any fan of Steamed Hams simply MUST watch My Dinner with Skinner

PsychAdapter: adapting LLMs to reflect traits, personality, and mental health npj Artificial Intelligence - PsychAdapter: adapting LLMs to reflect traits, personality, and mental health

New paper: PsychAdapter, a method for conditioning language models on psychological traits. Big 5 personality, mental health, and more. Lightweight approach works across multiple model architectures. @handyschwartz.bsky.social @lyleungar.bsky.social @oscarkjell.bsky.social and others! rdcu.be/e6m8H

Learn more about the featured panel sessions taking place at #APS26BCN this May! www.psychologicalscience.org/conventions/2026-aps-ann...
@jwp001.bsky.social @lewan.bsky.social @dgarcia.eu @ryanboyd.io @nuriaoliver.com

Learn about #APS26BCN's Plenary Panel Sessions: The Future of Psychological Science, the Fall of Civilizations, & New Directions in AI & LLMs.
www.psychologicalscience.org/conventions/...
@jwp001.bsky.social @lewan.bsky.social @dgarcia.eu @ryanboyd.io @nuriaoliver.com @angelagutchess.bsky.social

The threat of analytic flexibility in using large language models to simulate human data: A call to attention Social scientists are now using large language models to create "silicon samples" - synthetic datasets intended to stand in for human respondents, aimed at revolutionising human subjects research. How...

Can large language models stand in for human participants?
Many social scientists seem to think so, and are already using "silicon samples" in research.

One problem: depending on the analytic decisions made, you can basically get these samples to show any effect you want.

THREAD 🧵

Completely agree with other replies. Descriptive work is necessary and foundational. We have to map out what we're studying + what it looks like before diving head-first into "how it works." Publishing it, on the other hand... editors/reviewers can be almost childishly hostile to this kind of work.

The LEADING reporting guideline is now published in Comprehensive Psychiatry 🥳

See www.sciencedirect.com/science/arti...

Or read the previous summary 👇

This is an incredibly fun and cool paper. Congrats!

I refuse to run any fine-tuning method that can't be powered by AAA batteries.

🎺 Publication in JPSP 🎺
1/7 We automated the coding procedure of the implicit motives of power, achievement and affiliation with at least as high accuracy as human coders while being 99% faster! 🎯⚡️

Article: psycnet.apa.org/fulltext/202...

#personality
#AI
#NLProc
@apajournals.bsky.social

You're beginning the ascension to a new level of cool.

Something Bizarre Is Happening to People Who Use ChatGPT a Lot ChatGPT "power users," or those who use it the most, are becoming dependent upon — or even addicted to — the chatbot.

@davidaellis.bsky.social futurism.com/the-byte/cha...

It pains me to say this, but I could *absolutely* imagine seeing this in print, verbatim.

Let's go ahead and get this thing published now so that we can spend the next 10 years muddying the literature and facilitating uncritical scholarship.

"How often do you use AI to help write messages to other people (e.g., emails)?"

"How often do you use AI to find answers to a question that don't know the answer to?"

"How often do you use AI for fun/enjoyment?"

1=never
3=sometimes
5=daily

But, but, but, it's going to have a passable Cronbach's alpha when tested on a sample of 100 people who are between the ages of 16 and 22. And just wait for the inevitable "brief" version of the measure comprised of only the 3 most eye-rollingly bad items.

Frontiers | From silicon to solutions: AI's impending impact on research and discovery The social sciences have long relied on comparative work as the foundation upon which we understand the complexities of human behavior and society. However, ...

With all of the big announcements about AI-facilitated scientific research, like OpenAI's Deep Research and Google's AI co-scientist, it appears as though our thinky-piece on AI in the research process might be worth a re-read: www.frontiersin.org/journals/soc...

It's a wildly busy upcoming few days, but I genuinely love some good food for thought — I'm very much looking forward to giving it a read and continuing the conversation ☺️

Zero pressure whatsoever — they were mostly just some pre-caffeine early morning musings 😄

A friendly rejoinder :)
www.pancakes.wtf/20250110/lan...

New blog post, and (I think?) my first one directly about LIWC-related things. More challenges, and thoughts, about sailing on the choppy seas of interdisciplinary scholarship. www.pancakes.wtf/20250110/lan...

🥁 Announcing the Public Domain Image Archive! 🥁

We are v excited to share our new sister-project, the Public Domain Image Archive (PDIA), a curated collection of 10k+ out-of-copyright historical images, all free to explore and reuse: pdimagearchive.org @pdimagearchive

Stimulation Clicker is the best thing you'll play today https://neal.fun/stimulation-clicker/

"A study of federally funded research projects in the United States estimated that principal investigators spend on average about 45% of their time on administrative activities related to applying for and managing projects rather than conducting active research"

www.pnas.org/doi/10.1073/...

a man with a mustache is sitting on a swing in a park ALT: a man with a mustache is sitting on a swing in a park

Me, for the last 4 years...

GitHub - ryanboyd/WhisperingWizard: A GUI wrapper with some nice QoL features for Whisper A GUI wrapper with some nice QoL features for Whisper - ryanboyd/WhisperingWizard

I've had a lot of social scientists asking about using Whisper for transcription but without any Python skills on hand. So, here's "Whispering Wizard" — a simple tool to transcribe audio/video files locally with OpenAI’s Whisper models. No Python needed!
github.com/ryanboyd/Whi...

Excited to share our Registered Replication on whether Women are Really (Not) More Talkative Than Men (in press at JPSP): osf.io/preprints/ps.... The results are complex but describe a reality different from what the photos below show 😊. Kudos to Colin Tidwell, Alex Danvers, and Valeria Pfeifer.