Hacker News new | past | comments | ask | show | jobs | submit
I'll just note that the ChatGPT version you used is 5.5 instant (with no thinking budget).

Unfortunately, this is the experience of most people with ChatGPT, which is why the broad population is so unaware of how intelligent and nuanced an AI response can be.

I wasn't signed in. Do people not get better models when signing in? I imagine a lot of people have accounts.
You need to be signed in to a paid account + change the default thinking effort.

I imagine that OpenAI will eventually roll out smarter models to the free unsigned version, but it's just a delay that also causes public perception delay