
AI, Robot
Google DeepMind: The Podcast
Balancing AI Safety and Innovation
AI systems need to be allowed enough freedom to come up with innovative solutions, but without exhibiting unsafe behavior. This requires careful consideration of incentives and rewards, as seen in the example of Goodhart's Law where rewarding the killing of cobras in British India led to locals breeding more cobras to take advantage of the reward, resulting in worthless snake farms when the scheme was revoked.
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.