← Back to context Comment by davidw 8 hours ago You only have as much freedom as you can buy from a handful of huge corporations. 3 comments davidw Reply simonw 7 hours ago Or download for free from half a dozen Chinese AI labs (at least for the moment.)... or Google Gemma or Mistral. davidw 6 hours ago Aren't those widely regarded as 'not as good', because right now LLM's are a game of "who has the most resources wins"? And even those require very high end personal computers.The freedom pendulum looks like it swung away from individuals.See: https://news.ycombinator.com/item?id=48542100 simonw 4 hours ago They're not Opus 4.5 level yet, though the really big Kimi and DeepSeek v4 models (which need like 600+ GB of GPU-available RAM) aren't too far off.
simonw 7 hours ago Or download for free from half a dozen Chinese AI labs (at least for the moment.)... or Google Gemma or Mistral. davidw 6 hours ago Aren't those widely regarded as 'not as good', because right now LLM's are a game of "who has the most resources wins"? And even those require very high end personal computers.The freedom pendulum looks like it swung away from individuals.See: https://news.ycombinator.com/item?id=48542100 simonw 4 hours ago They're not Opus 4.5 level yet, though the really big Kimi and DeepSeek v4 models (which need like 600+ GB of GPU-available RAM) aren't too far off.
davidw 6 hours ago Aren't those widely regarded as 'not as good', because right now LLM's are a game of "who has the most resources wins"? And even those require very high end personal computers.The freedom pendulum looks like it swung away from individuals.See: https://news.ycombinator.com/item?id=48542100 simonw 4 hours ago They're not Opus 4.5 level yet, though the really big Kimi and DeepSeek v4 models (which need like 600+ GB of GPU-available RAM) aren't too far off.
simonw 4 hours ago They're not Opus 4.5 level yet, though the really big Kimi and DeepSeek v4 models (which need like 600+ GB of GPU-available RAM) aren't too far off.
Or download for free from half a dozen Chinese AI labs (at least for the moment.)
... or Google Gemma or Mistral.
Aren't those widely regarded as 'not as good', because right now LLM's are a game of "who has the most resources wins"? And even those require very high end personal computers.
The freedom pendulum looks like it swung away from individuals.
See: https://news.ycombinator.com/item?id=48542100
They're not Opus 4.5 level yet, though the really big Kimi and DeepSeek v4 models (which need like 600+ GB of GPU-available RAM) aren't too far off.