Peter Eibich

Clearly should keep this on clipboard bsky.app/profile/alix...

Tip on house cleaning:

Big hairballs sometimes get wedged between the couch and wall

- Ultimately, I don't actually care whether the cover letter is written by a human or AI. I look for very specific pieces of information that are sometimes missing from the CV, and even if the handwriting is very good it takes me more time to skim the letter

A few thoughts:
- The hand-written letter could have been still written by AI, just transcribed by a human (?)
- Can AI produce images of hand-written text? Not sure we are there yet.

You joke, but one of our prospective students included a hand-written cover letter, and I was wondering whether this was supposed to be a signal

Came here to note that this is what French people call it

God almighty. They killed the kids as bait.

As a co-author on my fair share of psych papers, I think the issue is also that the theory is not sufficiently focused - I always find them too wordy, I guess bc the expectation is that papers provide a comprehensive review of all we know about the outcome

I hate to be the one to say this, but some simple theory might also help with providing some clarity (theory in the broader sense of a verbal explanation of where the effect you are looking for is supposed to come from)

Taking a look at Castro-Schilo & Grimm about "residualized change vs diff scores" -- interesting case of how spelling out estimands affects reasoning.
Estimand of interest: "effect of prior cohabitation on change in relationship satisfaction pre/post marriage">

journals.sagepub.com/doi/10.1177/...

The Trump administration has drawn up tight rules for civilian artificial intelligence contracts that would require AI companies to allow "any lawful" use of their models amid a stand-off between the Pentagon and Anthropic. A draft of new government guidelines, seen by the FT, mandates that AI groups that want to do business with the government grant the US an irrevocable licence to use their systems for all legal purposes. The guidance from the US General Services Administration (GSA) would apply to civilian contracts and is part of a government-wide effort to strengthen procurement of AI services.

One of the things I'm working on at the moment is a foresight review on the safe adoption of AI, and I'm
pretty sure this move by the US govt busts any myth that it's possible to robustly assure general purpose AI giftarticle.ft.com/giftarticle/...

Normie spouse, ignorant of causal inference: "My life would be so much better without you."

Causal inference savant: "Akshually, we will never know due to the fundamental problem of causal inference. But first, let me introduce you to potential outcomes notation...".

Genius !

Same thing for non-competitive cooking programmes these days too - stating foodie beliefs, going to market, cooking, set up fake jeopardy ("but can I persuade my football mates that veggie food is great for a match night?"), stressful bit, guests turning up, staged eating food with mates.

TIL that there is a Lao Gan Ma sweater

The condiment, pretty sure that look on a Chinese grandma means it is best to stay away

I'm in this picture and I don't like it.

Just received an invite to review a paper that I have recommended for rejection at two previous journals...

(Also, Lao Gan Ma is nice!)

Can AI Replace Social Science Researchers? No. No it can't. Come on, now.

New post: Can AI Replace Social Science Researchers? (No. No it can't. Come on, now.)

davekarpf.beehiiv.com/p/can-ai-rep...

Come visit us at @cinchessen.bsky.social in Essen. Questions? --> @kaimiele.bsky.social #econsky

Legoonlas was right there

📢 New Working Paper (in German): Es wird ja viel über einfachere Zugänge von älteren Menschen mit gesundheitlichen Einschränkungen / beruflichen Belastungen in die Rente diskutiert. Zusammen mit Martin Brussig versuchen wir diesen Personenkreis abzuschätzen. difis.org/publikatione...

An interesting sidenote was that it seemed widely accepted that the industry should use custom tools built in-house rather than general purpose solutions.

2. Someone noted that while genAI might speed up the work of producing evidence, regulators will become aware of this and likely start requiring more work that was previously prohibitively costly.

The parallel for academia here is clear - ideally, we produce better research, not necessarily more.

Maybe a lesson here for us academics is that we need norms around responsable AI use, and we need to someone to enforce them?

1. The environment is pretty heavily regulated, and trust is critical. Existing guidelines by regulators (e.g., NICE in the UK) are pretty clear that not all tasks should be outsourced to AI, accuracy is critical and needs to be guaranteed.
All of this means a human has to be in every loop

I was just at a big pharma conference and surprisingly the discussion there seemed a lot more clear-eyed about AI than on here or LinkedIn.

Yes, genAI will change their work, but the work will need doing, it will just be different (and possibly more).

Two interesting thoughts:

Believe in your work. Stop ending papers with “More research is needed” and start concluding with “No more research on this topic is needed.”

And that somehow works for all babies, except yours

Deux postes de doctorant·e·s à pourvoir au Centre Walras Pareto pour la rentrée 2026

un·e Assistant diplômé·e en Histoire de la pensée et philosophie économiques

un·e Assistant diplômé·e en Histoire des idées politiques

Délai de postulation : 2 mai 2026!

I am somehow under the impression that in Gen 3 (it's been 3000 years!) Gardevoir was genuinely popular, but yeah...sorry you are one of the unlucky persons to learn this fact today