--- title: 'Tech needs humanists more than ever' date: 2026-09-07 lastmod: 2026-09-07 last_updated: 2026-09-07 doc_version: '2026-09-07' description: ' Whoever speaks badly, thinks badly and lives badly. We must find the right words: words are important! —Nanni Moretti A reader reached out to me the other day to talk about the future of tech writing as a profession. They asked me a pretty good question: What skills would you prioritize in this age of constant change and uncertainty? Is it AI? Maybe coding? Well, not really. ' canonical: https://passo.uno/tech-needs-humanists-more-than-ever/ --- # Tech needs humanists more than ever > *Whoever speaks badly, thinks badly and lives badly. We must find the right words: words are important! —*[Nanni Moretti](https://www.youtube.com/watch?v=wAsaOs0iUUk) A reader reached out to me the other day to talk about the future of tech writing as a profession. They asked me a pretty good question: What skills would you prioritize in this age of constant change and uncertainty? Is it AI? Maybe coding? Well, not really. My suggestion is to go back to the humanities and learn a language or two (not the ones you use to code; the other kind). When AI can take care of retrieving facts and putting words together in reassuringly trite ways, your edge as a carbon-based lifeform lies in your ability to go *deep*, have stances, and know how to convey and defend them. You can do that if you train in the dojo of Plato and Chomsky. Let me offer an alternative to learning a trade like plumbing or carpentry. Cast the monkey wrench aside for a moment. Instead, start looking at the intellectual foundations of human culture and consider studying philosophy (of any kind), literature, linguistics (the computational flavor is OK), math, history, or anthropology. If you already hold a degree in that area, it's time to dust it off. And if going back to university is not your thing (I know it’s not mine), self-study is the way to go. If you’re an engineer, adding a bundle of Classics to your skill tree will do you good. Not that degrees guarantee you’ll become a thinker, mind you, but some of it might stick on you still. As counterintuitive as it might sound in this age of reckless tech and unstoppable progress, the only way forward is *down* to the roots of human thought and theory. A growing share of knowledge work is not far from becoming a polite exchange of statements between AI agents, mediated by that patina of human ritual that we call professionalism. Less charitable takes represent the humans in the loop as “[meat proxies](https://dontbeameatproxy.com/).” The way to break these self-imposed, lazy shackles is what it has always been: taking a long detour, paving new paths, and coming up with your own thoughts, however wrong they may feel to you. How to wire things together still matters, of course, but for most developers it has rarely been enough. Many developers are not great merely because of their knowledge of Java boilerplate and Bash tips and tricks, as many tech writers are not great due to their command of Markdown or DITA: professionals are great because they know how to learn fast, how to think hard, and how to [picture whole systems in their minds](https://passo.uno/posts/in-the-docs-team-i-d-build-roles-are-verbs-not-nouns/), including their hairy human factors. You can still decide to become highly proficient at purely technical skills, but this is no longer the only direction for growth in tech, and possibly not the most urgent to cover. Don’t fall into the trap of thinking this is the stuff of academia and ivory towers on leafy campuses. Philosophy of mind, philosophy of science, computational linguistics, cognitive science… They all helped supply the conceptual foundations of the AI systems we’re now building. Logic and ontology reign supreme in software, from database design to information architecture. UX design draws heavily on the arts and anthropological accounts of tools and their use. Projects such as [NOPE](https://nope.net/) turn clinical psychology into standards for safer AI behavior. Linguistics and literary theory can help us find [the source of beautiful docs](https://passo.uno/what-makes-docs-beautiful/). What does this mean in practice? A lot of different and exciting things. It could mean, for instance, learning to read project proposals closely enough to detect bullshit, constructing an argument that survives disagreement, separating evidence from inference, or recognizing when the wheel is being reinvented. None of these are exclusive to academia. You need these skills to question AI output, design information architectures that make sense to humans, and explain products whose hardest problems are more concerned with ideas than with technical challenges. They are the skills that will help you address [real user needs](https://7act.org/). I mentioned languages. Another thing that helps a lot when learning how to think is learning the languages in which thoughts have been created, or not yet created. After learning some basic Persian a few years ago, I’m now gearing up for the somewhat foolish task of learning Mandarin Chinese, not because I expect to be able to work or write in it (that’s wishful thinking), but because the mere act of attempting to think in a radically different language is a majestic Trojan horse carrying culture, history, and different worldviews. The more you expose yourself to different languages and cultural substrates, the easier you’ll navigate problems. So yeah, my advice is to begin your own [Grand Tour](https://en.wikipedia.org/wiki/Grand_Tour). Become the most flexible, [underscore-shaped](https://nested.substack.com/p/short-fat-engineers-are-undervalued) mind you can cultivate, the best-prepared proto-diplomat and translator of cultures, the family philosopher (because you no longer need someone to fix your printer). Become a humanist in tech, like the ones AI labs [are hiring](https://www.economist.com/science-and-technology/2026/06/24/why-big-ai-labs-are-hiring-so-many-philosophers), someone ready to explore the depths of tech products from novel angles. I used to say that the most useful thing I learned was how to find your way in a library. What I’m telling you here is the upgraded version of that advice: learn how to read the map of the world, and then be ready to navigate it using the tools at your disposal. The rest are implementation details. > We have to go back —Several characters from [Lost](https://www.reddit.com/r/lost/comments/1o4vf9x/we_have_to_go_back/) --- Canonical HTML version: https://passo.uno/tech-needs-humanists-more-than-ever/ ## Sitemap - [Full post index](https://passo.uno/sitemap.md) - [llms.txt](https://passo.uno/llms.txt)