Blog · ai · the-ai-times · video · · Trevor Heshka
Anthropic's Constitutional AI: The Machine That Judges Itself (The AI Times Ep. 039)
How Anthropic's Constitutional AI has a model critique and revise its own answers against written principles, and the real test: whether those principles actually change what it says.
Originally posted at youtu.be
How Anthropic's Constitutional AI has a model critique and revise its own answers against written principles, and the real test: whether those principles actually change what it says.
<iframe width="560" height="315" src="https://www.youtube-nocookie.com/embed/bI_w8p3MYjI" frameborder="0" allowfullscreen></iframe>
Chapters
- 0:00 Intro
- 0:17 The next model first reasoned about
- 1:13 The family tree has three branches
- 3:19 Two levers, both at inference time
- 5:23 The newest release is the GPT-5.6
- 7:01 Outro
Watch on YouTube · More from The AI Times, a daily 6-8 minute AI explainer.
Want your own AI bot team? Get the course for $49 CAD