Is it me or during the last 2-3 days models got dumber?
retained text
Guys, I don't know if it's just me, but over the last three days I feel like OpenAI's models have suddenly gotten dumber... I'm developing a WordPress plugin. I wrote a pretty detailed prompt and then had Astra xhigh create a comprehensive implementation plan. I then fed that plan to a mix of Sol 6.1 high/xhigh and Terra 6.1 xhigh, switching between them depending on the task, often using `/goal`. This is a pretty standard workflow for me, one I've used countless times over the past few months. And that's when the nightmare began. I had to fix all sorts of mistakes, ranging from minor issues to downright critical errors. Features that were supposed to be implemented were completely missing, while other unnecessary complications and options had been added out of nowhere. Oh, and then there's the UI/UX (which, to be fair, seems to be a recurring weakness of OpenAI's models). Absolutely atrocious! Overcomplicated, convoluted interfaces with seemingly zero understanding of even the most basic principles of user-friendly design! Long story short, I ended up manually fixing God knows how many mistakes, because otherwise I'd have burned through tons of tokens just asking the models to redo the same things over and over again, and still never getting them right! The weird thing is that, until just a few days ago, I'd never experienced anything like this. I've used this exact same workflow on projects far more complex than a WordPress plugin, and I've never run into these kinds of problems (apart from the consistently awful UI/UX, of course...). So, the obvious question is: **is it just me, or has anyone else noticed the same thing lately?**
document fetched 2026-11-10 00:27 · canonical https://reddit.com/r/OpenAI/comments/1xro8vq/is_it_me_or_during_the_last_23_days_models_got