Sayash Kapoor
Person on 3 sites
Sayash Kapoor is a computer scientist who studies how to evaluate AI agents and frontier systems.
-
6 facts on wiqqi (opens wiqqi.app)
- Education: PhD in computer science, Princeton University (July 2026); Porter Ogden Jacobus Fellow (opens wiqqi.app)
- Next: Assistant professor, UC Berkeley School of Information, from July 2027 (opens wiqqi.app)
- Book: AI Snake Oil (with Arvind Narayanan), one of Nature's 10 best books of 2024 (opens wiqqi.app)
-
3 channels on ppll (opens ppll.app)
-
x.com: @sayashk (opens x.com)
“Yesterday, I had the privilege of testifying to the California Assembly Privacy Committee what AI legislation we need to help incentivize AI safety. One of my main takeaways was that there is immense enthusiasm in the…”
-
Newsletter: AI as Normal Technology (opens www.normaltech.ai)
“New paper: Emerging trends in AI agent evaluation”
- Site: Sayash Kapoor — Princeton (opens www.cs.princeton.edu)
-
x.com: @sayashk (opens x.com)
-
20 beliefs on korrents (opens korrents.com)
-
Evaluating an AI agent requires scrutinizing its entire trajectory, not just its final outcome. (opens korrents.com)
“The evaluator must scrutinize the entire trajectory, not just the final outcome.”
-
An AI agent's evaluation results at one level of scope are not guaranteed to hold at another. (opens korrents.com)
“Crucially, the results of evaluation at one level of scope are not guaranteed to translate to another.”
-
AI companies should be liable for what their agents do, and policymaking should make that clear. (opens korrents.com)
“We think AI companies should be liable for what their agents do, and this should be clarified through policymaking.”
-
Evaluating an AI agent requires scrutinizing its entire trajectory, not just its final outcome. (opens korrents.com)