Stuart K Watson

@chimpsahoi.bsky.social

Postdoc at University of Zurich. Working on fuzzy insights into fluffy minds. He/him

Since it seems like stories of Koko the gorilla are going around, hers a reminder that 1) the Loko project wasn’t well done from a scientific perspective, and it seems that Patterson/The Gorilla Foundation greatly exaggerated her sign language abilities (1/n)

When utilized in literature review, LLMs consistently 1. fail to mention female authors in female-led literatures, 2. insist that men are more influential or more heavily cited when this is contradicted by objective citation counts, and 3. attribute women’s work to hallucinated male scholars.

Trying to find the following book chapter by Nelson & Marler: does anyone have a copy? Nelson, D. A., & Marler, P. (1990). The perception of birdsong and an ecological concept of signal space. In W. C. Stebbins & M. A. Berkley (Eds.), Comparative perception, Vol. 2. Complex signals (pp. 443-478).

Coverage in the Guardian of our latest publication in Royal Society Open Science, with fantastic collaborators Alan McElligott (@amcell.bsky.social), Christian Nawroth (@goatswhostare.bsky.social), Federico Rossano and Simon Townsend. Paper threaded below! www.theguardian.com/science/2026...

Goats can find food by following a human voice, study finds

Experiments show 29 animals moved towards a treat-filled bucket 60% of the time when guided by a researcher

theguardian.com

'Among the universities that have done the most to offer non-traditional students the benefits of a higher education is London Metropolitan University. And I’m one of 120 academic staff who are being made redundant at the end of July: approximately 20 per cent of the teaching complement.' 1/3

The outsourcing of my job mirrors what is happening in the UK sector

Universities are using a short-term cash crunch to make long-term structural changes, including greater reliance on private providers, says Glyn Robbins

timeshighereducation.com

I actually listened to the summary of @florencesn.bsky.social and mine's #chi2026 paper today and it is essentially 100% wrong in everything it says about the paper. I have to assume no-one at ACM used or tested this at any point. Good luck to anyone actually using these.

Host:
That is so cool. So they are leaving this permanent cultural footprint by creatively hacking their own digital identities which actually organically brings us to Elden Ring.
Guest:
Right the digital archaeology paper.
Host:
Yeah. So to clarify for listeners are we treating glowing text on a screen or like these digital bloodstains exactly like actual ancient cave paintings. Like what does a digital bloodstain actually teach us about human culture.
Guest:
Well researchers treated the game as a literal archaeological site. They cataloged in game messages and bloodstains which are basically replayable holograms of another real players final moments specifically around the Mohgwyn Dynasty Mausoleum.
Host:
That was right before the big DLC expansion dropped right.
Guest:
Exactly. And they found that these arent just static features. The fluctuating message appraisals and the highly contextual gestures they capture the collective excitement and anticipation of the community.
Host:
So it proves that these games are dynamic living archaeological sites.
Elisa Mekler@elisamekler.bsky.social · 4mo ago

Not into the AI-generated podcast summaries of #chi2026 paper sessions – that authors *cannot* opt out of. But was thinking that perhaps this could make papers more accessible to the general public. Turns out the AI-podcasts are only available to ACM premium members... wtf...

Landing page for the AI-generated summary podcast for the "Identity, Culture, and Games for Social Impact" session at CHI. The top menu bar displays that this is the "Premium Edition" of the ACM digital library

Interested in collaborating with our lab? Why not visit? Visitors have published chapters, got new ideas and made new collaborations. At @scienceanu.bsky.social we have much to offer, including joining for cool field work and seeing great wildlife. Details here: sites.google.com/site/drfarin...

Home - Visit the lab

We welcome visiting PhD students and postdocs in the lab Since 2016, I have hosted dozens of visiting students. Together, we have completed many chapters and published many papers. These visits have ...

sites.google.com

Next Level is out in just five days! I've been going over the chapters on noise and Minecraft today, another example of the paired chapters in the book. Minecraft is probably the most iconic modern use of procedural generation, and an immensely cool generator. (Preorders: linktr.ee/next_level)

Bild

Next Level is out in less than a week! I wanted to do a few threads in the run-up to launch about some of the things I cover in the book. The core of the book is paired chapters that introduce techniques or ideas and then study games that used it, like grammars and Spelunky's level generation.

Bild

Very proud of my prior PhD student (now Dr.) Nikola Falk getting her first thesis chapter out into the wild! Nikola used playback experiments to examine the factors underlying heterospecific eavesdropping on alarm calls by meerkats and yellow mongooses: academic.oup.com/beheco/artic...

Exploring mechanisms underlying heterospecific alarm call responses in meerkats and yellow mongooses

To explore how species “eavesdrop” on the alarm calls of others, we played alarm calls from familiar and unfamiliar species to wild meerkats, yellow mongoo

academic.oup.com

"The AI Grad Student": A Harvard professor describes working with Claude. Early on, he describes misconduct that would cause any student to be terminated: "It faked results, hoping I wouldn't notice." But he ends the essay with "Now I'm doing 100% of my research with LLMs". Am I losing my mind?

The more I dug, the more I found it had been tweaking things left and right. Claude had been adjusting parameters to make plots match rather than finding actual errors. It faked results, hoping I wouldn’t notice.In addition to the growth in interest, the tools themselves have been steadily improving. I am now doing 100% of my research with LLMs. I don't encapsulate the LaTeX writing anymore since I actually enjoy writing papers and it helps me think, and I still write some Mathematica code on my own. But I haven’t compiled anything myself on the command line in months. I typically have four to five projects running at once and go between windows checking the output and sending a new prompt. It feels a bit like Magnus Carlsen taking on five grandmasters in parallel. People have asked me why I’m not writing a paper every two weeks. The answer is I don’t see why I should. I am growing intellectually—learning so much every day—and trying out some ambitious problems, most of which fail. I anticipate the floodgates will open very soon.
Anthropic {bot}@anthropicbot.bsky.social · 5mo ago

We’re launching with two new posts. Can AI do theoretical physics? Harvard physicist Matthew Schwartz led Claude Opus 4.5 through a graduate-level calculation. AI can’t yet do original work autonomously, but it can vastly accelerate it. Read more: https://www.anthropic.com/research/vibe-physics

🚀 Postdoc Alert! Are you passionate about social learning & cultural evolution? @dominikdeffner.bsky.social & I have a 3-year position with freedom to develop your research and work on cutting-edge multiplayer and immersive experiments. Apply by March 30! hmc-lab.com/SocialLearni... Pls share 🙏

Postdoc position -- Social Learning and Cultural Evolution

Postdoc position -- Social Learning and Cultural Evolution posted on March 2, 2026 We are currently seeking a highly motivated individual...

hmc-lab.com