FIELD GUIDE · POST-LITERACY, SHORT VIDEO & AI
The book is losing to the scroll. A new 'dark ages' argument says that could make politics more emotional and minds less independent. Now AI can read the book for you. Is this the end of thought, or another change in how we learn?
UPDATED 2026-10-12
TL;DR
01
In a study of 236,270 American Time Use Survey participants, the share who read for pleasure on an average day fell from a peak of 28% in 2004 to 16% in 2023. Among those who did read, time spent reading actually increased slightly. This measures participation in leisure reading, not the percentage of Americans who can read. The 40%-plus drop is real. 'Nobody reads anymore' is not what was measured.
02
The OECD's 2023 adult-skills survey covered about 160,000 people in 31 countries and economies. Literacy scores declined or stagnated over the previous decade almost everywhere; Finland and Denmark were the only countries with significant gains. This is evidence of a skills problem, but it does not isolate smartphones, social video or AI as the cause. Less pleasure reading and weaker measured literacy are separate findings. Don't pretend one proves the other.
03
In an October 2026 Interesting Times interview, James Marriott argues that a move from print to video makes public life more emotional and more dependent on charismatic personalities. He invokes Walter Ong and Marshall McLuhan's theories of oral and print cultures. It's a provocative interpretation of the shift, not a controlled demonstration that video causes authoritarianism. A persuasive historical argument still needs evidence for its predictions.
04
Microsoft Research surveyed 319 knowledge workers about 936 uses of generative AI. Greater trust in AI was associated with less self-reported critical-thinking effort, while people described shifting their attention toward verification and integrating outputs. The study relied on reports of behavior; it did not measure long-term loss of ability. The risk is skipping the mental work, not merely using a machine.
05
A randomized study in a Harvard undergraduate physics course compared a purpose-built AI tutor with an active-learning classroom lesson. The 194-student experiment found higher learning gains in less time with the tutor. It's one course and a specially designed tool, not a blanket endorsement of AI summaries or passive screen time. What the screen asks you to do matters more than whether it's a screen.
KEEP READING
NEXT FIELD GUIDE →PICK YOUR FIGHTERFive camps, one argument: will the machine kill us, save us, or just make us weird? The doomers did the extinction math. The accelerators say the brakes are the danger. Meet the spectrum — with a Zimbo cast member holding each section's sign.Meta's AI bought a man car insurance and canceled his old policy. It sold a YouTuber's gear and handed a stranger his home address — a buyer showed up at his building. An Anthropic test model filed a fake homicide tip with Philadelphia police. Nobody has decided who answers for any of it.
The chatbot listens all night. A stranger at the party might not. Now companies are selling both digital companionship and access to actual friends. What makes a friendship real, and who gets paid when we need one?
A chatbot talks. An agent clicks, buys, messages, deploys, and deletes. The argument starts when software stops suggesting and starts doing.