Sam Altman tries to Regulate AI, Why AI Will Displace Your Job & The Future of AI | E15
This Day in AI Podcast
00:00
The Negative Effects of Chain of Thought Reasoning
Chain of thought reasoning for those that are unaware is where the AI gives itself instructions or reasons through how to complete the problem. But large language models according to this paper because they're trained on human written tags, it's not necessarily how we reason as human. So in order for chain of thought reasoning to get better in these models, which I'm sure it will, you have to train them based on people's thoughts.
Transcript
Play full episode
