I posted this exact question here a while back about 3.5. OpenAI keeps neutering ChatGPT to find the lowest common denominator of what they think is "acceptable

I posted this exact question here a while back about 3.5. OpenAI keeps neutering ChatGPT to find the lowest common denominator of what they think is "acceptable" results, which makes the entire model suffer.