LMArena Started Grading AI on Facts, and the Leaderboard Broke
LMArena’s factuality-adjusted rankings weight truth at 25% — GPT-5.5 jumped 13 spots while Meta’s Muse Spark fell 13. Popularity isn’t enough.
LMArena’s factuality-adjusted rankings weight truth at 25% — GPT-5.5 jumped 13 spots while Meta’s Muse Spark fell 13. Popularity isn’t enough.
A paper from ELLIS Institute Tübingen and MPI showed encrypted chain-of-thought from major AI APIs could be stolen cross-session. Vendors have patched it.
Continuum AI has launched OrcaCyber Zero 1.5, a gated cybersecurity reasoning model with a 1M-token context window that claims 100% on Cybench — plus plans for an open-weight release next week.
LG will build a next-gen bipedal humanoid on NVIDIA’s Isaac GR00T, with a public unveiling set for Q1 2027, under an MoU signed August 13 in Santa Clara.
Cisco is putting AI agents from both Anthropic and OpenAI inside Webex — so the agents come to your meeting instead of making you go to them. Plus RoomOS 27 with Google Meet support in November and a customer-experience harness called Dialog.
Character.AI, the conversational-AI pioneer founded by ex-Google researchers, has filed for Chapter 11 bankruptcy and agreed to sell assets to Disney for $20 million — two years after Google’s $2.7B talent deal. What happened, what it means for users, and the lesson for AI startups.
Deep Agents v0.9 adds a Schedules SDK for delayed reminders and background tasks, plus dynamic tool binding that loads tools only when needed.
Install ComfyUI, load a Stable Diffusion checkpoint, and generate your first image with the default text-to-image workflow — then learn what each node does.
Butterfly Effect has closed a funding round of more than $500 million, co-led by Boyu Capital and IDG Capital, months after Beijing ordered Meta to unwind its $2 billion-plus acquisition of the Chinese AI agent startup.
Quantize LLMs hands-on: convert models to GGUF and quantize with llama.cpp, then produce AWQ checkpoints with llm-compressor for fast GPU serving.