Aurélien Allard

The alpha version of my new book "Optimal Transport for Machine Learners" is out, with in particular an online version with interactive figures www.gpeyre.com/ot4ml/

In the last few weeks i've been bombarded with emails from people who have minor tech support problems. ChatGPT is telling them to email me, writing their emails, and compiling "evidence packets" after convincing them they have stumbled on a major story: www.404media.co/chatgpt-keep...

In China, more than one in five workers is still employed in agriculture. China makes and exports more goods than any other country.

Social desirability bias is smaller than most people assume. In Blair, Coppock and Moor's (2020) metanalysis found clear effects only for (1) Support for auth regimes (in autocracies) and (2) vote buying. www.cambridge.org/core/journal...

Absolutely brutal: Over four years in review with Nature. Still published with a copy and paste error between experimental groups. Checked everything but the raw data I suppose. pubpeer.com/publications...

"I describe 3 directions: collecting trustworthy data from real people, formalizing psychological theory now that computational modeling is within reach of any researcher, and rethinking methodological education so that students learn to judge the analyses that AI will run for them."

What they say: "we have a lot of data" What they have: 11,000 variables for 18 people. "Maybe a cluster analysis?"

I actually don’t think you can be a substantive expert without a certain amount of methodological competence. Otherwise you don’t really have a sense of how to interpret your data or the studies of others. It’s just vibes.

26 papers featuring original experiment data published in SPQ (sociology’s social psych journal) in 2023-25. 6 used pre-registration (despite a near-uniform embrace of “hypothesis testing” language), 8 made data available. Oof.

Remains poorly understood that humans hallucinate more than frontier AIs at this point bsky.app/profile/matt...

I'm so so so excited that we at @zpid.bsky.social will host #SIPS2028! The SIPS conference was one of the first contacts I ever had with #OpenScience & influenced my decision to start my academic career greatly. See you in 2028 in Trier, Germany!🥰

Interesting. Of course, humans judge sample was only N=3, and they only read the reviews, not the paper itself. But I have the suspicious that the pattern will hold even with larger tests. The problem is how AI would be able to judge a truly novel idea, though the same is true for humans.

Gary King just dropped a study showing that AI peer reviews now dominate human peer reviews on every metric they tested. gking.harvard.edu/peerless/

This part was gold. What a chill character.

70 percent of the internet was vulnerable yesterday to very bad security flaws inside a coding language. Turns out, *one volunteer* is mostly responsible for maintaining that code & helping push updates. Yesterday Ondřej Surý saved the internet. So I called him up www.theatlantic.com/technology/2...

Dans la presse française, on parle beaucoup de "dette" et de "déficit" depuis septembre 2026, mais considérablement moins qu'en 2008-2011 (Le Monde, résolution mensuelle, www.gallicagram.com?word=dette+%...)

i am on a plane sitting next to a lady who looks 40ish and has informed me that this is her first time flying she is visibly excited. she asked the flight attendant how much a soda costs and was happy to learn that drinks are free on airplanes. she's taking lots of pictures. i love this for her

"Evidence that the substrate of the brain matters for its function is evidence against computational functionalism" is just not true. Any system will depend on its substrate - that doesn't speak to whether its functions could be realised in other substrates.

One thing I found a little weird is the decision to divide a person's impact for every paper by the number of co-authors linearlyl. Means that generous PIs who acknowledge the important work of RAs and students by including them as co-authors are being disadvantaged over greedy PIs.

My paper on the Varieties of Cultural Selection is now in press at @biologicaltheory.bsky.social. It's also about 30% longer than the original draft, with additions and clarifications inspired by the excellent comments I received from reviewers and colleagues. Preprint here: osf.io/preprints/so...

New preprint! Computational Humanities needs falsifiable theories. With 25 Paris-based researchers who share the same ambitions for the field and had an exciting conversation. zenodo.org/records/2349...

Everyone stop what you're doing and read my paper with @jacobmontgomery.bsky.social and Zion Little modeling preregistration as pre-play cheap talk. We ask how preregistration affects behavior like p-hacking when there is no enforcement and plans must be self-enforcing osf.io/preprints/so...

Thought I would make a quick timeline for some major events in 20th century evolutionary biology and quickly ran out of space. It's fascinating that so much of abstract population genetics was formalized before we knew genes are made of DNA

Immigration experts have never known more, but mattered less. In my new JEMS article, "Immigration for Realists," I argue that we have ourselves contributed to this gap and suggest ways to fix it. We misread the public, misrepresent the evidence, and lose our credibility as a result.

TIL that Auschwitz's emergence as a symbol of the Holocaust is an example of survivorship bias. Operation Reinhard, during which most perished in gas chambers at the death camps Belzec, Sobibor, and Treblinka, was the largest single murder campaign of the Holocaust 🧪 www.science.org/doi/10.1126/...

Halloween Animals Support Group.

At last, Claude manages to solve the deepest mysteries in the universe

I'm puzzled why the APS Conference only accepts papers in: Empirical Research Systematic Review Meta-Analysis & -Synthesis Methods & Applications Perspectives, an APS journal, had its theory crisis issue in 2021. How can this be solved without an audience? journals.sagepub.com/toc/ppsa/16/4

If you haven’t watched this one yet, this video by the European Commission (!) is an absolute gem and worth the minute.

In preparation for a discussion with the lab on AI use tomorrow morning, I thought I would come prepared and write down my thoughts in a blog post: How I use AI in my work daniellakens.blogspot.com/2026/10/how-...

www.psypost.org/when-women-s...

China reduced extreme poverty fast, but Indonesia hasn’t been far behind. China is often the poster child for rapid reductions in poverty, and for good reason. In the early 1980s, over 90% of its population lived in extreme poverty, but by the early 2020s, that number had dropped to nearly zero.

A theory of collecting. "Madness-as-strategy" & PTSD. The very first issue of Mind. Interview w/ Tom Nagel. Bingo card for bad AI-consciousness takes. How a philosopher & her husband picked up the pieces after she was brutally attacked. The Principle of Sufficient Reason & panpsychism...

New ACE audit report: Singh et al. (1997), one of the most cited trials of resistance training for depression in older adults. Misreporting of standard deviations and impossible statistics led to many meta-analyses having artificially inflated effects. auditclinicalevidence.org/reports/ACE2...

We cross-referenced the psychometric personality data of ~ 1 million people with their zodiac astrological signs: 1) There is a small but statistically significant effect of zodiac sign on personality controlling for demographic and attitudinal variables

We would like to thank 31 currently active PsyArXiv moderators and the leadership team who volunteer their time and energy to ensure that important psychological research is made available to the research community as quickly as possible. buff.ly/htX5xM0

The turkey 🦃, also known as the syphilis of birds.

Think “Fairness is important (Rawls, 1971).”

Currently, a diagram is making the rounds that is claimed to show a significant reduction in Russian disinformation bot activity after Ukrainian drones hit Yandex data centers. There is no evidence that this plot shows real data. I suspect that this image is AI-generated.

An extraordinary example of a paper that cannot logically address its estimand: effect of restricting under-18 playtime. But sample has no age data, and most of the sample is adults. Reviewer #3 even noted this, and authors agreed. See link in quoted for details.

It's out and open access baybeeeee www.degruyterbrill.com/document/doi...

Oh jfc

Check out our new paper! (1/5) www.pnas.org/doi/10.1073/... @err-ring.bsky.social @kkshimizu1.bsky.social @markstoneking.bsky.social @balthasarbickel.bsky.social @chiarabarbieri.bsky.social @isle-uzh.bsky.social @nccrlanguage.bsky.social

We show that global structural linguistic and human genetic diversity are inversely correlated. In other words: Genetically more diverse groups speak less structurally diverse languages and vice versa! (2/5)

fave new demonstration of survey items being really not that bad at tracking carefully-measured cognitive things: in 900,000 participants, a 5-item "how good are you at music" self-report (the colours) very directly tracks a 20min adaptive test of music perception ability (the x-axis)

New diamond open access outlet for developmental psych. Free to publish, free to read, and led by a great team. One more step away from for-profit publishers.

Illuninating post on the Mathocalypse scottaaronson.blog?p=10169