AI Leaders Warn: Recursive Self-Improvement is COMING FASTER! #shorts

Transcript

1. 0:00You're not there yet, and recursive self-improvement is not inevitable, but it could come sooner than most
2. 0:06institutions are prepared for. That sentence is something that Dario Amodei has been saying for the past 2 months.
3. 0:16And he didn't really say that before. Sam Altman's saying similar sort of things, and the the guys over at Google
4. 0:22are saying similar sort of things. Elon Musk is certainly saying similar sort of things. Um using public benchmarks and
5. 0:29previously unreported data from within Anthropic, the Anthropic Institute is showing that AI is already accelerating
6. 0:35the development of AI systems. To take just one example, today Anthropic engineers on average ship eight times as
7. 0:44much code per quarter as they did from 2021 to 2025. This technical trend The technical trends discussed in this
8. 0:52piece suggests that AI systems are going to become much more capable in coming years. AI that can build itself would be
9. 0:59a major development in the history of technology. One that could bring enormous good for the world in science, health care, um
10. 1:07but recursive self-improvement also might increase the risks of humans losing control over AI systems. If say
11. 1:15systems are capable of fully building their own successors, the ways we secure them, monitor them, and shape their
12. 1:21behavior all grow much more important. >> Watch the full replay at community.thesalon.ai.