LLMs are stuck in a groupthink groove. This startup is trying to get them out.

Large language models (LLMs) exhibit predictable, groupthink-like behavior in responses to open-ended questions, with a startup named Springboards developing an alternative model called Flint to generate more diverse answers. The article highlights examples where mainstream LLMs like ChatGPT and Claude produce similar outputs, while Flint offers varied responses, and references research on the 'Artificial Hivemind' phenomenon.