Deep Questions with Cal Newport Ep. 370: Deep Work in the Age of AI
1212 snips
Sep 15, 2025 A new study questions the assumption that AI tools universally boost productivity, revealing they can actually complicate deep work. Cal discusses how distractions and idle time may hinder focus even as technology assists cognitive loads. The impact of AI on environmental sustainability is examined, urging for responsible use. In a case study, the role of Wi-Fi access in a West Virginia school is scrutinized, challenging the notion that lack of internet makes students less capable, prompting a deeper look into tech's educational influence.
AI Snips
Chapters
Books
Transcript
Episode notes
AI Can Slow Experienced Developers
- A METR study found experienced open-source developers were ~20% slower when allowed to use generative AI.
- This surprising result challenges the assumption that AI always boosts programmer productivity.
Programming Tasks Are Deep Work
- The tasks studied required deep work: focused, cognitively demanding effort to create or fix code.
- Deep work is the primary source of valuable output in knowledge work, not shallow tasks.
Cybernetic Collaboration Alters Work Rhythm
- Developers used interactive, back-and-forth prompting with AI and spent more time reviewing outputs than actively coding.
- That review-and-wait loop replaced sustained coding focus with fragmented attention.
