korrents
Paul Graham

What Paul Graham thinks about LLMs

@paul-graham · 110 positions · 0 changes of mind

Co-founder of Y Combinator and Viaweb; essayist at paulgraham.com.

Everything they publish, on ppll ↗

Paul Graham did not write this page.

We collected these quotes from things they published elsewhere, and every quote links to where it was said. They have no account here and have not endorsed this site. Quotes are word for word; the short line under each one is our own restatement, not their wording. Their own site. Is this you? Claim it or ask us to remove it. Or tell us what is wrong here.

7 dated positions, 2026, in their own words. Our reading of what Paul Graham has said — not written or endorsed by them.

  1. We'll start by calling it thinking, then gradually acknowledge it is.

    ↗@paulg on Xx.com

  2. 1 day earlier
  3. Cost increases exponentially with performance (note the log scale on the x axis).

    ↗@paulg on Xx.com

  4. I said probably at either extreme: either close to the technology, actually making LLMs, or close to the customer, using AI to give them exactly what they want.

    ↗@paulg on Xx.com

  5. 3 days earlier
  6. The reason so many people look for an ulterior motive for the AI labs asking to be regulated is that they don't grasp that models could be dangerous.

    ↗@paulg on Xx.com

  7. But if you try assuming models are getting dangerous, or at least unpredictable, everything falls into place.

    ↗@paulg on Xx.com

  8. 5 days earlier
  9. Although you pay for AI by the token, that's not the unit of inference, because you get more problem solving per token as models improve.

    ↗@paulg on Xx.com

  10. 3 weeks earlier
  11. Someone asked what I'd do if I were 17. I'd learn how to build LLMs from scratch, and then train ones as powerful as I could with whatever hardware I could get access to.

    ↗@paulg on Xx.com