Why Paul Graham's essays from 2004 still matter
Paul Graham stopped writing prolifically around 2020. His most influential essays are from 2004 to 2010. And yet they circulate constantly -- passed around in engineering team chats, cited in company memos, quoted in job descriptions.
There is a reason for this. The best of them are not trend pieces. They are attempts to think clearly about things that are structurally hard: how startups die, how good ideas get suppressed, what it means to do work that matters. These problems have not changed.
What makes an essay age well
Most writing about the tech industry is about conditions that are specific to a moment: valuations in a particular cycle, the culture of a particular company, the capabilities of a particular tool. These pieces are useful when they are published. They decay quickly.
An essay ages well when it is about human behavior, institutional dynamics, or the structure of problems -- things that change slowly or not at all. Graham's best work is mostly about these things.
"Keep Your Identity Small" is not about software. It is about how identifying too strongly with any belief makes it harder to think clearly about it. This is a timeless problem. It is as relevant to a technical decision today as it was when he wrote it.
"The Anatomy of Determination" is about the relationship between stubbornness and direction. It applies to anyone doing hard work. It will continue to apply.
The ones that matter most for engineers
Some essays are specifically about the craft of building software and companies. These have aged better than they should, given how fast the field moves.
"Holding a Program in One's Head" is about the cognitive demands of programming -- specifically, how much gets lost when you are interrupted, context-switched, or forced to work in large, poorly factored codebases. The solution he describes is small, coherent programs with few people who understand them deeply. This remains correct and remains ignored.
"Schlep Blindness" is about the way people systematically avoid the hard, boring, unglamorous work that actually creates value. This dynamic has if anything intensified as the tools have made it easier to produce impressive-looking outputs without doing the underlying work.
"What You'll Wish You'd Known" was written as a high school commencement speech. It is about the relationship between curiosity and learning, and the failure mode of optimizing for markers of accomplishment instead of the actual thing. Every generation needs to read this, because every generation falls into the same trap.
What he got wrong
He underestimated how much access to funding and networks matter compared to raw quality of ideas. The model of "build something people want and the rest will follow" turned out to be incomplete in ways he later acknowledged.
His early writing on culture and hiring had blind spots that became more apparent over time.
And some of his generalizations were too confident -- about founders, about ambition, about what distinguishes people who succeed. The categories were real, but the edges were blurrier than he suggested.
Why it still matters
Reading Graham is useful not because he was always right, but because he was doing something specific: thinking carefully about hard problems and writing down what he found. The essays reward slow reading. They reward disagreement. They reward returning to them with more experience than you had the first time.
The ones that have aged best are the ones where he was most careful -- where he was reporting what he observed rather than issuing verdicts. Those are worth your time.
Not everything from 2004 deserves to still be circulating. The best of these do.