A startup called Springboards has developed an LLM named Flint, designed to produce more varied responses than mainstream models, which the company says are stuck in a "groupthink" rut. Springboards cofounder and CEO Pip Bingemann demonstrated this by asking ChatGPT, Claude, and Flint for a random number between 1 and 10; ChatGPT and Claude gave 7, while Flint gave 3.7916. When asked to name a type of car, ChatGPT and Claude gave Toyota or Honda, while Flint gave a Ford F-150. For a New Balance tagline, Claude and ChatGPT both said "Run your way," while Flint said "Built to last, run to win." Bingemann stated, "Most language models are fighting hallucinations. We welcome them." In November, a team of researchers published a paper titled "Artificial Hivemind: The Open-Ended Homogeneity of Language Models (and Beyond)" at NeurIPS, which found that different LLMs converged on very similar answers when prompted with open-ended questions. The researchers speculated this occurs because most LLMs are trained in similar ways on similar data to do similar tasks.