Scott Penberthy
Real AI, Verified
Day 247 of documenting the Approximation Era.

Never fall for AI hype again.

I verify what's real in AI. 5,000+ deployments checked against earnings calls, one more every day, so you know what's actually working before you bet money, budget, or your career on it.

5,000+ verified, and counting

By Scott Penberthy. MIT at 16, PhD in AI, four computing waves since 1977. Built Google Cloud's AI business, advised IBM's Lou Gerstner and PwC's Bob Moritz, now working on generative AI at Alphabet. About →


The problem
Every company says it is "doing AI." Almost none of it moves a number. The press releases, the demos, the hype cycles: most of it is theater. And you are being asked to bet real money, budget, and reputation on which parts are real.
How it works
So I built a scanner. It reads every earnings call and SEC filing across the major indexes, extracts every AI deployment, and validates each one against independent sources. What survives goes in the ledger: 5,000+ verified deployments across 3,338 public companies, one more every day. A running, checkable map of where AI is actually working, and where it is just window dressing.
AI Radar visualization of thousands of companies and verified AI use cases
The AI Radar. 3,338 companies. 5,000+ use cases. Extracted and scored by AI agents.
The pattern underneath has a name: the Approximation Era. AI is a universal approximator, and it is remaking one industry after another. Read the full thesis →

What the ledger reveals
See the full ledger →

Get the patterns before you bet.

One email a week. The verified deployments that moved a number, and what they mean for where the money is going. No hype.