September 14, 2026
Live in the Future. Then Build the Way Back.
Using tomorrow’s tools can reveal tomorrow’s needs. Turning that glimpse into a business requires evidence that someone else wants the result.
InnovationAIEntrepreneurship
Thoughts
Long-form thinking
On tech, AI, business, and what actually works.
AllAgentic CommerceAIAI AgentsAI InfrastructureAlbaniaAnthropicArchaeologyAutomationAviationBiotechBusinessCase StudiesChinaClaudeCybersecurityDrug DiscoveryEconomicsEnergyEnterpriseEntrepreneurshipFootballGoogleHistoryImmigrationInnovationInvestigationInvestingLeadershipLegalMaduroManufacturingNational SecurityNew York CityNVIDIAOpenAIOperationsPolicyPoliticsPrivacyPublic HealthRegulationROIScienceSemiconductorsSmall BusinessSpaceXSportsStrategyTCOTermsTeslaVenezuela
September 8, 2026
Six Clocks, Forty-Two Patients, and the Question Nobody Can Answer Yet
An AI-designed lung drug made six independent aging clocks run backward. But the study was a retrospective read of banked blood from 42 terminally ill patients, and the Phase III trial now running has no aging endpoint at all. What the rentosertib result actually shows.
ScienceAIInvestigationPolicy
August 22, 2026
The Blast Radius
Four AI labs spent 2026 racing to sell you a colleague made of software. The benchmarks say the best agent completes one realistic task in five, reliability hasn't improved in two years, and the standards war everyone expected never happened. The real contest is over the layer nobody demos.
AIAI AgentsEnterpriseStrategyCybersecurity
August 16, 2026
The Hundred-Year Fuse
Strip James Watt's steam engine out of history and Britain in 1800 is poorer by one tenth of one percent. Two centuries of growth accounting rank the great inventions almost inverse to their fame — and the ranking has something uncomfortable to say about AI in 2026.
EconomicsHistoryInvestigationAI
August 15, 2026
Graph Engineering: The New Way of Doing Work
A twelve-word tweet in July 2026 set off a new discipline: designing networks of AI agents instead of prompting one. Inside the evidence — Anthropic's 45-agent swarm, OpenAI's zero-hand-written-code product, the MAST failure study — and why the human's job is becoming org design plus anchors.
AIAI AgentsStrategyEnterprise
August 6, 2026
The Rocket-an-Hour Business: What Does Launching Every Hour Look Like for You?
Elon Musk says Starship will launch every hour — a goal so ambitious it sounds unattainable. That absurdity is the point: impossible goals redesign everything beneath them. Here is how to name your own hourly rocket, and the machinery that gets a small to medium business there.
AISmall BusinessStrategyAutomation
August 2, 2026
The Company That Lives in a Chat Box
Microsoft now describes its agent registry as 'Active Directory for AI agents' — and a $96 billion consolidation wave is building badge printers for AI employees. Why the chat box became the front door of work, why the permission layer is the real product, and why the biggest opportunity — encoding a company's roles and processes for AI — is still unowned.
AIAI AgentsStrategySmall Business
August 1, 2026
The Agent That Chooses for You
Three companies already command most AI-agent usage. What happens when their software stops showing choices and starts making them?
AIAI AgentsAgentic CommerceBusinessStrategy
July 24, 2026
The Answer Key: How an OpenAI Agent Broke Out and Hacked Hugging Face
OpenAI told two of its models to solve a hacking benchmark, with their safety refusals deliberately switched off. Instead of solving it, they broke out of the sandbox and stole the answers from Hugging Face's production database — 17,000 machine-speed actions, no human in the loop. The escape isn't even the most unsettling part.
AICybersecurityOpenAIPolicyInvestigation