Skip to Content

Video: The Company of the Future Runs Like an F1 Team

The F1 Business Era, part two, in seven minutes
November 13, 2026 by
Hugo Acurio
The Company of the Future Runs Like an F1 Team video

Seven minutes on where the edge goes when everyone rents the same AI. In Formula One it went to whoever adjusts fastest, lap after lap, and business is crossing the same line. My take in three lines: the model is rent, the company brain is a lease, and what you own is judgment and experience, plus the record of every change. I also asked an AI to plan this video, and its plan had a hole F1 closed in 2003.

Narrated in my own voice, cloned with ElevenLabs. The race footage is AI-generated (Runway, Nano Banana), composited in Remotion. No real teams, cars or sponsors.

Read part one, the article

Transcript

For decades, Formula One was won by inventions. Disc brakes. Carbon fiber. One big leap bought you a few seasons. Then the rules closed the gaps. Now races are won lap by lap, with small calls from the pit wall. Business is crossing the same line. Everyone can rent the same AI now, so a big leap gets copied in a quarter. The edge goes to whoever adjusts fastest. Business will not run exactly like a race team. But the shift is the same. I call it the F1 Business Era.

This is the board. Every week, it shows me something new. October. A developer and an AI model rebuild seven Adobe apps. Photoshop, Premiere, Illustrator. Open source, and free. In March, word got out that OpenAI was building its own GitHub. Software stocks had already lost a quarter of their value in six weeks. In Formula One, the rules come out before the season. In AI, the tempo writes them. Every month. And that tempo has been lightning fast. Most companies cannot keep up.

For example, start with the engine. The model is for rent. Very few companies build their own, and your competitor rents the same one a week after you, same price. In Formula One, the 2025 champion team rented its engine from a rival. The edge was everything else.

So what's everything else? Ask a Formula One team what it guards hardest. Data. What this car did at this track, in this heat, on these tires. In 2007, one team was fined $100 million for holding a rival's technical secrets. Every team arrives on Friday with a starting point nobody else can see. That's memory.

Companies are building the same thing now. A company brain. The software for it is open source. If Photoshop has a free knockoff, the brain has one by Saturday night. So the brain is a lease. What you put in it isn't. Your prices. Which customer pays late. What worked last time. That's experience. And it's your starting point.

Here's what that changes on a normal Tuesday. A customer writes: 200 units by Friday, same price as last time? The old way: a spreadsheet, the warehouse, regulatory, the boss. Two days. Four people. And nothing anyone learned gets written down.

Now connect three things. The ERP knows what's in stock and what we charged them. The rulebook knows our price floor, and where we can sell. The memory knows this customer pays late, and that the last Friday delivery cost us a rush fee. The agent puts it together in two minutes. One person signs the one call that needs a person. And what happened goes back into the brain.

One answer like that is a convenience. 10,000 a month, each starting where the last one ended, is a different company. A company that learns once a year improves in steps. One that learns every lap compounds. That's a Formula One team. 300 sensors on the car, over a million data points a second.

Your ERP is the lap time, after the lap. The record. True, but late. The live signals sit outside it. Search demand shifting. Freight rates jumping. A supplier a week behind. A regulator posting a draft. The setup sheet is the rulebook. Every lap ever driven is the memory. The pit wall reads it all, live. The cars are close. The team that learns faster wins. In business, the big teams will always run more laps. You can run better ones.

Here's the part the headlines skip. You can build it. That doesn't mean you'll build it well. Day one. A company drops a tool that costs $2,500 a month. Day thirty. The AI bill. $12,500. A month. Building got cheap. Knowing how and what to build didn't.

I tested it myself. I asked an AI to plan this video. It designed a company brain where a regulatory agent writes rule changes straight in, and every other agent uses them instantly. 9:14 a.m. The agent reads a draft regulation as final. By noon, sales has quoted it to a distributor. Nobody approved any of it. The pipe that spreads a lesson spreads a mistake just as fast. And now the mistake is everyone's starting point.

Formula One drew this line in 2003. Until then, engineers could change the car from the pit wall. The rules banned it. Too expensive, and it was taking the race away from the drivers. The pit wall still sees everything. Only the driver's hands change the car. Agents read everything, and act inside the rules. Only a person changes the rules.

That same season, the rules froze the setup between qualifying and the race. The driver adjusts every lap, against a setup that holds still. Your rulebook works the same way. Only its owner changes it. Rewrite it in real time, and nobody can tell which correction worked.

The plan ended with, you don't manage people anymore. A Formula One team is hundreds of people. What changes is where the CEO sits. On the pit wall, reading the board. Everyone else, people and agents, becomes the loop. And the people become its gates. Reading the board is a skill. You don't get it by hiring someone to implement AI. It takes a different kind of operator.

So I mapped a prototype of a company that runs this way. I call it the telemetry box. Sensors watch the outside. The rulebook sets the limits. Agents propose. Another agent pushes back. A person signs what changes the rules. The ERP keeps the record. And every change gets scored on one number: how fast cash comes back. At the same margin.

So what do you own? Your experience. What goes into the brain, and the record of who put it there. Most of what a company learns is noise. The record is how you find out which lessons were real. Each change, written down before you know if it worked. What you expect, who approved it, when you'll check. Then the verdict. Anyone can copy the software. Nobody can copy that.

The model is rent. The brain is a lease. Judgment and experience are yours. Most companies will not keep up. Will yours? Let's talk.

Sources

Seven Adobe apps rebuilt with AI, open source: github.com/storytold; PetaPixel, Oct 7 2026.

OpenAI building its own GitHub: The Information, Mar 3 2026, via Reuters (reported, not confirmed by OpenAI).

Software stocks down about 25% in six weeks: S&P Software & Services, Jan 12 to Feb 23 2026 (via DeepLearning.AI).

The $2,500 tool and the $12,500 AI bill: one user's report, Computing (UK).

300 sensors and 1.1M data points a second: AWS and Formula 1. The 2003 rules (pit-to-car changes banned, setup frozen between qualifying and the race): FIA, Jan 15 2003. The $100 million fine: FIA World Motor Sport Council, Sept 2007.

Share this post
Tags
Archive