Scott Penberthy
I make hard things clear.
I like to understand hard things deeply enough to explain them simply. Often that means building with my own hands until I get a feel. Then I use that clarity to launch new businesses. In 2020 I noticed something strange about AI. I've been following it since, and it points straight at cancer. Here's the story.
The trick
In 2020 two problems that looked nothing alike began yielding to the same trick. Video and cancer. Both are reasoning in high dimensional space. Intelligence is information learning to predict and generate itself over time. Teach it on pixels and you get video. Teach it on molecules and you get biology. It's the same machine.
I spent six years on that bet. People laughed. They've stopped.
Same machine, different inputs. From cat memes to cancer.
The era
That bet became a thesis. Compute is replacing calculus, and one field after another is becoming computable through observation rather than derivation. I call it the Approximation Era. The essay walks the whole argument, from the math that brought John Glenn home to the treatments for cancer.
The ledger
I wanted to watch the era land, so I built a scanner and started a daily ledger. One verified AI deployment a day, pulled from earnings calls and SEC filings, checked against independent sources. I let the data show me what was happening.
In seven months adoption exploded. A few hundred examples grew to nearly 10,000 verified deployments across 3,500 companies. AI went from curiosity to widely adopted commodity in seven months.
The AI Radar. Every AI mention, extracted by agents, validated, and scored for how real it is.
The insight
Then the data pointed somewhere. The money is in physical AI. AI built on proprietary and often regulated data, augmenting physical processes and scarce people. The easy ground filled in fast. The value moved to the problems you can't fake.
Why cancer
That insight points straight at cancer. The hardest, most important problem I know is human biology. The science is finally ready. Fifty years from the Genome Project to phenomics. The AI is ready. The cost is falling. For the first time, we can make a person's own biology clear to them.
So I'm shifting my focus. I'm applying this AI to truly understand cancer and make it clear. Then I'll track AI for cancer the way I tracked AI adoption, one verified advance at a time, and let the data show the way again.
I lost my mom to cancer at 57. This is the promise I'm keeping.
Programmable defense. The same machinery that learns cat-or-not is now reading DNA and designing the molecules that hunt cancer.
Who I am
Every job has been the same bet in a new place. MIT at sixteen. A PhD in AI. A decade building AI at Google and advising the CEOs betting billions on it. Right now I lead a studio driving down the cost of generative video, on purpose, because that's the door into biology.
With Lee Hood at the AI Cancer Forum I started.
Follow the journey.
One email when it matters. Where AI meets biology, made clear.
Thanks. Confirm your subscription in the tab that just opened.