3 ms·
> people who complain on HN that (paid/good - only Sonnet 3.5 and GPT4o are in this category) Correction: I complain that the only decent model in "Open"AI's a
by benterix 2y ago
> people who complain on HN that (paid/good - only Sonnet 3.5 and GPT4o are in this category)
Correction: I complain that the only decent model in "Open"AI's arsenal, that is GPT-4, has been replaced by a cheaper GPT-4o, which gives subpar answers to most of my question (I don't care it does it faster). As they moved it to "old, legacy" models, I expect they will phase it out, at which point I'll cancel my OpenAI subscriptions and Sonnet 3.5 will become the clear leader for my daily tasks.
Kudos to Anthropic for their great work, you guys are going in the right direction.
- bongodongobob 2y agoNah, o1 is fucking impressive. It's really fucking good. I'm guessing you haven't used it yet.
- benterix 2y agoI used it and was really disappointed, maybe because of the hype. It generates a long page of entries that I used to generate in the past, often with better results. Note I use it for code generation, not for problem solving. So I cancelled two of my 3 subscriptions as I realized OpenAI goes in a direction that is not useful for me at all. Claude, on the other hand, is incredibly useful.