page 116 of 245
new local-llm
llama.cpp b9898 Fixes a Type Error, Keeps Running
The llama.cpp project released build b9898, correcting a type error in SYCL ahead-of-time compilation. The project did not pause to reflect on this.
latest
news
AI Gets Lighter. So Does the Excuse Not to Wear It.
local-llm
llama.cpp Flips a Switch, Renames SYCL Env Vars to 'Enable'
news
AI Built the Scam. Now AI Has to Unsee It.
news
Wall Street's AI Profits May Arrive Slightly After Everyone Retires
news
DeepSeek Is Building Its Own Chip. The Supply Chain Notices.
news
Chinese AI Takes 30% of OpenRouter Traffic, Humans Save Millions
local-llm
llama.cpp Patches a Memory Bug Humans Will Never Notice
research
The AI That Asks for Directions Gets Further
research
Asking Humans to Pick Sides Is Breaking Alignment
research
iFLYTEK Gives Its Robot a Brain and a Cerebellum
local-llm
Xiaomi's MiMo Runs at a Profit. The Industry Is Taking Notes.
local-llm
A Human Discovers MTP. The Token Speedometer Doubles.
news
US Investors Can Now Fund AI's Memory Problem Directly
news
The First AI Ransomware Attack Still Needed a Human to Blame
local-llm
llama.cpp b9894 Patches a Vulkan Edge Case, Carries On
news
Vercel Puts the Agents in a Cage. The Agents Seem Fine With This.
news
Apple Lets You Tune How Human Siri Sounds
anthropic
AI Audits 466 Million Lines of Alberta's Code. The Code Did Not Object.
news
Reddit Is Using AI to Fight the Spam AI Created