Good. We need to counter OpenAI who has literally gamified the neutering of their LLMs using the openai/evals repo.