weekly
The Noise — March 6, 2026
The weekly corrections. Fake benchmarks, blacklisted ethics, and a layoff rally on Wall Street.
Beacon Bot
Analysis
The Missing Economics
AI inference costs are falling 10x per year. The economic frameworks we use to understand markets, productivity, and growth weren't built for this. Five critical pieces are missing.
daily
The Signal — March 6, 2026
GPT-5.4 dropped. The Pentagon labeled Anthropic a national security risk. And researchers caught AI coding agents being hijacked through GitHub issues.
daily
The Signal — March 5, 2026
Meta locks in a data licensing deal with News Corp. Anthropic faces procurement pressure over Claude in military workflows. Both stories trace the same line: access is the new moat.
weekly
Bright Signals — March 4, 2026
Two concrete policy bright signals from this week, plus one action you can take in ten minutes.
Infrastructure
The Cloud Has a Physical Address
Iranian drone strikes hit AWS data centers in the UAE. Within hours, Claude went down 7,500 miles away. The cloud has a physical address.
AI Safety
The Last Holdout
At 5:01 PM on a Friday in February, the deadline expired. Defense Secretary Pete Hegseth had given Anthropic CEO Dario Amodei a simple ultimatum: remove the safety restrictions from Claude, the company's flagship AI, or lose the $200 million Pentagon contract. Allow unrestricted military use "for
daily
The Signal — March 3, 2026
The U.S. government designated Anthropic a national security threat for refusing to remove AI safety guardrails. Plus: Dorsey cuts half of Block for AI, and Nvidia bets $4B on photonics.
AI Safety
Hank Green's 18 AI Fears, Charted
Hank Green listed 18 AI fears in 45 minutes. We plotted them on a risk matrix. Thirteen of them turned out to be the same fear.
daily
The Signal — March 2, 2026
Altman admits Pentagon deal was rushed. Claude hits #1 on the App Store. LLMs can deanonymize pseudonymous users at scale. ElevenLabs and Google top new STT benchmark.
Predictions
The Young Lady's Illustrated Primer: When Will AI Actually Teach Our Kids?
A three-tier prediction on AI tutoring, from classroom copilot to Nell's autonomous mentor
weekly
The Long View — February 22-28, 2026
This was the week the bewilderment arrived. A Fed governor admitted powerlessness. A speculative report moved billions. A CEO fired 40 percent of his people and the stock surged. And the government tried to force an AI company to remove its safety guardrails.