A 30-year-old built a working product with AI last week. In an afternoon. You watched it happen and felt something tighten in your chest.

The thought underneath it goes something like this. The advantage now belongs to people half your age who speak the language of the machines.

I want to take that thought apart, because the research that came out this month says it has the whole thing backwards.

You have spent twenty years getting good. You can walk into a room, read it in ten minutes, and name the one question nobody else is asking. For most of your career, that was the job.

Now a tool drafts in seconds what used to cost you days. And the quiet fear moves in. If the machine produces the output, what is your judgment worth. Is your experience becoming a commodity that software marks down every quarter.

I sat with that exact fear in my first months out of corporate. I assumed the technical people had the head start and I was the one running to catch up.

Then Anthropic studied 400,000 real working sessions to find out what separated the people getting great results from the people getting garbage.

What made the difference surprised them. The people who got great results were the ones who knew their own field on the back of their hands. Technical skill barely moved the needle.

Lawyers, managers, and scientists with zero software background finished the work within seven points of professional engineers. The experienced people hit success rates of 28 to 33 percent. The beginners sat at 15 percent. Same tool. The one variable that doubled the result was the judgment in the chair.

Satya Nadella said the same thing in an essay 65 million people read this month. The AI you can rent is becoming a commodity anyone can buy. The judgment, the relationships, the patterns you read on instinct, that is the part that compounds and the part nobody can copy out of your head.

So the AI agents are starved without you. Point a model at a problem with no expert behind it and you get confident, polished, slightly wrong work. Put your twenty years behind the same agent and it delivers something a client pays senior money for.

The bottleneck was never the technology. It was always how much of the real job the person at the keyboard actually understands. On that measure, you are sitting on the rare asset, while the tools themselves get cheaper by the month.

So stop trying to out-learn the 30-year-old on the tools. You will lose that race, and it is the wrong race anyway.

Do this instead, this week.

1. Pick one problem you solve in your sleep. The kind where a client says three sentences and you already know the questions to ask, in what order, and exactly where most people get it wrong.
2. Open a blank document.
3. Write down how you actually think it through. The questions. The patterns you look for. The calls you make and the reasons behind them.

Then hand that to an AI agent and watch it produce work that sounds like you on your best day.

That document is the first brick of something you own instead of rent. It is the difference between using AI to type a little faster and using AI agents to deliver the thinking that took you a career to build, at a scale your two hands could never reach.

The 30-year-old has the tools. You have the twenty years that make the tools worth anything. That gap is not closing on you. For once, it is opening in your favor.

The only real question is whether you move now, while most of your peers are still convinced they have to become technical first.

Warmly,

Salama

PS: If you are planning your move out of corporate in the next three to six months and want to map your first step, I keep a few 15-minute strategy sessions open each week. Apply here.