Paul Graham
Co-founder of Y Combinator
Y Combinator co-founder and the startup world’s essayist-in-chief. His accelerator now funds overwhelmingly AI startups, and his 2024 essay “Writes and Write-Nots” accepts AI’s dominance while warning it will split society into people who can think in writing and people who cannot.
Why this score
How the formula works →- +4.6proInvests in AI companies
- −1.2anti2 anti-AI statements (measured)
- +0.7pro1 pro-AI statement (measured)
- ±0nuancedHolds a documented two-sided view
Weights, statement counts, the ×1.25 multiplier for having a stake in an AI-impacted field, and the formula’s smoothing are already baked in — the points add up to the score, with evidence on the other side subtracting (direction-free evidence pulls toward the middle instead): Firmly Pro-AI · 4.1/10
In his words
A world divided into writes and write-nots is more dangerous than it sounds. It will be a world of thinks and think-nots.
In a couple decades there won’t be many people who can write.
Biography
Paul Graham (born 1964) sold his startup Viaweb to Yahoo in 1998, then reshaped how technology companies get built: his essays became the founding literature of startup culture, and Y Combinator — which he co-founded in 2005 with Jessica Livingston, Robert Morris, and Trevor Blackwell — became the world’s most influential accelerator, hatching Airbnb, Stripe, Dropbox, and OpenAI’s Sam Altman, whom Graham picked as his successor at YC.
Writes and write-nots
Graham’s stake in AI is structural — recent YC batches have been dominated by AI startups, and he has marveled publicly at founders coding with LLMs — but his most-discussed contribution to the debate is the October 2024 essay “Writes and Write-Nots.” Its prediction: because writing is hard and AI has made it optional, most people will stop learning to do it. Like physical strength after industrialization, clear writing will persist only in those who choose to train it.
Writing is thinking. In fact there’s a kind of thinking that can only be done by writing.
The essay is notable for what it does not say: Graham proposes no ban, no lament, no rollback — he treats AI’s takeover of routine writing as settled fact and simply maps the consequences, closing that he knows which half of the divide he wants to be in. Critics, including many writers, read the piece as fatalism from a man whose fund profits from the disruption; admirers read it as the rare AI take that is neither hype nor panic.
He is just as quick to deflate the boom he profits from. In August 2025 he posted on X that the two most impressive companies in the current YC batch were not working on AI — adding that the lesson was not that AI is unimportant, which he called “very important,” but that founders matter more than the idea. It is a consistent pattern: broad conviction that the technology matters, paired with steady scepticism about the uses and fashions built on top of it.
Quote sources
- paulgraham.com, 2024(paulgraham.com)
Sources & further reading
Canonical record: https://battlelines.ai/topic/paul-graham
