Uploaded December 2025 | Updated September 2026, 3 weeks ago
As AI systems become more capable, rule-based safeguards, hard-coded restrictions, and simple alignment strategies start to break down. Buck Shlegeris talks about some tactics we might use as detailed in a recent paper.
The referenced paper: arxiv.org/abs/2504.10374
Computerphile is supported by Jane Street. Learn more about them (and exciting career opportunities) at: jane-st.co/computerphile
This video was filmed and edited by Sean Riley.
Computerphile is a sister project to Brady Haran's Numberphile. More at bradyharanblog.com
As AI systems become more capable, rule-based safeguards, hard-coded restrictions, and simple alignment strategies start to break down. Buck Shlegeris talks about some tactics we might use as detailed in a recent paper.
The referenced paper: arxiv.org/abs/2504.10374
Computerphile is supported by Jane Street. Learn more about them (and exciting career opportunities) at: jane-st.co/computerphile
This video was filmed and edited by Sean Riley.
Computerphile is a sister project to Brady Haran's Numberphile. More at bradyharanblog.com
