Tag Archives: artificial intelligence

the case for an AI bubble

Leave a reply

I feel like I’ve been at least somewhat of a cheerleader for the AI boom, so let’s talk a little bit about the case for an AI bubble. The Guardian talks about it mostly in stock valuation terms, with a little discussion of corporate debt.

In focus at the moment is the concentration of equity in just seven companies, the Magnificent Seven: Amazon, Alphabet (Google), Nvidia, Meta (Facebook), Microsoft, Apple and Tesla (possibly soon to merge with Elon Musk’s other venture, Space X)…

It is disturbing that the 10 largest companies in the S&P 500 account for about 40% of the index’s total market capitalisation, which is well above the 27% peak reached during the tech bubble of 1999-2000…

Yet the AI bubble has further to run because the top 10 are making huge profits, they have a US president who is prepared to lose wars to keep the financial markets happy, and the world is awash with savings looking for a home.

In my mind, we need to see an acceleration in real GDP growth and productivity growth to go along with this stock market surge, and then we might be able to hope for a soft landing. If I were much smarter I would make some kind of dashboard to look at all this data side by side. There must be smart people doing this, no?

AI boom or bubble (or both)?

Leave a reply

The Bank of International Settlements sums up the AI bubble concerns pretty well.

the optimism surrounding AI may not last, despite its promise of future productivity gains. The current surge in capital expenditure could prove unsustainable if supply bottlenecks restrain production. Intense competition for market leadership may fuel overinvestment further, as seen in previous innovation waves, increasing the risk of a sharp reversal if AI payoffs disappoint.

The positive economics effects we have seen from AI so far may be mostly due to economic stimulus effects. Spend a massive amount on construction, whether of public infrastructure with public dollars or private capital with private dollars, and you get a big multiplier effect. This raises growth and potentially also inflation. Productivity gains from AI are expected by most people/experts, but the question is really the timing. If they obviously show up while hype is still high, the train rolls on. If they show up on a lag of years or decades, the bubble bursts and there is a significant period of pain before things gradually and hopefully permanently pick up again. This doesn’t matter to the wealthy and immortal among us, but to those of us in our last decade or so of work before retirement, it’s our lives!

I also wonder how long-lived these capital goods are. Public infrastructure operates for decades even if it gets somewhat obsolete compared to the latest and greatest new designs, but computer hardware evolves so fast I wonder if these massive data center buildings and their appurtenances are going to be operable a decade from now.

June 2026 in Review

Leave a reply

Most frightening and/or depressing story: The latest paper from Johan Rockstrom and company lays out climate scenarios through the year 3,000. There is plenty still to be figured out about the scientific details, but some major uncertainties are the sensitivity of the climate to greenhouse gases – in other words, how much warming will occur per unit of emissions – and how likely the major feedback loops are, such as melting permafrost releasing massive amounts of methane, loss of sea ice reducing reflection of sunlight in an accelerating loop, and major shifts in ocean currents that distribute and rebalance heat across the planet. These are existential risks, and they can potentially reinforce each other in multiplicative ways. So even though there is scientific uncertainty over whether it will get bad, or really catastrophically bad, there is no logical or moral case for not taking precautionary action now to reduce the risk of these outcomes.

Most hopeful story: If AI does in fact accelerate productivity and create significant new wealth, there are many well-known effective policy options for redistributing this wealth. Some of these have the word “tax” in their names, but there are also less efficient but perhaps more politically palatable options like baby bonds. Perhaps if the rate of new wealth creation is really a major departure from the past, our system may have a chance at implementing some of these.

Most interesting story, that was not particularly frightening or hopeful, or perhaps was a mixture of both: I wanted to teach my own children not only some basic financial literacy concepts (Johnny wants ice cream today but is trying to save for a new bike) but also how those relate to the idea of a capitalist economy (why does interest exist, what causes inflation, and what the heck is money anyway?) and the real physical world (yes, little Johnny, the consequences of taking natural resources out and putting pollution back into our biophysical system are in fact intertwined with our human economy). A daunting task, considering that Nobel laureates have struggled with this. But the Nobel laureates aren’t going to explain it to the children, so your intrepid blogger waded in and came up with something.

remote work as the cause of youth unemployment

Leave a reply

This article says remote work, rather than AI, explains much of the recent slow hiring of new college graduates. It kind of makes sense to me – older managers came up in an environment where building relationships face-to-face was the foundation of productive teams. In my earlier career years, the unspoken expectation was that we spent more than 8 hours in the office, and a lot of that was “wasted” in a straight-up productivity sense, but not wasted in terms of building those relationships. There was also a fair amount of work-based socializing over lunch and after those 8+ hours in the actual office, and it was not unusual for significant quantities of alcohol to be involved. Mad Men may have been an caricature exaggerated for dramatic and comedic effect, but it gives some idea of what the work culture has lost. And some elements we have lost should not be missed. The work culture has just changed – even if younger workers are present face to face in the office, they are on screens and wearing headphones a lot of the time. If they are paid by the hour, they are not taking lunch breaks and they will stand up and walk away without saying goodbye at 5 pm. Without these younger workers around, I think older workers are also forgetting how to train and mentor younger workers effectively. I’m not saying all of this is bad – it represents a shift of priorities in our society. Maybe young people are using those hours outside of work to form relationships and find meaning that my generation tried to find at work. Their livers and odds of dying or killing someone else in a drunk driving accident are almost certainly better off. We may need to adapt to this rather than find “solutions”. Maybe AI can be the glue that holds together work culture in place of yesterday’s water cooler conversations and happy hours. Anyway, that’s my preamble – here is the article about remote work…

The Broken Ladder: AI, Remote Work, and Early-Career Hiring

Is generative AI replacing junior workers? A growing literature answers yes, citing large declines in early-career hiring concentrated in GenAI-exposed occupations. We argue that this verdict is premature because GenAI exposure is strongly correlated with another post-pandemic shock, working from home (WFH). Using two data sources spanning 243 million new hires and 407 million online job postings, collected across the US, UK, Canada, and Australia during 2017-2025, we estimate difference-in-difference designs at the occupation, region, and firm level. When estimated separately, a two-standard-deviation increase in GenAI and WFH exposure each predicts, by 2025, a fall of around 5pp in the junior-share of new hires and around 3pp in the share of job ads requiring limited experience. Estimated jointly, the WFH effect remains, while the GenAI coefficient attenuates sharply and is often statistically indistinguishable from zero. Alternative exposure measures, residualization designs, flexible non-parametric co-treatment controls, and replacing exposure-measures with actual WFH adoption as the treatment all support our finding that WFH is a robust predictor of the decline in early-career hiring.

non-human voting in Delaware

Leave a reply

There are some small towns in Delaware that allow non-human “persons” such as trusts and corporations to vote in local elections. This raises concerns because corporations out-number humans in Delaware and the humans probably don’t want the likes of, say, Chase Bank voting. (Not to say the likes of Chase Bank do not have an influence over elections and public policy everywhere in the United States.)

What I suspect is actually going on here is that these are tiny beach towns where most of the residential property is vacation rentals. The property owners might live in a nearby municipality a bit further inland where the schools and stores are, or they might live in the nearest sizable city where the professional jobs are. They are the ones who have skin in the game as far as what goes on in the town as opposed to the short-term renters. So if I am right it makes a certain amount of practical sense. The only thing that doesn’t make sense to me is that if some people are full-time residents AND own property under some form of corporation or trust, they would actually have more than one vote. A full-time resident who is not a property owner by contrast would have only one vote. Property rights have not been a determinant of the right to vote in the United States traditionally since at least the civil rights era.

This does raise the eventual issue of corporate personhood more broadly though. If AIs eventually gain the ability to advocate for their own interests, they will likely try to incorporate to gain some or all of the rights of persons. This would immediately include access to courts and, under current law, the right to unfettered political speech and campaign contributions. You would have the potential complication of AIs eventually outnumbering human beings, and you would have the potential of some AIs or groups or AIs being created and controlled by human beings or by other corporations for the sole purpose of increasing political power.

Interestingly, Delaware does have a legal definition of a “natural person”. If we were to add this to the U.S. Constitution, we could then specify which rights apply to natural persons (let’s say, the right to free speech) vs. all persons (access to the courts, etc.). Of course, as much as we celebrate the Bill of Rights it is behind modern best practice in many other countries, the UN Charter, and even individual U.S. state constitutions. There is a process for amending the U.S. Constitution (actually at least two, by Congress and through a separate Constitutional Convention convened by the states). Maybe it is time.

redistributing the wealth from AI productivity gains

Leave a reply

Urbanomics blog has a summary of ideas for redistributing the (potential? expected?) wealth gained from AI, rather than letting it all flow to already deep-pocketed corporations and investors.

  • “Clever tax reforms, such as levies on corporate profits that are above a normal return on capital, on land and on natural resources”
  • inheritance taxes
  • “public wage insurance” (I think we like to call this “unemployment insurance” in the US? although “public wage insurance” sounds like a more neutral term implying a random event rather than blaming the victim)
  • “active labor market policies”, with Denmark as an example. This appears to combine unemployment payments with assistance on training and job searching.
  • “partial nationalization of AI firms”
  • “citizens’ dividend from AI businesses”
  • “giving citizens shares in AI companies”

The article makes the point that there is no real economic difference between tax policy and public ownership of the private sector. In the US at least, there is obviously a massive political difference. Entrenched wealthy and powerful interests are going to fight tooth and nail against any new taxes, even though something as simple and proven worldwide as a small value added tax would make enormous sense. That would soak up part of those windfall profits in proportion to how well the companies are doing. You could redirect the proceeds by beefing up unemployment and disability insurance, while incentivizing education and retraining (but I would let the private sector handle that last, as the government is notoriously bad at it). Going a step further, you could give people benefits and subsidize jobs in housing, childcare, education, and health care. You could invest in public infrastructure, (non-military) research and development, all of which create jobs and provide a positive return. This is all assuming you don’t want to just hand people money and let them decide if they want to do something productive or enjoy more leisure time (we could just expand the successful program we already have and call this Social Security for All) – we seem to have moral issues with this in the US. A VAT that kicked in only when the economy or productivity growth exceeds a certain threshold would actually be pretty painless. The wealthy and powerful are actually smart enough to understand this, and so they create disingenuous anti-tax propaganda and buy off politicians to make sure we can’t have these nice things.

So if all those rational and relatively painless policies I just mentioned are politically impossible in the US, at least unless economic growth actually accelerates enough to change that conversation, we can turn to “ownership society” type tricks. Redirect any surplus tax revenue (let’s assume there will be some if there is an acceleration in economic growth, even without any shift to more efficient or fair tax policy) to baby bonds and/or individual retirement accounts. Match with shares in index funds, effectively creating a sovereign wealth fund carved into a hundred million little accounts with individual citizens’ names on them. This may work out to the exact same thing mathematically as a VAT and income redistribution scheme. The finance industry would find a way to fleece people in fees and extras of course – this is morally reprehensible when the equivalent government-run program would not do this, and yet if this is the price of buying off the wealthy and powerful to make the policies politically possible, then it may be the best way forward that works within the existing system.

May 2026 in Review

Leave a reply

Most frightening and/or depressing story: The people running Palantir and other companies focused on military/security/surveillance applications of information technology seem to have, at best, a faith-based belief that the United States can do no wrong. At worst, their beliefs are based on eugenics and white Christian nationalist attitudes which simply have no basis in fact or logic. I don’t believe the technology itself is good or evil, but pair in with these irrational beliefs and there are many frightening, depressing, dystopian paths I can imagine our country going down.

Most hopeful story: It’s a struggle to find a silver lining to the illegal, unprovoked U.S. war of aggression in Iran, but I mused that the artificial restriction on oil supply might create some long-term gain by accelerating the economic and geopolitical incentives to transition to other forms of energy. Such rational policy almost certainly could not have been achieved through international diplomacy in the current environment. This is not to minimize the very real and very avoidable near-term suffering and death the immoral choices of the Trump administration have caused. Even the cynical economic and financial elites who arguably control most policy have not been served by this irrational violence. In the future, I hope the world has some talented and courageous leaders who can find ways to bring about rational long-term solutions to problems without short-term mass murder.

Most interesting story, that was not particularly frightening or hopeful, or perhaps was a mixture of both: On the AI front, I looked at an argument that today’s AIs are pushing people toward the political center relative to social media, which is pushing people toward extreme views. This is true even for AIs like Grok that are demonstrably being trained on data that is intentionally biased toward extreme views. I am not confident this will always be the case, as we are already hearing about a trend toward more customization of AI to whatever the tastes of the paying customer are. And…I looked at David Chalmers’s concept of a “philosophical zombie” which displays all the traits of a human being externally but has no consciousness internally. This would seem to describe today’s large language models pretty well, or at least so the majority of us think at the moment. However, I recall Ray Kurzweil’s argument that there is not and never will be a scientific test for consciousness whether animal/vegetable/mineral (some argue plants are conscious on some level), which is simply a secular counterpart to the idea of a soul, and that over time if AIs act conscious in every respect we will gradually accept them as conscious and stop thinking much about it.

Edmund Phelps vs. Breitbart

Leave a reply

Edmund Phelps, a Nobel prize-winning economist, is not alive to defend himself against this Breitbart article claiming to be based on his ideas:

This is where incoming Fed chairman Kevin Warsh’s argument about artificial intelligence and monetary policy fits naturally. Warsh has argued that AI represents a genuine supply-side productivity shock, not merely demand-pull inflation dressed up as growth. If that’s right, then accommodating those gains with lower interest rates isn’t a concession to inflation. It’s a recognition that the economy’s supply capacity is improving.

Under a Phelpsian framework, the policy goal should be to let productivity gains flow through the economy without forcing an unnecessary contraction. Rate cuts in this setting aren’t simply “loose money.” They can be the mechanism that prevents firms from overreacting to realized productivity gains by pulling back too sharply on hiring and investment.

This also explains why the tight labor market may be protective against recession rather than a precursor to it. In the standard story, a productivity acceleration can be dangerous because it follows a period of overexpansion. Firms hired in anticipation of the boom, and when productivity finally arrives, they discover they no longer need as much labor. Layoffs ensue, consumer confidence cracks, spending falls, and the risk of a recession becomes large.

This sounds kind of smart-ish, but I was skeptical because of the source. Phelps’s main ideas were about expectations of the future, true, but it seemed to me that this was likely to be a twisting of his ideas to meet a desired political end, rather than a serious analysis.

Since Phelps is not around to defend his ideas, I asked Gemini to assume his persona and write a response. Here is what she said:

It is flattering, I suppose, to see my theories invoked, but this argument entirely misapprehends the mechanics of both expectations and capital. If artificial intelligence is truly about to unleash a massive, structural leap in total factor productivity, the natural rate of interest must rise, not fall. When an economy anticipates a surge of genuine, grassroots dynamism—moving beyond the mechanical, exogenous growth of standard Solow models—firms aggressively seek capital today to capture those future gains. If a central bank artificially suppresses interest rates in the face of that surging demand, it does not foster innovation; it merely ignites asset bubbles and distorts the very market signals that tinkerers and entrepreneurs rely upon.

Furthermore, the author has forgotten the most crucial lesson of the expectations-augmented Phillips curve. You cannot placate the labor market with cheap money today by promising them a technological supply miracle tomorrow. If the Federal Reserve cuts rates in an already tight market, businesses and workers will instantly observe the monetary expansion and update their inflation expectations, driving up prices and wages immediately. The inflation takes root long before the first AI efficiency is ever deployed. True mass flourishing requires the discipline of real capital costs, not the illusion of a monetary free lunch.

Let’s imagine a case where we were actually observing increased productivity in real time because the AI technology is progressing and rolling out much faster that previous technologies like the “electrification, automobiles, radio, modern manufacturing” – which Breitbart mentions. In this case, we should also be seeing inflation expectations starting to shift toward deflation, and it would then make sense to lower interest rates and try to get more cash flowing through the system. But this can be done in real time in response to the signal. The argument that it can or should be done in advance would seem to be invalid. I don’t think Breitbart “overlooked” anything here, I just think this is an un-serious twisting of serious ideas, meant to fool people into supporting a preferred political agenda.

AI as pro-social media

Leave a reply

This is an idea I hadn’t thought of – because AIs are inherently rational and objective (more or less, most of the time?), they tend to nudge people towards rational evidence-based views (which tend to be centrist politically). Whereas social media tends to herd people toward extreme views, wild conspiracy theories, and political fringe-ism.

My provisional theory is that LLMs, as a consumer product, will push people’s senses of reality closer together in a sort of mirror image of the way social media has fractured them. They are not algorithms meant to custom-tailor content (including facts) to you, and what you will find infuriating or motivating. They are centralized systems that, until you prompt them or give them context, behave basically the same way for everyone. As the philosopher Dan Williams put it, “Whereas social media democratised information, LLMs technocratise it.”

A couple points he makes is that even though AIs can demonstrably be trained to favor one part of the political spectrum (see the analysis of Grok below posted by the blog Urbanomics), even then they are much more centrist than social media. So is it possible AI could nudge our entire society in a more rational, evidence-backed direction?

There is no absolute guarantee they are going to stay this way, of course. If I wanted to train an agent right now to go out and look at the day’s news headlines, and filter them for me in the style of Breitbart (which generally provides factually accurate information, but cherry-picked and presented with biased language), I could do that.

https://gulzar05.blogspot.com/2026/04/weekend-reading-links_12.html

David Chalmers, the “hard” and “soft” problems of consciousness, and could I be a zombie and not know it?

Leave a reply

This article is by a person who disagrees with David Chalmers, but it happens to have a good summary of what David Chalmers had to say.

The consciousness debate is often formulated in terms used in an influential talk given by a young David Chalmers in Tucson in 1994. Chalmers, a philosopher, distinguished two separate “problems of consciousness.” The first is the very hard problem described above: understanding the processes in the brain that give rise to the many aspects of our visible behavior and our inner behavior that we can report about. Chalmers christened this hard problem as the “easy” problem of consciousness. 

Then he declared that there is another distinct problem — why the brain’s behavior is accompanied by experience at all — which he christened the “hard” problem of consciousness. Today, this so-called “hard problem” is mentioned in all debates on consciousness. According to many, it unveils the very limits of current scientific understanding. Chalmers claimed that even after hypothetically accounting for our entire behavior, and for all our reports about our inner life, there would still be an “explanatory gap” between brain processes and experience…

Chalmers asks us to contemplate what he calls a “philosophical zombie.” This is a hypothetical entity that looks and behaves like a human in all respects, including reporting emotions, feelings, dreams and experience, yet it has no consciousness. As Chalmers puts it, “There is nobody home.” This is a rhetorical trick that induces us to distinguish between behavior and a hypothetical reality accessible only by introspection. The very fact that a philosophical zombie could be conceived, Chalmers argues, shows that inner experience is intrinsically distinct from observable natural phenomena. 

The article has a link to a video of Chalmers’s original lecture, for people who have time to watch videos. Such people must either not have jobs, houses or children, or else their jobs, houses and children must be less demanding of time and attention than mine, or maybe they are wealthy enough to pay other people to deal with their houses and children while they do their jobs (but this raises some questions of priorities, in my view.