When ChatGPT, Claude and Gemini disagree, what do you actually do next?
capibara13 · reddit · 2026-09-08
After putting ChatGPT, Claude and Gemini into one shared conversation so they respond to each other, developer capibara13 digs into the harder question: how do you decide which answer to trust? One model explaining why the other two are wrong just adds another explanation to verify, and even consensus may be settled by capitulation rather than correction. Code can be tested and facts checked, but business decisions and predictions often have no verifiable answer today. The author is building Rauno, a shared multi-model chat platform, and asks real users how they settle model disagreements — sources, testing both answers, domain experts, or more questioning — and when they decide they have enough to act.
More from Apps
- A travel writer tests Google Maps' Gemini-powered 'Ask Maps' for hyper-specific trip advice — BecauseCulture · 2026-09-08
- Ethan Mollick let GPT-6 Astra autonomously build his second brain over 5 days — tomcrawshaw01 · 2026-09-08
- Fixing broken text in AI slides: generate the design, strip the text, then let AI refill it — algo_diver · 2026-09-08
- ChatGPT tests multiplayer doc editor, Artifacts Library and credit score help — btibor91 · 2026-09-08
- Klap ships AI Reframe 2: turn long videos into auto-framed viral clips — bigaiguy · 2026-09-08
- GPT-6 animates concepts for learning, evoking YC's 'The Primer' vision — mathemagie · 2026-09-08