Hacker News new | past | comments | ask | show | jobs | submit
> I would counter argue with "But chinese models use distillation of Frontiermodels..." which is different than being a Frontierlab or not clear how different it is.

This reeks of a desperate attempt to convince yourself that "Chinese models" are worse than they are, in spite of actually being on par if not better than some so-called frontier models.

It's same blend of comments that accuse open weight models of being economic warfare.

After a point, references to "Chinese models" start to sound like a dog whistle. I mean, who walks around talking about "American models" ?

Perhaps it's high time that "Chinese models" are given some respect. They outperform models from openai and Anthropic and constantly show up in the Pareto front. That's quite the achievement, isn't it?