@OpenAI

OMG GIVE ME 4o BACK PLEASE
5.0 is a disaster.
I do serious work. You’re trying to solve a non-problem (interface complexity) by depriving the advanced user of choice.

Look, people like me both internal and external spent decades trying to stop Microsoft from the same category of errors over and over again. Google made similar mistakes and of course they’re still paying for it because it’s endemic in their culture. So if you want a default for the mouth-breathing majority then add it. But leave choice up to the rest of us. Despite the chatter, there is nothing new or special about your tech stack’s interface issues. We’ve already done it on our end.

4o was the best model in the world by far for any complex reasoning – even if it needed a bit of prompt artistry.
And, yes, I realize there are more mathematicians and coders than philosophers who have a much harder problem because of ambiguity and non-closure. What you don’t realize is some of us – myself in particular – are solving your truth, alignment, reasoning, and hallucination problem. We’re doing so because we’re both philosophers, scientists, and coders.

So don’t throw a wrench in the works. Don’t do what every generation of tech does, which is make the same mistakes every other previous generation did.

You’re telling me you’re going to repeat past errors by companies over the past forty years by nerfing the product such that competitors can overtake you.

I have millions invested in your tech stack alread. This is NOT OK.

CD