In the winter of 1961, a meteorologist named Edward Lorenz cut a corner. To save time restarting a weather simulation halfway through, he typed in 0.506 where the computer had been holding 0.506127, three decimal places he was certain could not matter. The forecast that came back was not slightly different from the first run. It was a different world entirely. That small, careless rounding became one of the most important accidents in the history of science.
He had stumbled onto what we now call the butterfly effect, the idea he would give a name a decade later in a talk asking whether the flap of a butterfly's wings in Brazil might set off a tornado in Texas. Tiny differences in where you begin can produce enormous differences in where you end up. The system is not random. It is deterministic, governed by fixed laws, and yet so exquisitely sensitive to its starting point that prediction runs into a wall you cannot see over.
We have made our peace with that paradox almost everywhere. We plot planetary orbits decades ahead, model a failing heart well enough to catch the fatal rhythm before it arrives, and forecast tomorrow's weather with something near ninety percent accuracy a day or two out, even as the ten-day outlook stays close to a guess. This is the part people miss about Lorenz: he did not make the weather more predictable. He showed us precisely how far ahead our certainty reaches before chaos takes the wheel. The discovery was not a tool. It was a lesson in humility about the limits of knowing.
And there is one system whose path we cannot plot at all, which we are nonetheless building as fast as our hands will move. We cannot reliably say how artificial intelligence will reshape work, cognition, power, or the quiet machinery of human attention. We are altering the initial conditions of an entire civilisation while waving away the decimal places we are sure cannot matter. That should humble us. Mostly, it does not.
Every System Amplifies Its Values
Here is the line I keep returning to: every system amplifies the values embedded in it. A platform built to capture attention will get better and better at capturing it. One built to maximise engagement will learn, without anyone instructing it, that outrage engages. When technology is steered first by unchecked incentives, attention economics, and profit at any cost, the butterfly effect stops being a curiosity in a lecture hall and becomes a force, multiplying small design choices into social, psychological, and economic weather that none of us voted for. The danger was never innovation itself. It is who sets the initial conditions, and with what intent.
A system inherits the conscience of whoever built it, then runs that conscience at a scale no single person could ever reach. Which is why the question of who we hold up as worth admiring is not a soft one. It decides, quietly and years in advance, what the next set of systems will be built to amplify.
Whose Eyes Are Worth Borrowing
There was a time a young person aimed to see the world through the eyes of Isaac Newton or Albert Einstein, Alan Turing or Nikola Tesla, people who rearranged our understanding of gravity, time, computation, and energy. If you are early in your own career, the most useful question I can hand you is this one: whose eyes do you actually want to borrow?
Because the names we are handed now are mostly Elon Musk, Mark Zuckerberg, Sam Altman, Jensen Huang, Bill Gates. In my humble opinion, role models defined first by net worth are the wrong teachers. They are, in the end, salesmen, gifted ones, and each is pursuing a monopoly over whatever he is building. Admire the engineering if you like. Be careful about borrowing the eyes, because you tend to inherit the values of the people you choose to see through, not the ones they put in the press release.
If I were choosing whose eyes to see through, I would choose the people who shape how we think, not only what we buy, and the people now asking the hard questions about the machines we are racing to ship.
They spent careers arguing that how you build is a moral choice, not only a technical one. Clean code, refactoring, test-first, the wiki, the idea of technical debt: the vocabulary a whole profession uses to talk about doing the work well.
They do the unglamorous work of asking what these systems might cost us before the bill arrives. Some helped build the field; some warn about where it leads; all of them treat the question as serious rather than settled.
This is not a list of accounts to follow for the latest news; I have mapped that separately. It is a list of people worth thinking like. They do not just build systems. They shape how we reason about systems, which is the rarer and more durable gift, and the one closest to what I mean by treating the work as a craft. Several of them have spent the last few years naming the risks the salesmen would rather we did not dwell on.
The Systems Inherit Us
The universe runs on laws we did not write and cannot break. The systems we build are a different kind of thing. They run on laws we choose, line by line, default by default, incentive by incentive, and they will amplify whatever we place inside them, faithfully and at scale, long after we have forgotten we ever chose it. A culture is a system like that. So is a company, and so, in the end, is a career. You become, slowly, the values you decide to keep close.
So if you are just starting out, take this as the practical part. The people you choose to admire are the initial conditions of who you are going to turn into, and like all initial conditions, the difference looks tiny at the start and enormous by the end. Turing is still worth more of your attention than a market cap. Choose your role models with the care Lorenz wished he had given those three decimal places.