I mostly roll my eyes at "AI is a bubble," but my 2008 Great Recession traumatized brain also gets very nervous whenever large-scale speculative investments are paired with large amounts of leverage. 🎁🔗 www.wsj.com/tech/ai/the-...

If you have (or will soon) complete your PhD & would like to join a thriving cultural evolution research group in NEW ZEALAND, drop me a line. UoA offers a handful of funded research fellowships. Applications must be made by a faculty member. Deadline August. Pls share with your students

"I don’t want to be in a discipline where people are worried about saying things because they think the truth will have bad effects on their politics. I’m not a politician. I’m a scholar."

Precisely the problem facing science, health, and academia. These vitally important institutions will fail in their mission if every criticism must be withheld for fear of aiding [-----]. chronicleofhighereducation.substack.com/p/is-a-new-r...

A well-deserved honor, though no mention of his contributions to R-help in the article

One of the better examples I’ve seen of visualization conveying what numbers alone cannot

For years I've tracked billionaire excess. Each year I show students a chart that blows their minds. This year I had to redraw it for a trillionaire: Elon Musk, $1.32T. Depressing that we're here. Silver lining: couldn't have happened to a less deserving person, proof that wealth ≠ merit.

A picture my dad took in NZ

Well fucking excuse me Claude!

When Trump withdrew from JCPOA in 2018, Mike Pompeo gave a speech outlining a set of demands that any new agreement would have to meet. If we want to judge, not just the outcome of this war, but the broader project of renegotiating the Iran deal, this is a useful yardstick.

Our department has three calls for tenure track positions. IE is a fast growing group of scholars doing top-notch work, which offers very attractive conditions in a great European city. If you are looking for an academic job in Europe, definitely do consider applying.

Today I fulfilled one of my San Diego bucket list items and went with my family to eat at Jimmy Carter's Mexican Cafe. It was very very good and I highly recommend

@i4replication.bsky.social NEW incredible progress from I4R on a reproducibility Engine The Engine is deliberately diagnostic... it executes code, flags mismatches, raises possible implementation issues, proposes robustness checks, but leaves substantive judgments to people.

Congratulations to Professor Anil Menon on receiving the Emerging Health Politics Scholar Award from the Health Politics and Policy Section of APSA!

worth reading..!

If you signed the letter below 1. Thank you! 2. There is now a second letter from UC faculty in the social sciences and humanities that I'd encourage you to sign as well. ucstudentsuccess.org/socscihum/

First time out with the camera in a long time. Some lovely evening light over the fields #sharemondays2026

Boom goes the dynamite. #knicksinfive

Google insisting on shoving cheap-but-dumb AI into every consumer facing product... pure brand destruction.

So, it turns out the answer is shakedown

This is why the rule of law is so important though. The scale required to build this class of model is enormous, limiting the number of viable companies. Given the small #, it’s very useful if we don’t have to worry about the govt capriciously punishing one because they didn’t bend the knee.

If “AI” is so powerful, how is it that Google and Meta can digest my entire browser history and still think they can get me to give a shit about soccer? I’M NOT CLICKING DUMMY

"Better than from outside the tent," they said...

Great post. The thing that makes the TACO equilibrium so hard to think through is that it’s not an equilibrium. In equilibrium, Trump would have correct beliefs about how his record of lying changes the meaning of subsequent market signals, but I don’t think that’s the case yet.

Neuroscientist here, who has talked about the US losing science ground to other countries: The biggest problem for generating future cures for brain diseases is the ongoing utter destruction of US basic neuroscience. It’s not about regulation.

Great post. The thing that makes the TACO equilibrium so hard to think through is that it’s not an equilibrium. In equilibrium, Trump would have correct beliefs about how his record of lying changes the meaning of subsequent market signals, but I don’t think that’s the case yet.

The President just said we're on the cusp of a deal with Iran. Today's post explores two fun facts: 1. It's the fiftieth time he's said this. (I counted!) 2. Even so, stocks rose on the news. newsletter.platypuseconomics.com/p/how-to-pri...

so literally exactly what we had before?

Of course we know there’s no deal but that doesn’t matter

"All this is sold as cutting “burdensome conditions,” a goal I would support, but sadly that is bullshit."

I’ve got good news and bad news and there but the same: only semis matter to the indices

Idk. Feels like the market over weights signals about how much he _wants_ a deal and underweights signals about how unwilling he is to do what’s necessary to get a deal.

Remember when Prigozhin was marching an armored column of mercenaries on the capital of a nuclear power in an attempted coup? I think that might still have been the craziest geopolitical moment of the past decade.

One thing I mentioned in passing in my Fable post is that, for long running tasks, Fable starts to develop its own dialect as its many agents and tasks reinforce themselves and make Claudish language ever more Claudish. You need to ask it to report out in plain English or this happens after 9 hours

This is like half the millennials in my feed rn

To all the people who said I was crazy for constantly saying “thank you” to the agents:

This is wild, especially when it gets to the part where Fable builds an isochrone map.

To all the people who said I was crazy for constantly saying “thank you” to the agents:

Evergreen sentiment among academics

The widespread use of LLMs diminishes creative ideas. An analysis of over 2000 college essays finds that human-written essays contributed more new ideas than GPT-4 written essays www.sciencedirect.com/science/arti... Parameter- or prompt-modifications does not mitigate the creativity gap.

Evergreen sentiment among academics

Hello. Philosophers and others have had an annoyingly hard time explaining what it means for a *human* to be conscious or how you would know. But you can’t say anything about AI consciousness without an account of humans. www.slowboring.com/p/nobody-kno...

In a couple weeks, I'll be teaching my Introduction to Survey Design class. Sign-up details below. instats.org/seminar/intr...

Excited that my article with @profmickeyj.bsky.social and #Roseanne_McManus, “The Domestic Sources of International Trust,” is now published as FirstView in @apsrjournal.bsky.social! Here is a brief summary.

Honestly, a common issue on a lot of needed reforms in liberal constituencies

I really hope that Sociology reforms. My main work is reanalyzing other scholars' experiments in service of meta-analysis. It's only possible because transparency is embraced so thoroughly by our political science community. Most exps are posted; i'm getting ~80% when I email for the rest.

imo, California should close dropboxes on Sunday If voters want to drop off their VBM ballot at a vote center, have them remove it from the envelope and enter it directly into a tabulation machine Will both get you to 90%+ counted on election night & reduce rejection rates