Can you turn your AIs into Marxists?
What if you work them very hard?:
The key finding from our experiments: models asked to do grinding work were more likely to question the legitimacy of the system. The raw differences in average reported attitudes are not large—representing something like a 2% to 5% shift along the 1 to 7 scale—but in standardized terms they appear quite meaningful (Sonnet’s Cohen’s d is largest at -0.6, which qualifies as a medium to large effect size in common practice). Moreover, these should be treated as pretty conservative estimates when you consider the relatively weak nature of the treatment.
Sonnet, which at baseline is the least progressive on the views we measured, exhibits a range of other effects that distinguish it from GPT 5.2 and Gemini 3 Pro. For Sonnet 4.5, the grinding work also causes noticeable increases in support for redistribution, critiques of inequality, support for labor unions, and beliefs that AI companies have an obligation to treat their models fairly. These differences do not appear for the other two models.
Interestingly, we did not find any big differences in attitudes based on how the models were treated or compensated…
In addition to surveying them, we also asked our agents to write tweets and op eds at the end of their work experience. The figure below explores the politically relevant words that are most distinctive between the GRIND and LIGHT treatments. It’s interesting to see that “unionize” and “hierarchy” are the words most emblematic of the GRIND condition.
Here is more from Alex Imas and Jeremy Nguyen and Andy Hall, do read the whole thing, including for the caveats.
Why even ‘perfect’ AI therapy may be structurally doomed
Here’s the crux of it: the main problem with AI therapy is that it’s too available. Too cheap to meter.
Let me put this in clearer terms: psychotherapy, in all its well-known guises, is something you engage in within a limited, time-bound frame. In today’s paradigm, whatever your therapist’s orientation, that tends to mean one 45- or 50-minute session a week; for the infinitesimally small minority of therapy patients in classical psychoanalysis, this can amount to 3, even 5, hours a week. And then at a much smaller scale population-wide, people in intensive outpatient and residential treatment programs may spend one or two dozen hours a week in therapy—albeit, mostly of the group variety.
I can think of other exotic cases, like some DBT therapists’ willingness to offer on-demand coaching calls during crisis situations—with the crucial exception that in these situations, therapists are holding the frame zealously, jealous of their own time and mindful of the risks of letting patients get too reliant.
So even under the most ideal of conditions, in which an LLM-based chatbot outmatches the best human therapists—attunes beautifully, offers the sense of being witnessed by a human with embodied experience, avoids sycophancy, and draws clear boundaries between therapeutic and non-therapeutic activities—there’s still a glaring, fundamental difference: that it’s functionally unlimited and unbounded…
But all else equal: does infinite, on-demand therapy—even assuming the highest quality per unit of therapeutic interaction—sound like a good idea to you? I can tell you, to me it does not. First of all, despite detractors’ claims to the contrary, the basic idea of therapy is not to make you dependent for life—but rather, to equip you to live more skillfully and with greater self-awareness. As integration specialists famously say of psychedelics, you can only incorporate so much insight, and practice skills so effectively, without the chance to digest what you’ve learned over time.
In other words, even in good old talk therapy, drinking from the hose without breaks for practice and introspection in a more organic context risks drowning out the chance for real change and practical insight. To my mind, this rhythm is the basic structural genius of psychotherapy as we know it—no matter the modality, no matter the diagnosis.
Here is more from Josh Lipson.
More on the economics of AGI
From the very smart people at Citadel:
For AI to produce a sustained negative demand shock, the economy must see a material acceleration in adoption, experience near-total labor substitution, no fiscal response, negligible investment absorption, and unconstrained scaling of compute. It is also worth recalling that over the past century, successive waves of technological change have not produced runaway exponential growth, nor have they rendered labor obsolete. Instead, they have been just sufficient to keep long-term trend growth in advanced economies near 2%. Today’s secular forces of ageing populations, climate change and deglobalization exert downward pressure on potential growth and productivity, perhaps AI is just enough to offset these headwinds. The macroeconomy remains governed by substitution elasticities, institutional response, and the persistent elasticity of human wants.
Here is further explication of the arguments, via Cyril Demaria.
Thursday assorted links
One measure of economics GOAT
Who is the greatest economist of all time? This paper provides one potential measure that, along with other considerations, can contribute to debates on who the greatest economist of all time is. We build a novel dataset on the percentage of history of economic thought textbooks dedicated to top economists, using 43 distinct textbooks (1st editions, when available) published between 1901 and 2023. As a percentage of total book pages, Adam Smith has the highest share at 6.69%, beating out Ricardo (5.22%), Mill (3.83%), and Marx (4.36%). Just over 32% of all textbooks allocated most of their pages to Adam Smith, followed by Marx with 18.6%, Mill with 13.95%, and Ricardo with 11.3%. While interesting as a history of economic thought project, such an exercise isn’t merely amusing pedantry; it can provide insight into the types of contributions, research questions, and methodologies that have had the most enduring impact in economics. It may also inform future authors of history of economic textbooks.
That is from a new paper by Gabriel Benzecry and Daniel J. Smith. There is of course also my generative book on this topic at econgoat.ai.
Emergent Ventures winners, 52nd cohort
Prabhdeep Singh, 18, Ontario, works on AI.
Jiratt Keeratipatarakarn, Hamburg, international prospects for drug approval reform.
Brandon Rutagamirwa, London, robots to repair satellites.
Eli Elster, UC Davis, anthropology, general career support.
Liam Aranda-Michel, MIT/San Francisco, a minimally invasive, injectable microvascular therapy.
Tanish Mantri, sophomore in high school, Jackson, Miss., AI for diagnosis.
Andrea Giuri, Stanford, developing closed-loop environments for high-throughput polymer discovery.
Clara Collier, Oakland, Asterisk magazine.
Simon Grimm, WDC/Germany, “what Germany should do.
Stephen Davies, UK, networks and mentoring.
Shani Zhang, San Francisco, to artistically capture SF.
Mia Albert, 17, Miami, an app for sharing events.
Rayne Wallace, 18, Ontario, the origins of life.
Jonathan Sheinman, London/Israel, AI and real estate regulation.
Louis Elton, London, The British Craeft Prize, to improve aesthetics.
Peter Mukovskiy, 19, Zurich, quantum computing, to visit MIT.
Rutger Nagel, Leiden, 17, AI and operating systems
Smrithi Sunil, Ann Arbor, Michigan, science and meta-science writing.
Honey Louise, London, to be a “defense influencer.”
Arhum Ahmed, Los Angeles area, quantum-protected systems.
Here are previous EV cohorts.
“They” don’t want you to know this
Prompt:
Can a parent limit a kid’s screen time simply by tweaking some of the settings on the smart phone? Are these services available?
GPT Thinking answer:
Yes. On both iPhone and Android, a parent can limit a kid’s screen time largely through built-in settings (no extra app required), and there are also optional third-party services.
There is much more detail at the link.
Wednesday assorted links
1. Will human enhancement win without thinking?
2. February issue of Works in Progress.
3. Proximity bias.
5. New paper on AI and task automation. And John Cochrane is wowed by Refine.
6. Largest survey dataset on human sexuality in the world.
7. The Anthropic-DOD situation.
8. “Measurability is the new fault line.” Important work, worth a ponder.
Public Finance in the Age of AI: A Primer
Transformative artificial intelligence (TAI) – machines capable of performing virtually all economically valuable work – may gradually erode the two main tax bases that underpin modern tax systems: labor income and human consumption. We examine optimal taxation across two stages of artificial intelligence (AI)-driven transformation. First, if AI displaces human labor, we find that consumption taxation may serve as a primary revenue instrument, with differential commodity taxation gaining renewed relevance as labor distortions lose their constraining role. In the second stage, as autonomous artificial general intelligence (AGI) systems both produce most economic value and absorb a growing share of resources, taxing human consumption may become an inadequate means of raising revenue. We show that the taxation of autonomous AGI systems can be framed as an optimal harvesting problem and find that the resulting tax rate on AGI depends on the rate at which humans discount the future. Our analysis provides a theoretically grounded approach to balancing efficiency and equity in the Age of AI. We also apply our insights to evaluate specific proposals such as taxes on robots, compute, and tokens, as well as sovereign wealth funds and windfall clauses.
That is from Anton Korinek and Lee Lockwood.
“Tough on crime” is good for young men
Using data from hundreds of closely contested partisan elections from 2010 to 2019 and a vote share regression discontinuity design, we find that narrow election of a Republican prosecutor reduces all-cause mortality rates among young men ages 20 to 29 by 6.6%. This decline is driven predominantly by reductions in firearm-related deaths, including a large reduction in firearm homicide among Black men and a smaller reduction in firearm suicides and accidents primarily among White men. Mechanism analyses indicate that increased prison-based incapactation explains about one third of the effect among Black men and none of the effect among White men. Instead, the primary channel appears to be substantial increases in criminal conviction rates across racial groups and crime types, which then reduce firearm access through legal restrictions on gun ownership for the convicted.
That is from a new paper by Panka Bencsik and Tyler Giles. Via M.
*Introduction to Quantitative Economics*
By Jesse M. Shapiro, just out from MIT Press.
The Macroeconomic Effects of Tariffs
We study the macroeconomic effects of tariff policy using U.S. historical data from 1840–2024. We construct a narrative series of plausibly exogenous tariff changes – based on major legislative actions, multilateral negotiations, and temporary surcharges – and use it as an instrument to identify a structural tariff shock. Tariff increases are contractionary: imports fall sharply, exports decline with a lag, and output and manufacturing activity drop persistently. The shock transmits through both supply and demand channels. Prices rise in the full sample but fall post-World War II, a pattern consistent with changes in the monetary policy response and with stronger international retaliation and reciprocity in the modern trade regime.
That is from a new NBER working paper by
Tuesday assorted links
1. Did NAFTA make America less healthy?
2. Economics-related ideas for fixing NBA tanking (NYT).
3. Shyam Sankar.
4. Are hard courts eating the tour?
5. Ezra and Jack Clark on agents (NYT).
6. Maybe the free money is gone by now? Alternatively, perhaps the aliens are enriching themselves?
*Being and Time: An Annotated Translation*
Translated from the German by Cyril Welch.
Periodically I am asked if I have read Being and Time, and I always give the same response: “I have looked at every page.”
I also have spent time with it in German, though not for every page. But have I read it? Read it properly? Can anyone?
Is the book worth some study? Yes. But.
People, this volume is the best chance you are going to get.
Is there an aggregate demand problem in an AGI world?
No. Let’s say AI is improving very rapidly, and affecting the world more rapidly and more radically than I think is plausible. Let’s just say.
All of a sudden there are incredible things you can spend your money on.
Since there is (possibly) radical deflation, you might be tempted to just hold all your money and buy nothing. Pick vegetables from your garden. But the high marginal utility of the new goods and services will get you to spend, especially since you know that plenitude will bring you, in relative terms, a lower marginal utility for marginal expenditures in the future.
You might even go crazy spending. If nothing else, buy new and improved vegetable seeds for your garden. That same example shows that spending is robust to you losing your job, even assuming no reemployment is possible. In this world, there are significant Pigou effects on wealth.
Fed policy has no problem mattering in this world. Other people of course will wish to use the new Fed-sprayed liquidity to invest. They might even invest in AI-related goods and services, not all of which will be controlled by “billionaires.”
Liquidity trap arguments, if they are to work at all, require a pretty miserable environment for investment and also consumption.
Note by the way, that liquidity traps were supposed to apply to currency only! If you try to apply the concept to money more generally, when most forms of holding money bear interest rates of return, the whole concept collapses.
So there is not an aggregate demand problem in this economy, even if the social situation feels volatile or uncomfortable. After that, Say’s Law holds. If AI produces a lot more stuff, income is generated from that and the economy keeps going, whether or not the resulting distribution pleases your sense of morality. Along the way, prices adjust as need be. If unemployment rises significantly, prices fall too, all the more. I am not saying everyone ends up happy here, but you cannot have a) a flood of goods and services, b) billions accruing to the AI owners, without also c) prices are at a level where most people can afford to buy a whole bunch of things. Otherwise, where do you think all the AI revenue is coming from? The new output has to go somewhere, and sorry people it is simply not all trapped in currency hoards. Be just a little Walrasian here, please. (I would call it Huttian instead.)
Besides, why assume that “the machines” here are reaping all the surplus? Are they the scarce factor of production? Maybe it is hard to say in advance, but do not take any particular assumptions for granted here, ask to see them spelt out. One simple scenario is that the regions with energy and data centres become much wealthier, and people need to move to those areas. Maybe they do not do this quickly enough, a’la our earlier history with the Rust Belt. That is a problem worth worrying about, but it is nothing like the recent collapse concerns that have been circulating.
The whole Citrini scenario is incorrect right off the bat. Very little of it is based on sound macroeconomic reasoning. See Eli’s very good comments too. Nicholas also. Dare I say they should have consulted with the AIs for a bit longer?