4 ms·
"locally-trained models for everyone" Wouldn't there be a massive duplication of effort in that case? It'll be interesting to see how the costs play out. There
by adrianwaj 7mo ago
"locally-trained models for everyone"
Wouldn't there be a massive duplication of effort in that case? It'll be interesting to see how the costs play out. There are security benefits to think about as well in keeping things local-first.
- all2 7mo agoThere are multiple efforts for 'folding at home' but for AI models at this point. I get the impression that we will see a frontier model released this year built on a system like this.