I asked AI to roast me — and it asked for time off
A prompt for a friendly roast based on your chat history, and a look at what AI noticed in my projects.
Blog
Articles and notes on making digital things understandable.
How I sped up Qwen 27B on a Mac with M4 Max
355.48 tokens per second on a long input, cache behavior and quality limits. A detailed experiment with Kapitsa, measurements and a reproducibility kit.
No matches yet
Try a different search or reset the filters.