I've been meaning to write about this since [2026-03-20 Fri]. finally doing it.
Quick note on the title before I get into it. It's riffing on Linus Torvalds' old line, "talk is cheap, show me the code". Except these days even that doesn't hold up, talk is cheap sure, but code is cheap too now, anyone can get an LLM to spit out a working function in seconds. What's actually hard to fake, and what most of what I read (and write) is missing, is the source: the history, the reasoning, the "why" behind a piece of code or a life's work, not just the diff. That's what this essay is actually about, so show me the source felt more honest than show me the code.
To put it simply: Business gives us endless Talk for free, AI gives us endless Code for free, but the Source, the real reason behind any of it, can only come from us, Humans.
So I read this essay by Blain smith and it stuck with me for months. The whole point of it is that the people who actually built computers as we know them were not "just" engineers. Hoare studied classics and philosophy before he ever wrote quicksort. Lovelace called herself a poetical scientist. Turing published his most famous paper in a philosophy journal, not a computer science one. Dijkstra wrote actual essays, by hand, instead of just writing code comments. None of them were pure "tech pals", they came from somewhere else first.
Honestly the thing that makes the essay slap me is not just "look, philosophy guy also did tech", its that the stuff they built actually held up. Quicksort is still the sort you get taught first, decades later. Dijkstra's shortest path algorithm is quietly running every time you open maps and ask for directions. C is still the grandparent of basically every language people use today, and unix is still under the hood of most servers and phones on earth. hopper's idea, that code should read like english (I favor code as docs too), is why every language after her looks the way it does instead of staying pure math symbols. None of this stuff got replaced because someone found a "better" idea next year, it just kept being the right idea. That's rare. Most code I see in industry or tool these days doesn't survive six months, let alone sixty years, every few year they all get replaced with better alternative or solution (grander eg: rust hype for everything)
What got me thinking is, i'm still just a novice in this field. Most of what I read day to day is lobste.rs, mastodon, docs, and youtube. Nobody ever tells you to go read philosophy or history to get better at writing code. Its always "learn this framework", "learn this language", "here's how to solve this", heck nobody even bothers telling how to think of solving.
The essay also says something that annoyed me a bit, in a good way. It says a lot of today's tech, social media, algorithms, all of it, is built by people who never studied ethics or psychology or how to write clearly, and thats part of why it feels careless. Like nobody stopped to ask "should we build this" and only asked "can we build this", they should've asked "Does it solve a problem", instead it went into "We create a problem and then solve it". I don't fully know if i agree with that, but it does feel true when i think about doomscrolling or apps that are clearly designed to be addictive on purpose, economy & business…!
The ending is basically: you don't need a philosophy degree, you just need to actually read stuff outside of tech. Books, essays, history. Let it change how we think about what we build, not just how to build it technically.
I genuinely don't know where i land on all of this yet. But it did make me want to read more outside of programming books, and maybe slow down a bit before shipping things.
Some more amazing resource to check: Rethinking the Developer Career Path – Randall Koutnik (slides) ,Learning to Manage - Building a shared foundation, Taste is all thats left and Write for people.
Little cheat sheet of everyone the essay talks about, one line each, with a link if you want to actually go read them instead of taking blain's word (or mine) for it.