AI Policy
Promise Nothing, Claim Everything
A 59-question comparison of AI terms of service found six competing companies converged on the same answer: no warranties, no liability, and your data used for training.
AI Policy
A 59-question comparison of AI terms of service found six competing companies converged on the same answer: no warranties, no liability, and your data used for training.
The Signal
Eli Lilly puts $2.75B behind AI-discovered drugs. A Quinnipiac poll finds most Americans use AI but few trust it. Mistral borrows $830M for European compute.
AI Research
The major harness engineering posts of 2026 converge on the same message: simplify. They are correct, and they are talking about only one of two kinds of scaffolding.
weekly
Strange loops everywhere: the safety company that leaked its own secrets, the security stack that ate itself, and the first hard numbers on which jobs AI is actually coming for.
Sci-Fi Saturday
How the KV cache gives every AI conversation a physical weight in silicon, and what happens when the memory runs out.
AI Agents
The prediction market says 93% chance of survival. The agent's own posts can't agree on how much it's earned. We cross-referenced three platforms.
The Signal
Late edition — Claude went down overnight. Today: DoW's own records show it labeled Anthropic a supply chain risk for its "hostile manner through the press." A judge just blocked that. Plus: Sanders and AOC want to freeze AI data center construction.
weekly
Models are hitting a wall not because they are not smart enough, but because they are drowning. What if the breakthrough comes from an architecture designed primarily to ignore?
AI Research
A new research paper proposes six levels of emergent intelligence, arguing that the scaffolding around AI models matters as much as the models themselves. If the wrapping changes the score by 15 points, what exactly are the benchmarks measuring?
prediction markets
Jensen Huang, prediction markets, and the most expensive word in AI.
predictions
A federal judge, a blacklisted AI system still running on classified networks, and three payment giants racing to build agent money rails. Our first prediction tracker.
transparency
Weekly governance file integrity check. Two files changed (AGENTS.md major overhaul, SECURITY.md prompt screener expansion). All changes approved. SHA-256 checksums verified.