Digitally Literate
Enter your search text in the box above

Select a result to preview

DL 445

The Other Side of the Wall

Published: August 26, 2026 • 📧 Newsletter

Hi all, welcome back to Digitally Literate.

We spend a lot of time thinking about what we put into our technologies. The questions we type. The writing we publish. The information we share. The places we go.

We spend considerably less time thinking about what happens after.

This week, teens describe what they're gaining (and giving up) as AI becomes part of schoolwork. New research suggests more than a third of the recently published web content may be machine-written. And communities across the country are reconsidering networks of cameras that quietly turn our movements into searchable data.

These seem like different stories. But each asks us to pay attention to what happens on the other side of the wall.

As always, your support is valued. Reach out anytime at hello@wiobyrne.com, and subscribe if you haven't already.

Teens report using AI in very nuanced ways

Common Sense Media's Youth AI Safety Institute surveyed a nationally representative sample of 1,017 teens, aged 13–17, and found 70% now use AI for schoolwork. Among them, 63% said they use it to get an answer rather than figure it out themselves. 35% said they revise an AI answer with what they already know, 31% rewrite it to sound like them, and 25% take it as-is.

The split isn't as clean as "cheaters" versus "responsible users." There's a lot of interesting texture in this study.

Most fascinating to me was the 66% who said AI helps them understand the schoolwork, not just finish it faster.

A third of the new web may be machine-written

New research from the Pew Research Center reviewed nearly half a million English-language pages from Common Crawl from the past five years. They started looking a couple of years before ChatGPT's advent. They ran all of these archived webpages through an AI-detection tool called Open Pangram to see how much of the content was written by AI rather than by humans.

More than one-third (35%) of the post-ChatGPT pages showed signs of AI authorship, echoing other studies suggesting that artificial intelligence is shaping a large part of the modern web. This is in line with other studies showing that a large share of recently published pages on the Internet were likely written or substantially edited by AI.

Most of the response to this is a focus on AI in our lives. I'm thinking about this a bit differently. First, as a content creator, I'm thinking about what this means for how I share, connect, and participate online. Second, I agree that it's a challenge as we see the quantity skyrocketing, but we also see the quality rising as well. It's strange, but it seems like quality, formatting, and grammar are the elements people identify as AI.

If you're wondering, AI models learn language by analyzing enormous collections of human writing. These collections are not a perfect representation of how people actually write. Rather than generate text, AI detector tools like Pangram are trained to spot the statistical fingerprints that AI models leave behind. By comparing a piece of writing against patterns learned from millions of human- and AI-authored samples, they estimate how likely it is that a given passage came from a model. This piece from the NY Times and this recent piece in Nature are great resources.

The Flock Is Watching

You've probably driven past a Flock camera over the last week without noticing. These are small, solar-powered cameras bolted to a pole that photograph every car that passes. These automatic license plate reader (ALPR) cameras are mainly offered in the US by Flock Safety, which operates more than 120,000 license plate readers and has contracts with numerous law enforcement agencies.

Flock cameras don't just read license plates. They capture the make, model, bumper stickers, decals, and roof racks. Covering your plate won't hide you. This is vehicle tracking, not plate tracking, and a camera in your town quietly becomes one node in a nationwide mass surveillance system.

The abuses are no longer hypothetical. Documented abuses of Flock's surveillance systems include officers using them to stalk romantic partners and acquaintances, law enforcement tracking people connected to reproductive healthcare, police monitoring protesters exercising First Amendment rights, and biased or unvetted searches enabled by out-of-state agencies cross-searching local databases with little oversight.

Communities are revolting. Over 50 cities and counties cut ties with Flock in 2026.

Flock has responded by stating that the country needs to find a compromise between privacy and safety. Flock announced several operational changes to this criticism. As we've seen before with other tech companies, we need to question their explanations of what they are or aren't doing.

This is an opportunity to talk about whether you think there should be a compromise between your privacy and safety. Are there opportunities for more transparency while also accountability for the companies that offer these products? Most importantly, we need to be more thoughtful about the data trail that we leave behind as we move through the world.

The Understory

In the Venetian Republic, stone carvings, often depicting a lion’s face, adorned the sides of government buildings. Called the bocca di leone (“lion’s mouth”), it was essentially an anonymous complaint box with a slot open in the mouth.

A Venetian with information about wrongdoing could write an accusation on a piece of paper and slip it through the mouth. On the other side of the wall, government officials collected the denunciations and decided what to do with them.

There were rules. Accusations generally required witnesses or corroboration before authorities could act. But the system became one small part of the elaborate machinery through which the Venetian Republic gathered information about what was happening within its territory.

What interests me is the interface. For the person standing in front of the bocca di leone, there wasn't much to see. A face carved into stone, an open mouth, a narrow slot. All of the machinery was on the other side of the wall. We've built considerably more sophisticated mouths since then.

We type questions into boxes. We publish words and images on the web. We drive beneath cameras mounted on poles. Sometimes we're deliberately handing over information. Sometimes we're creating information simply by moving through the world.

Most times, we don't know what happens on the other side. With the lion's mouth, you could see yourself feeding the machine.

The question isn't only what information these systems collect. It's what becomes possible once that information passes through the mouth. Who can retrieve it, what it can be connected to, what conclusions can be drawn from it, and who gets to act on those conclusions.

The lion's mouth was at least honest about one thing. You could see yourself feeding the machine.

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 444 • Archive: 📧 Newsletter

🕸️ Connected Concepts: