DL 438

Poverty of Attention

Published: July 8, 2026 • 📧 Newsletter

Hi all, welcome back to Digitally Literate. I took off the last two weeks as I left town to visit family. It's good to be back.

Three threads this week, and they're all the same story at different altitudes. Who gets to keep thinking clearly, and who decides.

🌐 From Access to Governance

We spent the last couple of years asking whether schools should allow AI. That question is over. The one replacing it — who governs AI use, for whom, and under what conditions — is now playing out at the scale of nations, not just districts.

The UN's first major report on AI lays out the stakes bluntly. AI is advancing faster than humanity's ability to measure or govern it. "The more AI advances without clear rules, the less say governments and people will have in the outcome," UN Secretary-General António Guterres told reporters at the report's release.

The panel found AI remains concentrated in a small number of companies and countries, with the US and China dominating compute and model development. The report warns this concentration of wealth and capability could enable authoritarian capture and undermine democratic accountability.

The divide isn't just about compute. More than 2 billion people remain entirely offline, and most languages get little to no functional AI support. The report cites a real machine-translation error that turned "you have been given intravenous antibiotics" into "you have been given intravenous insecticides."

Its recommendation for narrowing the gap leans on one lever every country can pull without needing frontier infrastructure. Improving AI literacy in schools and the workforce.

🧾 The Literacy and Attention Crisis Has a Paper Trail Now

If the first thread is about who governs AI, the second is about whether the humans on the other end of that governance can still read and sustain attention long enough for it to matter.

New research from the Fischer Family Trust, built on over a million reading-fluency assessments across 231,000 English pupils, found that a third of disadvantaged white children leave primary school reading below 90 words-correct-per-minute. Put simply, white disadvantaged pupils have lower reading fluency throughout primary than their wealthier peers and disadvantaged pupils from other ethnic backgrounds.

The gap doesn't close during primary school, it's set early and stays set.

If we zoom out to adults, the pattern doesn't reverse, it continues. The 2026 Reuters Institute Digital News Report finds trust in news at a record low of 37%, with social and video platforms now the primary news source globally (54%), overtaking websites and TV for the first time.

The report also showed AI chatbots are creeping into news consumption (7% to 10% this year) but are trusted even less (20%), and only 42% of users click through to an original source.

Younger adults' habits look less like a phase and more like a permanent shift as 56% of 18-24-year-olds have never regularly read a newspaper.

Why this matters: Read together, these aren't two crises. They're one pipeline. A fluency gap that opens in primary school and hardens into an adult population that won't sustain attention on anything a chatbot didn't already summarize for them.

🚪 Wealthy Families Are Betting Against the Classroom

If the first two threads are about governance and decay, the third is about what the people closest to AI are actually choosing when no one's watching the policy debate.

Alpha School, founded twelve years ago in Austin, is expanding fast. Eight new locations in 2025 including San Francisco and New York, with nearly two dozen more planned for fall in places like Palo Alto and Malibu. The model is two hours of AI tutoring that adjusts in real time to student engagement, followed by project-based workshops with "guides" instead of teachers. Tuition runs up to $75,000 a year.

The case for switching isn't hypothetical. A study of 26,000 students found AI-assisted homework scored higher but exam performance dropped up to 24%. 81% of long-term users had simply outsourced their thinking. A UC Berkeley study found the same pattern. Traditional schools have no answer yet for using AI without letting it think for you. Alpha is selling itself as the fix, at a price only the already-wealthy can pay.

Why this matters: The people who can afford it or increasingly looking away from education and toward AI. More specifically, they're betting on supervised AI, deliberately structured to avoid the cognitive-outsourcing trap the research keeps finding. That's a very different bet than what most classrooms, or most AI-literacy policy, are currently making.

💭 Consider

What information consumes is rather obvious: it consumes the attention of its recipients. Hence a wealth of information creates a poverty of attention, and a need to allocate that attention efficiently among the overabundance of information sources that might consume it.

— Herbert Simon

🌿 The Understory

This week is one story about cognitive sovereignty. Who retains the capacity for independent thought, and who controls the conditions under which others get to.

The capacity to think clearly is stratifying by nation, by literacy, by wealth. For now, AI is the mechanism doing the stratifying at every level. Attention is the scarce resource; whoever governs its allocation governs cognition itself.

The sharpest point buried in this issue. The people closest to AI, who understand it best and can afford to choose, are betting against the unsupervised classroom model that AI-literacy policy is largely still selling everyone else. That gap, between what the informed buy privately and what the public is offered, is the real alarm.

See you next Wednesday. As always, my email is hello@wiobyrne.com.

Digitally Literate is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Previous: DL 437 • Archive: 📧 Newsletter