---
title: "A Team of One"
description: "Everyone asks what AI will take from us. Look at what it is doing right now instead. It is handing one person the reach of a whole team, and the ceiling on what a single human can do just moved."
canonical: https://www.askgib.com/articles/a-team-of-one
last-updated: 2026-09-28
lang: en
---
# A Team of One
> Everyone asks what AI will take from us. Look at what it is doing right now instead. It is handing one person the reach of a whole team, and the ceiling on what a single human can do just moved.
By Gibril Bachouchi. Filed 2026-09-28. 4 min read. Web page: https://www.askgib.com/articles/a-team-of-one
Languages: [en](https://www.askgib.com/articles/a-team-of-one.md) · [fr](https://www.askgib.com/fr/articles/a-team-of-one.md) · [ar](https://www.askgib.com/ar/articles/a-team-of-one.md) · [es](https://www.askgib.com/es/articles/a-team-of-one.md) · [ru](https://www.askgib.com/ru/articles/a-team-of-one.md) · [pt](https://www.askgib.com/pt/articles/a-team-of-one.md) · [it](https://www.askgib.com/it/articles/a-team-of-one.md) · [de](https://www.askgib.com/de/articles/a-team-of-one.md) · [ja](https://www.askgib.com/ja/articles/a-team-of-one.md) · [ko](https://www.askgib.com/ko/articles/a-team-of-one.md) · [hi](https://www.askgib.com/hi/articles/a-team-of-one.md) · [he](https://www.askgib.com/he/articles/a-team-of-one.md) · [zh](https://www.askgib.com/zh/articles/a-team-of-one.md) · [tr](https://www.askgib.com/tr/articles/a-team-of-one.md) · [kab](https://www.askgib.com/kab/articles/a-team-of-one.md)
Everyone is asking the same two questions about AI. Will it take my job. Will it kill us.
Fair questions. But they are questions about the future, and they are so loud that almost nobody looks at the present.
So look at the present. What is AI actually doing, today, to the people who use it well?
It is not replacing them. It is extending them.
A hammer does not build a house. A person with a hammer builds a house that a person with bare hands cannot.
That is the oldest deal in the human story. A tool does not have the idea. It removes the limit between the idea and the thing.
Every tool before this one extended the body. The lever, the plough, the engine. The computer extended memory and arithmetic.
This one extends something we never had a tool for. It extends the part of you that drafts, sketches, checks, translates, tries again. The part that used to require other people.
For all of history, doing something big meant convincing a group to do it with you. You needed the designer, the engineer, the translator, the analyst, the person who reads the contract. If you could not hire them or charm them, the idea stayed in your head.
Now you can have a third arm. An extra brain. A whole team around you that thinks the way you think, because you are the one directing it.
This site is an example. It is one person. It publishes in 13 languages, with figures that move when you scroll. Five years ago that was a small media company. Today it is me and a laptop.
You do not have to take my word for it. Count the people.
Lovable is a Swedish company in Stockholm. It lets you describe an app in plain words and get working software back. In July 2025 it reached **$100 million** in annual recurring revenue. In February 2026 it passed **$400 million**. Its revenue chief gave the headcount at the time: **146 full-time employees**. That is about $2.77 million of recurring revenue per person. It says it has around 8 million users, and it describes them as "non-technical people with great ideas."
Read that last part again. The customers are not developers. They are people who had the idea and never had the team.
Base44 is smaller and stranger. One founder, Maor Shlomo. No outside funding. Eight employees. Six months after he started it, Wix bought it for **$80 million** in cash.
Midjourney has never taken a dollar of venture money. By most estimates it makes around half a billion dollars a year, with a team of somewhere between 100 and 160 people.
Ten years ago, numbers like these needed a thousand employees and three funding rounds. The work did not disappear. It is being done. It is being done by a few people with very long arms.
Companies are the easy proof. Science is the better one, because in science you cannot fake the result.
Paul Erdős was a Hungarian mathematician who left behind a mountain of open problems. A mathematician in Manchester, Thomas Bloom, collected nearly a thousand of them on a public website. Some have resisted the best minds in the field for sixty years.
Since October 2025, about **100** of those problems have moved into the solved column with the help of AI tools. That count comes from a page kept by Terence Tao, one of the greatest living mathematicians.
Be honest about what that means. Most of that help was a very good literature search: the machine found that someone had already solved the problem in a forgotten paper. In at least two cases, though, a model built an original, valid proof of something nobody had solved, with little human input.
Now look at who was holding the tool.
Kevin Barreto is an undergraduate at Cambridge. Liam Price is a former maths student. Both are in their early twenties. In December 2025 they started testing AI on the list, one problem after another. By January 4 they had Problem 728. In May 2026 they were co-authors, with Tao, on a paper resolving Problem 1196.
Wouter van Doorn works in customer service. He is not a professional mathematician. He took six months of leave to do maths because he loves it. By November 2025 he was working with Tao on the forum, and Tao's reply to one of his arguments was short: "I see how your argument works now. Nice!"
A customer service worker and a Fields medalist, on the same problem, on the same page.
The tool did not replace the mathematician. It let more people be one.
I am not going to sell you a fairy tale, so here is the other side.
It does not give you taste. It does not give you direction. It does not tell you which problem is worth a year of your life. Point it at nothing and it produces a great deal of nothing, very fast. An extension of you is only as good as you.
And the fear about jobs is not stupid. Look at the same numbers from the other side. If 146 people can run a $400 million company, then a lot of people who would have been hired were not. The leverage is real in both directions. Pretending otherwise is how you lose the argument.
But notice what the fear gets wrong. It imagines AI as a rival, a second worker standing next to you, competing for your chair. That is not what the people above experienced. None of them was replaced by a model. Each of them became more than one person.
The undergraduate became a research group. The founder became a company. The customer service worker became a mathematician.
Here is what I think we had wrong the whole time.
We thought the limit on a person was talent. Mostly it was hands. Hours. The number of skills one life has room to learn. The number of people you could pay or persuade.
That limit just moved, for anyone willing to pick the tool up.
So the useful question is not what AI will take from you.
It is what you would build if you had a team, because you have one now.