korrents
Cal Newport

What Cal Newport thinks about LLMs

@cal-newport · 96 positions · On ppll ↗

Computer science professor at Georgetown and author of Deep Work, Digital Minimalism and Slow Productivity. He has written the blog Study Hacks since 2007 and contributes to The New Yorker.

Cal Newport 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.

9 dated positions, 2026, in their own words. Our reading of what Cal Newport has said — not written or endorsed by them.

  1. So for a lot of people, the reward shortcut being offered by generative AI tools is we can jump straight to feeling like we have acquired or produced new knowledge.

    ↗Why Are Harvard Students Lining Up for This “Brutal” Class?youtube.com 7th of 15 in this recording

  2. I know a lot of people that say why would I ever read a book anymore? I could get a summary from AI. But what they don't understand is the book is the optimal summary. The book is incredibly condensed.

    ↗Why Are Harvard Students Lining Up for This “Brutal” Class?youtube.com 11th of 15 in this recording

  3. 3 days earlier
  4. There was multiple vulnerabilities in PHP found earlier this year. But because an AI model was involved in this one, it caught a lot of attention

    ↗The Normie’s Guide to Thinking About Techyoutube.com 15th of 27 in this recording

  5. AI agents

    these new sort of personal AI agents which I think you got to be super careful about because again LLMs are very unpredictable and LLM driven agents be very careful about using them.

    ↗The Normie’s Guide to Thinking About Techyoutube.com 16th of 27 in this recording

  6. chat interfaceseducation

    if you go back and forth on things that have a uh like a very logical or clear like right and wrong answer and something it's trained on, you get tutoring out of it.

    ↗The Normie’s Guide to Thinking About Techyoutube.com 20th of 27 in this recording

  7. 8 weeks earlier
  8. AI is not the same thing as LLMs. Many of the most impressive AI systems that exist today - including those that beat humans at games, win Nobel prizes for studying proteins, and self-drive cars through crowded city streets - have very little to do with LLMs.

    ↗Superintelligence is a Fairy Tale. But Chasing It Can Still Cause Harm.calnewport.com 2nd of 2 in this piece

  9. 1 day earlier
  10. Most of the things people already like doing with LLMs, or hope LLMs will enable soon, don't require these haphazard and unpredictable setups.

    ↗Anthropic Just Threatened to Kill Billions of People. This Is Not Okay.calnewport.com 1st of 3 in this piece

  11. AI alignmentOpenAI

    I don't believe that LLM-powered agents, on their own, can become superintelligent and create existential risk, as LLM capabilities are more jagged and limited than the average person reading triumphant OpenAI press releases might realize.

    ↗Anthropic Just Threatened to Kill Billions of People. This Is Not Okay.calnewport.com 2nd of 3 in this piece

  12. AI alignment

    But you don't need superintelligence to create a super amount of harm. The combination of unpredictable, non-normative output from unrestricted LLMs and powerful digital tools will inevitably create issues.

    ↗Anthropic Just Threatened to Kill Billions of People. This Is Not Okay.calnewport.com 3rd of 3 in this piece