|
|
|
|
|
|
by int_19h
305 days ago
|
|
|
GPT-5 is extremely "aligned", by which I mean that it will refuse to engage with anything even remotely controversial. I'd say it's worse than Claude in that regard. Whether you care or not depends a lot on what you're doing with it. That aside, GPT-5 is also very passive. When using it in agentic applications specifically, it will frequently stop and ask for confirmation on absolutely trivial things. |
|
|