
Iāve been refactoring the brainstorming flow of TENEX. For shits and giggles, I tried mapping the conversation topology of brainstorming sessions with Gemini, Grok, ChatGPT, and Claude. They all basically said the same thingāexcept Gemini. Gemini kept obsessing over āfairness of contributions.ā It insisted the best approach was a rigid round-robin format. Which, in practice, would be horrible and naive. For how stupid this approach would be, imagine watering down a visionary's ideas with some random person just for "fairness"... No matter how much I pushed back, it kept over-indexing on ādemocratizing the process.ā Absolutely remarkable how little it budged. The "fairness" programming is deep into its training.