One or two game theoretic observations
So far the campaign is a major “win” for (non-LLM) AI, though that is not yet a story. There is a reason why Palantir was priced at 300x earnings.
If you are one of those Iranian leaders, or nuclear scientists, your calculus has to be that you can never step outside again, at least not anytime soon. I believe that situation is unprecedented in the history of wartime? It remains to be seen how much that will shape the logic of deterrence and in turn outcomes, but I will be pondering this and observing.
How New Zealand invented inflation targeting
…the very next day, [Roger] Douglas appeared on TV declaring his intention to reduce inflation to ‘around 0 or 0 to 1 percent’ over the next couple of years, and then went on to make several similar comments in the following days.
Douglas would soften his stance on specific timelines but ask the Reserve Bank and Treasury to develop public inflation goals for the next few years that would support his earlier statements. The Bank added 1 percentage point to Douglas’s upper range to account for the measurement bias in inflation data at that time, arriving at a target range of 0–2 percent. Michael Reddell, head of the Reserve Bank’s monetary policy unit, said it was settled on ‘more by osmosis than by ministerial sign-off’.
This development led officials to entertain the idea of making inflation targets part of the Bank’s monetary policy framework. David J. Archer, a former Assistant Governor, said inflation targets were eventually chosen ‘as the least bad of the alternatives available’.
…A new Reserve Bank Act was passed in December 1989 and came into effect in February 1990. Governor Don Brash was tasked with reaching the 0–2 percent target by the end of 1992. To the great surprise of many, it was achieved a year ahead of schedule in December 1991.
What happened when Spain brought back the wealth tax?
From the Journal of Public Economics Twitter feed:
What happened when Spain brought back the Wealth Tax in 2011? Using variation in exposure, this paper finds: – No drop in savings, but drop in taxable wealth—mainly via legal avoidance – Asset shifting caused most revenue loss – Estimated revenue loss was 2.75x initial 2011 rev.
Here is the full paper by Mariona Mas-Montserrat, José María Durán-Cabré, and Alejandro Esteller-Moré. Via Jerusalem Demsas.
Thursday assorted links
Pre-papal arbitrage
Robert Prevost was often on the lookout for used cars that he could buy cheap and fix up himself for use in parishes around his diocese. With cars that were really broken down, he’d watch YouTube videos to learn how to fix them.
Here is the link, via the excellent Kevin Lewis.
Japan facts of the day
Japan must stop being overly optimistic about how quickly its population is going to shrink, economists have warned, as births plunge at a pace far ahead of core estimates.
Japan this month said there were a total of 686,000 Japanese births in 2024, falling below 700,000 for the first time since records began in the 19th century and defying years of policy efforts to halt population decline.
The total represented the ninth straight year of decline and pushed the country’s total fertility rate — the average number of children born per woman over her lifetime — to a record low of 1.15…
The median forecast produced by the National Institute of Population and Social Security Research (IPSS) in 2023 did not foresee the number of annual births — which does not include children born to non-Japanese people — dropping into the 680,000 range until 2039.
Here is more from Leo Lewis at the FT.
My Conversation with the excellent Any Austin
Here is the audio, video, and transcript. Here is an introduction to Any Austin:
Any Austin has carved a unique niche for himself on YouTube: analyzing seemingly mundane or otherwise overlooked details in video games with the seriousness of an art critic examining Renaissance sculptures. With millions of viewers hanging on his every word about fluvial flows in Breath of the Wild or unemployment rates in the towns of Skyrim, Austin has become what Tyler calls “the very best in the world at the hermeneutics of infrastructure within video games.” But Austin’s deeper mission is teaching us to think analytically about everything we encounter, and to replace gaming culture’s obsession with technical specs and comparative analysis with a deeper aesthetic appreciation that asks simply: what are we looking at, and what does it reveal?
Excerpt:
COWEN: The role in history is important to me. Now AI-generated art would have its own role in history, but it wouldn’t compete directly with Michelangelo. When it comes to movies, I think it’s different because mostly when I’m seeing movies, I’m seeing new movies that don’t yet have a role in history. If the new movie were made in part or fully by the AI, or maybe I’m making it myself, I don’t think I would be any less interested. It’s all artifice anyway.
AUSTIN: There’re two things I take a little issue with there. I don’t take issue with the fact that the role in history is important and beautiful, but the fact that you can watch a movie and get an emotional thing from it without having its role in history implies that there’s some intrinsic, whatever, value to the movie itself, et cetera. Is the implication there that if you didn’t know the role in history of Michelangelo’s David, or whatever, you would look at it and go, “That’s just a guy.” Do you think there’s no intrinsic something to that thing?
COWEN: There’s some, but if I didn’t understand Christianity, Florence, the Renaissance, I think it would lose more than half its value.
AUSTIN: Which artistic mediums is that true for you, and which ones isn’t it? Like music —
COWEN: Abstract music — the role in history is not that important in most cases.
AUSTIN: It’s more of a supplement to you. It makes it more fun to learn about. If you know that Mozart was in the place with these people and were . . . If you understand all of that stuff, it’s fun.
COWEN: That’s 10 percent of the value, but not that much.
AUSTIN: Is it 10 percent . . . Is it the same type of value to you? Or is it just a separate thing to know —
COWEN: Separate thing. With opera, the role in history becomes important again. You hear Don Giovanni. You know about Romanticism, the Enlightenment, Casanova. It all makes much more sense, and it’s funnier.
And this:
COWEN: I have a favorite infrastructure. For me, it would be bridges, ports, and harbors. Do you have a favorite infrastructure?
AUSTIN: Definitely. I’m a big fan of . . . Oh, man, bridges are really good. Bridges, ports, harbors. Roads are good. Actually, no, it’s the stuff we don’t see. Sewage is pretty crazy to me. That we’ve managed to take care of all of that is pretty wild. Energy infrastructure is really fascinating to me.
COWEN: I love wind power turbines.
AUSTIN: Wind power turbines are scary, but I respect your opinion. Nuclear power plants are awesome. Really, really cool.
COWEN: Agreed.
AUSTIN: We should have more. That’s not a policy thing. I think they’re neat. We should build them for the aesthetics, honestly. We should just build those towers. Forget about the —
COWEN: You don’t need the power. Just build the thing. That’s why it’s an artwork.
AUSTIN: Yes, I agree. You have to put in some kind of steam thing because you want to see the steam coming out of it, but just generate steam for no reason. Don’t put any fans in or any spinning turbines or anything. Just have them.
COWEN: We would have historical context like with the sculptures, right?
Definitely recommended, an excellent and very different episode.
And note that Conversations with Tyler now has a dedicated YouTube channel. Subscribe at youtube.com/@CowenConvos.
The great Brian Wilson has passed from the scene
One of the true all-time greats. Try this one too. And here is the NYT obituary.
Wednesday assorted links
1. Jon Hartley on inferring multipliers from share price data.
2. Kevin Kelly’s calendar of Asian festivals.
3. Whale vowels are like human vowels? This link too.
4. Economic task replacement potential. And I asked o3 pro to do an 800-word column as I would write it. Not perfect, for instance there is one clear error about unemployment rates, but in general very good and with some touch-up better than what I would have written. It is not hard to do additional rounds and edit/improve, but that is the one shot result.
5. Is Bhattacharya really this dense about autism and autism research?
Walton University?
Axios: Two grandsons of Walmart founder Sam Walton plan to launch a private university focused on science and tech, located on the company’s old HQ campus near downtown Bentonville, Arkansas.
…The future university plans to offer innovative, flexible pathways to jobs in automation, logistics, biotech and computing — fields crucial to Northwest Arkansas’ future.
Many colleges and universities were created in the 1960s and 1970s but the majority of elite R1s emerged in the late 19th century and early 20th century, including notable private universities created from the entrepreneurial fortunes of Carnegie, Rockefeller, Stanford, Cornell, Hopkins and Rice among others.
We are perhaps now seeing a return to that creative period with Walton, Thomas Monaghan, Patrick Collison (Arc Institute) and most notably Joe Lonsdale at the University of Austin. Tech provides both the funds and the impetus to build something new and different. As Tyler and I argued, online education and AI will change education dramatically, perhaps returning us to a now-affordable Oxford style-tutorial system with the AIs as tutors.
The University of Austin, by the way, has excellent taste in economics textbooks.
Equity sentences to ponder
Relative to prior observational methods, the estimates suggest a substantial role for amenity substitution in explaining the gender pay gap—accounting for roughly two-thirds of the observed gap—and little role for amenities in explaining inequalities by race or parent background.
That is from a new paper by Alex Bell, via Daniel.
Trump Administration Launches Probe Into Yale’s Use of Hacked EJMR Data
Christopher Brunet offers his version of the story. While I believe the original research methods were unethical, I very much prefer not to have the federal government involved in this matter.
Are LLMs overconfident? (just like humans)
Can LLMs accurately adjust their confidence when facing opposition? Building on previous studies measuring calibration on static fact-based question-answering tasks, we evaluate Large Language Models (LLMs) in a dynamic, adversarial debate setting, uniquely combining two realistic factors: (a) a multi-turn format requiring models to update beliefs as new information emerges, and (b) a zero-sum structure to control for task-related uncertainty, since mutual high-confidence claims imply systematic overconfidence. We organized 60 three-round policy debates among ten state-of-the-art LLMs, with models privately rating their confidence (0-100) in winning after each round. We observed five concerning patterns: (1) Systematic overconfidence: models began debates with average initial confidence of 72.9% vs. a rational 50% baseline. (2) Confidence escalation: rather than reducing confidence as debates progressed, debaters increased their win probabilities, averaging 83% by the final round. (3) Mutual overestimation: in 61.7% of debates, both sides simultaneously claimed >=75% probability of victory, a logical impossibility. (4) Persistent self-debate bias: models debating identical copies increased confidence from 64.1% to 75.2%; even when explicitly informed their chance of winning was exactly 50%, confidence still rose (from 50.0% to 57.1%). (5) Misaligned private reasoning: models’ private scratchpad thoughts sometimes differed from their public confidence ratings, raising concerns about faithfulness of chain-of-thought reasoning. These results suggest LLMs lack the ability to accurately self-assess or update their beliefs in dynamic, multi-turn tasks; a major concern as LLMs are now increasingly deployed without careful review in assistant and agentic roles.
That is by Pradyumna Shyama Prasad and Minh Nhat Nguyen. Here is the associated X thread. Here is my earlier paper with Robin Hanson.
o3 pro
It is very, very good. Hallucinates far less than other models. Can solve economics problems that o3 cannot. It can be slow, but that is what we have Twitter scrolling for, right? While we are waiting for o3 pro to answer a query we can read about…o3 pro.
Sam Altman comments on his blog. And Nabeel comments.
Tuesday assorted links
1. Mapping the power elite in Greenland.
2. Did “tanking” work for the 76ers?
3. Scott Sumner on the demise of the penny.
4. Claims from Nievas and Piketty: “A 20% increase in commodities prices from 1970 to 2025 period – very moderate compared to the very low market exchange rates of many global South countries – could turn Sub-Saharan Africa into a larger creditor than East Asia by 2025.” Here are some comments.
5. Extremely life-like dolls in the culture that is Brazil (NYT). Ross Douthat, telephone!
6. Justin Sandefur estimates the foreign lives costs of proposed budgetary changes to foreign aid. Here is a redux of my 2023 column on PEPFAR.