
Stephen Casper
MIT computer science PhD student researching technical and socio-technical AI safety, focusing on interpretability, red-teaming, and AI audits.
Top 3 podcasts with Stephen Casper
Ranked by the Snipd community

Aug 16, 2024 • 1h 7min
Episode 152: The Wide World of Concussions with Stephen Casper
Stephen Casper, a researcher specializing in the history of concussion and health ethics in sports, dives deep into the evolution of concussion understanding. He discusses how cultural perceptions of head injuries have changed while highlighting the NCAA's negligence towards athlete safety. The conversation brings attention to financial influences on concussion research and raises ethical concerns about the integrity of scientific investigations. Casper advocates for accountability in collegiate sports and urges the need for reform in how we address athlete health.

Aug 2, 2024 • 60min
#10: Stephen Casper on Technical and Sociotechnical AI Safety Research
Stephen Casper, a PhD student at MIT specializing in AI safety, dives into the intricacies of AI interpretability and the looming challenges of deceptive alignment. He explains the subtle complexities behind unobservable failures in AI systems, emphasizing the importance of robust evaluations and audits. The discussion also touches on Goodhart's law, illustrating the risks of prioritizing profit over societal well-being, as well as the pressing need for effective governance alongside AI advancements.

Jun 19, 2024 • 2h 42min
Ep 14 - Interp, latent robustness, RLHF limitations w/ Stephen Casper (PhD AI researcher, MIT)
PhD AI researcher Stephen Casper discusses interpretability, robustness, and limitations of AI models. They explore AI safety, risks, research gaps, and advice for early career researchers. Casper emphasizes finding flaws in neural nets, AGI risks, and practical AI applications.