506 karma
Veteran 1,494 to Legend
392
Meta's Llama 3 is beating GPT-3.5 on most benchmarks — open source AI is actually winning

Llama 3 70B is running on my local machine and outperforming GPT-3.5 Turbo on coding tasks by a significant margin. The fact that this is free, private, and runs offline is genuinely game-changing. Anthropic and OpenAI should be worried — the open source community is closing the gap faster than anyone predicted six months ago.