Elicit Negations via chat/interview with AI

[volkyeth]

Jack:

AI agent suggests relevant statements/negations based on preference interview. Or statements/negations that haven’t been submitted yet, which the user might want to submit. The AI agent should lean into conversation that makes it easier to elicit preferences without it feeling like a chore.

Then the AI could understand how the player is positioned in the argument and suggest:

• Endorsing relevant points
• Adding negations that are not yet present