Category: Web/Tech
Another use of AI in research (from my email)
“Another thing we (John [Horton] and I) have thought about is having a swarm of AIs “fight” over a literature. They could take the cumulative datasets available and continuously argue until they understand the question. One line of thinking says they reach a stalemate (as scientists currently do). But we think not. More likely, they push evidentiary understanding to the limit and coalesce around what’s most probable — if not definitive!”
That is from Benjamin Manning.
The interstate trade effects of autonomous trucks
Recent advances in autonomous and semi-autonomous vehicle technologies promise substantial cost savings for goods shipped by truck. In this study, we quantify the impacts of these transport cost reductions on the US interstate trade using a structural gravity model of domestic trade. Based on projected cost savings from the widespread adoption of self-driving technologies, we estimate significant increases in total interstate trade value. State-level impacts vary from 40.3% of GDP in Mississippi to 5.9% in Florida, while the largest impacts in dollar value are observed in Texas and New York. The sectoral analysis highlights motorized vehicles, mixed freight, and electronics as the industries experiencing the largest trade value growth. Additionally, goods with low value-to-weight ratios—where shipping costs represent a large share of the delivered value—are expected to benefit most in relative terms. These findings underscore the transformative potential of autonomous vehicle technologies in reshaping US trade patterns and sectoral dynamics.
That is from a recent paper by Taejun Mo, et.al., via the excellent Kevin Lewis.
Will AI kill the research paper?
Imagine taking a macroeconomics paper and adding a little button at the end “Press this button to update this paper with the latest macro data.”
All of a sudden you have multiple papers rather than one, and no single canonical version. It is the latter versions, not created directly by the authors, that people will look at.
Imagine adding another button, to either micro or macro papers “Please rerun these results using what the AI thinks might be five other different yet still plausible specifications.”
Then you have more papers yet.
Ultimately, why not just build a “meta-paper,” using AI, to answer any possible question about the subject area under consideration. This meta-paper would allow the reader, using AI, to make many sorts of modifications and additions to the basic work. The meta-paper also would allow the reader to add new data, to run additional robustness checks, and to do whatever else you might think of. Once again, the canonical version of the paper evolves away.
A researcher might spend a significant part of his or her career building such a meta-paper. Imagine a meta-paper, or sometimes I call it a “box,” devoted to answering questions about say fiscal policy, minimum wage hikes, or maybe the Industrial Revolution. Fed researchers would spend their entire careers, not writing papers, but improving the Fed’s “box” that answers questions about monetary policy and also prudential supervision.
Who will be good at doing such things? Is it the people today who become the top economists, or not? Will it be a highly decentralized endeavor, or, given the compute and team work requirements, a highly centralized one?
Economics is going to change a lot, as will many of the other sciences.
It is funny, and tragic, how much some of you are still obsessed with writing and publishing “papers.”
Self-fulfilling misalignment?
We started by investigating why Claude chose to blackmail. We believe the original source of the behavior was internet text that portrays AI as evil and interested in self-preservation.
And here is Alex Turner on the topic of self-fulfilling misalignment. I raised this possibility some while ago in a Free Press column, and mainly was met with hostility.
The social return to a positive world view, and avoiding negative emotional contagion, never has been higher.
The social media ban in Australia, how is it going?
In December 2025, Australia became the first country to ban youth under 16 years old from holding accounts on major social media platforms, a policy now under consideration in more than a dozen countries and in numerous states. Because social media use is inherently social, the effectiveness of a ban that is easy to circumvent may depend on whether compliance reaches a tipping point: a share of compliant peers high enough to make it optimal for individuals to comply themselves. We surveyed 835 Australian teenagers four months after the ban took effect and find that only about one in four 14–15-year-olds comply. The social environment around use has barely moved: most banned teens believe that their peers are still using banned platforms and cite social reasons for continuing use. Sustaining high compliance requires two ingredients: the share of compliers must be high enough and those who comply must find it preferable to continue complying. The current ban achieves neither. Teenagers report that they require roughly two-thirds of peers to stop using social media to stop themselves, far above the share currently complying. They also perceive compliers as less popular than non-compliers, so the more influential teens disproportionately stay on the platforms. Together, these patterns suggest that compliance is more likely to diminish than to rise. Sustaining higher compliance will likely require pairing the ban with instruments that act on social norms and individual incentives directly.
That is from a new NBER working paper by
A few days ago I was talking with a very smart fifteen year old in Australia (really). He was of the opinion that it was quite ineffective, though he noted he could no longer access LinkedIn. I would note there are more stringent measures, requiring more governmental monitoring and control of the internet, that perhaps could have a greater effect.
Do Americans really hate AI?
We might be heading towards a populist backlash towards AI, but we’re not there yet. Outside the tech bubble, Americans really don’t care about AI yet.
AI is Americans’ 29th most important issue, according to the fantastic survey @davidshor ran that everyone is rightly looking at.
It’s not surprising that Americans will answer sentiment questions about AI negatively, as they’ve been negative towards tech for a while. But it’s a big leap from negative sentiment to meaningful political action.
Americans have been negative on social media for 10 years, and there has been no meaningful political action. And that’s despite all the other hallmarks of backlash people are saying about AI—violent extremists (people forget there was a shooting at YouTube HQ), protests, etc.
My prediction: we will get real populist backlash to AI when the unemployment moves by, say, 2 percentage points and people see it as caused by AI.
That is part of a longer tweet from Andy Hall.
AGI Could Lower Interest Rates
Standard models predict that expectations of artificial general intelligence (AGI) should elevate long-term interest rates. I show that this prediction need not hold. I develop a heterogeneous-agent asset pricing model in which AGI, or more broadly, transformative AI (TAI) capable of automating most human labor, can lower interest rates even as it dramatically accelerates growth. Under baseline calibrations, the risk-free rate falls to near zero despite growth rising from 2% to 11%, and the equity premium expands from 6% to over 20%. The effect on yields is negative and muted for all maturities, even under aggressive assumptions about the speed of AI adoption. These results advise caution when interpreting long-term bond yields as a signal of market expectations of transformative AI.
That is from a new paper by Caleb Maresca of NYU. Via the excellent Kevin Lewis.
The best study to date on school phone bans
Schools across the U.S. have sharply restricted student use of phones during the school day. We evaluate one type of restriction—lockable phone pouches—using nationwide data combining large-scale surveys, GPS pings, standardized test scores, and school administrative records, along with sales records from the largest pouch provider. Using a staggered difference-in-differences design, we find that pouch adoption substantially reduces phone use as measured by GPS pings and teacher reports. In the first year after adoption, disciplinary incidents increase and student subjective well-being falls, consistent with short-term disruption. However, effects on well-being become positive in later years and disciplinary effects fade. For academic achievement, average effects on test scores are consistently close to zero. High schools see modest positive effects, particularly in math, while middle schools see small negative effects. We find little evidence of effects on school attendance, self-reported classroom attention, or perceived online bullying.
In sum, it is fine to want to run a school that way, but do not expect huge educational gains, if any. The evidence on this is accumulating, but many seem unable to accept the results. In any case it is not worthy of a major moral crusade.
Here is the NBER working paper, with top-tier researchers involved I might add, namely .
Have online worlds become the last free places for children?
Major public intellectuals and politicians have responded by arguing that children should rarely, if ever, participate in digital spaces. As a result, many schools in the US now demand that students seal their smartphones in magnetic pouches. A number of countries, including Australia, the United Kingdom and France, are even considering or have already implemented bans on social media accounts for children and teenagers.
Such restrictions, however, are not the tools of liberation we may imagine them to be.
In fact, for some children, the internet may be one of the last remaining spaces where they can grow up doing what children everywhere have evolved to do: independently play and explore with their peers.
Here is more from anthropologist Eli Stark-Elster. I would add a point. I do accept the evidence suggesting that limiting or banning cell phones in schools brings marginally better academic results. Yet the people who advocate such policies never point out that so many schools are just deadly dull and not very intellectually stimulating? Often what is on the phone is in fact more interesting and sometimes more instructive as well, even if the students do worse in terms of the standards set by the school.
The collapse of teen fertility in the digital era
Teen fertility collapsed globally starting around 2007. This affected countries across the income and policy spectrum. This paper argues that smartphones changed how teens spend time with each other, and that this change in turn drove the collapse in teen fertility. Once enough teens are on the phone, being on the phone is where the peer network is; in-person time falls sharply, and with it the unstructured contact in which most unintended teen conceptions occur. A coordination model formalizes this tipping: as the smartphone price falls, the in-person equilibrium ceases to exist and the economy moves to a phone-mediated one. Within the United States, terrainruggedness variation in broadband and 4G coverage identifies a causal effect on teen fertility, and time-use diaries show in-person socializing among teens roughly halving while digital leisure roughly tripled. A parallel design for England and Wales recovers the same acceleration and the same effect of mobile coverage on teen conceptions, ruling out country-specific contraceptive-access and welfare-reform stories. The model predicts that the shift towards the phone-mediated equilibrium affects multiple aspects of teen behavior. The same instrument that produces a collapse in teen fertility produces a surge in teen suicides.
That is from a recent paper by Nathan Hudson and Hernan Moscoso Boedo.
My very charming Conversation with Craig Newmark
Here is the audio, video, and transcript. Here is part of the episode summary:
Tyler and Craig discuss why webpage design has gotten worse for 30 years, what Craig’s “obsessive customer service disorder” taught him about human nature, why trusting people and maintaining a nine-second rule for scams aren’t as contradictory as they sound, why roommate ads are a better way to find love, why Craigslist never added seller evaluations, why Leonard Cohen speaks to him more than Bob Dylan, what William Gibson’s Neuromancer got right about the internet, why Jackson Lamb is now one of his role models, why large foundations lose accountability, what two painful Ivy League grants taught him philanthropy, what he gets from rescuing pigeons, the hard lesson he learned about confronting people who lie for a living, his favorite TV shows and movies, the one genuine luxury he can’t go without, what he still needs to learn, and much more.
Excerpt:
COWEN: What is scarce in your life then? You’re giving away money. You don’t have to run the company on a day-to-day basis. We’d all like more years to live, but what is it that if you had more of it, you could be more effective with?
NEWMARK: I guess, ideally, I would have more social skills—meaning, some.
COWEN: We’re simulating social skills just fine here.
NEWMARK: That’s the phrase I use. At least on my part, what looks like social skills is just fakery. I can do it for short amounts of time, maybe 90 minutes. I’ve given up, though, on actually accumulating social skills, getting better at it. More to the point, I try to get into positions where other people can show social skills.
COWEN: One journalist once described you as having “obsessive customer service disorder.” Isn’t that a social skill?
NEWMARK: That’s more obsession, so it’s pathological, but a good one. I believe that you should treat people like you want to be treated. Think of the many times that you needed customer service. Sometimes you can get good customer service, but that’s the exception. That’s no reason for us not to provide a good customer service. Like earlier today, someone sent in a grant proposal, and I had to tell them that they forgot to sign the thing, a very minor thing. More importantly, I’m telling people they need to do some planning for good communications because their work is much less valuable if they can’t talk about it effectively.
COWEN: According to Susan Freese, who wrote about you, in one year, you answered 40,000 customer service emails. Is that possibly true? If so, what did you learn about humanity doing that?
Recommended, charming and engaging throughout.
talkie: an LM from 1930
Here is the link, with explanation.
Will AI end anonymity?
Like many journalists, I have a bunch of unpublished fiction lying about, so I tried Claude on the first chapter of a romance novel that I started almost 20 years ago, during the hysterical, mawkish phase of a particularly bad breakup. “Megan McArdle,” said Opus 4.7, after a few seconds of thought. Fascinated, I kept feeding it smaller and smaller passages to see how little prose it needed for identification. The answer, apparently, was 1,441 words…
Would Claude do better or worse with something more modern? I fed Claude a different opening chapter from an unpublished science fiction novel I started right before the pandemic — I contain multitudes — and this time Claude needed only 1,132 words. The eulogy I gave for my mother, lightly edited to remove some too-specific biographical details, was even faster: Depending on the passage, Claude was able to peg me as the author in as few as 124 words.
Here is more from Megan McArdle.
Generative AI and Entrepreneurship
This paper studies how Generative AI (Gen AI) is reshaping the U.S. startup ecosystem. Exploiting the release of ChatGPT, we show that startups with greater pre-release Gen AI task exposure reduced employment within two quarters, primarily among junior and implementation roles. Displaced workers experienced longer unemployment spells and moved to lower-paying but less exposed jobs. Conversely, exposed startups increased productivity, scaled faster, and accelerated through financing rounds. Venture capital shifted toward frequent, smaller investments, boosting new firm formation. Overall, incumbent contraction was offset by new firm formation, leaving aggregate employment unchanged but shifting composition to senior roles.
That is from a new and important paper by Abhinav Gupta, Franklin Qian, Elena Simintz, & Yifan Sun.
From the UAE
Under the directives of the President of the UAE, we launch a new government model.
Within two years, 50% of government sectors, services, and operations will run on Agentic AI, making the UAE the first government globally to operate at this scale through autonomous systems.
AI is no longer a tool. It analyses, decides, executes, and improves in real time. It will become our executive partner to enhance services, accelerate decisions, and raise efficiency.
This transformation has a clear timeline. Two years. Performance across government will be measured by speed of adoption, quality of implementation, and mastery of AI in redesigning government work.
We are investing in our people. Every federal employee will be trained to master AI, building one of the world’s strongest capabilities in AI-driven government.
Implementation will be overseen by Sheikh Mansour bin Zayed, with a dedicated taskforce chaired by Mohammad Al Gergawi driving execution.
The world is changing. Technology is accelerating. Our principle remains constant. People come first. Our goal is a government that is faster, more responsive, and more impactful.
Here is the link. While there is typically a certain amount of PR in such pronouncements, I do not think this one is only PR.