Lennart Heim on Compute Governance
Future of Life Institute Podcast
00:00
How Big Is the Training Run?
We don't always know what's going to arise from these machine learning models because we've been surprised before. We might also learn more about like certain aspects of training runs which are particularly dangerous right so once you maybe invent an architecture where you have like some sub-architecture that enables your AI systems to have goals this is not foolproof but we learn over time and we could get started on this right now just like yep those systems have an impact this makes you more responsible this makes it less responsible in the future.
Transcript
Play full episode
