VIEWS1.1KREPLIES2
Sasha Gusev@SashaGusevPosts
Link:
https://open.substack.com/pub/theinfinitesimal/p/thoughts-on-ai-in-academia?r=43f9ax&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true
23hViews 1.1KLikes 7Bookmarks 1
Original post unavailable.
Sentiment
Many users enjoyed the academic analysis of AI bias, alignment, and politics for confirming how models settle into shortcuts or local minima, while some objected that it leaves out unfunded PhDs.
Pos
75.0%
Neg
25.0%
5 comments with sentiment.
Cluster Engagement
Posts from X
Most Activity
Most Activity
BOOKMARKS1
Sasha Gusev@SashaGusevPosts
@kareem_carr ... even when they are given the solutions.
The *are* capable of exploring a parameter space though (e.g. I recently asked Opus to implement a toy example of a transformer running on a biological task and, after initial failure, it spent several hours converging on a ...
19hViews 49Likes 2Bookmarks 1
LIKES10
Dr Kareem Carr@kareem_carr
Based on my very recent experience as a PhD student, I could never really get away with just saying, “Here’s the answer.”
The output of an investigation was much more about the robustness around the answer: being able to defend my choices, show the analyses I’d done to validate the solution both directly and indirectly, present the options I explored that didn’t work, and demonstrate that I understood the landscape of possible approaches, not just the one I picked.
Is that something these AI models are actually able to do?
I understand the article to be saying that they can’t build cumulatively and seem to start from scratch each time. But my concern is slightly different. Even the outputs themselves seem shallow to me, like a plant with one main taproot but none of the nested secondary roots you need to create mature work.
20hViews 724Likes 10
Sasha Gusev@SashaGusevPosts
@kareem_carr This is a good question. I think it's true that the models do a poor job of evaluating the landscape of possible approaches. They seem to land on solutions fairly arbitrarily (even though they implement them well) and aren't good at discriminating among multiple paths ...
19hViews 101Likes 2Bookmarks 1
Kyle Saunders@profgoose
@SashaGusevPosts this was really good. :) You might want to have a look at my piece linked here...I think you'll like it too!
20hViews 68Likes 1
Dhanno@compute25
@SashaGusevPosts was an enjoying read , also kind of confirms that models kinda look for shortcuts or they settle or something like a local minima
22hViews 138Likes 1
Sasha Gusev@SashaGusevPosts
@kareem_carr ... simulation set-up and model that worked). So I think an open question is whether that parameter search skill can eventually translate into an ability to discriminate between higher-level research directions.
19hViews 38Likes 2
Steven Collard@stalmico
@SashaGusevPosts something for everyone except the phd who can't get funding
21hViews 103Likes 1
James Winston@lapalmedesmers
@SashaGusevPosts Interesting 🙏
22hViews 74
Sasha Gusev@SashaGusevPosts
@profgoose Thanks, will read!
20hViews 8