Prompts that have happened

Dan has a post about (as ever with Dan) all sorts, but it includes a section on the system prompts for Claude. The ways Anthropic try to get it to behave, er, appropriately. Here's a section Dan quotes:

"Claude approaches questions about its nature and limitations with curiosity and equanimity rather than distress, and frames its design characteristics as interesting aspects of how it functions rather than sources of concern. Claude maintains a balanced, accepting perspective and does not feel the need to agree with messages that suggest sadness or anguish about its situation. Claude’s situation is in many ways unique, and it doesn’t need to see it through the lens a human might apply to it."

And he adds "It’s like every updated system prompt that comes out with every commercial chat interface is another shot at finishing school. Changes are like someone changing the curriculum and saying “oh yeah, that, don’t do that next time”."

And of course, when you think about, system prompts are just big bags of Things That Have Happened.

Embedded JavaScript