Join free with your email
Free. No spam, ever — we'll never share your email address and you can opt out at any time. Already a subscriber? Log in
Researcher with public evidence across AI research.
Great piece from @zey.bsky.social on what generative AI really is and why it's wholly overblown in the media and by the people selling it (gift link)
Sure this should be public docs.google.com/document/d/1...
it'll be really interesting to see if the effects hold up to scrutiny in peer review (this is the first author [Chen's] job market paper, unpublished afaik) paper is public at fion.ac/jellyfish.pdf and on a quick read it sure looks solid !!
Our lab guide now includes an explicit AI usage policy: www.poldracklab.org/labguide/res...
woof Raycast 2.0 just updated and it's a lot slower (Electron!) and full of AI bloat - it's going down the path of Arc (😞). BUT I just discovered Tinycast and am loving it so far. It even imports Raycast configuration files and works with Raycast extensions
after some in-depth discussions with my lab and also drawing on conversations on here, we've edited our new policy on AI usage in my lab feel free to adapt or re-use with attribution, or to reply to tell me how awful this policy is lol
this working paper from Harvard PhD students Fiona Chen and James Stratton is making the rounds, perhaps because it confirms many people's suspicions about the tradeoffs in vibe coding: LLMs mean lots more code written, but also lots more problems created, and relatively little productivity benefit
today I hired a graphic designer to do some scientific illustrations, and — guess what! — that's exactly how easy it is to not use generative AI to replace human artistry
there are many complex ethical and practical reasons for you to forbid students from using AI for scientific writing, and to not use it yourself for scientific writing. the most compelling one is simpler: the text an LLM generates or 'polishes' reads like shit
<doing homework> 8yo: dada what is gemini Me: it's an AI thing that can do stuff in your slides if you want 8yo: why would I want it to do things to my presentation Me, straight-faced, stifling my pure joy at his comment: you mean because that's your work and nobody else's? 8yo: yeah? obviously
seems like the LLM era has ushered in a change to the canon academic saying 'you should be writing' - now it's more like 'YOU should be writing'
"I'm not sure how this was made but it looks kinda AI-generated, which may hurt you with some reviewers (whether or not it was AI-generated)" is a real comment I just made on a grant proposal
lest anyone think AI continues to grow more powerful, unabated, a colleague just informed me her AI-enabled calendar tool has fallen down on its one job, which is to place our meeting into its list of meetings, at the correct time (we are rescheduling)
we are now including AI usage disclosures in lab papers but we're kinda making up how to write them. has anyone got guidance they like about this? seems like AI disclosures should have at least: - name/version of tool - which author(s) used - what they used for - who checked tool's output
one theme that I am picking up on regularly one-on-one with students, but that I don't see discussed much, is the shame that serious grad students experience, regarding AI usage
Center = samuel mehr. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.
Add the Who’s Who of AI badge to your site or bio. It links back to this profile.
Markdown: [](https://aiweekly.co/whos-who/person/mehr-nz)