Hacker NewsFoundation Models & LLMs
Llama-modchip: 284B DeepSeek on 64GB RAM, no GPU
475 points · 204 comments
Real-time tracking of high-signal AI developments, tool releases, and founder conversations across Hacker News and developer communities.
475 points · 204 comments
75 points · 12 comments
1,612 points · 535 comments
119 points · 28 comments
110 points · 120 comments
622 points · 421 comments
1,527 points · 498 comments
509 points · 309 comments
1,308 points · 749 comments
51 points · 71 comments
189 points · 213 comments
17 points · 4 comments
478 points · 506 comments