Will Lowe

Interesting looking paper on what you can and can't identify from staggered adoption panel data. Also, I move that all important equations get subtitles:

"A root beer float, please" "I'm sorry sir, we're little-endian here" "oh, I'm sorry. Then a float root beer, please"

Etwas für dich, falls du beim ersten Mal noch keines hattest music.apple.com/de/ipod

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...

Cool-looking causal inference and/or machine learning job at BIPS Bremen. Just the ticket for those who like blustery north German towns and arguing about mediation models with Vanessa Didelez. Disclosure: I did that this summer and it seemed like a great academic environment.

Interviewer: can you explain this gap in your CV? Me: Sure, I used \bigskip

Interviewer: can you explain this gap in your resume? William Carlos Williams: C'mon man, I put it all in the note. They were so sweet and so–– oh, in the *resume*? Sure, that's easy...

For non-psychologists: the Individual is worth 50, the Microsystem of school, society & neighbours is 25, the Exosystem of media family and government is worth triple, and the Chronosystem double. Surprisingly, only the Individual and the Microsystem are called 'the bull'.

Interviewer: can you explain this gap in your resume? Imputation estimator: Of course! I was raising children, illegally employed, imprisoned or hospitalized, with a small chance of having joined the French Foreign Legion.

When people say 'send him to the Hague!' I always wonder who is doing the sending, because the Hague doesn't do pickups. Not the same country that refuses to try him for all that stuff at home, I hope.

One has identified one's estimand.

TFW you realize the fundamental problem of causal inference was put there for your own good.

The rule of thumb for foreigners is: Everything you think you know about Germany is true of Switzerland.

Wenn du dachtest, der Fall Block wäre schon verrückt gewesen... Was ist denn mit diesen BND-Chefs los?

A nice article, maybe better on US culture than the relationship between statistics & gambling. For example, yes, probability, expected utility etc. were invented for gambling problems. But, like algebra itself, the problems they solved were about fairness and value, not about the odds and winning.

A former student working at @sovereign.tech, here in Berlin tells me they are looking for a Director of Technology. Looks like a *great* job.

TFW you read about something resource-intensive, dependent, and built for social interaction, that's learning to predict what you say next, can only be controlled by reinforcement, and is inexplicably convinced it's a person, and you notice it could equally well be about a pet, a child, or an LLM.

His friends call him 'Moral Chuck'

Always happy to see push back against polarization as an analytical concept. Here's some more from @lilymasonphd.bsky.social

Just realized that the best summary of Dawid (1982) 'The well-calibrated Bayesian' JASA is "In theory there is no difference between theory and practice, but in practice there is"

"Today, at my direction, I'm establishing an office. It will report straight to the top, which is also me" [freelancers nod]

What is your estimand?

TFW you ask an LLM about how to derive the key result in a paper you've just read the abstract for, it seems to do so, but you can't check it because you can't access the paper. Well done to everyone involved in this farce.

We're going to be doing "Begriffe ohne Anschauung sind leer, Anschauungen ohne Begriffe sind blind" for ever, aren't we?

First Abnormal Form (1AF)

hurry up and get the phone bans in place so we can attribute this to them

If you find this art as delightful as I do, but have, in other threads on here, found yourself mired in amateur philosophy discussions of what LLMs are doing and who you're talking to when you use them, then here's a short story relevant to both. I think you'll enjoy it.

[very Ian Hacking voice] So... there's a continuous measure of a continuous trait that smoothly predicts health outcomes until it is thresholded to define different types of people, who are then treated differently enough that their health consequences become discontinuous in the underlying measure?

tired: left brain, right brain wired: forebrain/midbrain, hindbrain

Hauntology is what opportunity cost feels like.

"After years of engaging in systematic torture rationalized by pretextual legal arguments, the United States has systematically failed—or refused—to hold torturers accountable or provide remedies to the victims," writes @nkorpett.bsky.social. www.lawfaremedia.org/article/9-11...

If you mentally replace all references to 'robots' & 'AI' and 'off switches' to 'tech companies' and 'regulation' then these papers become potentially quite useful. This one seems to be about regulatory capture.

Tired: Owning the libs Wired: Owning the labs

There's 20 y/o me in this superego and I don't like it

You may not care for revealed preference theory, but it's the baseline to beat.