cross-posted from : https://lemmy.zip/post/69107935
Hugging Face CEO says China is winning the AI race and dominating on open models
https://www.cnbc.com/2026/08/03/hugging-face-china-ai-race-open-models.html
https://www.cnbc.com/2026/08/03/hugging-face-china-ai-race-open-models.html
cross-posted from : https://lemmy.zip/post/69107935
7 Comments
supersquirrel@sopuli.xyz · 3 pts · 36d
yawn
ComradePenguin@lemmy.ml · 2 pts · 35d
Why pay for really expensive models when you can get almost the same performance for a fraction of the cost?
The future is not about the most powerful models necessarily. There will be cases, and there will be a development in intelligence going forward.
My prediction is that the future of AI (LLM++) will be harnesses and tools, not the absolute best inference.
brucethemoose@lemmy.world · 2 pts · 33d
Yeah.
At this point, everything around the LLM is woefully behind the performance of LLMs themselves. If all LLM development stopped tomorrow, I think there would still be years of catching up to do.
ComradePenguin@lemmy.ml · 2 pts · 33d
Most definitely.
I am an extreme user of LLMs and other AI, and the limit is creativity and adaptability. It is no longer the "intelligence" of the models that counts, they are already insanely good now. I use Kimi K3 and it blows me away.
The limit is what my mind can conjure up.
brucethemoose@lemmy.world · 2 pts · 33d
Well even the human user aside, I'm thinking of agent harnesses, tools, prompt "hacking," constrained and exotic sampling, augmentations like control vectors or loras, RAG, databases, multimodel thinking...
The basic chatbot interface is hardly tapping into these models at all. They're very effective tools when specifically constrained, but most are still figuring out how to do that.
sage@lemmy.blahaj.zone · 1 pts · 35d
But Amodei says open models are bad.
Redditmodstouchgrass@lemmy.zip · 0 pts · 33d
Fucking "Hugging Face"?