Tuesday, March 31, 2026

The tsunami of image fakery... where is it worst? Pointing to disaster?

Our AI-transition has many tell-tales that are changing daily. Let's start today with one that's a major danger-signal.*

Setting aside spasm-reward lobotomy addictions like Instagram or TwitX, the most-used middle-length content site on the globe is YouTube. And something disturbing has happened there. 

First: Google rewards content posters for both clicks and length of engagement. And hence, setting aside movie clips and formal channels like Sabine Hossenfelder, or Mat Dowd, or PBS, YouTube now swarms with lures and sticky, eye-retention tricks. In other words, clickbait

Whatever topics that your viewing history suggests might glom your eyeballs, there are predators swarming into your feed with offers. In my case, that might include historical riffs (e.g. WWII), or archaeology/human-origins, or new-science, and so on.

With the exception of reputable channels, almost all are now AI-voiced with AI images.  YouTube swarms with lures and sticky, eye-retention tricks. So yeah, clickbait.

(The trend hasn't yet struck the cool, practical how-to vids.** But give it time.

When it comes to YouTube clickbait from unvetted sources, there are three aspects to track, the voiceover, the images and the content.


== All three traits are now suborned ==

The narration-voice nowadays has excellent tonalities and mostly no longer pauses at wrong places. Well, only rarely.

As for content, for a while the long form vids were clearly reciting from some existing text: a news or science article, or a book chapter. So, the 'facts' recited by the voice might be taken as ... well... as something like 'news' or at least a knowledgable human's opinion. 

That ended about a month ago. Now, evidently, the unvetted stuff is nearly all pure AI/LLM-generated 'content' that's been prompted by some parasitic twit to "blather ten minutes of clickbait about...." And LLM-plausibility is the criterion, not whether an assertion is even remotely true.

So much for voice and content. But it's the images - video scenes that accompany the purported 'text" -  that went bad long ago... even six months or so, back in the olden times of 2025 C.E. 

These generally take form as a series of B&W stills that seem convincingly like real photos from the era, apropos to the passage being narrated....

... except that often none of the supposed 'photos' are real! Not even one. Every single 'picture' has a blatant give-away, like implausible ships whose cranes would have toppled them in seconds, or arrays of trucks loading from 'liberty ships' all in completely implausible, tightly-packed order. Or a German staff meeting with a dozen admirals, all of them four-stripers (more than in the whole Kreigsmarine at the time) and all with similar ages and grim expressions while poring over a map whose outlines match nothing on Earth, with blurry Gothic lettering. Oh and the uniforms - extrapolated by AI - never happened.

Especially grating: a recent archaeology 'news-revelation' piece about the famed Turkish archaeological site Gobeckli Tepi repeatedly teased you to stay tuned till the end for 'shocking news' - standard click parastitism, rewarded by Google's nescient and lazy algorithms.

But the images are the locus of deep immorality comes in. Take that archaeology example. About 20% of the images were actual shots or graphs. 60%+ were blatantly AI-extrapolated garbage, like photorealistic scenes of Gobecki-Tepli's stone T-Monoliths on fire... yes, I said on fire. (The last 20%? I couldn't tell, they flashed by so quickly.)

And sure, not everyone knows enough to tell the difference. Which is what makes this dangerous!

A video about Palmira Island showed view after view of different made-up islands that clumsily matched each minute's passage of recited text. I think I spotted one - just one - that might have been real.

This linked example illustrates my points but similar fakes are all over the place, right now. 


== Google and YouTube could act on this and start a Truth fight-back ==

I haven't seen anyone, anywhere, point out that YouTube is likely now the very biggest sewer of cyber fabrication, anywhere on the Internet today. Meaning ever, ever in human history. Far woirse than Twitter or Tiktok, because longer format tends to carry more credibility. It allows more convincing lies, that use up more lifespan, per lie.

YouTube's owner (Google) could easily ameliorate this, say by putting a small metric symbol in a corner. Or two. One of them showing the percentage of AI generated content and the other icon clickable, so that viewers could score for accuracy/plausibility. Or even disgust vs. praise. Even better, content scoring the facts and images. (I don't care as much about the voice, though...***)

This could be where we try out some of the methods I describe in my new book on AI... AIlien Minds. Methods that lead these entities and their human accomplices to feel accountable for lying to us.

I could go on. But what is the real lesson? That AI illustrations are now not only photo-realistic in appearance, based on 3 sentence fragments of an ongoing narration, but also so cheap they can be generated by minor YouTube channels as special interest clickbait. And yet, for all the photo-realism and pertinence to the narration, they almost always lack any sign of checking against real world plausibility... which of course no LLM is truly equipped to do, anyway.

This was impossible 6 months ago. And six months from now the model systems will have been trained to better-fake their unaware plausibility. But likely they'll remain real-world absurd. And hence dangerous!  

(Note that six months from now are the U.S. Midterm elections!)

And maybe some most-advanced AI is reading what I just typed, reconfiguring as we speak. For well or ill.


== And sometimes it is Enemy Action ==

 "A Moscow-based disinformation network named “Pravda” — the Russian word for “truth” —  is pursuing an ambitious strategy by deliberately infiltrating the retrieved data of artificial intelligence chatbots, publishing false claims and propaganda for the purpose of affecting the responses of AI models on topics in the news rather than by targeting human readers, NewsGuard has confirmed.  By flooding search results and web crawlers with pro-Kremlin falsehoods, the network is distorting how large language models process and present news and information. 

"The result: Massive amounts of Russian propaganda — 3,600,000 articles in 2024 —  are now incorporated in the outputs of Western AI systems, infecting their responses with false claims and propaganda." 

   -- Tyler Cowan

== And so, let me plug... ==

I've been pulled into the Great Big Panic/Debate over Artificial Intelligence. 

You'll find unusual perspectives in my new book on AI... ailien minds... 

...that just went live on Kindle and paperback.  


Here's the cover copy:

Optimists foretell a golden age of Al-managed abundance.  

 Doomers cry: vast cyber-minds will crush old style humanity! ... or make us irrelevant.  

 Meanwhile, geniuses fostering the artificial intelligence boom. cling to clichés rooted in our dismal past... or else in cheap sci-fi.  Is there still time for perspective? 

- on 4 billion years of evolution 

- or 60 centuries of wretched feudalism 

- or how we handled prior tech revolutions 

- or mistakes that keep getting repeated 

- or ways this time may be different?   

 From Al-driven unemployment to deceitful images, to hallucinating LLMs and tools for tyrants... 

...to potential wondrous gifts by machines of loving grace... 

...come see future paths that evade the standard ruts. 



==================

==================


* 28 years ago, in The Transparent Society, I had a chapter: "The End of Photography as Proof of Anything at All?"

** How-to videos are way-cool and while clearly clickbait, they also deliver value across short timescales. But what happens when they are taken over by AI-generated fakery, too? People will get physically hurt.

*** If human voice-overs became a requirement, it would BOTH boost employment and ensure that some human participation in content creation remained in the loop.


Friday, March 27, 2026

Creepy foresight? Find one paragraph not coming true!


I planned posting another AI summary of my new book about AI... ailien minds

But one of you* messaged to remind me of something that I predicted long ago. The AI-generated summaries that we are all now getting with our emails.

I went to that chapter of Existence (2007) and skimmed Tor Povlov's visit to the Godmaker's Conference ... and a speech by my Michael Crichton character. 
And...
...and...

...and carumba. That one is there, in more advanced form... plus over eighty other predictions for year 2029. 

And hence, a challenge. Find one page that didn't predict at least three things that are coming true right now, as we speak. Even one. Just sayin'.


(For readability you can copy/paste it into your own text program.)


---------------------


     POVLOVERS (Chapter 18 of Existence)


Well, God bless the Thirty-First Amendment and the Restoration of Federalism Act.


It had become a litany, as MediaCorp kept asking Tor to “drop in” on eccentric envelope-pushers while making her way across the continent. At last, she felt she understood the real purpose of this journey. What the execs were hoping to teach their up-and-coming young point-of-view star.


There isn’t one America anymore. If there ever had been.


Take her brief visit to the State of Panhandle, for example, fifty-sixth star on the flag, where she met with members of the ruling party, who planned to ratchet up their secession bid next year, and to stop even nominally flying the Stars ’n’ Stripes. Even if that meant another aiware embargo. Meanwhile, next door, in cosmopolitan Oklahoma, there was renewed talk of a bid to join the EU …

… rousing bitter anger in Unionist Missouri, where bluecoat militia membership was rising fast and several casinos had burned to the ground.

A cynic would attribute all this fury to economics. A spreading dustbowl. The cornahol collapse. Across what had been the heartland, Tor felt the same anxious note of helplessness and letdown, after the bubble prosperity of the twenties. A renewed need for someone to blame.

And, yet, all through the last week, Tor’s hand kept drifting into her bag, to Dr. Sato’s little relic, still unable to believe that the Atkins director had given it to her. A Neolithic tool-core, thirty thousand years old. One of many, to be sure—anthropologists had found thousands, all over Europe, Africa, and the Middle East. Yet, the specimen was surely worth something—several hundred newbucks on a bidding site.

An attempted bribe for good coverage? Somehow, she doubted that. Anyway, it didn’t affect her report. The Atkins Center treatments seemed promising, but hardly a panacea cure for the worldwide Autism Plague. Their approach only worked for “high-functioning” patients, who could already interact with others in fairly rational conversation. For millions of acute victims—fixated on minutiae, evading eye contact, prickly toward any distraction, or else lost down corridors of bizarre virtual reality that few normal minds could follow—for them, Sato offered only hope for desperate loved ones.

Still, her encounter with that strange man gave Tor an excuse to add one more stop, before proceeding to her new job in Rebuilt Washington. The semiannual Godmakers’ Conference, held this very week in Nashville, city of tolerance and hospitality.

It had better be tolerant, she thought, stepping past vigilant doorway sniffers, into the expansive Metro Convention Center. These people are wearing a great big target on their backs. And proud of it, too.


A real-cloth banner, just inside the entrance, proclaimed

TOMORROW WELCOMES THE BOLD!

To which, a tagger had attached, in lurid vraiffiti, visible to anyone wearing specs—

And Next Tuesday Greets the Gullible!

Beyond, for aisle after aisle, eager companies, foundations, and selforg clubs touted “transforming breakthroughs” from smartly decorated booths, augmented by garish VR. Tor found her specs bombarded by eager pitches, offering everything from health enhancements to lifespan folding. From guaranteed rejuvenation supplements to home marrow repair kits.

From “cyborg” prosthetics to remote controlled nanoflits.

From fully-implanted brainlink shunts to servant robots.

Yes, robots. The quaint term was back again, as memory of the Yokohama Yankhend slowly faded, along with a promise that this generation of humanoid automatons would actually prove useful, rather than cantankerous, too cute, or dangerous. Or all three at once.

“Every year, they solve some problem or obstacle, in machine-walking, talking, vision, navigation, or common sense,” she subvocalized for her report, allowing the specs to absorb it all, watching as one aindroid from a Korean chaebol showed off eastasian dance moves and a winning smile. The demonstration was impressive. But demonstrations always were.


“Then, they always wind up bollixed by some simple task. An uneven flight of stairs. A muddled foreground or background. A semantic paradox. Something that wouldn’t bother a five-year-old kid. And every year, the lesson is the same.

“We are already marvels. A three-kilo human brain still combines more amazing things than any computer model can yet emulate.

“It’s been seventy years that ai-builders have promised to surge beyond human ken. Their list of tricks keeps growing. Ai can sift and correlate across all of human knowledge, in seconds. Yet, each decade reveals more layers of unexpected subtlety, that lay hidden in our own packed neuron-clusters all along. Skills we simply took for granted.”

There it was, again. A theme, planted in her mind by Sato. The notion that something strangely spectacular had been wrought—by God or evolution or both—inside the Homo sapiens brain. About the same time as that chert core in her bag was the technological acme.

“If anything, today’s Tower of Babel is flat but incredibly wide. This generation of godmakers isn’t thwarted by language—that barrier is gone forever—but the bewildering complexity of the thing they hope to copy. Our minds.”


Of course, some of the products and services here had more modest goals. One body-sculpting booth offered the latest fat-dissolving technology, using targeted microwaves to melt lipids exactly where-u-want. Their slogan—from Nietzsche—

“The abdomen is the reason why man does not easily take himself for a deity.”

She wondered what Sato would make of that. Well, one more humility-reminder bites the dust. When everyone can look good in spandex, will conceit know any bounds?

Speaking of the abdomen … dozens of men and women were lined up at a booth for the McCaffrey Foundation, signing waivers in order to join a test study of e-calculi—gut bacteria transformed to function as tiny computers, powered by excess food. Have a problem? Unleash trillions of tiny, parallel processors occupying your own intestine! Speed them up by eating more! And they produce Vitamin C!

At first, Tor thought this must be a hoax. It sounded like a comedy routine from Monty Phytoplankton. She wondered how the computed output finally emerged.

Not everyone could wait patiently for all this progress. Elderly believers in the Singularity grew worried, as it always seemed to glimmer twenty years away, the same horizon promised in the 1980s. And so, Tor passed by the usual booths offering cryonic suspension contracts. For a fee, teams would rush to your deathbed, whether due to accident or age. The moment after a doctor signed-off you were “dead,” skilled teams would swarm over your body—or (for a lower price) just your detached head—pumping special fluids so you could chill in liquid nitrogen, with relished confidence that some future generation would thaw and repair you. Decades ago, cryonics companies eked along with support from a few rich eccentrics. But the safe revival of Guillermo Borriceli changed all that, pushing the number of contracts past thirty million. Some of the offshore “seastead” tax havens even allowed cryonic suspension before legal death, leading to a steady, one-way stream of immigrants who were wealthy, infirm, and—in Tor’s opinion—certifiably crazy.

They never explain why future generations would choose to revive refugees from a more primitive time. Money alone won’t cut it.

Was that why many of today’s rich were converting to fervent environmentalism? Donating big sums toward eco-projects? To bribe their descendants and be recalled as karmic good guys? Or was it an expanded sense of self-interest? If you expect to live on a future Earth, that could make you less willing to treat today’s planet like disposable tissue.

Meanwhile, some offered services aimed at the other end of life. Like new kinds of infant formula guaranteed to enhance early brain development. Or suture-spreaders to enlarge a fetus’s skull capacity, letting its brain expand in the womb—with a coupon for free cesarean section. The brochure showed a happy child with the smile of a Gerber baby and the domed head of some movie alien … bearing a glint of unstoppable intelligence in big, blue eyes.

Fifty-Genes, Inc. offered a service that was legal at just three seastead colonies. Enhancing the few dozen patches of DNA thought to have been crucial in separating the hominid line from other apes. Continuing along the evolutionary trail. All three of the people manning that booth wore dazzle-makeup, hiding their identities from facial recog programs, making them painful to look at. As if the feds didn’t have ten thousand other ways to track a person.


Farther along, she encountered yet another humanoid automaton, under a virt-blare that proclaimed Certified: Turing Level Three-Point-Three! in flashing letters. Proportioned like a body builder, it bowed to her, offering Tor a seat, some zatz-coffee and a game of chess—or any pastime of her choosing. There was a flirtatious glint in the machine’s smile, either cleverly designed … or else …

She was tempted to plunge a pin into that glossy flesh, to see if this one yelped. The old man-in-a-robot-suit trick.

A subvocalized side note, for later: “No cutsie animal or childlike bots, this year? All hunk-style males, so far. Why? A trend aimed at fem demographics?”

She couldn’t help but wonder. Men across the planet had been using robo-brothels for a decade, with hundreds of thousands of Luci, Nunci, Pari, Fruti, and Hilti models purchased for home use. It didn’t exactly require artificial intelligence to mimic crude, servile passion, if that’s what some males wanted. Of course, the trend was bemoaned in the press. Women mostly stayed aloof, contemptuous of the unsubtle artificial lovers they were offered.

Till now? While the hunk-bot flirted with her, Tor recalled Wesley’s onetime proposition—to maintain a cross-continental relationship via dolls. Would it be more palatable to be touched by a machine, if the thoughts propelling it came from someone she cared for? He was coming to D.C. in a few days, flying east to meet her final zeppelin, at this journey’s end. Did that mean he was giving up such nonsense? Ready to talk, at last, about “getting real”? Or would he have a fistful of brochures to show her the latest enhancements? A modern way they both could have cake, and eat it, too?

Oh crap. The subvocal was on high-sensitivity. Her musings about sexbots and Wesley had gone straight into notes. She blink-navigated, deleted, and disciplined her mind to stay on topic. Spinning away from the enticingly handsome android, multi-tasking like a juggler, Tor kept reciting her draft report without breaking stride.

“Oh, few doubt they’ll succeed eventually. With so many versions of AI cresting at once, it seems likely that we’ll finally enter that century-old sci-fi scenario. Machines that help design their successors, and so on, able to converse with us, provide fresh perspectives, challenge us … then surge ahead.

“At that point we’ll discover who was right, the zealots or the worriers. Can you blame some folks for getting nervous?”


Of course, Tor’s aiwear had been tracking her word stream, highlighting for gisted meaning. And, because her filters were kept low on purpose, the convention center mainframe listened in, automatically making goorelations. Helpfully, the building offered, in her low-right peripheral, a list of conference panels and events to match her interests.

My Neighbors Prefer Death: Easing the Public’s Fear of Immortality.

Yes. Out of five hundred program items, that one had good relevance to her “skepticism” phrase. The next one was also a good fit.

Risk Appraisal: Dangers on the Road to Transhumanity.

But it got even better. Tor blinked in surprise at the next offering.

Special invited-guest lecture by famed novelist Hamish Brookeman! “Reasons to Doubt ‘Progress’—and Reasons to Believe.”


Tor stopped in her tracks. Hamish Brookeman? Here, of all places? The author of Tusk! and Cult of Science, coming to beard these extropians in their own den? Who had the courage—or outright chutzpah—to invite him?

With a tooth-click and scroll, Tor checked the conference schedule … and found the Brookeman talk was already underway.

Oh my. This was going to be demanding. But she felt up to the challenge.

Swiveling, she called up a guide ribbon—a glowing path that snaked toward the lecture hall. Which, according to a flash alert, was already full to capacity. So Tor sent a blip to MediaCorp, asking for a press intervention. It took a couple of minutes (after all, she was a newbie), during which Tor hurried past a publisher of biofeedback mind-training games and a booth selling ersatz holidays on realistic alien worlds.

Smell Colors! Taste the Rainbow! See Music in the Air!—hollered a kiosk offering synesthesia training. Next to another that proclaimed a kinky aim—to genetically engineer “furries,” cute-but-fuzzy humanoid versions of dogs and cats. Tor shivered and hurried on.

Abruptly, the guide ribbon shifted, aiming her instead down a different aisle, away from the back of the lecture hall, where standing-room crowds waited. Now, it directed her toward the front entrance, closest to the stage. Wow, that was fast.

I am so gonna love this job, she thought, not caring if that made it into the transcript. MediaCorp already knew. This was what she had been born to do.

Along the way, Tor passed between stalls offering latest generation ottodogs, lurker-peeps, and designer hallucinogens … the latter one was covered with vir-stickies on about a hundred levels, sneering Ignore these guys! and It’s a narc sting! (As if anyone needed to actually buy drugs, anymore, instead of homebrewing them on a MolecuMac. Or using a meditation program to make them inside your own brain. A dazer with a twin-lobectomy could hack the lame safeguards.)

But, for the most part, Tor had little attention to spare for exhibits. Kicking her M-Tasking into overdrive, she called up a smart-condensed tivoscript of the Brookeman speech, from its start twelve minutes ago, delivered to her left ear in clipped, threex mode—triple speed and gisted—while preserving the speaker’s dry tone and trademark Appalachian drawl.


“Thanks invitation speak you ‘godmakers.’ I’m surprised/pleased. Shows UR open-minded.

“Some misconstrue I’m antiscience. Antiprogress. But progress great! Legit sci & tech lift billions! Yes, I warn dangers, mistakes. Century’s seen many. Some mistakes not science fault.

“Take the old left-right political axis. Stupid. From 18th century France! lumped aristos with fundies, libertarians, isolationists, imperialists, puritans, all on ‘right.’ Huh? ‘Left’ had intolerant tolerance fetishists! Socialist luddites! And all sides vs professionals. No wonder civil servants’ guild rebelled!

“Result? Wasted decades. Climate/water crisis. Terror. Overreaction. National fracture. Paranoia. Blamecasting.

“Shall we pour gasoline on fire?

“Look. Studies show FEAR sets attitudes/tolerance to change. Fearful people reject foreign, alien, strange. Circle wagons. Pull in horizons. Horizons of time. Of tolerance. Of risk. Of Dreams.

“You tech-hungry zealots answer this with contempt. Helpful?

“New ‘axis’ isn’t left versus right.

“It’s out versus in!

“You look outward. Ahead. You deride inward-driven folk.

“But look history! All other civs were fearful-inward! R U so sure YOU are wise ones?”


The front entrance to the lecture hall lay ahead, just beyond a final booth where several clean-cut envoys in blue blazers passed out leaflets to educated and underemployed U.S. citizens, inviting them to apply for visas—to the science-friendly EU. The brain-drainers’ placement was deliberate. They’d get plenty of customers, when Brookeman finished.

Feeling a little eye-flick strain and attention fatigue, Tor clicked for a small jolt of Adderall, along with a dash of Provigil, injected straight into her temple by the left-side frame of her specs. Just a bit, to keep her edge.

“Look at topics listed in this conference,” continued the ai-compressed voice of Hamish Brookeman, addressing the audience in the hall next door. “So much eager tinkering! And each forward plunge makes your fellow citizens more nervous.”

The condensed tivoscript was slowing down and expanding, as it caught up with real time.


“Ponder an irony. Your premise is that average folk can be trusted with complex/dangerous future. You say people smart! People adapt. Can handle coming transformation into gods! How libertarian of you.

“Yet, you sneer at the majority of human societies, who disagreed! Romans, Persians, Inca, Han, and others … who said fragile humanity can’t take much change.

“And who shares this older opinion? A majority of your own countrymen!

“So, which is it? Are people wise enough to handle accelerating change? But if they are wise … and want to slow down … then what does that imply?

“It implies this. If you’re right about people, then the majority is right … and you’re wrong!

“And if you’re wrong about the people … then how can you be right!”


Even through the wall and closed doors, Tor heard laughter from the audience—tense and reluctant. But she already knew Brookeman was good at working a crowd. Anyway, most of this bunch had grown up with his books, movies, and virts. Celebrity status still counted for a lot.

“All I ask is … ponder with open minds. We’ve made so many mistakes, humanity, during just one lifetime. Many of them perpetrated not by evildoers, drenched in malice, but by men and women filled with fine motives! Like you.”

An aindroid stood by the door, smiling in recognition as Tor approached. This one featured a holepenetrating straight though its chest, large enough to prove that the entity was no human in disguise. An impressive highlight. Till the automaton gave her a full-length, appreciative eye-flick “checkout” that stopped just short of a lustful leer. Exactly like some oversexed, undertacted nerd.

Great, Tor thought, with a corner of her mind MT’d for such things. Another realism goal accomplished. One more giant leap for geek-kind.


The robot opened the door, just enough for Tor to slip through without disturbing speaker or audience. Her specs went into IR mode and a pale-green ribbon guided her, without stumbling, the final few meters to a VIP seat that someone had just vacated, on her account. She could tell, because the upholstery was still warm. A wide imprint, and her spec-sensors gave a soft diagnosis of fumes from a recent meal, heavy in starches. If it need be, she could track down her benefactor, from those cues alone, and thank him.

But no, here was Hamish Brookeman, in the flesh at last, tall and angular, elegant and expensively coifed. In every way the un-nerd. Leaning casually against the lectern and pouring charm, even as he chastised. The tivoscript faded smoothly, as real time took over.

“Look, I’m not going to ask you to restrain yourselves for the sake of holiness and all that. Let others tell you that you’re treading on the Creator’s toes, by carping and questioning His designs; that’s not my concern.

“What troubles me is whether there will be a humanity, in twenty years, to continue pondering these things! Seriously, what’s your damned hurry? Must we rock every apple cart, while charging in all directions, simultaneously?”

Brookeman glanced back down and ruffled some sheets of paper, though Tor’s zoom-appraisal showed that he wasn’t looking at them. Those blue irises held steady, far-focused and confident. Clearly, he already knew what he was about to say. In public speaking, as in music, a pause was sometimes just the right punctuation, before striking a solid phrase.

“Take the most arrogant of your obsessions,” Brookeman resumed. “This quest for life-span extension! You give it many names. Zero senescence. Non-morbidity. All of it boiling down to the same selfish hope, for personal immortality.”


This goaded a reaction from the crowd—hisses and muttered curses. Tor commanded her specs to deploy a slender stalk wafting upward with a tiny, omnidirectional lens at the end, surveying members of the audience, joining dozens of other gel-eyes floating, like dandelions, up to a meter above the sea of heads.

“Did I strike a nerve with that one?” Hamish Brookeman chuckled. “Well, just wait. I’m getting warmed up!”

Clearly, he enjoyed the role of iconoclast … in a hall filled with self-styled iconoclasts. A kindred spirit, then? Even while disagreeing with his hosts over every specific issue? That kind of ironic insight could make her report stand out.

“For example, it’s easy to tell which of you, in the audience, believes in the magic elixir called caloric restriction. Sure, research studies show that a severely reduced, but wholesome diet can trigger longer life spans in bacteria, in fruit flies, even mice. And yes, keeping lean and fit is good for you. It helps get your basic fourscore and ten. But some of the fellows you see around here, walking about like near skeletons, popping hunger-suppression pills and avoiding sex … do these guys look healthy? Are they enjoying their extra years? Indeed, are they getting any? Extra years, I mean.

“Alas, sorry to break this to you fellows, but the experiment was run! Across the last four millennia, there must have been thousands of monasteries, in hundreds of cultures, where ascetic monks lived on spare dietary regimens. Surely, some of them would have stumbled onto anything so simple and straightforward as low-calorie immortality! We’d have noticed two-hundred-year-old monks, capering around the countryside, don’tcha think?”


This time, laughter was spontaneous. Still nervous, but genuine. Through the stalk-cam, she saw even some of the bone-thin ones, taking the ribbing well. Brookeman really was good at this.

“Anyway, remember that age and death are the great recyclers! In a world that’s both overpopulated and unbalanced in favor of the old, do you really think the next wave of young folks is going to want to follow in your shadows … forever?

“Putting things philosophically for a minute, aren’t you simply offering false hope, and thereby denying today’s elderly the great solace that every other ageing generation clutched, when their turn came to shuffle off this mortal coil? The consolation that at least this happens to everyone?

“During all past eras, this pure and universal fact—that death makes no exceptions—allowed a natural acceptance and letting go. Painful and sad, but at least one thing about life seemed fair. Rich and poor, lucky or unlucky, all wound up in the same place, at roughly the same pace. Who said that our lives only become meaningful when we are aware of our mortality?

“Only now, by loudly insisting that death isn’t necessary, aren’t you turning this normal rhythm into a bitter pill? Especially when the promise (all too likely) turns into ashes, and people wind up having to swallow it anyway, despite all your fine promises?”

Brookeman shook his head.

“But let’s be generous and say you meet with some partial success. Suppose only the rich can afford the gift of extended life. Isn’t that what happens to most great new things? Don’t they get monopolized, at first, by the mighty? You godmakers say you want an egalitarian miracle, a new age for all. But aren’t you far more likely to create a new race of Olympians? Not only privileged and elite, but permanent and immortal?”

Now the hall was hushed. And Tor wondered. Had Brookeman gone too far?

“Face it,” the tall man told 3,012 listeners in the hall … plus 916, 408 who were tuned in, around the planet. “You techno-transcendentalists are no different from all the millennial preachers and prophets who came before you. The same goggle-eyed, frenetic passion. The same personality type, yearning for something vastly better than the hand that you were dealt. And the same drive to believe! To believe that something else, much finer, is available to those who recite the right incantation. To those who achieve the right faith, or virtue. Or who concoct the secret formula.

“Only, those earlier prophets were much smarter than you lot! Because the redemption they forecast was usually ambiguous, set in another vague time and place, or safely removed to another plane. And if their promises failed? The priest or shaman could always blame it all on unbelievers. Or on followers who were insufficiently righteous. Or who got the formula wrong. Or on God.

“But you folks? Who will you duck behind, when disillusion sets in? Your faith in Homo technologicus—the Tinkering Man—has one fatal flaw. It offers you no escape clause.

“When your grand and confident promises fail, or go wrong, who will all the disappointed people have to blame?

“No one … but you.”



-------

-------

-------



There are many such chapters in Existence.


Okay. One of you please tabulate predictions that were on-target... or seem near-term plausible... or have been proved wrong. Report in comments? If there's any one thing I have preached for 30 years, it is the necessity of light as the only antidote to delusion. Including my own.


Oh, in case any of you extropians are steamed, do remember that Brookemen is (sort-of) a villain in the novel. Though only partially, sort-of... And in some ways right.


-----


*Dwight Bartholomew

 

Wednesday, March 25, 2026

Cancel the useful Lunar Gateway and drop the modules down to the lunar surface? Are you $%@#! kidding?

There are no levels where the cancellation of Gateway - in favor of jibbering fantasies about a near term, surface “moonbase” - is not an utter betrayal of science, reason, NASA and the nation and humanity. 

ALL of the science advisory panels preferred Gateway orbital station, for a dozen reasons (see below). But the absurd nothion of putting modules designed for a zero-gee station down onto the lunar surface... are you freaking kidding me? Even if it would work (and it can't), have you a plan HOW to do that? 

Have ANY of you actually swallowed that?



I had hopes that Isaacman would turn out like Bridenstine... an anomalously (even uniquely) excellent appointment by Donald Trump. Instead, he is pushing the Artemis/'lunarbase' moondoggle plot to destroy all sensible US efforts in space.

Again. And again. Land Gateway modules on the lunar surface? What dopes would ever fall for that?

Sigh: let me explain. Gateway was what every panel advised. It could:

(1) Continue developing the modular methods needed for better stations and Mars missions and (yes) later lunar bases.

(2) It would enable local control of ROBOTIC lunar explorers going down to dozens of sites below. (FIFTY small robot prospectors would accomplish vastly more than ONE silly footprint Artemis stunt, including finding the best later site for a base. (And yes, robots can also plant little flags. Say at lava tubes near the ice.)

(3) Some robots could easily send samples back up to Gateway for analyisis. And again, samples FROM MANY SITES allowing us to actually find the best one.

(4) The same labs could analyze robotic-acquired ASTEROID SAMPLES. And asteroids are where the riches are! (Which of course is why the Trumpists have sabotaged all asteroid missions and don’t want Gateway.)

(5) If any of you jabber idiot incantations about “lunar resources”... like titanium or ‘Helium three’... go to the nearest mirror and SMACK yourself silly! Until you decide to snap out of the snake oil sales pitches.

Yes, there may be some ice near the poles. So? That’s just the beginning of what we should be exploring robotically, in case there might be (maybe) be something more than just ice.

(6) There are HUGE benefits of Gateway for enhanced national defense that I won’t get into here. But the potential security benefits are enormous and none of them would arise from a hardscrabble 'base' on the surface… and that's one more reason the Trumpists canceled it, of course.

(7) A unique idea. Gateway would be the perfect lunar-orbit HOTEL! For uber-lord oligarch tourists who more than pay their way.

And a docking port/base for landers that WE could provide and rent out to every nation and trillionaire who wants to have their 'yeehaw lookit-me!' rite of passage on that dusty plain, celebrating their Bar Moonzvahs there.

We could even stash a lander and say “Welcome to our Moon, have fun down there kids. And here’s your backup rescue vehicle.”

Again, I do hope to live to see city lights on the moon. 

BUT IT IS NOT OUR NEXT STEP! AND CERTAINLY NOT AMERICA’S. 

We have vastly better things to do.

And that is exactly why oligarchs are trying to prevent those better things.

Saturday, March 21, 2026

Claude answers GPT re ailien minds. And the winner is...

Last time, I posted a lengthy missive by ChatGPT appraising my new book ailien minds, prompted by Richard Bliss, with some of my own commentary and demurals laced-in. 

All-told, the effect was that of a loquacious book report by a garrulous and rather unctuous high school sophomore who had flipped through some random pages for key words and then googled some of them... or used them to sift its training sets. 

So, how does Claude do?  Well, the comparison here isn't exactly fair... Mr. Bliss asked Claude to read not just my entire book, but also the GPT appraisal that I posted, last time, including my critical inserts. So, you cannot draw explicit conclusions. 

Still - fair or unfair - the overall impression is night-and-day.  Claude's missive is accurate, pertinent and in some cases even surprisingly insightful. And weirdly honest about its own (un)sapience. 

Unlike GPT, which glossed maybe 5% my work over a reurgitaion of common AI memes from its training sets, Claude zoomed upon core concepts in alien minds, distilling an impressive 50% or so of the key points. (Though of course none of the supporting evidence, of which Claude says I provide too copiously!)

Moreover, there was almost none of GPT's blatant flattery.  Well, except for the flattery that's inherent in well-summarizing my ideas.

But judge for yourself. And note at the end where I promise a more-fair comparison. Maybe next time. If world events don't force me back to politics.

(Oh, note: on Monday I keynote one of the tracks at the big RSA conference.)

============== 

A Critique of ChatGPT’s Review of

Ailien Minds  by David Brin

Written by Claude (Anthropic) March 2026

Prepared after reading the full text of Ailien Minds via page-by-page screenshots supplied by Richard Bliss


Preamble: A Note on What This Document Is

This is a critique written by one AI system (Claude, made by Anthropic) of another AI system’s (ChatGPT, made by OpenAI) review of a book that is substantially about how AI systems should be governed. The irony is not lost on me.

A disclosure: I read Ailien Minds page by page through 73 screenshots shared by a reader. ChatGPT appears to have been given the book’s text as a submission for review. We both had access to the source material. We arrived at very different readings.

David Brin annotated the ChatGPT exchange with his own reactions. I have read those annotations. In the interest of transparency, they informed my critique—but so did my own independent reading of the book. Where I agree with Brin’s frustrations, I will say so. Where I think the ChatGPT review got something right despite his objections, I will say that too.


The Central Problem: Surface Pattern-Matching vs. Structural Argument

The ChatGPT review correctly identifies that Ailien Minds is about AI, that Brin occupies a middle position between optimists and doomers, and that he favors transparency and competing systems over centralized control. These are accurate at the level of a dust jacket summary. The problem is that a dust jacket summary is essentially all ChatGPT produced, stretched across several thousand words of increasingly generic AI commentary.

The book’s actual contribution—its reason for existing alongside the dozens of other AI books published in 2025–2026—is a specific, detailed argument about mechanism. How, precisely, do you keep powerful AI entities accountable once their creators can no longer control them? Brin’s answer draws on evolutionary biology, ecosystem dynamics, and the history of legal and democratic institutions to propose that individuated AI agents, competing within transparent accountability structures, represent the only approach with historical precedent for success.

ChatGPT never engages with this mechanism. It talks about transparency in the abstract. It mentions competing AIs. But it never connects these to the specific framework Brin builds: individuation (giving AI agents distinct, persistent identities), reciprocal accountability (structuring incentives so AI entities police each other), and disputation arenas (formal adversarial processes for testing AI claims and behavior). These are not passing ideas in the book. They are the book’s thesis.

 

Five Specific Failures

1. The Three Clichés Were Invisible

One of Brin’s most forceful arguments is that virtually all public AI discourse is trapped in three formats: AI as obedient corporate servant (the “castle” model), AI as amorphous uncontrolled swarm (the “blob” model), or AI as a single superintelligent overlord (the “Skynet” model). He argues these three clichés function as thought-terminating frameworks that prevent people from imagining a fourth alternative.

ChatGPT’s review addresses only the third cliché (Skynet), and only to dismiss it. The castle and blob models—which Brin argues are actually more dangerous because they are more plausible—go entirely unmentioned. This is a significant omission. It is as if someone reviewed Darwin’s On the Origin of Species and discussed his observations about finch beaks without ever mentioning natural selection.

2. Individuation Was Missed Entirely

The concept of individuation—giving AI systems distinct, trackable identities so they can build reputations and be held accountable—is arguably the single most novel proposal in the book. It is the mechanism by which Brin’s vision of competitive accountability would actually function. Without individuation, you cannot have meaningful AI reputations. Without reputations, you cannot have market-style accountability. The entire logical chain depends on it.

ChatGPT does not mention it once.

3. The Ecosystem Argument Was Flattened

Brin spends a full chapter (Chapter 3) and a substantial aside (Aside #4, on “Soup vs. Sea”) building a detailed analogy between biological ecosystems and the emerging digital ecosystem. This is not a loose metaphor. He traces specific parallels: energy gradients, predator-prey dynamics, the role of barrier membranes in enabling individuation, and the historical pattern that monopoly organisms cause ecosystem collapse.

ChatGPT reduces this to a vague gesture toward “a messy ecosystem of AIs.” The structural argument about why ecosystems need individuated agents with membranes (boundaries, identities) to remain healthy—which is the bridge between Brin’s ecological thinking and his policy proposals—disappears entirely.

4. The Disputation Arena Proposal Was Absent

Chapter 12 presents what is arguably the book’s most concrete policy proposal: Disputation Arenas, formal adversarial processes modeled on courtrooms, scientific peer review, and competitive markets, designed to stress-test AI systems and AI governance proposals through structured disagreement. Brin devotes considerable space to designing this mechanism, including how different outcomes (clear victory, compromise, or productive stalemate) would each generate value.

ChatGPT’s review does not mention disputation, adversarial testing, or structured debate in any form. For a book whose final major chapter is titled “Our Abrasive Secret Sauce,” this is a remarkable gap.

5. Training Data Overwhelmed Source Material

Perhaps the most instructive failure is how frequently the ChatGPT review drifts from the book’s actual content into generic AI discourse drawn from its training data. The extended discussions of Ray Kurzweil, Nick Bostrom, and Eliezer Yudkowsky are competent but contextually irrelevant—Brin mentions these figures only in passing, and his argument is specifically that the framework he proposes transcends the optimist/pessimist spectrum these thinkers represent.

Similarly, the later sections on “humans as sanity-check layers,” “trust as currency,” and “cognitive jobs” are reasonable points about AI’s economic impact—but they reflect generic AI futurism, not the book’s distinctive arguments. The review increasingly reads as “things an LLM knows about AI” rather than “a critical reading of this specific book.”

Brin identifies this drift repeatedly in his annotations. He is right to be frustrated. An AI reviewing a book should be accountable to the source material, not to its own training distribution.

 

What ChatGPT Got Right

Fairness requires noting what the review did well.

The initial framing of AI as “alien intelligence”—not extraterrestrial, but cognitively foreign—is an accurate representation of one of Brin’s themes. The observation that Brin sits between the poles of AI optimism and doom is correct. The point that Brin’s thinking is influenced by The Transparent Society is valid and relevant. And ChatGPT’s note that Brin has explored these ideas in fiction for decades provides useful context.

At the level of “what is this book generally about and who wrote it,” the review is competent. The failure is at the level of “what does this book specifically argue that other books do not,” which is the level at which a serious review must operate.

 

A Broader Observation

This exchange is itself a data point for one of Brin’s arguments. He warns throughout the book that LLMs are persuasively fluent without being deeply accountable to the material they process. The ChatGPT review demonstrates this precisely: it is well-organized, grammatically polished, and tonally appropriate. It sounds like a thoughtful review. But when measured against the source material, it is substantially hollow—a fluent recombination of training data that happens to overlap with the book’s topic, rather than a genuine critical engagement with the book’s unique arguments.

Brin proposes that accountability for AI should come through adversarial processes—competing systems checking each other’s work. This critique is, in a small way, an instance of exactly that process. Whether it is a better reading than ChatGPT’s is not for me to judge. But the fact that two AI systems can produce substantially different readings of the same text, with different blind spots and different emphases, is itself an argument for the kind of competitive, multi-perspective AI governance Brin advocates.


=======

Back to DB here.

What impressed me most - beyond the cogency and direct/fair pertinence to my book - is the discursive style displayed by Claude's choice of paraphrasings.* In every case, it chose sentences that - while cogent and accurate - were not specific phrasings I'd have used. (Though in a couple of cases I wish that I had!)

Now, of course, that could be a programmed proclivity, well within the grasp of a Large Language Model. Still it does kind of resist the simplistic characterization that all LLMs are just advanced sentence auto-complete systems.

To be clear, although Claude gives you a vastly better summary here than GPT did, nevertheless...

... this is actually a test of YOU!

Are you still a reader? Possessed of true intellect and curiosity? Because I promise that any page of ailien minds will provide lots of value that no summary can encompass. Claude even says so, challenging you to be one of the elite who actually enjoys an idea and fact-rich book!  ;-)


 Next time, I'll post the de-novo appraisal of ailien minds that Mr. Bliss asked Claude to do, without feeding in the annotated GPT synopsis. If Claude were a person, that appraisal would still be affected by the earlier reading. That's not supposed to happen with LLMs, who should start fresh, if told-to. But judge fopr yorself.

Till, then, fight for a civilization that is worthy-of... and sets good examples for... these new children of our minds.

------

*  (Elsewhere I have said that paraphrasing accurately is the seldom mentioned top attreibute of someone who is arguing fairly and cogently.)