No Man Is an Island, but Your Engineering Team Might Be

I read something last week that unsettled me more than most of the AI articles that fill my inbox these days, and it wasn’t about job losses or productivity charts. It was about motivation: why would anyone bother getting good at something, in public, if the thing they’re good at is about to become invisible?

Fernando Borretti’s Argument

The piece is No Man Is an Island, by Fernando Borretti, published on the 9th of October. His argument is narrower than most AI-and-the-future-of-work essays, and that’s what makes it land. I think his focus isn’t on whether AI can write code as well as a senior engineer, but on why anyone would want to get good at writing code, or designing a language, or publishing a library, once there’s no audience left for that work.

His case goes like this. Private intellectual work, the kind that takes years and doesn’t pay the bills directly, only survives because it sits inside a community. You need a shared body of work to build on, and you need some extrinsic reason to keep going when the intrinsic spark runs out, usually some version of being seen and recognised by people who understand what you’ve done. He writes that “private intellectual activity that is sustained, complex, and long-term requires an external intellectual community.” If you take that community away, he doesn’t think you get a smaller version of the same thing. You get nothing: “without community, you don’t get isolated individuals each working on their own things: you get nothing.”

AI, in his telling, dissolves that community from both ends. It can absorb the blog post, the library, the clever explanation, without crediting the author or even necessarily being trained on it in any way anyone could trace. And it can increasingly produce the thing itself, so the writing, the designing, the explaining starts to look pointless before anyone even gets to the credit question. He puts it plainly: “contributing to the commons of software engineering is increasingly pointless.” He’d expected, a year or so ago, to spend his days managing AI agents at work and keep his intellectual life going in his own time. He says that second part “has not quite worked out,” and that the discourse around software engineering has gone “narrower, shallower, and more repetitive” in the time since.

The Numbers Back the Worry

I don’t think Fernando Borretti is being precious about blog traffic, and a study published around the same time gives his worry some empirical weight. Annie Vella and Kelly Blincoe at the University of Auckland surveyed professional engineers who were actively using AI coding assistants, once in October 2024 and again in April 2025, and tracked how their work had shifted across that six months.

I think the real headline isn’t that AI made people faster, though most still reported feeling more productive, but that the shape of the work moved: 82% of respondents said they were spending less time writing code by the second survey, and the balance between creating code and verifying someone else’s output (increasingly the AI’s) shifted significantly toward verification. Flow state declined for more people than it improved for, and the share reporting a worse experience on at least one dimension of developer experience nearly doubled, from 14% to 27%. The one concern that rose sharply enough to be statistically significant was maintainability, up from 3% of respondents naming it as a top worry to 19%.

If you put those two pieces together, you get almost exactly the same argument from two different directions. Fernando Borretti is worried that nobody will build the deep, publicly visible expertise that used to come from writing things up and putting them out into the world. Annie Vella and Kelly Blincoe are measuring the early stages of exactly that: less time spent on the kind of work that builds deep understanding of a codebase, more time spent checking someone (or something) else’s.

Why This Is a Leadership Problem

Here’s where I think it stops being a philosophical worry for people who write essays in their spare time and becomes something every engineering leader needs to take seriously. A lot of how engineers used to build reputation and depth happened outside the building: writing it up on a blog, answering questions in public, shipping a side project, giving a conference talk, getting a pull request merged into something widely used. That external world did two jobs at once. It built the individual engineer’s skill, and it built their visibility, which is how a lot of hiring, promotion and internal trust decisions actually get made, whatever the job description says.

If that external commons keeps shrinking in the way Fernando Borretti describes, I don’t think you can keep assuming your engineers will carry on learning and demonstrating their skill in public on their own time, the way a lot of us did for the last fifteen years or so. If you take that assumption away, you’re left relying entirely on whatever structure exists inside your own walls to do that job instead. For most organisations I’ve seen, that structure is thin: an internal wiki nobody updates, a Slack channel that goes quiet after the first few weeks, a tech talk series that depends entirely on whoever’s keen that quarter.

Building the Commons Inside Your Walls

I think this means the internal version of that commons has to become a deliberate piece of engineering leadership, not a nice-to-have. That’s things like making code review a genuinely teaching-oriented act rather than a gate, giving public credit inside the organisation for the unglamorous library or internal tool that saves everyone else time, running an internal version of a tech blog or lunch-and-learn that actually gets read and attended, and protecting time for people to write things up at all when the default pressure is always toward the next ticket.

None of that fixes Fernando Borretti’s bigger worry, which is really about meaning and motivation rather than organisational design. But I think it’s the part of the problem that sits squarely inside a leader’s job, and it’s worth doing even if you’re sceptical of the rest of his argument. If the outside world stops rewarding people for getting good at this work in public, the only place left that can do it is the place they already spend most of their week.