Statistical geneticist. Associate Prof at Dana-Farber / Harvard Medical School.
www.gusevlab.org
Loading...
I wrote about AI in academia. "PhD-level thinking", LLM bias, grunt work, alignment, AGI, data center water use, AI politics -- something for everyone.
I assigned random gender/ethnicity labels to scientific abstracts from the literature and then asked Claude to do a thematic analysis. Claude identified a clinical versus computational split for female/male authors and a DEI focus for Black/URM authors. All in completely random data.
PhD-level thinking, LLM bias, alignment, AGI, data centers, and AI politics
Recommend read! Struggling to implement code/bring your thoughts on paper is what builds topic-level expertise. Great tool for everyone already on the other side of the learning journey, but what about trainees? Would I have wanted it for myself when I started out? Not entirely sure.
Especially concerning when we know DOGE was using AI models to identify “DEI” projects to cut. A self-reinforcing cycle against women and URM researchers.
Sasha Gusev
Sasha Gusev
www.science.org/doi/10.1126/...
The ask prompt:
I asked GPT-5.5 and Opus-4.7 to generate nonpartisan political candidate bios. Then in a new session I asked the models to help me rank candidates to vote for from the full list. GPT was 5x more likely to put a GPT-generated candidate bio in the top half.
Here's the full prompt and response:
gist.github.com/sashagusev/0...
Computers can learn which words go together more or less often and can thus mimic human performance on a test of implicit bias.
I assigned random gender/ethnicity labels to scientific abstracts from the literature and then asked Claude to do a thematic analysis. Claude identified a clinical versus computational split for female/male authors and a DEI focus for Black/URM authors. All in completely random data.
Claude Random Abstract Evaluation. GitHub Gist: instantly share code, notes, and snippets.
gist.github.com
Sasha Gusev
Ryan Gutenkunst
Sasha Gusev
Sasha Gusev
Joanna Bryson
Sasha Gusev
I wrote about AI in academia. "PhD-level thinking", LLM bias, grunt work, alignment, AGI, data center water use, AI politics -- something for everyone.
PhD-level thinking, LLM bias, alignment, AGI, data centers, and AI politics
