3 ms·
The current LLMs are also "magic" so anything is possible. AFAIK there is no proof that the current architecture is optimal. And we have our brains as a pretty
by YZF 5mo ago
The current LLMs are also "magic" so anything is possible. AFAIK there is no proof that the current architecture is optimal. And we have our brains as a pretty powerful local thinking machine as a counter-example to the idea that thinking has to happen in data centers.
- _heimdall 5mo agoI want to ask what makes them magic, but even those building LLMs don't really know what happens when they run inference... I have to assume current architectures aren't optimal though, the idea that we stumbled into the one and only optimal solution seems almost impossible.