zlacker

> ChatGPT is currently king of the mountain.

This is completely false. Claude Opus is significantly better than GPT 4.

> Mistral/Mixtral (decent with no censoring/guard rails)

These models have been heavily censored, I'm not sure what you're talking about. Community efforts like Dolphin to fine-tune Mixtral have some success, but no, Karen is definitely still hard at work in France, ensuring that Mistral AI's models don't offend anyone's precious fee-fees.

>>caeril+(OP)
I think you're missing the forest for the trees here. You're right that Claude Opus is better, which I hadn't known, but I think in your zeal to make that point you're completely forgetting what my comment was about.

It's nevertheless true that there is a coherent landscape of better and worse models, and Chat GPT really does have separation from the other models as I mentioned above. I even mentioned that ChatGPTs position would be subject to change. My understanding is that this most recent version of Claude has been out and about in the wild for perhaps 2 months.

I feel like with even a little bit of charitable interpretation you could read my comment in a way that accounts for the emergence of such a thing as a new and improved model of Claude. So I appreciate your correction but it's hard to see how it amounts to anything more than a drive-by cheap shot that's unrelated to the point I'm making.