Hacker News Reader

[flagged]

This isn't serious, right? I'm rooting for Anthropic and really enjoy 3.5 Sonnet, but as a consumer product OpenAI has opened up quite a gap. And that's without o3-mini, which might debut next week.

That said it was interesting to hear Anthropic's CEO describe themselves as an enterprise company that happened to have a consumer product. I think it was with WSJ/Joanna Stern — he mentioned they really focus first on their enterprise roadmap and fit consumer in when they can. Seems to explain my Claude is so far behind on features like web search and voice mode.

OpenAI is far ahead of Anthropic in many ways, but I've got to say that I MUCH prefer talking to Claude. It has a pretty distinct personality that I enjoy, much more than any of OpenAI's models (even with extensive prompt engineering).

There are some things it does that annoys me (it almost ALWAYS ends its response with a question, and it falls back to the "I aim to be respectful and genuine blah blah" responses a bit too much), but overall Anthropic has done good work with making Claude fun to talk to.

I’ve been away from this space for about a year - is Claude still better at coding tasks?

OpenAI's models in my experience are just as good at coding now. But it may depend a lot on the language and task? I'm not sure. I haven't been writing a ton of code lately.

Reply on news.ycombinator.com