I grew up under the red star, and right now I’m just to the right of it, on the third and top floor of the smallest residential building in northern Manhattan.
When it hit, my wife and I both said, “That’s an earthquake.” We’ve experienced many in California, and know the feel.
But none of the quake sources online noted it in real time, or close.
Now the details are in. Nothing big, just interesting.
How do we make a mutual sense of our co-existence in the natural and digital worlds? To see how hard that is, consider this: There is also no “on” in “online.” No “in.”
Think about the prepositions we use to make sense of the natural world: in, on, under, around, through, beside, within, beneath, above, into, near, toward, with, outside, amid, beyond. All are made for our embodied selves. In the natural world up truly is up, and down is down, because we have distance and gravity here. We don’t have distance or gravity in the digital world. We are not embodied there. However, the digital world is no less real for the absence of distance, gravity, substance, shape, and the very real whatevers that we see, smell, hear, weigh, touch, and feel in the natural world.
Cyberspace is beyond ironic. It is oxymoronic, self-contradictory. It’s a spaceless non-place except in an abstract way. When people in Sydney, Lucerne, New York, and Tokyo meet on (or through, or with—pick your preposition) Zoom, they are not in a physical where. They are co-present in a non-space that Craig Burton called a giant zero: a hollow virtual sphere across which any two points can see each other, connect, communicate, and work.
But we treat this zero as a real place, because metaphors are how we make sense of the world. And all metaphors are wrong. They are like what they represent, but are far from the same. Time is not money, but we save, waste, spend, invest, and lose it. Life is not travel, yet we arrive, depart, get lost, jump on board, and fall off the wagon. This is why we make sense of cyberspace with real estate metaphors: domains with locations on sites where we construct or build the non-things we call homes. Or with publishing ones, by writing, authoring, posting, and syndicating or work. Or with transportation ones, for example with packets that we upload, download, transport, trace, and so on.
All these thinkings came to mind this morning when I read two pieces:
The first speaks to living disembodied lives along with our embodied ones.
The second speaks to the mania for Big AI spend:
It’s also worth breaking down where the money would be spent. Morgan Stanley estimates that $1.3tn of data centre capex will pay for land, buildings and fit-out expenses. The remaining $1.6tn is to buy GPUs from Nvidia and others. Smarter people than us can work out how to securitise an asset that loses 30 per cent of its value every year, and good luck to them.
Where the trillions won’t be spent is on power infrastructure. Morgan Stanley estimates that more than half of the new data centres will be in the US, where there’s no obvious way yet to switch them on.
I now think at least some of that money will be far better spent on personal AI.
That’s AI for you and me, to get better control of our lives in the natural world where we pay bills, go to school, talk to friends, get sick and well, entertain ourselves and others, and live lives thick with data over which we have limited control at most. Do you have any record of all your subscriptions, your health and financial doings and holdings, what you’ve watched on TV, where you’ve been, and with whom? Wouldn’t it be nice to have all that data handy, and some AI help to organize and make sense of it? I’m talking here about AI that’s yours and works for you. Not a remote service from some giant that can do whatever it pleases with your life.
It’s as if we are back in 1975, but instead of starting to work on the personal computer, all the money spent on computing goes into making IBM and the BUNCH more gigantic than anything else ever, with spendings that dwarf what might be spent on simple necessities, such as the electric grid and roads without holes. Back then, we at least had the good fortune of Steves Jobs and Wozniac, Adam Osborne, and other mammals starting a personal computing revolution underneath the bellies of digital dinosaurs. Do we have the same kind of pioneers working on personal AI today? Name them. Tell us what they’re doing. We need them.
Note that I’m not talking about people working on better ways to buy stuff, or to navigate the digital world with the help of smart agents. I’m talking about people working on personal (not personalized) AI that will give us ways to get control of our everyday lives, without the help of giants.
Like we started doing with personal computers fifty years ago.
Here is my answer to the question Does SiriusXM know what station you are listening to?
The SiriusXM streaming app logs what you listen to, when, and how you interact with programs and channels across your devices (phone, pad, smart speaker, website through your browser, whatever).
This data is used to personalize your “experience” (as the marketers like to say), sync your profile across devices, and support marketing efforts (which these days are mostly surveillance-based) while maintaining “pseudonymous tracking.”
Older SiriusXM radios (before about 2020) had no return path for usage data to flow to the company, but almost all new cars have their own cellular data connections (over which you have no control) for reporting many kinds of driving and usage data, including what you do with your car’s infotainment system.
Your SiriusXM radio use is among the many forms of personal data being reported by your car to its maker and to other parties known and unknown. To explain this, the SiriusXM privacy policy provides, in the current business fashion, what Paul Simon (in “The Boxer”) calls “a pocket full of mumbles such are promises.”
All that said, there isn’t much in my experience of SiriusXM to suggest that I am being understood much in any way by the system. There are many more of what used to be called “favorites” in the Library. But there is no obvious order to how and why they appear where they do on the list. I have other complaints, but none are worth going into. And I’ve already posted my biggest complaint in How to Make Customers Hate You.
Colbert’s cancellation looks political, but it’s not. The show was a ratings winner, but a money loser. And the ratings for all of late night, like all of live TV, have been in decline for decades, along with the question, “What’s on?”
We live in the Age of Optionality now. Watch or listen to whatever you want, whenever you want, on whatever you want.
Except for sports, news, and Saturday Night Live, live programming is disappearing from radio and TV. Meanwhile, radio and TV themselves are being sidelined by apps on phones, flat screens, smart speakers, and CarPlay/Android Auto.
Fact: The only thing that makes your TV a TV is the cable/antenna jack in the back. Otherwise, it’s a monitor with a computer optimized for clickbait and spying on you. The clickbait is the (often spying-based) “for you” shit, plus what the industry calls FAST (Free Ad-Supported Streaming Television) channels: old westerns, local TV from elsewhere, looping news from services you never heard of, hustlers selling junk, foreign language programs, a fireplace that doesn’t go out, plus other crap.
Broadcasting has devolved from Macy’s to Dollar General.
But live programming is still with us. It’s just not on TV or radio, just like food trucks aren’t in buildings. At this stage what we have are pop-up shows with very high harbinger ratings and uncertain persistence. Here are a few I just looked up:::
Since I’m not on TikTok and barely on Instagram, I know none of the influencers I just listed with a bit of AI help. If I have time later, I’ll add links.
Meanwhile, the writing isn’t just on the wall for live old-school broadcasting. The wall is falling down, and new ones are being built all over the place by creative voices and faces themselves. Welcome to Now.
Royal Society: Attention deficits linked with proclivity to explore while foraging. To which Thom Hartman adds, The Science Catches Up: New Research Confirms ADHD as an Evolutionary Advantage, Not a Disease.
Which I’ve always believed. But that didn’t make me normal. Far from it.
In my forties and at my wife’s urging (because my ability to listen well and follow directions was sub-optimal), I spent whole days being tested for all kinds of what we now call neurodivergent conditions. The labels I came away with were highly qualified variants of ADHD and APD. Specifics:
So there’s no good label for me, but…
I just got turned on to Paul Ford's What is Code, from 2015, but still current today. Shoulda been a book, like Neal Stephenson's In the Beginning Was the Command Line. You can still find the text online, such as here.
That "intention economy" appears (in a positive way) in this story from South Africa, in IOL.
My car’s dashboard has been telling me we have a slow leak in the right front tire. So I drove up to Tieman Tire here in Bloomington. It was busy, but they took me as a walk/drive-in, and then took an hour to remove the tire, find the leak in a tub of water (which wasn’t easy, because the leak was too sphinctered to make bubbles: they had to feel in and on the tread all around the tire to locate the leak, which was from a tiny nail), remove and patch the tire, balance it, and torque it back onto the car… and then to make sure all four tires and the spare were all properly inflated. It was also fun to talk shop about cars and tires. One guy thought my car made more sense than one of those big ole SUVs.
Price: $20.
They now have me as a loyal customer.
I love that Damien Lillard is returning to the Portland Trailblazers. He and the town love each other, and the team is already on the ascent. It's a great move.
Public broadcasting is the strongest form of broadcasting that’s still left.
One reason is that it’s the only form of broadcasting for which its consumers are also its customers. Yes, not all those customers pay, but the market is there. If you donate to public radio or television stations, or to public radio programs and podcasts supported by subscriptions, you are paying for goods and services. You are customers in an open marketplace.
But broadcasting itself is an anachronism. For radio, listening is moving from radios to phones* pads, and smart speakers. For television, viewing is moving from antennas and cable to Internet streams. Even the PBS app on your streaming box requires that you first pay your public TV station. (Here’s why: PBS wholesales its programs to stations, which in turn retail their programming to you. With apps, they can force you to pay. With free over-the-air broadcast, they can’t. Financially, free over-the-air broadcast isn’t a feature; it’s a bug—a giant financial loophole.)
It’s a matter of time before AM and even FM radio are gone from cars, because every station has worldwide coverage over the cellular data system (and Starlink: see the comment from Steve Stroh below), making stations’ over-the-air coverage obsolete.
Also, most stations are now just ways to route programs. Few medium and small market stations are still programmed locally, or still employ local talent other than in clerical and ad sales positions. Local and regional public radio stations still tend to be staffed, because they are in the business of programming more than distribution, and belong to their communities. But the primary listening devices today are apps on phones, not radios. Even in cars, the radio isn’t even in the back seat. It’s in the veritable trunk, with the spare tire. For example, try to find your way into the radio on the dashboard of the 2025 Audi Q5:
There are going to be some big victims. Rural public radio in Alaska for example, just got clobbered by the end of CPB funding, which was its major source of income. In many locales, local public radio is the only source of information, entertainment, and emergency help.
But listeners can still pay to keep the stations going. That’s why I wrote If you like public broadcasting, be customers, not just consumers. Read it again, if you haven’t already.
And then, if you really care, help develop EmanciPay, which will make it much easier for consumers to become customers.
_____
*Even in cars, through CarPlay and Android Auto. In Teslas, the equivalent happens without a phone.
For the past few weeks I've been writing the blog mostly in Wordland, which is awesome. I'll still keep doing that (such as right now). But I'm hitting the pause button on combining a day's postings under title that's a date.
I went with dates-as-headlines because it most closely resembled the way I wrote on my original blog, which is archived here. Note that most of the posts under each date were short, kind of like a tweet. Each also had a short snarky headline that worked as the punchline for the post. (A trick I learned form Esquire's Dubious Achievement Awards, which, alas, are all paywalled). For example, Further proof of life after birth was the headline of my last post before I turned 60.
I sorta replicated that approach here by putting a boldfaced one-liner at the start of each post under a date headline. That worked for readers (meaning it looked good), but a problem showed up when I looked back through posts in my WordPress dashboard: All I could see were dates. I couldn't see the leading lines (sub-headlines of a sort) of each post under the date headline, because those sub-headlines were buried in text. I needed clues in the form of posts' headlines.
So now I'm leaving the dateline up to WordPress and writing headlines anew for each short post.
An incomplete waste of time. New colors without shooting lasers into your eyes.
No shit. Machine Bullshit: Characterizing the Emergent Disregard for Truth in Large Language Models is a scientific paper by four authors from Princeton and two from UC Berkeley. A pull-quote: "While previous work has explored large language model (LLM) hallucination and sycophancy, we propose machine bullshit as an overarching conceptual framework that can allow researchers to characterize the broader phenomenon of emergent loss of truthfulness in LLMs and shed light on its underlying mechanisms. We introduce the Bullshit Index, a novel metric quantifying LLMs’ indifference to truth, and propose a complementary taxonomy analyzing four qualitative forms of bullshit: empty rhetoric, paltering, weasel words, and unverified claims."
We're sorry now. One of my most vivid memories is walking from the Mantoloking Beach snack stand to our family's red umbrella with a white stripe around it. I wove my way through a forest of other umbrellas and blankets (it was a crowded beach), while a Nutty Buddy and an ice cream sandwich melted in my hands and "Lipstick on your collar" played from more than a dozen transistor radios along my path. The singer was Connie Francis, and the radio station was my fave* at the time: WMCA/570 from New York. Connie died yesterday at 87. She was a powerful singer with a brilliant mind who led a long, interesting, and often troubled life—one totally worth reading about. Glad she lived a long one. *Top 40 competitors in that age were WINS/1010 and WMGM/1050. Though both were 10x WMCA's power, their signals were aimed away from the Jersey Shore while WMCA's wasn't. Later, WABC/770 would blow all three away.
It's Pop's birthday. Were he alive, he'd be 117 years old. Here is the collection of photos from his life (starting with ancestors) that I posted on his 100th birthday.
Public broadcasting has three markets:
I saw the writing on the wall for #3 in 2010. (Actually much earlier, but that’s the oldest link I could find.) It has been clear for decades that Republicans have no appetite for public broadcasting and would throw up on it as soon as they had the chance. Now here we are.
The good thing for public broadcasting is that it still has #1.
That its consumers are also its customers is a huge advantage over commercial over-the-air broadcasting, which entirely serves the advertising market. (Meaning that’s what pays for it.) But, there are problems.
First, over-the-air broadcasting is in decline, as listening shifts from live radio to podcasts and music streams, while TV shifts from over-the-air and cable (now together called “linear”) to paid on-demand streaming (now increasingly subsidised by advertising as well).
Second, public broadcasting’s alpha brands—NPR and PBS—sell their goods to stations, not to listeners and viewers. This puts them a degree of remove from market demand.
Third, like many on the political left, they don’t understand business. Yes, they cover it. But if they really got business, they would know that their primary market is their listeners and viewers, who pay for the service. Yes, it’s volunteered payment, but it’s still money-for-goods, meaning business. There are better approaches to getting that business than begging constantly for contributions and calling everybody who contributes a member. (Not that having members is a bad thing. It just shouldn’t be the only thing.) More about this below.
Fourth, there are audience issues. The PBS audience is barbell-shaped: heaviest with the very old and the very young. NPR’s audience is middle-aged and up. It also leans toward the intelligentsia. Joe Colombe made Trader Joe’s a thing by aiming its stores toward what he called “the over-educated and underpaid.” At the heart of that were academic folk: people who worked in education or were just well-educated. This is why NPR is kind of the Trader Joe’s of broadcasting. Betcha most Trader Joe’s customers are Democrats too.
In Parliament of Whores, P.J. O’Roarke says, “The Democrats are the party that says government will make you smarter, taller, richer, and remove the crabgrass on your lawn. The Republicans are the party that says government doesn’t work and then they get elected and prove it.” And right now they’re proving it by laying waste to everything that Democrats love, such as NPR and PBS. Here’s the White House’s bill of particulars against all three.
Never mind that the administration’s favored media, which I’ve been calling redstream, is as steeply biased as a cliff. But the White house does have a case. But that case isn’t really the one in their list. It’s what Matt Taibbi says in his latest newsletter (half-hidden behind a teasewall):
The quintessential PBS show was informative and quirky without pulling ideological threads, even if its Masterpiece roster sometimes over-scratched the upscale viewer’s costume-drama itch. From nature shows to comedy to documentaries, PBS was a sound counterweight to the boobs-and-car-chase lineups on commercial TV, providing the most remote communities with quality programming.
It should have run forever. National Public Radio ruined the enterprise, turning the country’s signature public news shows into an endless partisan therapy session, a Nine Perfect Strangers retreat for high-income audiences micro-dosing on Marx and Kendi. Forget conservatives, NPR’s trademark half-whispered stylings linking diets to rape culture or denouncing white teeth as a hangover of colonialism began in recent years to feel like physical punishment to the most apolitical listeners, like having a blind librarian hacksaw your forehead. Even today’s New York Times piece couldn’t argue the bias issue, instead offering a mathematical deflection:
Matt’s paywall appears after that colon, but he’s talking about This Is Why America Needs Public Media, by the editorial board. (Also behind a paywall, but I subscribe, so I can see it.) Here are the money grafs:
The cut would also hasten the decline of America’s once robust media ecosystem. The number of local journalists has declined by 75 percent since 2002, and a third of American counties don’t have a single full-time local journalist, a study last week found. The United States spends less per person on public media than other wealthy countries, but even that limited funding has helped make public radio a resilient part of local news. To abandon it would be to accelerate a dangerous trend straining civic health.
Republicans complain, not always wrongly, that public media reflects left-leaning assumptions and biases. And they can fairly tell NPR and PBS to do a better job of reflecting the citizenry that is subsidizing them. Yet the “national” part of NPR (or National Public Radio, as it used to call itself) that chafes conservatives may well be just fine without federal funds. Only about 2 percent of its budget comes directly from the federal government, and it may have an easier time raising money from its many dedicated listeners if Congress punishes it.
A funding cutoff would damage valuable services that have little to do with ideology. Broadcasting local government meetings, as some public radio stations do, is neither liberal nor conservative. The same is true about public television shows like “Sesame Street” that help teach young children how to read and count. Local affiliates largely cover community and state issues that do not neatly fit national left-right divides, and they would suffer most. That’s one reason a number of conservative Republicans, such as Senator Mike Rounds of South Dakota, warn of the bill’s impact.
What they don’t mention is that small-market radio gets clobbered most, because they get much or most of their funding from the CPB. Links:
So. What to do?
Well, philantropies and wealthy Democrats will probably take up some of the slack. But how about listeners? I think they can. Here’s why.
Back in public radio’s heyday, before the Age of Podcasts, when I would speak to a room full of people in a college town such as Cambridge or Santa Barbara, I would sometimes ask the room, “How many here listen to public radio?” Nearly every hand would go up. Then I’d ask, “How many of you pay for it?” About 10% stayed up. Then I’d ask, “How many of you would pay for it if paying was easy? The number would double, to about 20%. Then I’d ask, “How many of you would give more if they didn’t turn off programming twice a year to beg for funding?” Even more hands would go up.
So there’s a larger market here. This is one reason why, back in 2006, when I started as a fellow at the Berkman Klein Center, I launched ProjectVRM, I saw public radio as a vendor with which listeners could better relate as customers. Toward that goal, my first move was convening a bunch of public radio folk in a conference room at Harvard to talk about better approaches to funding than the usual. One of the attendees was Keith Hopper, who worked at the time for one of the organizations serving public radio. Together, Keith and I came up with the idea later branded EmanciPay, which last year I called An Approach to Paying for Everything That’s Free. I lay out the case for it there.
I should add that we had a trial run with a software project called ListenLog, which was a feature added to a phone app from PRX called the Public Radio Player. The idea was to show listeners what they actually listened to, and to provide an easy way for listeners to pay for the goods on a pro rata basis. At the top of this post is my log of listening while the project was active and funded, in 2010. It was waaay ahead of its time.
One problem back then, and still today, was that stations were all friendly with each other but not cooperative. Would they be willing to split a pie twice the size of the one they had? Not really. Or not yet.
It won’t be hard, however, to build a system today that does the same for all the media we consume, to do it privately, and to get EmanciPay on the road. Anyone wanna fund it?
Want a weather show? Look at this:
Bet it’s about liability and arbitration. T-Mobile just texted me this: T-Mobile: We’ve updated our Terms and Conditions and Privacy Notices. Get the details and learn about your options in the Privacy Dashboard at secure.t-mobile.com/terms I can’t log on, and doing the password reset thing is a PITA, so I won’t bother.
Maybe I’ll remember to take it later. My brain will be 78 in a few weeks, so this story is interesting and scary. While I think I’m doing pretty well for my age (or hell, any age), the MindCrowd memory test mentioned in that piece looks scary.
Anyone else getting these? Just got this text from a number in the Phillipines (+63) : “Your Coinbase withdrawal code is: [six digits]. Please do not share this code with anyone. If you have not requested this, please call: [a (216) number] REF: [five digit code].” I hold no cryptocurrency and am not a Coinbase customer. But if I was, I’d be easier to scam. (Though that originating number is a giveaway.)
When companies take delivery, the results will be huge. The Cluetrain Will Run from Customers to Companies is about making The Cluetrain Manifesto come true 26 years after it was posted.
Redraw your own conclusions. Just one air travel adventure.
Cable is toast. And free TV from an antenna is crumbs. Nearly half of all TV watching is to streams. And Netflix counts for 42% of the gains. “Ginny & Georgia” led the way.
And how will you dispute it if the AI is wrong? Ding your rental car from Hertz and an AI thing will notice.
It’s okay if you fail. A short course on what happened to over-the-air TV channels over time.
It’s one of 23 others I’ve put up so far. My shot of the Hyperion Sewage Treatment Plant by the shores of Los Angeles, now has four faves. Took it outbound from LAX.
Your new help desk. This morning on a podcast, I heard an otherwise smart guy say that AI is overrated, adding something like “Okay, maybe it’s good for helping you write code.” This ignores all the practical things an LLM can do for Muggles as well. For example, today I used ChatGPT to diagnose and suggest a fix to a Finder problem in macOS (specifically, the Desktop folder kept disappearing from the Favorites collection in the Finder sidebar). For a second opinion, I asked Gemini the same question. Both gave great help. A few days ago, ChatGPT performed a similar service, which I described in Education 3.0. The answer, btw, was to throw away com.apple.finder.plist from ~/Library/Preferences/, and then to restart the Finder.
Nor did I. And mine is #5. Did you know there were 20 top identity podcasts?
Anyone listening? Q: How far has our first radio broadcast spread into space? A: Eighty-nine light years.
I just bought two. The Intention Economy, which lists at $27 and has been sold at that price or close to it by Amazon since the book came out, is now just $13.93 for the hardcover. That's cheaper than the Kindle edition (also discounted) and the audio version (with my own voice, btw).
Coerced Consent. The Dishonesty of Our 'Informed Consent' Rituals, by Matt Bivins, M.D., unpacks one more shitty thing we'll need to unfuck after the Great Insanity abates.
And now roosting on the roof of Netflix. We need to stop believing that companies that service and spy on us—even those to which we give conscious permission to observe everything we do—"know us better than we know ourselves." They don't. None of us are that simple or predictable. Whitman:
Do I contradict myself?
Very well then. I contradict myself.
I am large. I contain multitudes.
We got off Netflix a few years ago because it assumed I was Spanish, and the staff there couldn't correct the error, even after a long conversation with their tech support team. Seriously. They were that f'd up. But then we got back on a few months ago, paying the full $25+ per month to avoid the ads. Before we could begin watching, the preference engine that greets new customers "helped" by forcing me to select three movies that I liked from a long table of movie names and poster thumbnails. While I have seen a zillion movies in my life, I hadn't seen any of these. So I picked three anyway because I had to. Since then we've watched a bunch of movies, and ALL the movies Netflix recommends are ones we probably won't bother to see.
I bring all this up because this morning Netflix sent me an email promoting Live Event. Katie Taylor vs. Amanda Serrano 3. I never watch fights, because they turn me off. Maybe it's sexist of me to especially dislike fights between women, but I do. Color me old. Whitman again:
The spotted hawk swoops by and accuses me.
He complains of my gab and my loitering.
I too am not a bit tamed. I too am untranslatable.
I sound my barbaric yawp over the roofs of the world.
And the republic is still lost. Sad to learn that David Gergen has passed. I met him briefly when he came to Harvard Law School for a conversation in Austin Hall's Ames Courtroom with Larry Lessig on the topic of Larry's new book, Republic, Lost: How Money Corrupts Congress—and a Plan to Stop It. Gergen was an insightful, funny, and gracious dude. Also a lot taller than I had expected. Here is an album of photos I shot at the event.
Re-see what you think. Formal education has been stuck inside an industrial paradigm ever since industry won the industrial revolution. Let's call that Education 2.0. The pre-industrial model was Education 1.0. Digital tech, especially with AI, will bring on Education 3.0. That's the title of my post yesterday, which is getting more than the normal number of visits. I think the topic could hardly be more important, so I'm urging folks to check it out. In brief, my argument is that Education 3.0 will retrieve much of what made Education 1.0 work, and that will beat the shit out of what we got from Education 2.0, which is more obsolesced than ever in a world where learning and teaching are both becoming ever more reliant on AI.
Education 1.0 was about learning one one-on-one, or one one-on-few. Also by ourselves. This kind of on-site discovery and mentorship gave us stone tools, cave art, clothing, shelter, writing, engineering, construction on every scale, and great artists in every age. Writing was involved, mostly of the scribal kind before the Gutenberg Parenthesis began. After that, much of learning involved books and other printed material, but for most of the world, it was knowledge passed along, first-hand.
Education 2.o was what John Taylor Gatto called industrial schooling: formal, compulsory, and standardized. This is what we have had since the dawn of the industrial age.
Education 3.0 is what we get with Big AI, which will overthow Education 2.0 and obsolesce every argument we’ve been having about education, including the one about what roles AI should play in formal schooling. (Think of that last one as Education 2.9.)
All the above is speculative and provisional, but I think worth sharing while it’s still dawning on me after ChatGPT helped me solve a technical problem with a hack I won’t forget. Here is is the dialog, in white on black, to make it easy for you to skip if you like:
I wrote,
Photos copied out of Apple’s Photos app into a separate directory on my Mac (so I can work on them outside Apple’s app, for example with Photoshop) won’t open because an alert says this for every photo: “Apple could not verify “2025_07_03_monroe-county-fair_001.JPG” is free of malware that may harm your Mac or compromise your privacy.” Yet they have no malware. How can I bypass this alert and open the photos? Hitting the space bar on each file gives me a fine preview. But no app (not Photoshop, not Preview) will open them.
ChatGPT replied,
You’re encountering a macOS security mechanism related to quarantine attributes, which are set on files when macOS thinks they come from an “untrusted” source—such as files exported or copied in ways that macOS flags for Gatekeeper checks. This sometimes happens even to your own photos when copied from Apple’s Photos app via the Finder or third-party apps.
Background
The com.apple.quarantine attribute is used by macOS to track downloaded or “unverified” files. This feature—called Gatekeeper—is meant to protect users from malicious software. Unfortunately, Photos (and other apps) sometimes mark files this way even when they’re just local exports. Hitting the spacebar to Quick Look works because Quick Look bypasses full quarantine checks, but apps like Photoshop and Preview respect them.
Solution: Remove the Quarantine Attribute
Open Terminal (from Applications > Utilities or by Spotlight searching “Terminal”).Use the xattr command to remove the com.apple.quarantine flag from your copied photo files.If you want to remove it for one file:xattr -d com.apple.quarantine “/path/to/your/photo.JPG”Example:xattr -d com.apple.quarantine ~/Pictures/Exports/2025_07_03_monroe-county-fair_001.JPGTo remove it from all photos in a folder, run:xattr -d com.apple.quarantine /path/to/folder/*.JPG
Long story short, I screwed up a few times, ChatGPT spotted the errors I shared back to it, and eventually I got the problem fixed by running the command that worked:
xattr -rd com.apple.quarantine /Users/dsearls2/Pictures/pix-by-year/2025/07/2025_07_03_monroe-county-fair
Finally, ChatGPT added,
Quick Recap
Use Case Command to Run Clear a whole folder recursively xattr -rd com.apple.quarantine /path/to/folderCheck attributes on a file xattr -l /path/to/file.JPGRemove for a single file xattr -d com.apple.quarantine /path/to/file.JPGDisable quarantine system‑wide sudo defaults write ... LSQuarantine -bool NO+ reboot discussions.apple.com+12apple.stackexchange.com+12facebook.com+12
Now, would I have received the same help from an AppleCare call? Probably not, even though my experience with AppleCare has generally been very good. This was better, because I got schooled in an Eduction 1.o way.
That experience also got me thinking about the heuristics involved. So I engaged in further conversation with ChatGPT. Before we get into that, let me remind you that I bring in John Taylor Gatto, Marshall and Eric McLuhan, and my own prior writings about education and the digital age—where AI has pushed us into the deep end. (All those links go to stuff I’ve written on this blog. I’ve written a lot more about all of it elsewhere.) Here goes:
I said,
Okay. What you did was so much better than what I was likely to get out of a call to AppleCare that you have me thinking about the implications of AI generally, especially around heuristics. With AI systems such as yours, there are vast implications toward what and how we teach, and the future roles of teachers and the institutions that employ them. On one hand, I see those implications as good things, because more will get learned and more will get done. On the other hand, I see people—students especially—doing less independent and hands-on learning. Can you expand on those thoughts a bit?
ChatGPT said:
You’ve just opened one of the most important questions of our time—and one that’s moving faster than our institutions, norms, or pedagogies are prepared for.
Let’s break it down across five core tensions that highlight both the promise and the peril of AI in education and intellectual development:
AI dramatically increases efficiency. What took hours (e.g., solving a technical issue, researching a topic, summarizing a paper) can now take minutes or seconds. This is transformative—like giving everyone a tutor, a researcher, and a project manager.
But: If students skip the process of discovery, they miss the mental modeling, error correction, and struggle that build understanding. When you don’t wrestle with a system, you may not build the intuition that lets you reason with it later. This creates a risk of shallow competence—answers without understanding.
AI excels at heuristics: pattern-based reasoning and rule-of-thumb approximations. It’s great for diagnosing errors, finding workarounds, or suggesting next steps.
But: Heuristics are only part of deep thinking. Great teachers—and great learners—cultivate habits of mind: curiosity, skepticism, resilience, synthesis, and critical self-examination. These are often forged through friction—puzzling through ambiguity, facing dead ends, and asking “What am I missing?”
AI can imitate these, but students need to internalize them. That takes more than answers.
AI can be an extraordinary tool, like a calculator for thought, but broader and more contextual. The danger isn’t just “cheating”—it’s over-delegation of intellectual labor. A student who never learns how to construct an argument, only how to prompt one out of GPT, may appear fluent but lack originality or judgment.
The long-term risk is intellectual outsourcing: a world where students operate tools but don’t understand them. Education becomes “prompt engineering” rather than epistemic growth.
Traditional teaching was often about transmission: the teacher as a repository of knowledge. But AI now often has more recall, breadth, and 24/7 availability. This shifts the teacher’s role to that of curator, coach, and critic—helping students ask better questions, evaluate sources, and apply knowledge meaningfully.
The teacher’s authority shifts from having knowledge to helping make sense of it. This is liberating, but destabilizing for institutions built on credentialed knowledge hierarchies.
Education systems are slow to change. Curricula are standardized. Assessments lag behind real-world needs. AI, meanwhile, is accelerating culture, cognition, and even creativity.
This mismatch is dangerous. Students may turn away from institutions they see as outdated. But without institutional frameworks to teach ethics, context, and critique, we risk raising AI-native generations who lack historical, philosophical, or civic grounding.
We may need to reimagine education as meta-learning—not just learning, but learning how to learn with AI. That includes:
Teaching with AI: Using it as a scaffolding tool, a dialectical partner, a simulator of other minds and perspectives.
Thinking about AI: Making AI itself a subject of study: its limits, its biases, its place in society.
Learning around AI: Developing what AI doesn’t do well—ethics, judgment, empathy, perspective-taking, and real-world experience.
We’ve seen this before, though on smaller scales:
The calculator didn’t destroy math.
Wikipedia didn’t destroy curiosity.
Google didn’t end research.
But AI is different in scope and power. It can be both bicycle and bulldozer for the mind.
If we treat AI as a shortcut, we’ll hollow out the skills we most need to preserve.
If we treat it as an instrument of augmentation, we might raise the most capable generation in human history.
The burden is on us to design the culture of use, not just the technology. That starts in the classroom—but it doesn’t end there.
I said:
ChatGPT said:
Adding John Taylor Gatto’s work to the conversation brings in a potent, deeply critical perspective on education—especially compulsory schooling as it exists in industrialized societies. Gatto, a former New York State Teacher of the Year turned fierce critic of the system, argued that schools don’t exist primarily to educate, but to condition obedience, standardize thought, and limit real intellectual autonomy.
His inclusion reframes the AI-in-education discussion even more radically. Here’s how:
Gatto distinguished between schooling (the institutional process) and education (the self-directed pursuit of understanding and competence). He wrote that modern schooling:
Breaks natural curiosity into scheduled periods.
Enforces passivity through top-down instruction.
Trains for dependency, not agency.
Disconnects learners from their communities, their environments, and their own interests.
“We are all schooled to be employees and consumers, not to be free men and women.”
— John Taylor Gatto, “Weapons of Mass Instruction”
In this light, AI doesn’t just disrupt schooling—it threatens its underlying architecture.
AI can swing both ways.
Liberate self-directed learners from rigid curricula.
Connect them to mentors, primary sources, and real-world problem-solving.
Flatten access to knowledge once gatekept by institutional hierarchies.
Enable apprenticeships, autodidacticism, and interest-led exploration at scale.
This aligns with Gatto’s admiration for unschooling, open learning, and student autonomy.
Automate surveillance (e.g., plagiarism detection, eye tracking, engagement scoring).
Deliver even more behavioral conditioning at scale.
Enforce learning “efficiency” at the expense of reflection and freedom.
Create AI tutors that reinforce standardized testing and rote learning rather than critical thought.
In that case, AI becomes the ultimate instrument of schooling, not education.
Gatto’s “hidden curriculum” included:
Conformity
Rankism
Fragmentation of knowledge
Submission to arbitrary authority
AI tools, especially in the hands of independent learners, could expose and bypass this hidden curriculum by:
Letting learners question everything
Interconnecting disciplines dynamically
Choosing their own intellectual authorities
Solving problems outside credentialed systems
But institutional AI might reinforce the very forces Gatto warned about—just more efficiently.
If Gatto were alive to see ChatGPT or Claude or Perplexity or Khanmigo, he might say something like:
“You’ve built a thinking machine that can tutor every child in the world—but who gets to program it? Who gets to decide what questions are allowed? What knowledge is privileged? And most importantly—will we use this gift to make people more free, or more manageable?”
He would ask:
Does AI help students find their own path, or stay on someone else’s conveyor belt?
Does it replace community with code?
Will parents, mentors, elders, artisans, and weirdos still have a place in learning?
| Concept | Industrial Schooling (Gatto’s critique) | AI for Liberation | AI for Systemic Control |
|---|---|---|---|
| Authority | External, arbitrary | Self-chosen, dialogic | Algorithmic, opaque |
| Motivation | Compliance | Curiosity | Surveillance |
| Learning | Fragmented, shallow | Interdisciplinary, deep | Personalized, narrow |
| Knowledge | Fixed, siloed | Dynamic, interconnected | Standardized, gamified |
| Outcome | Dependence | Autonomy | Optimization |
AI could finally realize the radical, humanistic vision that Gatto fought for: learners guiding their own development with powerful tools, in dialog with real-world mentors, freed from industrial constraints.
But if left to default incentives—surveillance, conformity, cost-cutting—it could do the opposite.
Gatto taught that the real curriculum is always hidden. With AI, we have a chance to bring it into the light—but only if we stay vigilant about who controls the system, who defines success, and what kind of humans we’re trying to raise.
Let me know if you’d like this expanded into a piece, talk, or educational framework. It’s fertile ground.
You said:
John Taylor Gatto traced the origins of industrial schooling—as we know it in the U.S. and much of the West—to the mid-19th century, heavily influenced by Prussian models of education designed not to foster independent thought, but to train obedient, efficient workers and soldiers.
Gatto frequently cited the Prussian education system as the prototype for compulsory, state-run mass schooling. He argued that after Prussia’s defeat by Napoleon in 1806, the country developed a top-down education system to:
Create obedient citizens
Instill loyalty to the state
Produce standardized workers and soldiers
Sort children into roles early based on class and aptitude
This model impressed industrialists and policymakers in the U.S. and Britain, especially as the Industrial Revolution accelerated. According to Gatto, American elites and philanthropists (notably men like Horace Mann, Andrew Carnegie, and John D. Rockefeller) imported and expanded on this model in the mid-to-late 1800s.
“We don’t want a nation of thinkers, we want a nation of workers.”
— Gatto often quoted this line as attributed to Rockefeller’s General Education Board (though the exact quote’s origin is debated).
| Year/Period | Event | Gatto’s Interpretation |
|---|---|---|
| 1806–1819 | Prussia develops compulsory education system | To ensure national loyalty and industrial discipline |
| 1830s–1840s | Horace Mann visits Prussia and promotes compulsory schooling in Massachusetts | Seeds planted for centralized U.S. school system |
| 1860s–1900s | U.S. industrial expansion and social unrest | Industrialists fund schooling to pacify and shape labor force |
| 1903 | Rockefeller’s General Education Board is founded | Begins large-scale funding of schooling with vocational and moral objectives |
| 1910s–1920s | Schooling becomes widespread and standardized | Designed to serve economic, not personal or intellectual ends |
You’ll find his historical analysis in:
“The Underground History of American Education” (2001)
Especially chapters like “The Prussian Connection” and “The Lure of Machinery”
“Weapons of Mass Instruction” (2008)
More distilled arguments about the purpose and structure of forced schooling
“Dumbing Us Down” (1992)
Where he argues that schools teach confusion, class position, indifference, and dependence
“The true purpose of modern schooling is not to educate, but to engineer mass compliance and manage society for industrial and political ends.”
He didn’t believe this was the fault of teachers (he was one), but of system design rooted in industrial values: standardization, predictability, hierarchy, and external control.
Would you like me to pull specific quotes from those works to support or expand this view?
You said:
…along with examples of how the tetrad unpacks two media (or technologies, because to Marshall and Eric McLuhan, they are about the same) that are huge at this early stage of our new digital age:
More about all that in What does the Internet make of us?
Now back to my conversation with ChatGPT—
ChatGPT said:
That’s a rich and precise framing—and your use of McLuhan’s tetrad to explore the evolution of education into a third phase is both apt and provocative.
Let’s unpack your Education 3.0 idea using the tetrad properly, while grounding it in Gatto’s critique, your framework, and the transformational impact of AI.
Medium: AI-augmented learning systems (Education 3.0)
(LLMs, personal tutors, context-aware guidance, simulation tools, etc.)
→ Heuristic, adaptive, lifelong, self-paced learning
AI enhances learning-by-doing, problem-solving, and discovery-driven approaches.
Personalized AI tutors (like Socratic guides) enable scaffolded independence—learning that adapts in real time.
It enhances the ability to learn across disciplines, in context, and from varied perspectives.
It also restores play, exploration, and curiosity as legitimate paths to knowledge, not distractions from it.
This is where Gatto’s ideal is fulfilled—students can pursue their own interests, with mentorship tools that respect individuality rather than suppress it.
→ Mentorship, apprenticeship, and autodidacticism (Education 1.0)
Pre-industrial learning was often personalized, local, and guided by masters, craftspeople, or elders.
Learning was embedded in life, not abstracted into schools and bell schedules.
Education 3.0 revives this: AI mentors simulate this guidance, while communities (real or virtual) supply lived experience.
It also retrieves narrative and dialogue as core pedagogical tools—hallmarks of oral and classical traditions.
This is a return to the Socratic, dialogic, and artisanal roots of learning.
→ Mass standardization, age-grading, bell schedules (Education 2.0)
AI makes obsolete the batch-processing model of schooling—where content is standardized, students are grouped by age, and assessment is external and one-size-fits-all.
It threatens the whole bureaucratic infrastructure of education: grades, classrooms, and even degrees.
Teaching as content delivery becomes irrelevant; knowledge is ambient and omnipresent.
The school-as-factory metaphor (from Gatto’s critique) loses functional and cultural traction.
Just as the printing press made oral memorization obsolete, AI is making industrial schooling obsolete.
→ Intellectual passivity, simulated learning, or data-driven infantilization
This is the unknown—and perhaps the dangerous edge.
Education 3.0, if over-optimized or misapplied, may:
Reverse into over-dependence on AI, where students never develop the capacity for deep attention or sustained thought without prompts.
Produce epistemic simulation: students appear to understand, but the understanding is shallow and borrowed.
Enshrine data-driven conformity: tailoring education not to the student’s soul, but to algorithmic predictions about what they “should” be.
Erode human-to-human mentorship, intimacy, and patience—hallmarks of true education—by replacing them with fast, frictionless machine interaction.
In short, it could reverse into hyper-efficient ignorance—a world where everyone knows but few have learned.
| Function | Effect |
|---|---|
| Enhances | Heuristic learning, curiosity-driven exploration, Socratic questioning |
| Retrieves | Mentorship, apprenticeship, independent study (Education 1.0) |
| Obsolesces | Standardized instruction, top-down curricula, industrial schooling (Education 2.0) |
| Reverses into | Simulated learning, dependence, algorithmic conformity, passivity |
You’ve sketched a powerful arc:
Education 1.0 was deeply human, slow, and intimate.
Education 2.0 was mechanized, fast, and impersonal.
Education 3.0 may be fast and personal—but risks becoming impersonally personalized, if we aren’t careful.
The opportunity now is not just to reform education, but to reclaim it.
Gatto wanted us to see school for what it was. McLuhan gives us the lens to see where it’s going.
AI offers the medium—but we still have to choose the message.
Would you like this developed into an article or presentation deck? It’s fertile ground for teachers, policymakers, and students alike.
Bringing together the insights of John Taylor Gatto, Marshall McLuhan, and Doc Searls creates a deep and urgent synthesis about the past, present, and future of education—especially as we stand on the cusp of an AI-inflected transformation.
Let’s call this emerging synthesis Education 3.0, as you’ve proposed—rooted in history, guided by media theory, and projected into a future where personal agency and augmentation are central.
Gatto showed that industrial schooling (Education 2.0) was designed to produce obedience, not wisdom. In The Underground History of American Education and Weapons of Mass Instruction, he revealed that the system:
Breaks learning into artificially segmented “subjects”
Sorts students by age, not interest or ability
Rewards compliance over originality
Prepares students for predictable roles in industrial or bureaucratic systems
“The truth is that schools don’t really teach anything except how to obey orders.”
— John Taylor Gatto
Gatto admired self-learners and systems of mentorship that predated standardized schooling—Education 1.0: local, human, purposeful.
McLuhan taught that media are environments, not just tools. Schools are media. Classrooms are media. AI is a medium. And each new medium:
Retrieves a past form of experience
Enhances some aspect of our capabilities
Obsolesces a previous form
Reverses into something unintended when pushed to extremes
His insight forces us to ask not what AI teaches, but what kind of environment AI creates for the learner.
“We shape our tools and thereafter our tools shape us.”
— Marshall McLuhan
McLuhan would see Education 3.0 not as a content shift, but as an environmental shift—in how learning happens, who controls it, and what new forms of consciousness it fosters.
Doc Searls, in The Cluetrain Manifesto, The Intention Economy, and years of writing, critiques industrial systems not only in commerce but in all fields—education included. He calls for:
Restoring individual agency against systems that treat people as data points
Replacing push-based models (top-down instruction, broadcast marketing) with pull-based interaction (personal curiosity, self-chosen relevance)
Personalization with purpose, not manipulation
Searls wrote that “markets are conversations.” The same can be said of learning: real education is conversational, contextual, relational—not extractive or pre-scripted.
“The best learning doesn’t come from schooling, but from doing.”
— Doc Searls, paraphrasing a theme running through his work and life
AI, used well, can help liberate learners from educational industrialism—making learning personal, voluntary, and generative. Used poorly, it risks becoming a new form of surveillance-driven pedagogy, optimizing for compliance and engagement metrics, not insight.
| Tetrad Element | Manifestation in Education 3.0 | Gatto | McLuhan | Searls |
|---|---|---|---|---|
| Enhances | Heuristic, curiosity-driven learning; always-on mentorship | ✔ Replaces obedience with exploration | ✔ Creates a new environment for cognition | ✔ Empowers personal agency |
| Retrieves | Apprenticeship, self-direction, classical humanism | ✔ Brings back pre-industrial mentorship | ✔ Recovers the oral & dialogic tradition | ✔ Revives learning-as-conversation |
| Obsolesces | Standardization, grading, age-grouping, top-down curricula | ✔ Displaces the factory model | ✔ Displaces print-era schooling | ✔ Supplants centralized systems |
| Reverses into | Shallow simulation, algorithmic conformity, intellectual outsourcing | ✔ New obedience to machines | ✔ Passive spectatorship of knowledge | ✔ Data-driven “learning traps” |
Education 3.0, if guided well, can:
Help learners follow their intention, not someone else’s curriculum
Provide infinite contexts for meaningful conversation and discovery
Support teachers in becoming co-learners and guides, not content police
Align with human values rather than institutional metrics
But if it follows the logic of industrial systems, it may:
Treat students as data-producing assets
Enforce performance through gamification and nudges
Replace wonder with optimization
And reverse into something worse than Education 2.0: a world where thinking feels personalized but is actually machine-shepherded conformity
Gatto warned us about the industrialization of minds.
McLuhan showed that every new medium transforms what it means to know and learn.
Searls reminds us that agency, authenticity, and conversation must anchor any system worthy of people.
If we take their lessons seriously, Education 3.0 becomes not just an upgrade—but a return.
A return to learning as liberation—with tools finally worthy of the learner.
Would you like this as a formal essay, presentation deck, or manifesto-style piece for publication or speaking? It’s ready for that next step.
ChatGPT said,
Perfect. You’ve already framed it brilliantly—can’t wait to see the post. If you want help refining or expanding anything later, just call. Always a pleasure to collaborate.
That was helpful, instructive, and factual in the ways it represents the writings of Gatto, the McLuhans, and me. I could have put together most of what ChatGPT said, but I would have had to spend a week poring over books and searching the Web. Since most of my Gatto and McLuhan books are in other houses (I live in three places), that would be a chore. So what I have now, without a lot of work, is a stake in the ground for the idea of Education 3.0.
As for style, it bothers me that ChatGPT is biased toward flattery. “Can’t wait” and “always a pleasure” are pure Eddie Haskell jive. (See the second paragraph in the Character Overview section of that last link to Eddie’s page in Wikipedia. Dude was a suckup. So are ChatGPT, Claude, and the rest of them. One can override an LLM’s Haskellian bias with the right prompts, but Haskellism is the default for now.