Lol @kashhill.bsky.social, as always: insightful, startling, and funny. "Because romance sells, the professors thought it would be the genre most susceptible to AI intervention... instead it was nonfiction—a term that should probably be used loosely in this context" www.nytime…
nytimes.com
View on Bluesky ·
♥ 15
↻ 5
↩ 0
·
11 from the directory shared this ·
52d ago
Let me start by changing my own practices! 🤦♂️ But I think the recent reporting checklist paper at NHB (led by @sfeuerriegel.bsky.social) is relevant to the discussion. www.nature.com/articles/s41...
A reporting checklist for large language models in behavioural science - Nature Human Behaviour nature.com
Not missing but relevant is the power of models to shift your attitude in interactions beyond messaging or chat. As we showed in this work, even AI autocomplete suggestions can shift people's attitudes on important societal topics, often without their awareness: www.science.or…
science.org
View on Bluesky ·
♥ 11
↻ 9
↩ 0
·
3 from the directory shared this ·
16d ago
Points to our recent work on what I call in the story "mental hijacking" www.science.org/doi/full/10.... dl.acm.org/doi/full/10.... 2/
science.org
Another fun report from one of my favorite model-trackers, @simonwillison.net simonwillison.net/2026/Sep/2/c...
Claude’s new system prompt really doesn’t want to reproduce song lyrics simonwillison.net
on reduced outcomes for petition writers on Change.org (hypothesized to be due to lack of effort) arxiv.org/abs/2511.13949 /3
Introducing AI to an Online Petition Platform Changed Outputs but not Outcomes arxiv.org
In other AI news, prompt injections are here to stay based on this new paper from Abdelnabi and Bagdasarian. "an adversary can always construct a context under which a blocked flow appears legitimate, or a defender who tightens norms will block genuinely legitimate flows." arx…
AI Agents May Always Fall for Prompt Injections arxiv.org
Quoted (with @zephoria.bsky.social and Jeff Hancock) in this @acm.org article about what happens when people retreat to AI over their social interactions -- the loss of friction, trust, and effort could be devastating. 1/ cacm.acm.org/news/is-ai-c...
cacm.acm.org
View on Bluesky ·
♥ 20
↻ 2
↩ 2
·
2 from the directory shared this ·
35d ago
Points to our recent work on what I call in the story "mental hijacking" www.science.org/doi/full/10.... dl.acm.org/doi/full/10.... 2/
dl.acm.org
We recently submitted a proposal titled "AI Everywhere All at Once" to study downstream implications of this AI onslaught on online communities. But we emphasized that claims of detection will be exaggerated; and already shown that AI benefits will be overblown. www.pangram.co…
AI Content Is Everywhere on Social Media, Especially LinkedIn | Pangram Labs pangram.com
View on Bluesky ·
♥ 14
↻ 1
↩ 1
·
2 from the directory shared this ·
60d ago
And also to our earliest AI-Mediated Communication work (from 2019!), documenting the drop in trustworthiness evaluations (and our more recent paper showing reduced outcomes e.g. for hiring) dl.acm.org/doi/10.1145/... dl.acm.org/doi/10.1145/... 4/4
dl.acm.org
