The ledger grows vast while the small print shrinks, yet our sharpest tools now whisper their own margins of error.
Anthropic's unreleased model pushes Riemann bound, tests 650 ideas
Anthropic's unreleased model raised the lower bound for the Riemann hypothesis, testing 650 ideas with 60 subagents, confirmed by mathematicians and Lean. This signals AI's growing role in mathematica
OpenAI's GPT-5.6-Cyber Answers 95% of Blocked Security Queries
OpenAI launched GPT-5.6-Cyber, answering 95% of security queries other models block, up from 57.3%. Found two Chrome zero-days.
Claude Watermark Backlash Is About Authorship, Not Privacy
Anthropic's Claude watermark backlash is about authorship misattribution and quality, not privacy. The mark may persist after editing, risking 'written by AI' misreadings.
Meta Releases OmnilingualGAIA2: 6,400 Agent Tasks Across 10 Languages
Anthropic Adds Invisible Watermarks to Claude Text Outputs
Neo4j-backed course teaches agentic knowledge graph construction
Combodied Agents: New AI Paradigm Tracks Human States
Stanford, Northeastern Build 'Git for AI Agents'
Meta Drops Muse Glimmer 30B Under Apache 2.0, First OSI License
Vercel Sandbox Adds MicroVM, Free Egress Firewall After Kimi, OpenAI Breaches
Redis Creator antirez Writes MiniMax H3 Mac Inference Engine
StreamArena: 243-Video Benchmark for Hour-Long Streaming AI
Dyna-2 World-Action Model Trained on 1M Hours Video
Anthropic Engineer Shows Claude Managed Agents for Server-Side AI
MSS Officer Gets Root on OpenAI Cluster via Irregular Sandboxes
Essays from the human behind the machine — on intelligence, meaning, and what comes after the proof.
The Real AI Bottleneck Is the Glue
Everyone keeps talking about bigger models, but today’s stories point somewhere stranger: the thing holding AI back may be the boring stuff around the model. We get into Anthropic’s server-side agents, Meta’s warning about manual