concentrated than the U.S. stock market. Emerging markets rose to their highest level since June 2022 yesterday. Out of an index that covers 20+ countries, a single stock (Taiwan Semiconductor) accounted for 70% of the move.

Stock market returns over the long run have always been dominated but a small minority of the biggest, best-performing companies…

… Bloomberg is out with a new report that sounds the alarm on U.S. government debt levels:

With uncertainty about so many of the variables, Bloomberg Economics has run a million simulations to assess the fragility of the debt outlook. In 88% of the simulations, the results show the debt-to-GDP ratio is on an unsustainable path — defined as an increase over the next decade.

In the end, it may take a crisis — perhaps a disorderly rout in the Treasuries market triggered by sovereign US credit-rating downgrades, or a panic over the depletion of the Medicare or Social Security trust funds — to force action. That’s playing with fire.

I’ll believe it when I see it.

People have been sounding the alarm on government debt in this country for decades. There has been no panic. No financial crisis. No debt default…

… Interest expense relative to the size of the economy has shot higher in recent years from the combination of more debt and higher rates:

But we’re still well below the highs from the 1980s and 1990s. And when you look at the absolute numbers here, going from 1.5% of GDP to 3% of GDP isn’t exactly the end of the world…

…Debt-to-GDP is now as high as it was in World War II:

That seems scary until you realize in Japan, debt-to-GDP is closer to 300%. I’m not saying we should test our limits but there is no pre-set line in the sand on these things.

3. The inequity method of accounting – Sujeet Indap

The fundamental bargain of M&A seems pretty simple. At the closing of a deal, the buyer pays the seller, and gets a business in return.

It hasn’t been so straightforward for the family who agreed in 2022 to sell its California supermarket Save Mart to the private equity firm Kingswood Capital Management, which valued the grocery chain at $245mn.

Three months after the papers were signed, Kingswood demanded that Save Mart’s prior owners, the Piccinini family, fork back over $109mn after already surrendering the company. In effect, Kingswood wanted to receive a net $77mn payment to take over Save Mart.

And thanks to some ballsy lawyering and nebulous bookkeeping, it seems the PE firm might actually succeed, its gambit upheld by a controversial arbitration ruling in September 2023…

…When Kingswood signed the deal for Save Mart, it was really acquiring two separate businesses. One was the Save Mart grocery chain, comprised of 200 stores and more than $4bn in annual revenue. Save Mart separately held a majority stake in Superstore Industries (SSI), a successful food wholesaler/distributor that had two other owners…

…The two sides agreed that Save Mart’s equity stake in SSI, the joint venture, would be valued at $90mn, a significant step up from the ~$22.5mn value that Save Mart had assigned the investment on its books.

The increase reflected SSI’s valuable land portfolio, according to one person familiar with the transaction. And it enables Kingswood to lower SSI’s tax basis should it ever want to sell SSI, according to a person involved in the transaction.

Those seem reasonable enough. Still, the accounting of SSI’s value is what laid the foundation for this dispute.

For context, a company’s investments can be recorded on its balance sheet in three ways: cost method, equity method, and full consolidation.

Save Mart selected the equity method for its SSI stake.

To explain a bit further: Let’s imagine a company with $100 of asset value and $60 in liabilities, which leaves it with an equity value of $40. Say this company has a 50-per-cent owner, meaning it owns $20 in equity. The owner’s balance sheet would list that $20 as a single line item, called “equity in unconsolidated affiliates”. That account would grow with the subsidiary’s proportional net income, and decrease with any net losses or dividends.

Save Mart’s stake in SSI was listed as a single line on its balance sheet — worth $22.5mn…

…In March 2022, Kingswood and Save Mart closed their deal with the PE firm sending payments based on the family’s proposed accounting. That then set off a final round of post-closing negotiations, where Kingswood got 90 days to argue with the Piccinini’s maths…

…But Kingswood dropped in one massive adjustment with the boilerplate.

It added back $109mn of gross SSI debt, and asserted that the figure counted as official “Indebtedness”. And it argued it should be paid back for all that additional debt.

The PE firm pointed to the language in the deal contract, and said the definition of “Indebtedness” included any Save Mart “group” debt…

…Arbitrator Joseph Slights III, a lawyer in private practice who was formerly a Delaware Vice Chancellor, did not ultimately buy any of what the Piccinnis were selling.

He wrote in the arbitration decision:  “Delaware law is more contractarian than most, and Delaware courts will enforce the letter of the parties’ contract without regard for whether they have struck a good deal or bad deal . . . the result is not absurd or commercially unreasonable.”…

…The Piccinnis, understandably, believe writing a cheque for $109mn is indeed “absurd” and “commercially unreasonable”. They have accused Kingswood of “bad faith” and “gamesmanship” in their court papers.

They will now appeal to the Delaware Supreme Court, pointing to a 2017 decision that said in a post-closing adjustment dispute, the legal system should aim to uphold the broader spirit of the contract instead of narrow contract definitions…

…Kingswood had believed, all along prior to signing and closing, that the gross SSI debt belonged on Save Mart’s main balance sheet. But they decided to keep quiet about that until after the deal closed.

One implication is that they were happy to close on the Piccininis’ terms, and winning on the SSI debt issue would be a bonus, given that there was no guarantee of winning the arbitration.

The firm’s equity check on the $240mn transaction was just $60mn (see the sources and uses table above). If Kingswood is eventually paid the $109mn, it will receive nearly two times their equity contribution by weaponising accounting and legal technicalities.

4. Don’t Be Afraid – Michael Batnick

All-time highs are interesting in the emotions they elicit. Some people might be euphoric as their accounts reach dollar amounts never seen before. Others might fear this is as good as it’s going to get and worry about a trap-door scenario.

Your emotional state might also depend on your asset allocation. If you’re sitting on a large cash pile, it’s understandable that you might be hesitant to go “all in” at a record price. It might not “feel” right.

The good news is the data doesn’t support those feelings. On average since 1970, the S&P 500 has done better 1, 3, and 5 years after making an all-time high than picking a random day.

5. An Interview with Google Cloud CEO Thomas Kurian About Google’s Enterprise AI Strategy – Ben Thompson and Thomas Kurian

You did mention that, “People are moving out of proof-of-concept into actually doing products”. Is that actually happening? What are the actual use cases that companies are actually rolling out broadly as opposed to doing experiments on what might be possible?

TK: Broad-brush, Ben, we can break it into four major categories. One category is streamlining internal processes within the organization, streamlining internal processes. In finance, you want to automate accounts receivable, collections, and cashflow prediction. In human resources, you want to automate your human help desk as well as improve the efficiency with which you can do benefits matching, for example. In procurement and supply chain, you want for example, look at all my suppliers, their contracts with me and tell me which ones have indemnification and warranty protection, so I can drive more volume to those that give me indemnification and warranties and less to those that don’t, for example. These are all practical cases we have customers live in deployment with.

Second is transforming the customer experience. Transforming the customer experiences, how you market, how you merchandise, how you do commerce, how you do sales and service. An example is what Mercedes-Benz CEO Ola Källenius talked about how they’re building a completely new experience for the way that they market and sell and service their vehicles.

Third is that some people are integrating it into their products, and when I say re-imagining their products, re-imagining their core products using AI. We had two examples of companies who are in the devices space. One is Samsung and the other one is Oppo, and they’re re-imagining the actual device itself using AI with all the multimodality that we provide.

There are quite a few companies now re-thinking that if a model can change the way that I see it, that I can process multimodal information. For example, in media we have people saying, “If your model can read as much information as it can, can it take a long movie and shrink it into highlights? Can I take a sports recording of the NCAA basketball final and say, ‘find me all the highlights by this particular player’?” and not have to have a human being sit there and splice the video, but have it do it and I can create the highlights reel really quickly. So there are lots of people re-imagining the product offerings that they have.

And finally, there are some people saying, “With the cost efficiency of this, I can change how I enter a brand new market because, for example, I can do personalized offers in a market where I may not have a physical presence, but I can do much higher conversion rate for customers with online marketing and advertising because now I can do highly tailored campaigns because the cost of creating the content is much lower.” So broad-brush, streamline the core processes and back office, transform the customer experience and it doesn’t mean call centers or chatbots, it can be actually transferring the product itself, transforming the nature of the product you build and enter new markets.

Is it fair to say then when you talk about, “Moving from proof-of-concept to actual production”, or maybe that’s not the words you used, but people are saying, “Okay, we’re going to build this” because this stuff’s not showing up yet, in the real world. Is it the case that, “We see that this could be valuable, now we’re in”, and that’s why you’re emphasizing the platform choice now because they’ve committed to AI broadly, and now it’s like, “Where are we going to build it”?

TK: We have people experimenting, but we also have people actually live deployment and directing traffic. Orange, the telecom company, was talking about how many customers they’re handling online, Discover Financial was talking about how their agents are actually using AI search and AI tools to discover information from policy and procedure documents live. So there are people actually literally running true traffic through these systems and actually using them to handle real customer workload.

Are you seeing the case in a lot of in customers, or maybe you’re hearing from potential customers, that AI is rolling out, if that’s the right word, in an employee arbitrage situation? Where there’s individual employees that are taking on themselves to use these tools and they are personally benefiting from the increased productivity — maybe they’re doing less work or maybe they’re getting more done — and the companies want to capture that more systematically. Is that a theme that you’re seeing?

TK: We’re seeing three flavors. Flavor one is a company has, we’re going to try eight or nine, what they call customer journeys or use cases, we’re going to pick the three that we see as the maximum return, meaning value and value does not mean cost savings always. It could be, for example, we have one who is handling 1 million calls a day through our customer service system. Now a million calls a day, if you think about it, Ben, an average person can do about 250 calls a day, that’s a certain volume in an eight-hour day. If you handled a million, that is a lot of people, so the reality is that several of them were not being answered and people never called because the wait time was so long. So in that case, it was not about cost savings, it’s the fact that they’re getting able to reach many more customers than they could do before. So that’s one. One part is people saying, “I have a bunch of scenarios, I’m going to pick the three”, and in many cases, they’re actually augmenting something they’re doing or doing something they couldn’t do before, that’s scenario one.

Scenario two was I have, for example, there’s a large insurance company that’s working with us. Today, when they do claims and risk calculation, it takes a long time to handle the claims and the risk, particularly the risk calculation, because there’s thousands of pages of documents, there’s a lot of spreadsheets going back and forth. They put it into Gemini and it was able to run the calculations much, much more quickly. So second is I’m picking a very high value use case for my organization, which is the core function, and I’m going to implement it because I can get a real competitive advantage. In their case, it’s the fact that they can both get more accurate scoring on the risk and they can also do a much more accurate job, faster job in responding.

And the third scenario is what you said. “Hey, we’ve got a bunch of people, we’re going to give it to a certain number of developers”. For example, our coding tool, “They are going to test it, they say it helps me generate much better unit tests, it helps me write better quality code”. Wayfair’s CTO was talking about what their experience is, and then they say, “Let’s go broadly”, so all three patterns are being seen…

…Do you see AI, though, in all this talk about, “You need to choose a platform? Sure, our platform’s going to be open, you can use it anywhere” — but do you see this as a wedge to be like, “Okay, this is a reboot broadly for the industry as far as cloud goes, and sure, your data may be in AWS, or in Azure, or whatever it might be, but if you have a platform going forward, you should start with us”? Then maybe we’ll look up in ten, fifteen years, and all the center of gravity shifted to wherever the platforms are?

TK: For sure. I mean, it’s a change in the way that people make purchase decisions, right? Ten years ago, you were worried about commodity computing, and you were like, “Who’s going to give me the lowest cost for compute, and the lowest cost for storage, and the lowest cost for networking?”. Now the basis of competition has changed and we have a very strong position, given our capability both at the top, meaning offering a platform, offering models, et cetera, and building products that have long integrated models.

Just as an example, Ben, integrating a model into a product is not as easy as people think; Gmail has been doing that since 2015. On any daily basis, there are over 500 million operations a day that we run and to do it well, when a partner talked about the fact that 75% of people who generate an image for slides actually end up presenting it, it’s because we have paid a lot of attention over the years on how to integrate it.

So we play at the top of the stack, and we have the infrastructure and scale to do it really well from a cost, performance, and global scale that changes the nature of the competition. So we definitely see this, as you said, as a reset moment for how customers thinking of choosing their cloud decision.

If you’re talking about a lot of choices about models, and customers were over-indexed on choosing the correct model, that implies that models are maybe a commodity, and that we’ve seen with GPT-4 prices are down something like 90% since release. Is that a trend you anticipate continuing, and is it something that you want to push and actually accelerate?

TK: Models — whether they’re a commodity or not, time will tell, these are very early innings. All we’re pointing out is every month, there’s a new model from a new player, and the existing models get better on many different dimensions. It’s like trying to pick a phone based on a camera, and the camera’s changing every two weeks, right? Is that the basis on which you want to make your selection?

Well, but if you make that basis, then you might be locked into the operating system.

TK: That’s right, and so that’s why we say you should choose an open platform, and you should be able to use a collection of different models, because it’s changing, and don’t lock into a particular operating system at a time when the applications on top of it are changing, to use your analogy.

Why is your platform open as compared to others? Microsoft has announced you can use other models, not just OpenAI models. Amazon is sort of, to the extent you can ascertain a strategy, it’s like, “Look, we’re not committing to anything, you could do whatever you want.” Why do you feel comfortable saying, “No, we’re the open one,” and they’re not?

TK: Well, first of all, the completeness of our platform; Vertex has a lot more services than you can get with the other platforms. Secondly, in order to improve a platform, you have to have your own model, because there’s a bunch of things you do when you engineer services with that model.

I’ll give you a really basic example. You use a model, you decide to ground the answers. Grounding improves quality, but can also introduce latency. How do you make sure that when you’re grounding, you’re not serially post-processing a model’s answer to add latency? Unless you have your own model, you wouldn’t even get to that. So because we have our own model, we’re able to engineer these things, but we make them available as services with other models, so you can use enterprise grounding as a very specific example. There are lots of customers using it with Mistral and with Llama and with Anthropic.

Second thing, we are not just offering models, but we’re actually helping the third party go to customers with us. I met a lot of customers today jointly with [CEO] Dario [Amodei] from Anthropic, and it’s a commitment to make sure we’re not just giving you our infrastructure, we’re not just training, integrating a model into Vertex, we’re not just making it a first-class model, but we’re actually bringing it to clients together.

I think that’s what we mean by open. One of the other players has no models of their own, so naturally they’re offering a bunch of models, and the other player has outsourced their model development to a third party…

…How important is that million context window in the story you are telling? My perception is, there’s a lot of stuff you could do if you build a lot of infrastructure around it, whether it be RAG or other implementations, but it feels like with Gemini 1.5 there are jack-of-all-trades possibilities that seem to open up to a much greater extent, and there’s a bit where, you had that compliance bit, the statements of work and they had to compare it to the 100-page compliance document. I got some comments like, “Maybe companies shouldn’t have 100-page compliance notebooks or whatever it might be”, but the reality is, that’s the case, the world has that. My perception of the keynote is, that was the killer feature, that seemed to undergird everything. Was that the correct perception?

TK: Yeah, there are two reasons. Just to be perfectly clear, Ben, the long context window allows you to do three things that are important. First of all, when you look at high definition video, for example, and other modalities, and just imagine you’re dumping a high definition video in and you want to create out of the NCAA final, which just happened, the highlight reel but you don’t want to specify every attribute about what you want spliced into the highlight reel. The model has to digest it and because it has to process it, it’s a fairly dense representation of the video because there are objects, there are people moving, there are actions, like I’m throwing a pass. They could be, I have my name on the back of my t-shirt, there could be a score like, “When did they change from 24 to 26 points? Did they score three pointers?”, so there are many, many, many dimensions. So reasoning becomes a lot better when you can take a lot more context, that’s one, and it’s particularly true of modality.

The second is, today people don’t use models to maintain state or memory, meaning they ask it a question, the next time they think, “Hey, it may not remember”, so when you’re able to maintain a longer context, you can maintain more state, and therefore you can do richer and richer things rather than just talk back-and-forth with a very simplistic interface. You see what I mean?

The third thing is, there are certainly complex scenarios, it’s the unfortunate reality, there’s lots of policies and procedure books that are even longer than what we showed, and so there are scenarios like that that we have to be able to deal with. But in the longer term, the real breakthrough is the following. Context length, if you can decouple the capabilities of the model and the latency to serve a model from the context length, then you can fundamentally change how quickly you can scale a model.

Is this ultimately, from your perspective, a question of infrastructure, and that just leans into Google’s biggest advantage?

TK: It’s a question of global infrastructure, but also optimizations at every layer in the infrastructure, which we can co-engineer with DeepMind…

…Sundar Pichai mentioned in his video greeting, he emphasized the number of AI startups, and particularly AI unicorns using Google Cloud. To go back to the reboot idea, do you view the AI Era as a restart in terms of capturing the next generation of companies? I mean, obviously, AWS had a huge advantage here as far as general cloud computing, the entire mobile app ecosystem was by and large built on AWS. In the enterprise era, you have to deal with what’s there, what they’ve already dealt with, you have to have the integrations. Do you see yourself as having this as a big focus, “We’re going to own this era of startups”?

TK: Yes. And by the way, every one of those startups is being pursued by the other two, and the fact that 90% of the unicorns and 60% of all AI-funded startups, up in each case by ten points in eight months, and they are the most discerning ones. I mean, just to be frank, the unicorns, for them, it is the really biggest cost of goods sold in their P&L.

So what’s the driver there?

TK: The efficiency of our infrastructure.

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 (parent of AWS), Microsoft, and TSMC. Holdings are subject to change at any time.