Skip to Content

Essay

AI Is a Skill

The best AI models are open to the entire world.

Same Engine

The question I hear most about AI is which one to buy, usually meaning which assistant or subscription the whole team should be using. Buying is how most businesses have always gotten new technology, and hundreds of AI products are competing to be the answer.

Which one you pick changes less than people expect. Nearly every AI product runs on models from a handful of companies, so your competitors can use exactly what you use, on the same day. It's foolish to believe that having access to AI means anything if you don't know how to actually wield it.

Anyone Can Start a Business

Anyone in the United States can start a business for a filing fee and an afternoon. Everything there is to know about running one is a search away, and the idea itself is simple: sell something for more than it costs you to make. People who have built one know how much harder it is than that. Roughly one in five new businesses doesn't make it through its first year, and about half are gone within five. All of them started with that access and that information, and the results came down to the person running it.

Anyone can own a basketball, too, and almost nobody who does plays in the NBA.

Practice is part of the difference. So are talent, intelligence, creativity and acumen, the judgment to know what to do next, all coming together in one person. Time is part of it too: the hours and energy a person can pour into getting better, and the resources to keep doing it year after year. None of it is spread evenly, in business or with AI.

The Engine Is for Sale

Formula 1 shows it with machinery. A team is allowed to buy its engine, and the rules require it to design its own car: the chassis and bodywork that decide what all that power actually does on a track. That's why several teams often race with identical engines.

In 2024, four teams used the same Mercedes power unit. McLaren won the constructors' championship with it, its first since 1998. Mercedes' own team finished fourth. Aston Martin finished fifth, and Williams finished ninth, with 17 points to McLaren's 666.

The Person at the Helm

It holds across the whole automotive world. An auto mechanic and an engineer at a carmaker work from the same physics as the team that wins in Formula 1, and they can read the same books. On paper they share a skill, and the results are radically different, because the person at the helm is the difference.

The most successful designer in the sport's history is Adrian Newey. Cars he designed have won more than two dozen world championships at three different teams, starting with Williams in the early 1990s and running through McLaren and Red Bull. The championships followed him from team to team.

The Craft Around the Model

AI is that kind of skill. The models are the engine. What separates one company's results from another's is the car: the software and rules built around the model, and the talent and judgment of whoever designs them.

Average results from AI take very little skill, which is why so many people get them. Used carelessly, it can even make a team slower while everyone feels faster, which I wrote about in Feeling Fast Isn't Being Fast. Exceptional results come from knowing what the models do well and where they fail, which decisions belong to a machine and which should be made once and written down as rules, how to give AI the information it needs while keeping control of everything it touches, and how to test its work before anyone relies on it.

Like any craft, it compounds. Someone with the talent for it who designs with AI every day for a year ends up so far ahead of someone who bought a subscription that they're no longer in the same race.

Off the Rack Is How You Keep Up

You can buy a road car with a Formula 1 engine in it. Mercedes built 275 of them, the AMG ONE, and sold them for about $2.7 million each. It's an extraordinary machine, and it will never start a Grand Prix.

Most of the AI products sold to businesses are built like that: a great engine inside a car designed for everyone. A product sold to millions of businesses has to work for the average one, so it's built for the average job, and its settings mark the limit of how far it will bend to fit yours. Own the Platform goes into why the big products are built for the average. They're priced for the vendor's business too, usually per seat and per month for every person on the team, whether they use it once a week or all day, so you pay for the engine plus everything the vendor needs to earn on top of it.

And anything you can buy, your competitor can buy tomorrow. An off-the-rack tool makes you faster at the same rate it makes everyone else faster, so the best it can do is help you keep up. Keeping up is worth something, and in a race it gets you a finish somewhere in the middle of the field.

This Happened With the Internet

The early internet looked like something a business could buy, too. In 2001, Borders, then the second-largest bookstore chain in America, handed its website to Amazon. Amazon ran the online store and kept the customers. Borders took its website back in 2008, and in 2011 it closed for good.

Plenty went wrong for Borders by the end, and one of those things was a seven-year decision to let the most important new skill in its industry grow up inside a competitor.

The Most Valuable Person on Your Team

The most valuable person a company can have right now is someone who is legitimately one of the best in the world at engineering and designing software with AI. That person takes the engine and designs a car around one company and the way it actually runs.

They also make the AI itself cost less. Someone who knows the engine sends each job to the least expensive model that can do it well, and turns every decision that only needs to be made once into a rule that runs at no cost at all, the idea behind Decide Once.

And the car they build is the one part of your AI that a competitor can't buy, which is the whole definition of an unfair advantage.

Who Designs Your Car

Start by treating the models as the engine. Use the best ones you can find and switch whenever a better one comes out.

If you're running a small business and trying to make a living, off the rack makes sense. You're not going to be able to afford anyone with real, legitimate talent at this, and the tools sold to everyone will keep you in the race.

If you're trying to build something truly special, or even change the world on some level, average won't cut it. It takes brilliance, amplified by technology that lets you scale now, and the right people in place to keep you ahead at all times. Finding them is hard. The internet is rife with "AI experts" who don't actually build anything, have never built a legitimately successful business, don't understand consulting or how to adapt technology beyond what they already know, and have never had to live with the results of their own advice. Judge anyone you're considering by what they've built and what's running today.

It's the reason I'm brought into every venture I'm part of. I know it sounds self-serving. I own and partner with companies in some of the largest industries that exist, alongside some of the smartest people in the world, and in the last year I've built and run more businesses than the overwhelming majority of entrepreneurs launch in a lifetime.

Teaching this is a big part of what Timatsu, my platform, is built for, so you can start wherever you are: as an entrepreneur looking to give yourself an edge, or as someone who wants to learn the strategy and the mindset it took to build what I've built.

Get Early Access. Timatsu 2.0 opens in January 2027. Leave your name and email and I'll tell you the moment the doors open.

You'll see the price before anything charges.