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 18 October 2026:
1. How Personal Agents Get Paid – Tanay Jaipuria
Over time, in certain categories, there may be a world where agents decide that they want to capture more of the take rate that aggregators do. For example, Booking’s might make 14% to 15% of GMV, but the affiliate fees they provide might only net out to ~5-7%.
To earn that extra fee, agents could get more involved in directly providing inventory access, accurate prices, reliable bookings, potentially being the merchant of record and support for changes and cancellations, rather than using the aggregator itself for these things.
As one note, it may be significantly easier to “compete” or try to displace aggregators that are more digitally oriented than for aggregators/marketplaces that do a lot of work in the physical world (such as an Uber or a DoorDash)…
…So far, we’ve largely covered the potential for agents via fees from merchants/partners, but there’s one other opportunity, which is around advertising (i.e., specifically trying to drive incremental demand for merchants).
It is possible that these don’t look like ads at all and in an agentic world, could just be paid through higher take rates, but I figure it’s still worth quantifying since we may also see other forms of ad units as well.
A merchant might pay the agent (in addition to the customer directly) to have their offer considered when someone asks for a hotel or a pair of shoes.
2. Models as Insider Risks in the Super Intelligence Era – Satya Nadella
We can’t treat Super Intelligence as a set of nested black boxes and simply accept or reject its recommendations, answers, and actions. We must build contained systems whose behavior we can observe, limits we can test, and actions we can always contain.
In other words, we need to separate the supply of intelligence from the authority over it.
Setting aside the hard problem of alignment, we need to start with an engineering approach to containment and governance. We need to surround non-deterministic models with strong, deterministic system design, human controls, and reliable operating procedures, and establish industry standards where existing ones are insufficient.
Treating frontier closed and open weight models like insider risks is a way to build such a system. Not because they are necessarily malicious, but because any sufficiently capable actor with access to important systems can make mistakes or be compromised, and the architecture of containment and control must account for that.
The good news is we have learned a lot about how to handle powerful actors inside the enterprise. This isn’t new! We’ve established best practices and refined them over decades (establish identity, limit privileges, log activity, create containment boundaries, etc.)…
…We should therefore design these systems around principles of observability…
…The most trustworthy Super Intelligence system will not be the one with the model we trust most. It will be the one that enables us to trust the model the least.
3. Room at the Top – Abdullah Al-Rezwan
I came across this astonishing chart from Michael Batnick yesterday showing that the combined market cap of the bottom 434 companies in the S&P 500 equals that of the top 5. It is the time series, though, that drives home just how unusually concentrated the index has become over the last 30 years. Even at the height of the tech bubble, the number was below 350 and following the crash, the number went closer to 200. And we have been at or above tech-bubble levels of concentration consistently since 2019, save for a brief pause in 2022…
…In a follow-up tweet, he noted: “Interestingly, the TTM net income of the top 5 stocks is also equal to the bottom 436 companies.” …
…Technology distribution is now truly global, a handful of companies have aggregated demand on top of it, and with near-zero marginal cost, much of every incremental dollar of revenue falls straight to the bottom line. Concentrated market cap hence looks more like the market faithfully mirroring where the profits have actually gone.
But wait a minute, aren’t we leaving behind the world of zero marginal cost? Perhaps, but given the allegedly lofty inference margin in the API layer for the labs, one wonders whether we haven’t quite diverged from the zero marginal world as much we thought we did. More importantly, I sometimes wonder what it would imply if all this capex buildout actually yields attractive returns…
…I am a naturally optimistic person. I share the unease about a world in which a handful of companies generate much of its profits, but despite the differences, railroads can still be quite instructive. The same UBS report had this rather mind boggling stat:
“…railroads, despite being a declining industry over the period of the study (falling from 63% of the US market in 1900 to less than 1% today), actually outperformed both the US stock market and their newer technology competitors since 1900.”
Ultimately, the economy the railroads helped make possible grew far larger than the rails themselves. Refrigerated railcars turned a St. Louis brewery into Budweiser. A national rail network turned a mail-order catalog into Sears. Railroad express contracts turned a parcel-and-cash courier into American Express. Grain sheds at the end of a rail line turned an Iowa grain flathouse into Cargill.
AI may inflict plenty of creative destruction, but it will likely also usher in a Cambrian explosion of small businesses that can hire and scale with AI agents faster than ever before. Much of their spending may well flow straight back to the top, so the “434” may not come down anytime soon. But perhaps the most hopeful version of this story doesn’t require the top 5 to shrink at all; it’s one where companies not yet founded grow large enough to join them.
4. Insurance claims to test Altman and Amodei liability for ‘rogue’ AI – Lee Harris
Insurers are preparing for multimillion-dollar claims stemming from AI agents going “rogue”, with growing concern that top executives such as OpenAI’s Sam Altman and Anthropic’s Dario Amodei could be held liable for the actions of their models…
…Insurance broker Aon analysed more than 300 AI-related legal cases and found that insurers could also be on the hook to pay claims under policies covering crime, intellectual property, media liability, cyber security, and technology errors and omissions.
Top industry figures also said AI executives could be sued for the actions of models, with insurers potentially facing claims under “directors and officers” insurance policies, which cover executives if they are taken to court for their decisions or statements.
“Ultimately, the OpenAI CEO is liable [for the Hugging Face incident], because there’s an absence of control in their business,” said Tim Rayner, UK head of underwriting and claims at Verisk. “So from a directors’ and officers’ perspective, it would come back to him.”
If OpenAI had purchased D&O insurance, Rayner added, the ChatGPT maker could seek to claim on the policy to cover potential future losses from lawsuits targeting its chief executive Sam Altman…
…Kevin Kalinich, a broker at Aon focused on digital risks, said that the strength of a legal claim against the head of an AI lab would depend on factors such as whether the executive had used “reasonable business judgment” in public statements about the company and its products.
However, Aki Hussain, chief executive of insurer Hiscox, said it was too soon to know how US courts would treat liability for AI agents: “It’s the US, so [a claim] could be under anything — whatever insurance they’ve got.”
5. Will AI replace workers? Not if we build it right – Daron Acemoglu
Many people have been forecasting that AI will drive remarkable improvements in productivity and economic output.
But reality seems to be closer to the great economist Robert Solow’s 1987 quip about computers—“everywhere but in the productivity statistics”. AI advances aren’t in the productivity statistics either, at least as of now.
Achievements of current AI models in writing computer code and designing software or in radiology diagnostics are truly remarkable. But that is in the lab, where context is clear and tasks are neatly delineated. The evidence from firms that adopt these tools for the same tasks “in the wild” isn’t as impressive. Often, engineers have to spend a lot of time debugging AI-written code and radiologists and doctors cannot combine their expertise with AI tools, leading to suboptimal outcomes. Several studies and reports have failed to find much in the way of productivity gains from adoption of AI by most companies. Of course, things may be different, and much better, with new AI agents, but so far the disappointing productivity performance isn’t confined to the aggregate. It is also there for firms that are experimenting with AI.
None of this should be surprising. Every new technology takes time to be productively adopted, and typically, the more transformative the technology, the slower this process is. Some draw parallels between AI and fire. A less hyperbolic comparison might be to electricity.
Famously, electricity, despite all of its revolutionary potential, took several decades to be adopted and used widely, even in manufacturing. Already in 1881 New York and London had central generating stations and many were foreseeing that electricity would completely transform manufacturing, homes and transport. But as late as the 1920s, only about half of factories and homes were using electricity. We may in fact expect productive adoption of AI to be even slower because of the necessary reorganization of firms. Job tasks will have to be reallocated between AI models and workers, and many specialists will have to learn how to work with AI. Managers will have to develop their own understanding of what it is that they want from AI and how they can work towards getting it. Worse, many firms may attempt to go for full automation of certain occupations, which is likely to remain impossible in the near future.
All of this will not just delay the productivity gains from AI, but imply that job displacement effects will be slow as well. Several rounds of forecasts predicted the end of radiology or drivers or accountants: none of that has happened yet. Hospitals still need radiologists (in fact more than before), there are still journalists, paralegals, accountants and office workers.
Two years ago, I estimated that within 10 years, we shouldn’t expect much more than about 5% of what humans do to be replaced by AI.
Where does this 5% number come from? Though it is admittedly no more than a guesstimate, it was based on an assessment of what AI models could achieve then and how these capabilities would develop in the second half of the 2020s.
To put it simply, my forecast was that by the end of the 2020s AI would be in a position to successfully automate simple office and cognitive tasks, making up about 20% of work in the US economy. I predicted that the impact on more complex cognitive and office jobs would be slower in coming, because it would take AI much longer to acquire the judgment, multi-dimensional reasoning and social skills that most jobs require. Nor would society be comfortable with AI fully taking over air traffic control, mental health therapy or the management of companies.
AI-based automation of jobs that involve heavy interaction with the physical world will also have to wait, because it will take time for flexible robotics, AI models with a full sense of spatial-causal relations. It may be even longer for computer vision technologies to reach the point where AI could be seamlessly combined with robots to perform tasks in manufacturing and construction.
I also estimated that, given the diffusion of other technologies, such as electricity, computers and previous computer vision technologies, only about one quarter of the 20% that can be fully automated would be automated within 10 years. This gave the 5% number.
My forecast (with the full warning that it was and remains impossible to know how quickly these models would advance) was that GDP would increase by about 1.5% over 10 years. Since then, advances in models have been more rapid than many, including me, predicted. On the other hand, we haven’t seen many applications that can easily be adopted and transform the production process, and, as I explain below, the magic of AI needs to happen in the applications for it to have a sizable impact on the economy.
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 Microsoft (Satya Nadella is the CEO of Microsoft). Holdings are subject to change at any time.