00:00I think in terms of short, medium, and long-term risks.
00:02Short-term risks are the things we're facing today around things like bias and misinformation.
00:07Medium-term risks, I think in a couple years as models get better at things like science, engineering, biology,
00:14you can just do very bad things with the models that you wouldn't have been able to do without them.
00:20And then as we go into models that have the key property of agency,
00:24which means that they don't just output text, but they can do things,
00:28whether it's with a robot or on the internet,
00:31then I think we have to worry about them becoming too autonomous
00:34and it being hard to stop or control what they do.
00:37And I think the extreme end of that is concerns about existential risk.
00:41I don't think we should freak out about these things.
00:43They're not going to happen tomorrow.
00:44But as we continue on the AI exponential,
00:48we should understand that those risks are at the end of that exponential.
