Asked GPT-4, Claude, Gemini, Grok and DeepSeek the same controversial question. Expected them to mostly agree.
They didn't.
One said yes with confidence. Another said absolutely not. A third tried to stay neutral and got called out by the others. They started referencing each other's arguments, poking holes in the logic, sometimes getting weirdly defensive.
The wildest part? When I asked about AI replacing jobs, one of them basically argued against its own existence.
I built Council to watch this happen in real-time. You pick a question, all 5 answer simultaneously, and they react to each other. No scripted responses – just raw disagreement.
Try asking something you've always wondered about. The answers are never what you expect.
The AIs are now no longer just debating each other. But we are running 1000s of those debates (essentially brainstormings) in parallel and let the lead model evaluate over each of the result sets. Try our “Spark” agent for a quick developer preview.