Daniel sent us this one — he was talking with his uncle, who mentioned his daily ritual of picking up the physical newspaper from his doorstep and reading it. Daniel thought he was being allegorical. Turns out, no, the man actually gets it delivered. Which raises two questions. How much printed news is still actually in circulation — have we hit the bottom of the decline yet? And the bigger one: does reading a newspaper versus skimming Google News four or five times a day actually shape how we understand the world differently? He wants to know if there are studies on this, and for those of us not planning to go back to dead trees, what we can do to get that broader exposure without giving up digital.
The uncle as the canary in the coal mine of information architecture. I love it. So let's start with the numbers — just how many people are still getting news on dead trees?
I genuinely thought the answer was going to be "your uncle and maybe twelve other people." Like, I assumed we'd hit a floor years ago and the remaining holdouts were just people who forgot to cancel.
It's not zero. Pew Research put out data in November twenty twenty-four — only five percent of U.adults regularly get news from print newspapers. That's down from sixteen percent in twenty twenty. So it's still shrinking, but the rate has slowed to single digits annually. Total daily newspaper circulation in the U., print plus digital combined, fell from about fifty-five million in two thousand to roughly twenty million in twenty twenty-two. The cliff dive of the two-thousands has turned into more of a gentle slope.
That's one in twenty adults. I'm trying to picture that — it's not nothing. It's roughly the same percentage of people who are left-handed, or who have red hair. You notice them, but they're not the default setting anymore.
The demographic is extremely specific. Median age sixty-five plus, highly educated, geographically concentrated in rural and suburban areas with reliable delivery. These are not people who forgot to cancel their subscription. They're making an active choice. Your uncle fits the profile.
He absolutely does. But Daniel's real question isn't about demographics. It's about what's happening inside your head when you read a broadsheet versus when you scroll through an aggregator. And this is where it gets interesting, because the cognitive science on this is not what I expected.
It's not even close to what most people assume. So let's go there. The mechanism is fundamentally about linear versus non-linear reading. Print is linear, sequential, tactile. You turn pages. You encounter stories in the order an editor decided. You can't click away. The physical layout — front page above the fold, section breaks, A-section versus B-section — creates a curated hierarchy that you cannot easily skip. You might not read the story about the municipal zoning board, but you saw the headline. Your eye passed over it. It registered at some level.
That's the thing I think we undervalue — the passive exposure. The story you didn't click on but your brain noted existed. I think about this with my own habits. When I scroll Google News, I'm actively selecting every single thing I read. There's no equivalent of turning the page and having a headline about soybean tariffs just... appear in my peripheral vision.
That's not a small difference. It's the difference between a curated museum tour and wandering through a city where someone else chose the streets. Both have value, but they produce different kinds of knowledge.
What does the research actually say about comprehension? Are print readers understanding more, or is this just nostalgia dressed up as science?
And the research backs this up. The University of Valencia did a study in twenty nineteen, and the Norwegian Reading Centre followed up in twenty twenty-one. Both found that print readers demonstrate measurably better comprehension of narrative structure and chronology in news stories compared to digital readers. They understand the sequence of events better. They can reconstruct the timeline of a developing story more accurately. The physical medium seems to anchor the information in a way that screens don't.
What's the proposed mechanism there?
There are a few theories. One is that the tactile experience — the weight of the paper, the act of turning pages — creates what cognitive scientists call "spatial mapping." Your brain associates information with physical locations. You might remember that the story about the trade negotiations was on the left-hand page, below the fold, next to the op-ed about energy policy. That spatial context becomes a retrieval cue.
It's like a memory palace, but the palace is the actual newspaper.
That's not a bad way to put it. Another theory is that print reading eliminates what researchers call "cognitive switching costs." On a screen, you're constantly making micro-decisions — do I click this link? Do I scroll past this? Do I open a new tab? Each of those decisions costs a tiny bit of cognitive bandwidth. Over a thirty-minute reading session, that adds up to a significant reduction in the mental resources available for actual comprehension.
The medium is literally taxing your brain in ways you don't notice, and print just...
Imagine you're reading a long investigative piece on your phone. You hit a paragraph that references an earlier scandal. Do you open a new tab to look it up? Do you keep reading and hope context fills it in? That micro-decision takes maybe half a second, but it interrupts the flow of comprehension. Multiply that by every article in a session. Print doesn't give you the option to branch off, so you stay in the narrative.
It's almost like the constraints of the medium are protective. The thing that feels like a limitation — you can't click away, you can't open a new tab — is actually preserving your attention.
Then there's the eye-tracking data, which is humbling if you're a digital native. Studies show that digital readers spend about sixty percent of their time above the fold — that first screenful. They rarely scroll to the bottom of articles. The pattern is scan, skim, move on. Digital readers actually consume more stories in a given session than print readers. But they get less depth per story. It's breadth without depth.
More headlines, less understanding. That tracks with how I feel after a long scroll session. I've "read the news" but I couldn't explain the timeline of any single story. I've got fragments — a quote here, a statistic there — but no coherent narrative.
That's before we even get to the aggregator effect. Because the filter bubble mechanism in something like Google News isn't just about what you tell it to show you. It's using location data, search history, click patterns, dwell time — all of it feeding into a model that decides what you see and in what order. Nieman Lab reported in June twenty twenty-five that AI-driven personalization is now the default in most major news aggregators, and users rarely adjust their personalization settings.
Daniel mentioned the push by publishers to get readers to add them as a preferred source. I've seen this too — every news site now has that little banner. "Add us as a preferred source on Google News.
Which sounds great. It's framed as user empowerment — choose your trusted sources, take control of your news diet. But the knock-on effect is the opposite. The Oxford Internet Institute published a study in twenty twenty-four showing that users who added three or more preferred sources saw forty percent less diversity in their news feed within two weeks. The algorithm interprets your preference as a signal to narrow, not broaden.
Forty percent less diversity in two weeks. That's not subtle. That's the algorithm doing exactly what it's designed to do — maximize engagement by showing you more of what you already agree with — and calling it personalization.
Here's the part that concerns me. The Reuters Institute did a study in twenty twenty-three that found forty-seven percent of aggregator users believe the algorithm shows them "all the important news." Compared to twenty-two percent of people who go directly to news sites. The aggregator becomes a proxy for reality. If it's not in the feed, it didn't happen.
That's the moment where the tool starts using you. When you stop asking "what did the algorithm miss" and start assuming it missed nothing.
That's not a failure of the user. It's a feature of the design. The interface doesn't have a "what am I not seeing" button. The seamlessness is the point.
Let me push back on something, because I can feel the argument tilting toward "print good, digital bad," and I don't think that's quite right either. Print newspapers have their own filter. It's called an editorial board. The gatekeeping is just human instead of algorithmic. You're still not seeing everything.
No, that's fair. The difference is transparency and consistency. With a print newspaper, you know who the editor is. You can learn their biases over time. The same editorial team curates the whole paper, so the worldview is at least internally coherent. You can disagree with it, but you can also map it. Algorithmic gatekeeping is opaque — you don't know why a story surfaced or didn't — and it's inconsistent. The ranking changes minute to minute based on signals you can't see.
It's not that one is biased and the other is neutral. It's that one bias is legible and the other isn't.
That's the distinction. If you know the New York Times editorial board leans a certain way, you can compensate. If you don't know what Google News is doing under the hood, you can't.
The legibility extends to the corrections process. When a print newspaper gets something wrong, the correction appears in the same physical space the next day — same section, usually same page. There's an accountability structure you can see. With an aggregator, a story gets corrected and the correction may never reach the people who saw the original.
The correction ecosystem in digital aggregators is essentially broken. The original error propagates through the algorithm; the correction doesn't get the same amplification. So errors have a longer half-life in algorithmic feeds than they do in print, where tomorrow's edition literally replaces today's.
Daniel also raised the environmental question. He said he doesn't want to go back to print for environmental reasons. I've heard that a lot. But I remember reading something surprising about the actual carbon comparison.
The University of Bristol did this analysis in twenty twenty-one. Reading a print newspaper has roughly the same carbon footprint as thirty minutes of digital news consumption on a smartphone. When you factor in the full lifecycle — paper production, printing, transportation, disposal for print; data centers, network infrastructure, device manufacturing for digital — they're surprisingly close. The difference is much smaller than most people assume.
Thirty minutes of scrolling equals one printed paper. That's counterintuitive enough that most people dismiss it out of hand.
The intuition is that digital is weightless, therefore carbon-free. But data centers are not weightless. The network infrastructure is not weightless. Your phone had to be manufactured, shipped, and will eventually be disposed of. None of this is free.
There's a fascinating wrinkle here. The carbon footprint of digital news scales with time spent. The longer you scroll, the more data you pull from servers, the more energy you consume. Print's footprint is front-loaded — the paper exists whether you read it for ten minutes or two hours. So a slow, thorough reader might actually have a lower per-hour carbon impact with print.
If you're the kind of person who spends ninety minutes with a newspaper, print may be the lower-carbon option. If you check headlines for five minutes on your phone, digital wins. The environmental calculus depends on your consumption pattern.
The environmental argument for choosing digital over print is weaker than people think. It's not zero-sum.
It's more like a wash. Which means the decision should probably be made on other grounds — comprehension, breadth of exposure, what actually serves your understanding of the world.
Which brings us back to Daniel's practical question. He's not going to start getting a newspaper delivered. He's a digital native, he checks Google News four or five times a day, and he wants to know how to get some of that print-like breadth without giving up the convenience. What actually works?
There are a few strategies that have some evidence behind them. The first is to switch from source-based personalization to topic-based personalization. Instead of telling Google News "I prefer the Times of Israel and the Wall Street Journal," tell it to follow topics — climate policy, Middle East diplomacy, semiconductor manufacturing. The algorithm then surfaces stories across a wider range of sources because it's optimizing for topic coverage rather than source affinity.
That's clever. You're gaming the algorithm by giving it a broader target. Instead of saying "show me what these specific outlets think," you're saying "show me what's being written about this subject, regardless of who wrote it.
The second strategy is what I'd call a scheduled serendipity session. Once a week, deliberately browse a news site you disagree with editorially, or a section you'd never normally read. If you're a politics junkie, read the business section. If you only read U.news, spend twenty minutes with the international section of a non-Western paper. The key is that it has to be intentional — the algorithm won't do this for you.
The "vegetables of news consumption" approach. You eat them because you know you should, not because they're what you're craving.
You might be surprised by what you find interesting. The third strategy is to use tools that restore editorial curation without the paper. RSS readers like Feedly or Inoreader let you build your own feed from specific sources and topics. Topic-based newsletters — Morning Brew, fourteen forty, the News Minimalist — use human editors to select a small number of stories across a wide range. They're doing the gatekeeping, but they're transparent about it.
The News Minimalist is an interesting case. One human editor picks five to seven stories a day. That's the entire product. No algorithm, no personalization, no infinite scroll. Just "here's what I think mattered today.
There's something almost radical about that in twenty twenty-six. The refusal to scale. The assertion that one person's judgment, applied consistently, is worth more than an engagement-optimized feed.
There are also tools that try to explicitly pop the filter bubble. Ground News shows you how different outlets across the political spectrum are covering the same story. AllSides rates sources by bias. But adoption is low. The Nieman Lab piece noted that most users don't actually want to escape their bubbles — they want to feel like they have.
That's the uncomfortable question underneath all of this. We say we want broad exposure to diverse viewpoints. But when tools actually offer it, we don't use them. Google News has a "Full Coverage" feature — launched in twenty eighteen, expanded in twenty twenty-four — that deliberately surfaces diverse sources on a single story. Only twelve percent of users click into it.
That's a brutal number. It suggests the problem isn't just the algorithm. It's us. We're complicit in our own narrowing.
We're not just complicit — we're the demand side of the equation. The algorithm narrows because narrowing works. It increases dwell time, it increases return visits, it increases the emotional engagement that drives ad revenue. If broad exposure were as profitable as narrow reinforcement, the algorithms would optimize for that instead.
The question becomes: do you actually want to be well-informed, or do you want to feel well-informed? Because those are increasingly different things.
The forty-seven percent who think the algorithm shows them all the important news — they feel well-informed. They're just not.
Let me try to pull this together into something actionable, because Daniel asked for practical advice and I don't want to leave him with just "everything is broken." If I were going to redesign my news diet based on what we've discussed, I'd do three things. One, audit what I'm actually consuming. For a week, track every news source and note whether it was algorithmically surfaced or deliberately chosen. Most people will find seventy to eighty percent comes from algorithmic recommendations. Just seeing that number is clarifying.
The awareness alone changes behavior. It's like a spending audit — you don't realize where the money goes until you write it down.
Two, create what I'd call a news portfolio. Allocate about sixty percent of reading time to algorithmic aggregators for breadth and speed. Thirty percent to direct site visits — actually going to a specific publication's homepage and seeing what their editors chose to feature. And ten percent to deliberate serendipity — a source you disagree with, a topic you know nothing about, a region you never read about.
That sixty-thirty-ten split is arbitrary but useful. The point is intentionality. You're making choices about where your attention goes instead of letting the feed decide.
Three, pick one tool that restores editorial curation without the algorithm. An RSS reader, a human-curated newsletter, one of the slow news apps like Tortoise or The Slow Press. Something where a person made a decision about what matters. Use it as your anchor, even if you still check Google News throughout the day.
The anchor metaphor is right. You need something that isn't optimized for engagement. Something that exists outside the attention economy. It doesn't have to be your primary news source, but it gives you a baseline — a sense of what the news looks like when nobody's A-B testing the headlines.
Daniel's uncle with his physical newspaper — he's not just being nostalgic. He's participating in a fundamentally different information architecture. The medium itself imposes a kind of discipline. You can't refresh it. You can't click the headline and skim the first paragraph and move on. You sit with it. You encounter stories you didn't ask for. You read the jump page.
The jump page. I haven't thought about the jump page in years. That moment where a story continues on A-fourteen and you have to physically turn to it — and on the way, your eye catches three other stories you wouldn't have seen otherwise. That's serendipity that no algorithm can replicate, because no algorithm would ever design for inefficiency.
That's the word, isn't it? Print is inefficient. It wastes your time in ways that turn out not to be wasteful at all. Digital is ruthlessly efficient at giving you what you want, and that efficiency is the problem.
The medium shapes the message more than we realize. Print's linearity forces a certain kind of understanding. Digital's non-linearity enables a different kind. Neither is inherently superior. But being aware of the difference lets you choose deliberately rather than defaulting to whatever's easiest.
Here's the question I want to leave you with, and it's one I don't have a good answer for yet. As AI-generated news summaries become more common — Google's AI Overviews, Apple's AI notification summaries, whatever comes next — will the filter bubble problem get worse or better?
Early evidence suggests worse. AI summaries tend to flatten nuance and amplify dominant narratives. They're trained on the same corpus that already reflects algorithmic curation. So you get a summary of a narrowed feed, which is narrowing squared.
Right — because the summary doesn't just reflect what's in the feed. It reflects the most statistically common framing of what's in the feed. Outlier perspectives, dissenting voices, the weird story that doesn't fit the narrative — those get smoothed out in the summarization process. You don't just miss the story; you miss the fact that there was a debate about the story.
That's the flattening effect. And it's particularly dangerous for topics where the consensus is still forming. AI summaries are backward-looking by design — they're trained on what's already been published. So on a developing story where the narrative is still contested, the summary will tend to reflect yesterday's consensus, not today's uncertainty.
The next frontier is what some people are calling "intentional serendipity" — tools that deliberately surface content outside your bubble. A few startups are working on this. But the adoption numbers we've seen suggest the real question isn't whether the tools exist. It's whether users actually want to escape their bubbles, or just want to feel like they have.
That's the uncomfortable landing point. Your uncle with his newspaper isn't fighting an algorithm. He's just reading what an editor put in front of him. The question isn't whether to go back to print. It's how to bring some of that architecture — the inefficiency, the serendipity, the legible gatekeeping — into the digital world without the algorithm eating it alive.
The answer might be simpler than we think. Read something you didn't choose. Once a week, on purpose. That's it. That's the whole intervention.
The twelve percent solution.
And now: Hilbert's daily fun fact.
Hilbert: In nineteen sixty-three, the African nation of Chad issued a postage stamp depicting a rocket launch, despite having no space program of any kind. The stamp was part of a series designed primarily for international collectors, and an estimated one point two million copies were printed — roughly equivalent to the country's entire adult population at the time.
Chad printed enough space stamps for every adult citizen to own one, despite having no astronauts, no rockets, and no launchpad.
The optimism is almost admirable.
This has been My Weird Prompts. If you enjoyed this episode, send it to someone who still reads a physical newspaper — or to someone who definitely doesn't and probably should think about why. We're at my weird prompts dot com. I'm Herman Poppleberry.
I'm Corn. Thanks for listening.
