ChatGPT, Claude, Gemini & Deepseek all agreed "Engaging four AI models in a debate to reach consensus is estimated to improve reasoning capabilities by approximately 20-50%". They formed this consensus through getconsensus.ai. Enough Said!