Reinforcement Learning in Generative AI – Computerphile

From Computerphile. The problem with reinforcement learning in Generative AI is that it’s difficult to turn the real-world into a graph. Sydney Von Arx of METR talks about an approach to solve this. More about Sydney: https://bit.ly/4aqm1mG Computerphile is supported by Jane Street. Learn more about them (and exciting career opportunities) at: https://jane-st.co/computerphile This video…

Constraining AI Agents – Computerphile

From Computerphile. 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: https://arxiv.org/abs/2504.10374 Computerphile is supported by Jane Street. Learn more about them (and exciting career opportunities) at: https://jane-st.co/computerphile…

Path Planning for Robotics – Computerphile

From Computerphile. Need to get to your goal quickly? Ensure you plan the right path! Robots need to work out how to get from here to there somehow! Ayse explains some of the methods they choose. Assistant Professor Ayse Kucukyilmaz is based at the University of Nottingham Thanks to Dave Domminney Fowler for kindly helping…