Based on a true story.
Survivors’ names have been anonymized to protect their identities.

zzz...ZZZ...zzz...

That’s the sound of my personal hell. Not the howling wind outside the tent. Not the distant creak of glacier ice shifting in the dark. It’s the ringtone waking me up to the reality of reliving this same Everest summit attempt for the seventh time.

Tyler’s fucking snore.

zzz...ZZZ...zzz...

Three beats. Short, Long, Short, fading like an unanswered call into the void. It vibrates through the tent fabric, to my skull.

I open my eyes to the orange ceiling of my tent. There’s a patch above my head—seven stitches, the third one crooked. I know this because it’s the first glitch within my eyesight.

The tent smells like kerosene and old sweat. My sleeping bag crinkles when I move. Outside, prayer flags snapping in the wind.

My watch says 4:47 AM. In thirteen minutes, Tyler will unzip his tent. In fifteen, he’ll say his line.

I close my eyes and mouth the words: “Rise and shine, team! Time to push for the Summit!”

Loop Seven, and I’ve memorized the script like it’s my own.

【Memory Fragment: Loop Six】

Base Camp to Camp I: four hours, 800 meters up.

Camp I to Camp II through the Western Cwm: three hours if nobody stops to adjust gear.

Camp II to Camp III up the Lhotse Face: seven hours of blue ice at 50 degrees.

Camp III to Camp IV, the death zone: oxygen tanks mandatory.

At 7,850 meters, Lena’s oxygen regulator froze.

Fifty meters ahead of me. Less than 5 seconds if I were playing Usain Bolt on the Olympic track.

I watched her kneel in the snow. I ran to her through air so thin it felt like my chest was being pressed against the wall.

“I can make it,” she wheezed behind her mask.

I radioed Tyler: “You need to climb down. Now.”

Static. Then his voice: “We’re at 97% of the target, Cecil. Cost-benefit analysis says we continue.”

Cost-benefit analysis.

Three minutes later, Lena stopped moving. Her lips turned blue.

Tyler summited before sunset. We scored 58 points.

Then, everything reset.

【End Fragment】

zzz...ZZZ...zzz...

The snore continues. I focus on details to stay online: the way my camera strap has worn a groove in my thumb. The lightning-bolt scratch on the LCD screen. Three stuck pixels forming a tiny constellation.

Zip.

Right on schedule. Tyler’s tent opens. Boots going on. Double knot on the left, single on the right because the right lace is shorter. He hasn’t noticed in seven loops.

I count to sixty.

“Rise and shine, team!” His voice cuts through the chilling air like a megaphone. “Time to push for the Summit! Weather window looks perfect!”

Perfect. In Loop Two, “perfect” meant Dee lost three toes to frostbite. In Loop Five, “perfect” meant nine hours pinned in the death zone while the jet stream dropped to 8,400 meters.

I unzip my tent—zzzzzip—and the Himalayas punch me in the face.

The air is so cold it feels unreal. Breathing feels like sucking from a blocked pipe. The sky is a saturated blue that only exists above 5,000 meters—so intense it looks AI-generated. Everest rises to the north: 8,849 meters of indifferent stone and ice, bigger than every photo you’ve ever seen.

Jennie’s already up, checking her medical kit. Dee’s doing her morning routines. And Lena.

Slim yet athletic, sitting on a rock adjusting her crampons.

In six loops, I’ve watched her die three different ways.

“Cecil!” Tyler waves me over. He’s made drip coffee—smells sour and bitter at the same time. “Big day! You ready?”

I take the cup. The metal rim burns my lip. I let it sit there until the pain drops below my sensory threshold.

“Sure,” I say.

“I’ve been thinking about your photography,” Tyler says, leaning in with his boardroom voice. “Could you capture our expedition. Show the struggle, the triumph—”

zzz...ZZZ...zzz...

I took a peek at Tyler’s tent. It’s empty. He’s standing right in front of me, mouth moving.

But I can still hear the snore.

It’s coming from inside my head.

Two hours later. Strategy meeting.

Tyler stands at a whiteboard, drawing our route with the confidence of someone who’s never failed.

“Based on role assignments and objective analysis,” he says, “optimal strategy is: Day One, push to Camp II. Day Two-Three, reach Camp IV. Day Four, summit. Day Five-Six, descend.”

“What if someone gets sick?” Jennie asks. She’s the only one who challenges him. “This plan has zero margin for error.”

“That’s why we stay together,” Tyler says. “Medical resources available to everyone. You’re with the main group, Jennie. You can help whoever needs it.”

“But if someone needs to rest, I should stay with them, not keep climbing—”

“We need to consider overall efficiency,” Tyler cuts in, voice tightening. “If we stop every time someone’s uncomfortable, we’ll never summit. You have to think about the bigger picture.”

The room goes quiet.

This is Tyler’s pattern: when challenged, he deploys words like “efficiency” and “big picture” to crush dissent. Then ends discussion with “I’m in charge, I take full responsibility.”

A few loops ago, I would’ve jumped in here.

【Memory Fragment: Loop Three】

“Wait, Tyler,” I’d said. “Jennie’s right. Medical resources should stay with whoever needs them, not prioritize ‘efficiency’ over individual safety.”

He turned to me: “So what’s your suggestion?”

“Split into agile groups. Those who can move, move. Those who need rest, rest. Jennie stays with anyone who needs care.”

“That’s not agile,” Tyler said. “That’s chaos.”

“It’s flexible coordination—”

“Cecil,” he cut me off, “I’m the leader. I need to manage the whole team. Your plan would scatter our resources, complicate communication, increase risk.”

“Your plan will get someone killed!” I lost it at the top of my voice.

Silence.

Everyone is quiet. Looking at me.

Tyler’s face went red: “Fine. You lead then.”

“I didn’t say—”

“Since you think my decisions are incorrect, you make them.”

The atmosphere shattered. Tyler stormed off with Dee to prove his point, pushed too fast in brutal cold. Dee nearly died of frostbite between Camp III and IV.

Nobody summited. Zero points.

【End Fragment】

Now, Loop Seven, I watch the same tension build between Tyler and Jennie.

And I do nothing.

I look down at my notes, pretending to study my photography assignments.

Jennie glances at me. Her eyes hold a question, waiting for me to speak.

I don’t.

Because I’ve learned: speaking doesn’t help. Tyler won’t listen. The team will fracture. We always fail.

So, forget it.

“Any other questions?” Tyler asks.

“Yes,” Lena says quietly.

I tilt my head to search for her sound—Lena rarely spoke in the last 6 loops.

“I don’t feel well,” she says, with an ASMR-like voice. “A little dizzy this morning.”

Tyler frowns. “Altitude sickness?”

“Maybe.”

Jennie checks her immediately—oxygen levels, heart rate. “Mild altitude sickness symptoms. Recommend rest today. No further ascent.”

Tyler thinks. This variable disrupts his perfect plan.

“Lena, you stay at Base Camp and rest. The rest of us push to Camp I.”

“Wait,” Jennie says. “If I don’t stay with Lena, and something happens—”

“We’re only going to Camp I,” Tyler interrupts. “not summiting. Should be fine.”

Jennie wants to argue. But she sees Tyler’s “I’ve decided” expression and swallows her words.

I watch this scene. A voice in my head says: You should say something. What if someone gets hurt along the way?

“I will stay.”

Everyone looks at me.

“What?” Tyler frowns.

“My role requires specific photography from Base Camp and Camp I. If I rush ahead with you, I’ll miss these shots. So I’ll stay at Base Camp and complete my assignments.”

This is true. My individual tasks do require me to stay at lower altitude.

But it’s also an excuse.

The real reason: I’m tired.

Six failures. Six conflicts with Tyler. Six times watching the team collapse from bad decisions.

This time, I’m opting out.

Tyler stares at me. He knows exactly what I’m doing—using “complete my assignments” as an excuse for “I quit.”

“Fine,” he says finally. “You stay. Dee and Jennie will continue.”

Meeting ends.

Two hours after they leave, Lena comes out of her tent.

“You’re not resting?” I ask.

“Felt better,” she says, sitting on a rock. “Thought I’d get some air.”

In six loops, Lena has never stayed behind for personal reasons. Never.

We sit in silence as the sun rises over the Himalayas. Not the Instagram sunrise—the raw kind that makes you understand why ancient people thought mountains were sleeping titans. Light hits ice crystals in the air, turning everything into gold and blue and white that hurts to look at.

I press the shutter button out of muscle memory. Only then do I realize the data will all be wiped in the next loop.

“Can I ask something?” Lena says softly. “Do you ever feel like you’ve done this before?”

My hands freeze.

“Like dreams,” she continues. “Fragments. I dreamt about summiting last night. About not being able to breathe. About someone—” she glances at me “—trying to save me.”

The camera slips from my hands.

“How do you—”

The wind picks up. In the distance, two tiny figures move up the ridge, roped together like prisoners.

“I don’t know,” she says. “But I remember your face. When I... when it happened last time. You looked destroyed.”

“I’m tired,” I admit. “I’m so fucking tired of proving myself to be right.”

“Then don’t.”

She turns to face me, and I see it—a pixel-like glitch in her left eye, there and gone.

“Look around, Cecil. Really look.”

I look at the cloud formations with every breath I take. At the crystalised ice fractals on her eyelashes. At the distant figures moving with mechanical precision.

“This isn’t about reaching the summit,” she says. “It never was.”

We spend the day doing nothing. Two people at 5,364 meters, just sitting.

Around 2 PM, my watch buzzes:

The pixels shuffle, the world shifts. Like a stable diffusion model revealing content in front of me.

I’m now standing in a plant-filled office with standing desks.

My mouth open and a familiar voice is heard through my ears.

“Rise and shine, team!” Tyler’s voice from my mouth with muscle memory. “Let’s push for the record!”

I’m holding a coffee mug instead of a camera. Dee types on a laptop behind me. Jennie glances at me from the VP’s office wearing glasses, checking her phone.

I spin around, searching for something reflective to see my own face. Instead, I find Lena—new grad hire, twenty-three, name tag and all—sitting across the room with that knowing look.

“Cecil” she mouths.

I pause my speech. “Lena?”

“Welcome to Level 2”

English | 中文版

The Invitation

The wechat message arrived on a Tuesday morning, unassumingly. The sender was Yaoyao, from Peking University’s National School of Development. Her ask was simple:

Represent the PKUxUCL class and deliver a speech at the opening ceremony for the 2025 MBA program.

My finger hovered over the phone, my mind already drafting the polite refusal. My internal architecture, the wiring of an ISTP, is built for observation, not the stage. I had been the man in the back of the room who sees the detail everyone else misses, not the one at the podium.

But a different instinct, a newer one, took over. I replied “Yes.”

To understand the weight of that single word, you have to understand the blueprint. I am the product of an ordinary family, if “ordinary” can describe the collision of two remarkable forces. My father, the firstborn son of a Red Army footsoldier turned farmer, and my mother, a rebel who ran away from her “cleansed by cultural revolution” home in Anhui, both found their way to Beijing just years after the Gaokao resumed—a testament to their sheer will and intellectual. They forged a life on their own terms, and from them, I inherited a potent mix of independence, a quiet curiosity about the world, and a sense that wisdom is something you build, not something you’re given. This upbringing created a self-sufficient, analytical introvert, a fate I thought was fixed.

A Human Future

策看世界 | CeTheWorld

Please excuse my random lightbulb moments and goofy shower thoughts about AI, Tech, Money, Films, Books, and this thin slice of the Universe where we co-exist.

The speech I delivered that day was titled “The Future is Human (even in the age of AI)”—a theme that felt deeply personal as I witness a neo-Luddite wave of machines replacing human workers through my work at TikTok. The title was sparked by an article from Professor Zhang Weiying, who wrote that the future is not a predetermined book waiting to be revealed of its pages but a story written by our collective decisions.

My presence on that stage would be one such decision. So was the choice to be at PKU at all—investing my family’s and my own hard-earned money into a future that, in this VUCA world, is never guaranteed. The most human decision of all was the one I made every weekend: leaving my four-month-old daughter in another city after a long week of work, trading precious time with her for these classes. Each choice was a small act of defiance against a predetermined path.

To Defy Fate

This string of choices was part of a conscious project that began in October 2024, when the news of my wife’s pregnancy landed. The thought of a daughter—of the legacy I would leave her, or the lack thereof—sparked a mandate for change. I decided to defy my own nature—a concept I call 逆天改命, to rewrite my own destiny. The goal was to reshape my ISTP wiring into that of an ENTJ, a personality type that, perhaps conspiratorially, seems to command the highest levels of influence and success in this version of the world.

My self-audit revealed a complex picture. I build trust easily, my eye for detail catching the subtle cues others miss. I am competent, the one you call for a critical task, and I remain emotionally stable and rational even in a crisis. And underneath it all was my nature: open-minded, yes, but fundamentally introverted. Yet, my reliability has its limits, especially when I improvise. I once took my wife to a restaurant in Saga, Japan, a spot I’d chosen on a whim while cycling. It was run by a ninety-year-old calligrapher grandma who, charmingly, served us cold food. My wife was not impressed.

The Turn

Then came the surprise. Just earlier today, in class, I took another MBTI test. For months, I had been consciously pushing myself—speaking up, leading teams, accepting invitations I would have once dodged. I had aimed for ENTJ. The screen loaded. The result: ENTP. Not the commander, but the debater. Close, but not the target. It wasn’t the destination I had mapped, but it was proof that the journey had begun. The internal architecture was, against all odds, being renovated.

The Road Ahead

Still, the new blueprint is far from complete. I am an avid reader of science fiction, but I don’t possess a Steve Jobs-like vision for the future; the best I can do is follow his advice to “keep looking.” This search is fueled by a deeper anxiety, a fear of becoming like Mr. Stevens in Kazuo Ishiguro’s The Remains of the Day—a man of immense competence, but ultimately an attachment to someone else’s legacy.

My goal is entrepreneurship—a yellow brick road toward something of my own. I find a strange paradox in this though: in conversations, I can easily spot the resources and opportunities for others, sketching out a business plan for them on a napkin. I can see their path to becoming their own boss. But for myself, the road remains shrouded in a fog. Perhaps, as Socrates put it, the ultimate task is to “know thyself.”

Written on G27 from Beijing to Shanghai, 19 October 2025.

Maybe?!

Summer 2024, 2 AM, somewhere in the Canadian prairies where the only thing open at this hour is Jayden’s laptop. He sits with headphones on, fingers hovering over his RGB keyboard. This Canadian guy who’s never touched an instrument is about to complete a song that will appear on a Times Square billboard.

The catch: strictly speaking, he didn’t write a single note.

“I always felt like I couldn’t do it,” Jayden says. “I don’t play instruments, never studied music, never performed live—it was intimidating.” Fourteen months ago, he discovered Suno.ai—a platform where you type a few words and AI generates complete songs. Within hours, he was creating professional-quality music with his own lyrics. Months later, he beat thousands of competitors to win Suno’s inaugural “Summer of Suno” creation contest.

⬆️“Maybe?!“ on Times Square, New York

Up to 12 million people worldwide already use Suno AI to make music. Some post their creations on TikTok, building millions of followers. These AI-generated songs accumulate hundreds of millions of streams on Spotify, sometimes becoming chart-topping hits. This is beginning to shift the status quo of music creation, promotion, and consumption that has long been dominated by the “Big 3.”

For decades, three companies have dominated the global music landscape. Universal Music Group, Sony Music Entertainment, and Warner Music Group—the “Big Three”—collectively control approximately 70% of the $30 billion global recorded music market (IFPI, 2025). Every megastar you can name is signed to one of them. The system operates like an elegant feudal structure: they control not just the artists but the entire value chain—recording studios, production houses, media promotion, playlist placement, distribution networks, and the master recordings themselves.

Lisbon “Gothic Girl”

Thirty-year-old Italian programmer Tagliuz who lives in Lisbon Portugal created Alice—a pink-haired, orange-eyed gothic metal singer.

“I’m sick of those garbage,” Tagliuz tells me with his Italian accent, “empty songs are everywhere, repetitively boring.” So he decided to create music he wanted to hear himself, channeling it through a 2D character inspired by Alice Cooper.

Tagliuz first uses Leonardo AI to generate a consistent Gothic visual imagery for Alice, repeatedly fine-tuning prompts to maintain character trait consistency. For audio, he relies on Suno’s Persona feature, training samples to maintain her vocal recognition. “The whole process requires back and forth human intervention,” Tagliuz explains to me defensively. “It’s not just pushing buttons. You have to actively participate in AI’s creative process.” AI creators often need to defend their creative contributions against the misconception that AIGC is just pushing buttons.

Since the 1970s, Alice Cooper has shocked the rock world with provocative stage performances. His heavy makeup, distinctive styling, and theatrical shows not only redefined rock aesthetics but became symbols of creative freedom and self-expression. For many fans, Alice Cooper represents the courage to challenge social norms and refuse to be bound by predetermined identities—which is why his music and image continue to inspire those seeking authentic self-expression.

Then everything came to a sudden halt. Alice Cooper made controversial comments about transgender issues in an interview, shattering his image for many fans. Coincidentally, Tagliuz has stopped updating all content—Alice’s Suno, Spotify, and YouTube pages have remained inactive since July 2025.

Brazilian Hit Factory

On the other side of the Atlantic in São Paulo, Brazil, 23-year-old Raul achieved what many musicians dream of: 2 songs on Spotify Viral 50 chart at the same time. He created his own Blow Records label, specializing in using AI to convert contemporary Brazilian hits into retro blues and disco versions.

“I never release without communicating with original artists first,” Raul says. His strategy is to first create AI-generated clips under one minute, and if they go viral, use that success as leverage to request formal authorization from original creators before official release.

Raul’s success—millions of TikTok videos using his songs, people creating choreography to his tracks, hundreds of thousands of Spotify listeners—proves AI music’s commercial viability to every stakeholder in this game.

When AI Work Gets Stolen

YouTube creator Hashem Al-Ghaili, with millions of followers worldwide, created the soundtrack “Mushroom Cloud“ for his short film “Hiroshima“ made with a suite of AIGC tools. But before he could officially release it, third parties pirated it and uploaded it to major streaming platforms under their names. Hashem received copyright claims on Facebook—which he disputed, though they keep recurring. When he asked the platforms to remove unauthorized uploads, they called it a “gray area” and declined to act.

To protect original works and meet “sufficient human creative input” requirements, AI creators must take extreme measures. For another satirical short film about a clone musician called “Kira,” Hashem saved 600 pages of ChatGPT conversation records documenting the entire creative process—from story conception to character design, technical implementation to post-production optimization. He told me that Hollywood production companies have already approached him to purchase his original short film rights for full movie development.

TikTok’s Rapping STEM Teachers

Unlike overseas hit songs, Douyin (a.k.a Chinese TikTok) creators took a different path with “knowledge made catchy” short videos:

“It struggles forward with its load/The cargo’s size makes progress slow/Hundreds of times its own head/But the little protein never shows dread”—biology knowledge about proteins transformed into rhyming lyrics, set to an addictive melody reminiscent of the BGM in “Shawarma Legend.” Douyin has numerous similar accounts: “Hu Xiaoxian“ covers science and engineering history, “Rap & Science“ popularizes math concepts, “English Enlightenment UXA“ teaches Zoology with English lyrics, and the “Shun Hua Sings Herbs“ series focuses on plants.

Douyin platform’s algorithm seems to encourage educational content, making these videos highly viewed. More importantly, they make knowledge enter your brain in a relaxed way. Comments say: “Did I just learn something? /s [Confused emoji]” Beyond humor, users rarely discuss whether videos and music are generated by AI, rather they focus more on the content itself, showing knowledge-infused rap actually effective for educational purposes.

The relative simple workflow-based production of these viral content format mean more people joining in. However for monetization, besides platform revenue sharing, general knowledge fields have low conversion rate—users won’t pay for this. AI education rappers on Douyin told me that they haven’t found a reliable path to scale up the revenue stream.

The Voiceless Singer’s Voice

Perhaps no story better embodies AI creation tools’ democratizing potential beyond commercial value than Japanese user Nyxen. Due to speech impairment, she can only communicate with me through text.

“AI gave me a voice [pray emoji],” she types in Discord, occasionally throwing in an anime emoji. For her, AI isn’t just a music-making tool—through Suno, she gives an anime-style virtual character a unique voice, accumulating thousands of monthly Spotify listeners. Nyxen’s creative process reflects many AI musicians’ experimental qualities. “I originally just wrote lyrics for fun, trying to make them sound perfect on Suno. Never thought deeply about becoming a professional musician,” she types to me. This casual attitude—creating for joy rather than livelihood—is common in the Suno community.

Swiss “Godfather”

Even more remarkably, Swiss creator TongMick became a spiritual leader in the Suno community. He uses thousands of tweets to promote other Suno platform music creators. Since last August, his “My Favorite Suno Artists” series has featured 241 creators and continues updating.

“My own music has no fixed style because I always get new inspiration in the co-creation process with AI.” He deliberately maintains anonymity, choosing a cute white cat as his avatar. For him, what matters isn’t money but speaking for this community of tens of millions of users.

However, TongMick has growing concerns about platform changes. His frustrations with Suno’s direction culminated in leading a community open letter to the Suno team in September 2025, signed by dozens of creators who raised issues about trending chart manipulation, broken iOS features, and lack of transparency in discovery algorithms. “The global trending charts are increasingly dominated by accounts using artificial engagement,” the letter states, “pushing genuine creators out of visibility.”

TongMick’s day job is in the audio-visual equipment space. Having witnessed the evolution from cassette player to iPod and streaming services, he thinks AI music is just starting and predicts that “within two to three years, there will definitely be charts specifically for AI.” His vision isn’t AI replacing human creation but coexistence—two parallel ecosystems serving different audiences.

Suno Goes to Court

While creators immerse themselves in AI music’s creative pleasure, Suno’s management team faces court battles against the Big 3. Last June, Universal, Sony, and Warner jointly sued Suno, alleging unauthorized use of copyrighted music to train AI models. Almost simultaneously, independent musicians also filed class action lawsuits claiming Suno-generated music directly infringes their work copyrights.

In its August 2025 motion to dismiss, Suno‘s lawyer Andrew M. Gass argued that copyright law only protects against copying “actual sounds.” Since Suno “exclusively generates new sounds, rather than stitching together samples,” the company claims “no Suno output can infringe the rights in anything in the training set, as a matter of law.” It’s a technical argument that sidesteps the question of whether training on copyrighted music is legal, focusing instead on whether the output itself contains copied audio.

The case is still pending, but the stakes extend far beyond Suno. California federal judges have recently sided with AI companies in several intellectual property disputes, suggesting that training on copyrighted material could constitute “fair use,” basically giving other AI companies a green light to go ahead. Still, Suno faces international lawsuits from music copyright organizations—even an American victory won’t end the global legal battle.

While the lawsuit plays out in the public eye, Suno is also quietly positioning itself as more than just a music generation tool—it’s building an army of creator-advocates. These everyday users, who have experienced firsthand how AI democratizes music creation, provide powerful testimonials against claims that Suno merely “stitches” existing works. This grassroots community—teachers creating educational raps, voice-impaired individuals finding expression, and hobbyists creating viral hits—reinforces the narrative that Suno enables genuine creativity rather than facilitating copyright infringement.

Rewriting Game Rules with UGC

As legal proceedings continue, this global community’s diverse creations may serve as compelling evidence that AIGC music empowers original expression rather than plagiarism—potentially Suno’s strongest defense in a landscape still adapting to AI’s creative potential.

Before AIGC, TikTok already proved that 15 second clips could make or break a song—bedroom producers went from nobody to millions of streams because their music caught a viral moment on TikTok—disrupting Big 3’s dominance on music promotion.

In September 2025, Suno quietly began testing “Hooks,” and on October 1st, both Hooks and OpenAI’s Sora 2 launched publicly on the same day. The coincidental timing reveals a shared philosophy that’s reshaping how AI companies approach product development: from applying models to existing products to building models specifically for user products in mind.

What’s truly remarkable about both of these products—beyond their TikTok-like interfaces—are the “Remix” features that allow content to be audio-visual, social and endlessly reinterpretable.

Suno Hooks lets users pair their AI-generated songs with 10-20 second video clips and publish them to a TikTok-style feed where other Suno users can remix from the sound. Sora 2 takes it one-step further—users can “remix” any published video by prompting, inserting their own faces via “cameos,” or swapping visual styles while keeping the original structure. Both platforms share the same premise: content isn’t meant to be finished. It’s meant to inspire the next User-Generated Content to be watched by more users on the platform.

This shift has profound implications. When TikTok lowered the bar for video creation, it didn’t make everyone a filmmaker—it changed what counted as video worth watching. Now AI is pushing further: creation and consumption are merging into a single social loop. You don’t just watch a Hook; you remix it. The line between audience and creator dissolves because the content isn’t complete without engagement.

As for Jayden and the other Suno creators, after his moment of fame, Jayden returned to his normal life in Lloydminster. He still writes songs when inspiration strikes. The “Summer of Suno” competition he won was just the beginning—the contest has since reached its ninth iteration, still drawing thousands of new creators who dream of their own Times Square moment.

English version | 中文版

5min Audio Digest - by NotebookLM

1. Collateral Damage: The Price of Technological Progress

AI breakthroughs pour relentlessly into our feeds—each day another glittering billion-dollar promise—but when I meet former tech colleagues for drinks, the mood is far less triumphant.

The internet gold rush, it seems, has lost some of its luster. The hours are longer, the pay is stagnant, and the layoffs keep coming. The past couple of years have been particularly merciless: 2024 alone saw 150,000 job cuts in the tech sector, and this unsettling trend is expected to continue beyond tech throughout 2025, with “restructuring” and “optimization” recurring endlessly, keeping everyone perpetually on edge.

Some companies dangle voluntary severance packages, while others allocate quotas to each department—like homework assignments for managers to fulfill—only for those same managers to discover they've been laid off in later rounds. Then there are the ones that keep pivoting—a convenient euphemism for another round of house-cleaning. Quite a few friends of mine have been caught in the crossfire.

For Chinese workers, even relocation abroad isn't the escape hatch it once was. I know people who poured everything into securing an overseas transfer—only to have their entire teams dissolved before he could even settle down his family. "Business restructuring," the bosses called it, as these workers packed up their dreams and flew home at their own expense, forced to start job hunting all over again.

The fresh graduates have it worst. Many of them spent years earning a master's degrees away from home, couldn't scrape together enough immigration points to qualify for a visa, came home, and now cannot even land a decent internship. It's giving me serious flashbacks to Japan's "Employment Ice Age"—that period where young people who graduated into economic stagnation and never quite recovered.

After a few drinks, the conversation inevitably turns to the same lament:

We're living through the worst possible timing.

We are the generation of workless, caught in the AI revolution before anyone understood its implications—becoming obsolete before schools could develop proper curriculum to prepare us. So here we are, frantically subscribing to AI newsletters, following dozens of tech influencers on our social media of choice, getting bombarded with updates about the latest models and applications,—desperately trying to keep up with a revolution rewriting itself faster than we can proofread.

It's precisely among this crowd of "AI enthusiasts" that a darker question has started to emerge:

What if technological progress has a price? And what if our entire generation are the ones to pay it?

It sounds like a very modern anxiety. But two centuries ago, another group asked the exact same question—and they didn't just tweet about it. They grabbed hammers and went to war.

They were the Luddites—and they had some opinions about technological "progress."

Subscribe now

2. The Real Story of the Machine Breakers

Picture this: Nottingham, England, 1811. The nights are getting restless. Groups of workers slip through the darkness like guerrillas, their faces blackened with coal dust, carrying axes and sledgehammers. They move with military precision toward a clear target: the machines that stole their livelihoods.

These were the Luddites, and they took their name from Ned Ludd, a stocking frame apprentice whose story became legendary through newspaper reports that began circulating in 1811.

Ludd had been an apprentice in Leicestershire, trapped in what the records called the "confinement" of the stocking frame—perhaps both the physical imprisonment standing in the same spot for hours on end, but also the psychological cage of mind-numbing repetitive work, the kind that slowly grinds away at your sense of self. When Ludd's performance started slipping, his master complained to the local magistrate, who prescribed “a little whipping” to straighten out the young man's attitude.

But instead of breaking Ludd's spirit, the punishment lit something else entirely. In a moment of pure rage, he grabbed a hammer and obliterated the machine that had stolen his freedom. Just like that, Ned Ludd became a legend.

But the textile workers who rallied under Ludd's banner had a more practical grievance: the new machines were making their skills worthless. Anyone could operate them, which meant that decades of apprenticeship and expertise suddenly counted for nothing. The numbers were stark—between 1796 and 1828, mechanization reduced the labor needed to produce broadcloth by seventy-five percent——work that once required four master craftsmen could then be done by one person babysitting a machine.

This wasn't just about money; it was about identity. These artisans watched their social status evaporate while factory owners enriched themselves.

Different groups experienced technological disruption differently:

• Skilled craftsmen saw their social status disappear overnight. They became the face of the Luddite resistance, fighting to preserve a way of life that was becoming obsolete.

• Unskilled workers flooded into the factories, at first grateful for the roofed work opportunities, only to discover that oversupply made their working condition worsen over time. Some eventually joined the Luddites.

• Factory owners profited enormously by replacing skilled labor with machines and cheap, unskilled labor—including children.

The Luddites were not anti-technology zealots. Before resorting to violence, they exhausted all proper channels—petitions, peaceful demonstrations, and letters to Parliament. They gathered thousands of signatures demanding minimum wage protections and industry regulations. But the aristocrats turned a blind eye to their pleas.

When peaceful protest failed, violence felt like the only remaining option. Between 1811 and 1816, the Luddites organized nighttime raids on factories, methodically destroying the machines. They called their trademark sledgehammer "Enoch" and had a motto: "Enoch made them, Enoch shall break them."

Authorities responded swiftly and brutally. Factory owners formed vigilante committees and lobbied Parliament. In March 1812, the Destruction of Stocking Frames Act made machine-breaking punishable by death.

Violence escalated. In April 1812, Luddites assassinated William Horsfall, a factory owner leading efforts against them. The government declared open war on machine-breakers. The resulting arrests, trials, and executions crushed the rebellion by 1816.

In the decades afterward, the Industrial Revolution advanced rapidly in Britain, enriching factory owners, while workers’ conditions improved slowly. Only twenty years after the initial Luddite uprising did the Factory Acts of 1833 begin to address working conditions, especially for children and women. It wasn't until 1871, with the passage of the Trade Union Act, that workers gained legal recognition to organize collectively.

The Luddites weren't fighting technological progress itself—they challenged the distribution of its benefits. Having spent their lives mastering intricate skills, they questioned why automation's advantages should benefit only a privileged few.

It's a question strikingly relevant to the AI revolution today.

3. The Knowledge Worker Dilemma

Fast-forward two centuries, and we're living through our own machine revolution. AI isn't just another technology—it's what economists call a General-Purpose Technology, the kind that reshapes entire civilizations. And this time, it's coming for the knowledge workers—the people who “think for a living” with their college degrees and cognitive skills.

Looking at today's landscape through the lens of the Luddite era, the parallels are unsettling:

• The bosses haven't changed much. Like the factory owners of 1812, they control the means of production and see AI as a way to cut costs and increase profits.

• Tech workers face their own version of the skilled craftsman’s nightmare. Software engineers who've spent years mastering programming languages now watch AI write cleaner, faster code—and iterate itself through human feedback to improve relentlessly. Financial analysts find themselves outpaced by algorithms processing data at superhuman speeds, flawless and tireless. These are the new Luddites—highly educated, well-compensated, yet suddenly expendable.

• Traditional professionals, such as doctors and lawyers, occupy an interesting middle ground. Like the skilled artisans of old, the wealthy and well-connected among them lobby legislatures for protection, citing AI’s lack of explainability and safety concerns as reasons to restrict its reach into their fields. Others embrace AI, leveraging accumulated capital to transition from practitioners into employers, benefiting directly from efficiency gains. Yet a third group, driven by ethical conviction or loyalty to traditional craftsmanship, joins what amounts to a new Luddite movement, drawing on shared anxieties to rally wider opposition.

• Everyone else should, theoretically, benefit from cheaper, higher-quality AI-driven services. Yet the future feels less optimistic, because the AI revolution differs fundamentally from its industrial predecessor. AI-powered automation threatens nearly all forms of work, leaving fewer jobs untouched—and fewer people unaffected—than the machines of two centuries ago.

• Finally, parallels echo in the plight of today’s students. Ned Ludd’s rebellion was sparked by realizing his painstakingly learned skills had become worthless overnight. Today's university students grapple with a similar dread: Why spend years studying when AI may render their chosen fields obsolete before they even graduate?

The question that tormented Ned Ludd now haunts our entire generation:

When machines can do our jobs better than we can, what becomes of us?

4. The “Free Money” Experiment

With mass technological unemployment looming, thinkers and policymakers have been exploring what comes next. The most discussed solution is Universal Basic Income (UBI)—essentially, a government check sent to every citizen, no strings attached.

The idea goes back to Thomas Paine, one of America's founding fathers, and it's gained serious momentum lately. Elon Musk has been talking about it for years, arguing that when AI takes over most jobs, UBI becomes a necessity, not a luxury.

The obvious criticism is that free money will turn us all into couch potatoes, reminiscent of the filmWall-E. Why work if you don't have to? Conservative economists have pushed this argument for decades, and intuitively it makes sense.

But a fascinating experiment in Germany recently challenged this assumption.

A Berlin nonprofit called "Mein Grundeinkommen" tracked 122 people for three years, giving them €1,200 per month, no questions asked. The results were eye-opening:

• People kept working anyway. The participants worked an average of 40 hours per week—exactly the same as the control group that didn't get free money. As one researcher put it, "We found no evidence that people enjoy doing nothing."

• Their work got better. With financial security as a backstop, people were more likely to change jobs, start businesses, or go back to school. They took better risks and avoided desperate choices.

• They got happier and healthier. Stress levels dropped, mental health improved, and people reported feeling more autonomous and purposeful. Women especially felt more in control of their lives.

Similar experiments in Alaska (which has been paying dividends to residents since 1976), Stockton, California, and Finland have reached similar conclusions. Matthew Johnson, a public policy professor at Northumbria University, summed it up: "There's no evidence that basic income reduces economic activity. On the contrary, it gives workers security to make better choices."

It turns out that when people aren't terrified of losing their health insurance or becoming homeless, they don't just lie around watching Netflix. They do more interesting, productive, meaningful work.

5. “A World without Work”

As AI continues its march through the economy, we're moving toward what Stanford Economist Erik Brynjolfsson calls "The Second Machine Age"—where the critical scarce resource isn't labor or capital, but ideas. The people who can dream up new products, services, and business models will do incredibly well. Everyone else... well, that's the question.

"In the future, Ideas will be the real scarce inputs—scarcer than both labor and capital."— Erik Brynjolfsson, Andrew McAfee, and Michael Spence

Oxford economist Daniel Susskind, author of A World Without Work, suggests we’ll need more than unconditional UBI. He proposes Conditional Basic Income—financial support tied to expectations of meaningful social contributions, such as creating art, caregiving, teaching, or community service.

Work offers more than money—it provides purpose, social bonds, and a sense of contributing to something larger than oneself. As AI overtakes traditional jobs, we'll be forced to redefine what it means to be valuable and useful.

This leads to difficult questions that our generation will likely face within our lifetimes:

• If you no longer had to work for money, would you still choose to work?

• If you did, what would constitute a meaningful societal contribution using your unique talents?

• If AI continue its the rapid pace of progress, how far away are we from this future?

The Luddites faced a future that arrived faster than they could adapt—machines triumphed, factory owners flourished, and worker protections lagged decades behind. Today, we have a chance to chart a different course: to share the benefits of technological progress more equitably, leaving future generations a legacy that fundamentally redefines the purpose of work.

Author: Cecil Li
Blog: Cethe.World

References

I. Books

1. Marchant, B. (2023). Blood in the machine: The origins of the rebellion against big tech.

2. Susskind, D. (2020). A world without work: Technology, automation and how we should respond.

3. Thompson, E. P. (2013). The making of the English working class.

II. Academic Papers/Reports

1. Brynjolfsson, E., McAfee, A., & Spence, M. (2014). New World Order: Labor, Capital, and Ideas in the Power Law Economy. Foreign Affairs, 93(4), 44-53. https://www.jstor.org/stable/24483556

2. 金炳彻. (2023, July). 全民基本收入的实践探讨与案例分析 [Universal basic income practice exploration and case analysis]. CAOSS. https://www.caoss.org.cn/upload/file/23091916474798.pdf

3. Kangas, O., Jauhiainen, S., Simanainen, M., & Ylikännö, M. (Eds.). (2019). The basic income experiment 2017–2018 in Finland. Preliminary results (Reports and Memorandums of the Ministry of Social Affairs and Health 2019:9). Ministry of Social Affairs and Health. http://urn.fi/URN:ISBN:978-952-00-4035-2

4. McKay, J. P., Hill, B. D., & Buckler, J. (n.d.). The lives of rural textile workers. In Understanding World Societies (2nd ed., Ch. 17, Section 12). Macmillan Learning. https://digfir-published.macmillanusa.com/mckayunderstanding2e/mckayunderstanding2e_ch17_12.html

5. RAND Corporation. (2024, October). The AI playbook: What other sectors can learn from the creative industry's fight against AI. Retrieved July 12, 2025, from https://www.rand.org/pubs/commentary/2024/10/the-ai-playbook-what-other-sectors-can-learn-from-the.html

III. News Articles

1. Business Insider. (2025). The list of major companies laying off staff this year including Intel, Scale AI, Morgan Stanley, and Meta. https://www.businessinsider.com/recent-company-layoffs-laying-off-workers-2025

2. Gurman, M., & Griffin, R. (2025, July 10). Meta poached Apple's Pang with pay package over $200 million. Bloomberg. https://www.bloomberg.com/news/articles/2025-07-09/meta-poached-apple-s-pang-with-pay-package-over-200-million

3. Nguyen, D. (2025, June 12). American Medical Association wants explainable AI. POLITICO Future Pulse. https://www.politico.com/newsletters/future-pulse/2025/06/12/american-medical-association-wants-explainable-ai-00402237

4. NPR. (2021, March 4). California program giving $500 no-strings-attached stipends pays off, study finds. https://www.npr.org/2021/03/04/973653719/california-program-giving-500-no-strings-attached-stipends-pays-off-study-finds

5. TechCrunch. (2024, December 31). A comprehensive archive of 2024 tech layoffs. https://techcrunch.com/2024/12/31/a-comprehensive-archive-of-2024-tech-layoffs/

6. Varanasi, L. (2025, February 27). Will AI replace human jobs and make universal basic income necessary? Here's what AI leaders have said about UBI. Business Insider. https://www.businessinsider.com/universal-basic-income-ai

IV. Web Resources/Encyclopedia Entries

1. Alaska Permanent Fund. (2024, July 12). In Wikipedia. https://en.wikipedia.org/wiki/Alaska_Permanent_Fund

2. Destruction of Stocking Frames, etc. Act 1812. (2024, July 12). In Wikipedia. https://en.wikipedia.org/wiki/Destruction_of_Stocking_Frames,_etc._Act_1812

3. Employment Ice Age. (2024, July 12). In Wikipedia. https://en.wikipedia.org/wiki/Employment_Ice_Age

4. General-purpose technology. (2024, July 12). In Wikipedia. https://en.wikipedia.org/wiki/General-purpose_technology

5. Luddite. (2024, July 12). In Wikipedia. https://en.wikipedia.org/wiki/Luddite

6. Neo-Luddism. (2024, July 12). In Wikipedia. https://en.wikipedia.org/wiki/Neo-Luddism

7. Pilotprojekt Grundeinkommen. (2024, July 12). Labour. https://www.pilotprojekt-grundeinkommen.de/en/labour

8. Universal basic income. (2024, July 12). In Wikipedia. https://en.wikipedia.org/wiki/Universal_basic_income

9. Yang, A. (2020). The freedom dividend. Yang 2020. https://2020.yang2020.com/policies/the-freedom-dividend/

For You Page's New Favourite Kid

On May 20, 2025, Google unveiled Veo 3 at its annual conference Google I/O. Within weeks, over 500,000 videos tagged with #veo3 appeared on TikTok, with viral content being shared across all major social media platforms. The content ranged from time travel, satirical interviews, cutting a planet with a knife videos, and AI characters breaking the fourth wall to ask for help from the viewers.

There are so many viral videos created and watched that people started curating daily Veo3 video rankings—just like NBA’s Top 5 plays of the night—sharing the most creative viral videos everyday. Videos with millions—even hundreds of millions—of views have given global users a taste of the creative revolution brought by AI-generated videos.

What sparked this UGC creativity boom was the Bigfoot vlogs. It’s hard to trace patient zero who started it, but within days of going viral, different Bigfoot characters with various accents, furs, and pets emerged everywhere. I personally followed one with a unique comedic "personality" —by the name of "Speedilla."

The Bigfoot Economy

The story of Bigfoot Speedilla begins in Hamburg, Germany, where his creator Mo, a Syrian immigrant, discovered that AI-generated Bigfoot vlogs could serve as both artistic expression and potential income during his spare time from work.

"Speedilla, he is part of my inner self," Mo explains when describing the character. "I'm trying to create some kind of dark, satirical comedy."

In Speedilla's virtual world, comedy is the surface of his content, while satirical reflections form its core. As Bigfoot gets rejected by a woman he's trying to court and subsequently turns her into steak as a revenge, this 20-second piece receives hundreds of thousands of likes—it's a collective “middle finger” to mainstream cultural taboos from a group that empathizes with Mo’s creation.

It only took a couple viral videos for Speedilla to build tens of thousands of followers and attracted brand collaboration offers. But the good times didn't last long. Mo soon discovered his content being stolen and replicated by others. He asked his followers to report the “infringement” to the platform, but TikTok did not respond to the claims.

As Lifehacker's Stephen Johnson observed, driven by the profit motives of overnight viral success, countless creators are "putting Bigfoot in all kinds of ridiculous scenarios." This extends far beyond Bigfoot—TikTok has seen an explosion of AI-generated content featuring Star Wars, Harry Potter, and other film franchises, creating a vast but copyright-ambiguous UGC content ecosystem.

The Copyright Conundrum

These vast copyright gray areas haven't gone unnoticed. Entertainment giants Disney and Universal recently sued AI company Midjourney, calling its AI image generator a "bottomless pit of plagiarism" that blatantly infringes on their intellectual property libraries.

In this 143-page lawsuit, Disney catalogued Midjourney's blatant copying of their signature characters, including Storm Troopers and Darth Vader from Star Wars, Elsa from Frozen, and the Minions from Despicable Me, just to name a few.

The lawsuit not only accuses the AI model of directly copying character appearances but also charges that its outputs constitute "infringing derivative works." Disney argues that these images generated from simple prompts are not users' original works, but unauthorized adaptations of their copyrighted characters, directly violating copyright holders' exclusive rights to derivative works.

This accusation directly speaks to the core problem creators face: How to define the originality of AI-generated works? If simple AI outputs are "infringing derivative works," what must creators do to make their works legally recognized as copyrightable original creations?

While this case is still in the process without a final ruling, there are already multiple AI copyright cases serving as precedents. For instance, cases that successfully registered AI-generated content copyrights with the U.S. Copyright Office in 2025: The key lies in proving that human creators exercised "creative control" over the AI generation process, forming "sufficient human creative input". Creators can:

1. Provide original input instructions: As in the previously successful "Rose Enigma" case, the creator uploaded an original hand-drawn rose artwork as a foundation, then used text prompts to have AI visually process it. The composition and creative conception of this hand-drawn draft were considered key evidence of human creative control over the final product.

2. Extensively modify and arrange AI-generated content: In another successfully registered case, "A Single Piece of American Cheese," the creator performed 35 image detail redraws on the AI-generated initial image, adding elements like a third eye and melting cheese, and recomposing the overall picture. The U.S. Copyright Office recognized this active "selection, combination, and arrangement" of AI-generated materials as demonstrating sufficient human originality.

These precedents offer some guidance for AI creators: While the legal framework is still forming, it's clear that copyright is meant to protect human creativity, not AI's computational results. For individual creators like Mo, this means simply generating a character is far from enough. When a generic Bigfoot character was replicated by others, the creator was virtually helpless because he couldn't prove he had invested sufficient, legally protectable "human creativity" in the character's generation.

The Misinformation Minefield

Veo 3 faces similar challenges to Midjourney not only in copyright issues, but its realism and physical accuracy creates information risks that cannot be ignored. Testing by TIME magazine found that Veo 3 can generate convincingly realistic "fake news", including Pakistani crowds burning Hindu temples, Chinese researchers handling bats in virus laboratories, and election workers shredding ballots—extremely inflammatory and controversial content.

In fact, in the first week of Veo 3's release, online users posted fake news segments in multiple languages, including TV anchors announcing J.K. Rowling's death and fake political press conferences. This ability to easily manufacture content has pushed both creation and consumption platforms to the forefront, making transparency assurance and prevention of misinformation an urgent "labeling war" requiring deep involvement from both regulatory bodies and tech companies worldwide.

The Labeling War: A Three-Way Game Between Creators, Platforms, and Regulators

Facing escalating misinformation risks, platforms have implemented inconsistent and often inadequate strategies. As TIME magazine reported, only after they contacted Google about misleading videos generated by Veo 3 did Google, as the content creation platform, take reactive measures: adding a tiny, easily croppable visible watermark to videos.

On the other side where the content is distributed and watched by billions of people, TikTok, the short video platform famous for viral content, has since September 2023 required creators to actively label AI-generated content to avoid misleading audiences. But in reality, when scrolling through TikTok feed today one will find very limited coverage of this "AI-generated" label, with only a small portion of content displaying this label due to creators' voluntary disclosure.

The reason behind this is simple: creators all want their content to appear as realistic as possible. Having a prominent label telling viewers "this is AI-generated" would presumably impact video performance heavily, so relying solely on creators' voluntary reporting isn’t adequate.

Globally, regulatory requirements for AI content labeling are inconsistent. China and the EU take the lead in this space, respectively legislatied a mandate that require content platforms to implement metadata for machine and visible watermarks for humans to identify AI content. The U.S. hasn't legislated at the federal level, but states like California have taken the lead, with multiple bills in the works. Othe regions including Australia, ASEAN, and the Middle East have yet to legislate, details see Table 1.

With incomplete regulatory frameworks, social media platforms are proactively becoming de facto rule-makers. China's Douyin, for example, strictly enforces national mandatory requirements, requiring both standardized metadata and fixed-position visible watermarks for all AI-generated content. Meta’s Instagram and Facebook, Douyin’s sibling TikTok, YouTube, and Snapchat are also actively advancing dual strategies combining metadata (like C2PA standards) and visible labels (like "AI Info," "Imagined with AI"), but implementation varies. In contrast, some platforms like X (formerly Twitter) have been slower to respond, not yet introducing AI content labeling policies that promote information transparency. This inconsistent status across platforms reflects tech companies' attempts to balance user growth, commercial success, and content compliance in the AI race.

Table 1: AI Content Labeling Requirements by Country/Region

Beyond Viral Content

However, despite the complexity of copyright, misinformation, and platform regulation, Veo 3's potential is also being explored by professional directors. While countless content creators chase the next "Bigfoot"-style viral moment, Turkish self-made director Öner S. Biberkökü chose a completely different path. His goal wasn't to create a quick viral moment that would spread on social media, but to use this new technology to tell a story that could deeply move hearts.

Öner's collaborator is Turkey's household name "Queen of Pop" Sezen Aksu, a legendary figure spanning five decades in music, who released a new song Doğrucu. "I want people to remember their childhood," Öner mentioned in the interview with me, "I listened to her songs as a child, her voice has special meaning for Turkish people."

This project cost tens of thousands of dollars in Veo3 credits alone plus substantial human resources, completed by Öner's team (Pepperroot Studio) working around the clock. During creation, Öner and his team deeply explored Veo 3's technical limitations. He frankly admitted that achieving character consistency was “not an easy feat“ using this tool alone, forcing them to combine multiple generation methods through careful planning and repeated experimentation to finally achieve the desired outcome.

This approach differs from another music video Öner created 4 months earlier, "Uchigatana" (which I often refer to as "Reborn as a Samurai in Edo"), which pursued spectacle moments, style transfer, lipsync by AI character that showcased AI’s capabilities. This new music video with Sezen Aksu deliberately maintained restraint and simplicity. "I wanted a simple music video with only one magical moment—the instant when sparrows lift the woman from a fall" Öner said. He prioritized emotional delivery over AI technical showmanship. Ultimately, this "magical moment" served the song's narrative, completing the work's concrete expression to aspire for "hope."

The same creative tool, in different creators' hands, some people can produce comedic viral content, some people use it to fabricate misinformation, and some would wield its power to call for a nation's emotional response.

Looking Forward: The AI Renaissance

In addition to Google's Veo3, the AI video model field is highly competitive with numerous participants. There are platforms like Kling (by Kuaishou) and Dreamina (by ByteDance) that possess vast video training data from their short video platforms, as well as companies like Runway, Higgsfield, and Minimax etc. continuously iterating products in their respective niche sub-fields. New technological innovations are announced almost every few days. Even Reddit already has rumors circulating that Veo 4 will be released in December 2025, though no one can accurately predict what capabilities these models will achieve by then.

However, as Google CEO Sundar Pichai said in a podcast interview, the trend from these advanced models bring is clear—they will enable a dramatic increase in content creation, empowering a broader range of creators—democratizing video creation. He predicts: using AI video tools will become as commonplace as “using Google Docs is today."

The Veo 3 viral moment represents more than technological progress—it's a mirror reflecting the essence of human creativity. When a Syrian immigrant in Germany creates artistic expression with an AI Bigfoot, when a Turkish team produces emotionally complete stories at minimal cost, when the boundaries between real and synthetic blur, when you and I can use AI tools to manifest "What You Think Is What You See" we're not just witnessing a tool's evolution—we're observing the transformation across the entire content industry, what some call the AI Renaissance, an era where AI is accelerating the re-discovery and re-definition of humanity’s place in the world.

This emerging AI Renaissance brings us back to our fundamental principles. The question we need to answer isn't whether AI will change content creation—because it already has. The question is whether we can establish a framework that embraces these tools' democratizing potential while protecting the value of human creativity.

I believe the core of content creation has never been, and should never be, determined by how advanced our tools are, but by the human spirit with which we wield these tools. In that future of AI Renaissance, the stories we choose to tell—and how we tell them—will matter far more than the technology used to bring those stories to life.

Subscribe now

References

I. Interview Materials

1. Öner S. Biberkökü and Cansın Çetin Kuşluvan (Co-founders of Turkish creative studio Pepperroot Studio). Video conference recordings with the author, June 2025.

2. Mo a.k.a. "Speedilla" (AI content creator). Video conference recordings with the author, June 2025.

3. Lex Fridman. (June 5, 2025). Transcript for Sundar Pichai: CEO of Google and Alphabet | Lex Fridman Podcast #471. Retrieved June 29, 2025, from https://lexfridman.com/sundar-pichai-transcript/

II. Legal Documents and Reports

1. ASEAN. (2025). Expanded ASEAN Guide on AI Governance and Ethics for Generative AI. Retrieved June 28, 2025, from https://asean.org/wp-content/uploads/2025/01/Expanded-ASEAN-Guide-on-AI-Governance-and-Ethics-Generative-AI.pdf

2. Disney Enterprises, Inc., et al. v. Midjourney, Inc. (2025). Complaint for Direct Copyright Infringement and Secondary Copyright Infringement. U.S. District Court for the Central District of California.

3. European Parliament. (June 8, 2023, updated February 19, 2025). EU AI Act: first regulation on artificial intelligence. Retrieved June 29, 2025, from https://www.europarl.europa.eu/topics/en/article/20230601STO93804/eu-ai-act-first-regulation-on-artificial-intelligence

4. Future of Privacy Forum. (2024). U.S. Legislative Trends in AI-Generated Content: 2024 and Beyond. Retrieved June 28, 2025, from https://fpf.org/blog/u-s-legislative-trends-in-ai-generated-content-2024-and-beyond/

5. U.S. Congress. (2023). S.2691 - AI Labeling Act of 2023. Retrieved June 28, 2025, from https://www.congress.gov/bill/118th-congress/senate-bill/2691/text

6. King & Wood Mallesons. (March 10, 2025). Clearing the Fog: Analysis of AIGC Copyrightability in the US and Case Studies. Retrieved June 28, 2025, from https://www.kwm.com/cn/zh/insights/latest-thinking/copyrightability-of-ai-generated-works-in-the-us-and-typical-cases.html

7. Digital Policy Alert (2025). China: Mandatory National Standard Labeling method for content generated by artificial intelligence (GB 45438-2025). Standard enters into force on September 1, 2025. Retrieved June 29, 2025, from https://digitalpolicyalert.org/change/10930

III. News Reports and Analysis

1. Chow, A. R., & Perrigo, B. (June 3, 2025). Google's New AI Tool Generates Convincing Deepfakes of Riots, Conflict, and Election Fraud. TIME. Retrieved June 27, 2025, from https://time.com/7290050/veo-3-google-misinformation-deepfake/

2. Johnson, S. (June 23, 2025). The Out-of-Touch Adults' Guide to Kid Culture: Bigfoot Vlogs. Lifehacker. Retrieved June 27, 2025, from https://lifehacker.com/entertainment/the-out-of-touch-adults-guide-to-kid-culture-bigfoot-vlogs

3. Montgomery, B. (June 11, 2025). Disney and Universal sue AI image creator Midjourney, alleging copyright infringement. The Guardian. Retrieved June 28, 2025, from https://www.theguardian.com/technology/2025/jun/11/disney-universal-ai-lawsuit

4. Odilov, S. (December 22, 2024). 2025 Is Poised To Usher In The AI Renaissance—Are You Prepared? Forbes. Retrieved June 28, 2025, from https://www.forbes.com/sites/sherzododilov/2024/12/22/2025-is-poised-to-usher-in-the-ai-renaissanceare-you-prepared/

5. Ismail, K. (April 17, 2024). Snapchat to Watermark AI-Generated Images. PetaPixel. Retrieved June 28, 2025, from https://petapixel.com/2024/04/17/snapchat-to-watermark-ai-generated-images/

6. Tolentino, J. C. (September 21, 2024). TikTok is watermarking AI-generated content. Here's what that means. Mashable. Retrieved June 28, 2025, from https://mashable.com/article/tiktok-watermarking-ai-generated-content

IV. Platform and Policy Documents

1. China Law Translate. (2024). A Glimpse of the Future? Why China's Labeling Law May Signal a Global Shift in AI Governance. Retrieved June 28, 2025, from https://www.linkedin.com/pulse/glimpse-future-why-chinas-labeling-law-may-signal-swaner-jd-aigp-homqc

2. Douyin. (2024). 抖音关于人工智能生成内容的标识的水印与元数据规范 (Douyin's Watermark and Metadata Specifications for Labeling AI-Generated Content). Retrieved June 28, 2025, from https://www.chinalawtranslate.com/en/抖音关于人工智能生成内容标识的水印与元数据规/

3. Legal123. (n.d.). How to Guide: Legal Issues with AI & ChatGPT in Australia. Retrieved June 28, 2025, from https://legal123.com.au/how-to-guide/legal-issues-ai-chatgpt/

4. Meta. (February 2024). Labeling AI-Generated Images on Facebook, Instagram, and Threads. Retrieved June 28, 2025, from https://about.fb.com/news/2024/02/labeling-ai-generated-images-on-facebook-instagram-and-threads/

5. PBS NewsHour. (June 13, 2023). New bipartisan bill would require labeling of AI-generated videos and audio. Retrieved June 28, 2025, from https://www.pbs.org/newshour/politics/new-bipartisan-bill-would-require-labeling-of-ai-generated-videos-and-audio

6. TikTok Newsroom. (September 21, 2023). New labels for disclosing AI-generated content. Retrieved June 28, 2025, from https://newsroom.tiktok.com/en-us/new-labels-for-disclosing-ai-generated-content

7. Thomson Reuters. (n.d.). How is AI regulated in the UAE? What lawyers need to know. Retrieved June 28, 2025, from https://insight.thomsonreuters.com/mena/legal/posts/how-is-ai-regulated-in-the-uae-what-lawyers-need-to-know

Should I rent or buy a home?

Lately, this topic has come up a lot in my conversations with friends who are in their 30s. Some of them are based in China, watching house prices decline for the past 2 years and wondering if it’s bottomed. Some in Australia, UK and EU, seeing the significant drop in mortgage rates led by central bank decisions, and contemplating becoming a homeowner rather than "helping the landlord to pay off their mortgage".

This is a fundamental question that troubles many many people, and I’m not an expert that can give an one-size-fits-all answer. Today we'll explore 10 common misconceptions about Renting vs Buying by examining the research of Professor Eli Beracha, the world's 3rd ranked Real-Estate-Academic-Leadership author in this field.

This article draws from his three influential papers and two in-depth podcast interviews, plus other supporting research. Through his analysis of over 30 years U.S. real estate data, he uncovered a mind-boggling truth:

While conventional wisdom about homeownership may be right, it does so for the wrong reasons.

The 10 Major Misconceptions

Misconception 1: "Buying is always better than renting"

Contrary to the common belief that renting is "throwing money away”, from a purely financial perspective, renting can lead to higher wealth accumulation on an unadjusted basis. The research analysed over 30 years of data from the 1980s to 2010s found that if viewed purely from an investment return perspective, renters who invest their savings in stocks and bonds often accumulate more wealth than homeowners. This makes theoretical sense—

"Because you spend less on housing costs through renting, and then by investing that difference wisely, you can get ahead or at least reach an equivalent level."

But the key lies in the premise of "wisely investing."

Misconception 2: "Renting is just paying your landlord's mortgage"

This is the most misleading statement, causing countless people to ignore the real costs of buying.

In reality, both buying and renting have 'disappearing money,' but most people selectively ignore the sunk costs of buying.

"Disappearing money" refers to money you pay out that you'll never get back—essentially pure consumption rather than investment. For renters, the "disappearing money" is simple: rent. But the sunk costs of homeowners are severely underestimated: mortgage interest, property tax, insurance, maintenance, transaction costs, and the opportunity cost of the initial capital.

The danger of this misconception is that it creates the illusion that buying a house "doesn't cost money."

Misconception 3: “House prices only go one way—up.”

Typical cognitive bias leads people to overestimate real estate investment returns. It's precisely this widespread perception that gives real estate certain "rational bubble" characteristics. People only remember the purchase price and selling price, completely forgetting the inputs and time costs in between. In reality:

"Long-term house price increases basically equal the inflation rate, with almost zero growth after adjusting for inflation."

From the data:

• U.S. 1890-2020 (130 years): Annual house price growth rate approximately 3.2%

• Inflation rate same period: approximately 2.9%

• Real growth after inflation: only 0.3%

  

Misconception 4: Houses are safer than stocks

Many people believe real estate is safer than stock, and this view is widespread:

"Houses are tangible and safe, stocks are just numbers on a screen."

But this completely ignores the real risks of a single property. Real estate actually has many hidden risks: Using a typical American freestanding house as an example, from daily issues like mold and pests, to neighbor building an extension that blocks your view, school rezoning, and even earthquakes and floods—these events impact single property far more than the impact on a diversified investment portfolio.

In comparison, stock funds consist of hundreds or thousands of the world's best companies across various industries. This diversification is a much better protection to uncertainties than that of a single asset.

Misconception 5: Homeowners are wealthier than renters, so buying makes people rich

This is a textbook example of "treating correlation as causation." Data indeed shows homeowners have on average 40 times the wealth of renters, but:

It's not buying homes that makes people wealthy, but rather wealthy people are more likely to buy homes. It's like saying people who wear Rolex watches are more successful, but the Rolex didn't make them successful.

Key finding from research: A wealth threshold effect

Data modeling shows surprising results: When one's net worth is below the value of one house, renting actually provides higher long-term returns. Only when you already have sufficient wealth then buying truly become advantageous.

Misconception 6: Since renting is better, you should never buy any property

Theoretical result and practical experience are completely different. Human’s inherent behaviour weaknesses determine that buying is better for the vast majority of people.

Because buying is essentially "forced saving"—even if you don't understand investing or lack self-control, you can passively save money through mortgage payments.

To put it simply:

"You don't have a choice between paying your mortgage and not paying the mortgage, but you do have a choice between saving and spending."

Why is "forcing" so crucial? Renters can theoretically invest their savings in stocks, bonds, gold, crypto and other diversified asset portfolios for higher returns, but in reality, most people can't do this—they'll pause, reduce, or withdraw investments for various reasons, ultimately ending up with zero savings. Not paying a mortgage means losing the house, and this "forced saving" mechanism helps buyers overcome human weaknesses.

Unexpected advantages of real estate: The "disadvantages" of real estate—opaque pricing, poor liquidity, high transaction barriers—actually prevent impulsive decisions. Stock price updates every second, easily triggering emotional behaviors like FOMO (Fear of Missing Out) and FUD (Fear, Uncertainty, Doubt), leading to emotional trading activities. Houses only truly reveal their exact value when put on the market, and selling requires a complex 2-3 months process. These "hassles" prevent emotional decisions that have significant financial implications. This is similar to buying physical gold—precisely because trading is troublesome, people tend to hold long-term, avoiding emotional behaviours.

In summary:

The seemingly negative aspects of real estate actually become advantages for saving—working with human weaknesses instead of fighting against them.

Misconception 7: Everyone should buy a house

Three types of people should avoid buying houses, as homeownership would work against their interests. The professor clearly identifies:

1. people with unstable financial situation that can't afford a 15-year mortgage

2. young professionals who frequently change cities, and

3. high-net-worth individuals with sufficient investment discipline.

Do you fit the bill for any of these 3 categories?

The first group would be dragged down by the house, the second would miss opportunities, and the third doesn't need “forced saving”.

Misconception 8: House prices are unpredictable

Actually, real estate markets are more predictable than stock markets; the key is finding the right method. Professor Beracha just won first place in a "prediction" competition among 150 real estate economists, earning the somewhat ironic "Crystal Ball Award."

His model incorporates various metrics but the core is quite simple: Rent-to-price ratio = Annual rent/Total house price

Principle: When house prices in a city are too high relative to rents, the market self-corrects. Like stock P/E ratios, when rent-to-price ratios deviate too far from historical averages, they would revert to normal:

• U.S. reasonable level: 3-4%

• China reasonable level: 2% (due to lower interest rates)

But note:

"This model doesn't predict absolute gains, but relative performance—comparing horizontally, which markets will outperform others."

Misconception 9: Tier-1 cities are "too expensive" to buy

In ultra-high-priced cities with extremely low rent-to-price ratios, traditional buying logic fails. Why?

1. Severely imbalanced rent-to-price ratios: Renting is almost always more financially sensible, and the opportunity cost of buying is enormous—investing the same money in other assets could yield far better long-term returns than real estate itself.

2. Liquidity trap: All wealth locked in real estate limits career choices, geographic mobility and investment decisions, lacking flexibility when facing emergencies/opportunities.

But when is it still worth buying?

1. Ultra-high income groups: Low mortgage payment pressure with sufficient funds for other investments

2. Clear long-term demand: Those who need a long-term residence for education etc., willing to pay premium for certainty

3. Special professional needs: Work strongly tied to specific locations, unstable rental market

Many people buy because they worry "if I don't buy now, I'll never afford it," but in ultra-high-priced cities, buying is often an emotional rather than rational decision. It's precisely because of widespread public irrationality that these cities’ house prices remain "bubbly". If you really want to buy, ensure it's a well-considered rational choice, not an impulse decision driven by social pressure.

Misconception 10: The decision making process is the same for everyone

The logic of buying for the rich and the rest differs dramatically. For high-net-worth individuals, real estate is a hedging tool in investment portfolios that can improve overall risk-return ratios. But for the middle class, from a rational perspective, if real estate comprises too high a percentage of net worth, risk becomes overly concentrated. This means while both are buying houses, the wealthy are allocating their wealth in different assets but for ordinary people this is going all-in on a singular asset.

But beyond rational analysis, we can't deny that real estate differs fundamentally from stocks—it's not just an asset but it brings many values that can't be measured in money, such as:

• Emotional value: Sense of belonging, stability and control

• Practical value: Renovation freedom, pet-friendly, storage space

• Social value: Children's education, neighborhood relationships, family gatherings

So my recommendation is:

"Don't let others' success models dictate your choices. There’s no standard answer—only the one that suits you best."

References

1. Beracha, E., & Johnson, K. H. (2012). Lessons from Over 30 Years of Buy versus Rent Decisions: Is the American Dream Always Wise? Real Estate Economics, 40(2), 217–247. https://doi.org/10.1111/j.1540-6229.2011.00321.x

2. Beracha, E., & Skiba, H. (2013, August). Findings from a Cross-Sectional Housing Risk-Factor Model. The Journal of Real Estate Finance and Economics. https://doi.org/10.1007/s11146-011-9360-x

3. Beracha, E., Skiba, A., & Johnson, K. H. (2017). Housing Ownership Decision Making in the Framework of Household Portfolio Choice. Journal of Real Estate Research, 39(2), 263–289. https://doi.org/10.1080/10835547.2017.12091472

4. Case, K. E., & Shiller, R. J. (1989). The Efficiency of the Market for Single-Family Homes. The American Economic Review, 79(1), 125–137. http://www.jstor.org/stable/1804778

5. David Siddons Group. (2024, November 27). The Fundamentals Driving Miami’s Real Estate Resilience with Prof. Eli Beracha [Video]. YouTube. Retrieved from David Siddons Group YouTube channel.

6. Linping Residential Big Data Research Institute [麟评居住大数据研究院]. (2024, September). Rent-to-Sale Ratio in Different Chinese Cities and Global Comparative Analysis [中国不同城市租售比 及全球对比分析].

7. Money. (2025, June). Homeowners are now 43 times wealthier than renters. https://money.com/homeowners-wealthier-than-renters-net-worth/

8. Saginor, J. (2025). The Real Estate Academic Leadership (REAL) Rankings for 2020–2024. Journal of Real Estate Literature, ahead-of-print(ahead-of-print). https://doi.org/10.1080/09277544.2025.2476876

9. The Rational Reminder Podcast. (2024, June 18). Eli Beracha: An Academic Perspective on Renting vs. Owning a Home | Rational Reminder 358 [Video]. YouTube. Retrieved from The Rational Reminder Podcast YouTube channel.

Are you worried that your skills are becoming obsolete?

Are you concerned that the money in your pocket is becoming worthless?

Do you feel small, unable to change the social status quo?

Are you worried that your children will face an even more difficult world in the future?

Have you been feeling a sense of "out of control" from time to time in the past 1-2 years, as if being pushed around by an external force?

If your answer is Yes. Don't worry, you're not the only one.

Subscribe now

This book co-authored by three accomplished authors, "Permacrisis: A Plan to Fix a Fractured World," (Gordon Brown, Mohamed A. El-Erian, Michael Spence, 2023) is a policy book for business leaders, academics, and politicians. While it may not give individuals direct answers as a self-help guide, it might help you and me "understand" what’s happening to our lives.

First, Why Does Everything Feel Like It's Falling Apart?

This isn't your delusion, nor is it your incompetence. Our world is undergoing structural changes. Terms like "the East rising and the West declining", "great changes unseen in a century", and "Make America Great Again" are all specific manifestations of this. You should know that this isn't just cyclical fluctuation, but rather a structural transformation. What we feel personally is actually a microcosm of global systematic problems.

The book proposes three fundamental shifts:

• The end of American hegemony — leading the world from unipolar to multipolar

• Politics taking priority over economics — leading globally from neoliberalism to new nationalism

• Risk management prioritized over efficiency optimization — leading capital from hyper-globalization to managed globalization-lite

These transformations have brought us into a new norm of "permacrisis," but for individuals, this crisis feeling could be the outcome of being online 24/7, cameras in every corner of the globe capturing every incident, and endless content amplified through generative AI on social media.

As the authors put it:

"Don't let the perma prefix fool you; there's nothing permanent about a permacrisis."

Throughout history, humanity has advanced civilization and technology in a cyclical pattern. However when crises arise, humanity always manage to unite at critical moments to solve the problems.

Particularly for people of Gen Y and younger generations, because human’s feelings are subjective and often stem from personal experience and memory. The relative stability and growth of the past 30 years is actually "abnormal" in the long timespan of human history. The instability we are witnessing today is actually a mean-reversion to historical normalcy.

Whether you live in China, the US, Europe or any other countries, talk to the elders, they will tell you that decade-long periods of peace with stable economic growth are indeed rare from their experience.

So, What Can Individuals Learn From This Policy Book?

1. Actively embrace change, value "new things," and maintain lifelong learning habits — Do you know anyone around you who aren't old yet but don't know how to operate ordinary household appliances, or even use basic phone functions? Try not to decay like them. Actively adapt to change, simply because returning to past states is an unrealistic fantasy.

2. From an investment and financial perspective, the book mentions structural changes to a supply-constrained economy, meaning higher long-term inflation (Note: This especially targets the US with its high trade deficit; China is relatively better off due to its over-supply economic structure). Diversify investments in "assets" with long-term intrinsic value, deal “bubbles” cautiously and with rationality, and practice sound risk management to ensure peace of mind during uncertain times.

3. In daily life, avoid making impulsive decisions based on what you see while scrolling through the feed (as some of the content are meant to trigger emotional responses through your “System 1”), and try not to be overly anxious either with your “System 2”. Following the principle of "naming is managing", by acknowledging our current state as a "permacrisis," policymakers would have already begun studying and working to manage this situation.

4. Become a long-term optimist. If you're a parent, you can explain the current situation to your children like this:

"Dear kids, we adults messed up the world, but we're working hard to fix it. We might not succeed immediately, but we'll keep trying. This is the best gift our generation can give you: not a perfect world, but the human spirit of never giving up."

P.S.

I was fortunate to attend a lecture today by Professor Spence, one of the book's authors and Nobel Prize winner in Economics, at Peking University.

During the Q&A session, one of the questions perfectly echoed my own concerns as a new father of a 2-week old.

"I have a three-year-old daughter, and I need to balance busy work with parenting responsibilities," she said.

"In this AI era, how should I help my child better adapt? Could you give me some advice, preferably simple—something that can be explained in one sentence?"

The entire audience burst into spontaneous applause, as if this mother had spoken everyone’s mind — people want simple answers to navigate the complexity.

Professor Spence wasn't surprised by the question. He answered by addressing the overwhelming AI coverage in social media, where outlets compete for attention and traffic. He emphasized a crucial point: parents should act as information gatekeepers, ensuring children access valuable information from reliable sources rather than letting them use social media unrestricted. This, he added, is both a parental responsibility and a crucial task for all countries in governing AI and internet space.

Since it was the day of national college entrance examination for all Chinese students in the whole country, Professor Huang Yiping, the host, followed up:

"What advice do you have for today's exam students regarding selecting a major for university? Is studying economics still useful in the AI era?"

"Never choose economics!" Professor Spence quipped, drawing laughs. Then, turning serious, he noted that while many parents are steering their children toward AI and computer science majors, this isn't necessarily ideal. He emphasized a fundamental truth: discovering each child's unique interests and allowing them to make authentic choices about their majors—choices that will give them the inner motive and fulfillment to "wake up from bed every day". In terms of learning AI technology specifically, he compared AI’s societal impact to that of the printing press, explaining that as a General Purpose Technology, AI will have impact across the board. And that individual success lies in finding the intersection between one's chosen discipline and AI.

After the group photo session, I quickly made my way to the front row to get Professor Spence's autograph which I'm sharing with everyone now, as it might actually give a simple one-sentence answer to that mother's question:

"Don't Panic."

This wasn't my first time in Turkey. What was meant to be a brief stop-over for work unexpectedly turned into a historical pilgrimage because of the conversations I had with local friends.

1. Constantinople: Capital of Three Empires

The Hagia Sophia is a must-see for any visitor to this city. Not far from this magnificent cathedral stands the Obelisk of Theodosius—an ancient Egyptian monument rising majestically in the center of the square.

My Turkish friend Aylin told me, "Every stone here is history. You know this obelisk? It was specially cut and transported from Egypt in 330 AD by Emperor Constantine the Great. He ordered this monument with ancient inscriptions to be erected here, and named this city after himself—Constantinople. If you look closely, the base of the obelisk has carved illustrations showing how it was transported from Egypt."

Indeed, looking at the map, Constantinople spans two continents, and its geographical prominance gave this city unparalleled strategic value, making it a crucial hub for Eurasian trade.

"You know, this was the capital of three great empires of the old world," Aylin explained proudly. "The Roman Empire for several centuries, the Byzantine Empire for over a thousand years, and the Ottoman Empire for six hundred years. Even China, with its thousands of years of civilization, doesn't have a single city that can compare."

Especially the Byzantine Empire—leveraging Constantinople's strategic position at the center of the Silk Road, connecting East and West, it created the most successful economic system of the medieval period. This name continued from the Roman Empire until the early 20th century, when the city was renamed Istanbul as it expanded—the word derives from a local phrase "eis tin polin," literally meaning "to the city."

2. The Ottoman Empire: The Art of Inclusion as a ruler

Walking casually with Aylin through the courtyards of Topkapi Palace, I posed a naive and somewhat rude question: "Why don't you look like what I expected a Turkish person to be?" Aylin’s skin tone was pale-white and had very defined facial features—if she hadn't told me herself, I would have mistaken her for European.

Aylin chuckled, seemingly not offended at all. "That's the charm of Turkey," she said, stroking an ancient palace column with one hand. "We are Turks—but it's hard to define clearly. On the street you can see people with black hair, blonde hair, blue eyes, green eyes, and very dark skin. Turkish people are a historical mixture."

Subscribe now

She paused, gazing toward the distant Bosphorus: "You know, the origins of major Eurasian peoples are all mysterious. They say the Celts originated right here—a tribe from Anatolia crossed the strait, migrated north, and eventually became the Gauls and Celts of Europe."

"So you're saying Europeans actually come from Turkey??" I was somewhat surprised.

"Some of my European friends have done DNA testing and found they all have some Anatolian ancestry, because for thousands of years people have continuously migrated through here. This has always been humanity's crossroads."

Following this topic, I curiously asked how Turkey historically managed such diverse peoples.

She explained that after the Ottoman Empire conquered the Mediterranean and established its capital in Constantinople, it adopted a system called the “Millet system”. The rulers said, “Although we are Muslims, I allow you to maintain your own religion and culture. Greeks had Greek courts, Armenians had Armenian schools, Jews had Jewish communities. You just need to pay some extra taxes for protection."

This inclusive and pragmatic policy created an unprecedented multi-ethnic empire. At its peak in the 16th century, Europe's total population was about 90 million, while the Ottoman Empire alone had 28 million people.

"But," Aylin's tone became somewhat hesitant, "this diversity also planted the seeds for later fragmentation. When nationalism became popular in 19th-century Europe, various ethnic groups within the Ottoman Empire began to awaken, all learning from the spirit of the French Revolution and pursuing their own nation-states."

She led me to the palace's observation deck, looking out at the modern city skyline in the distance. "The Greeks were the first to start independence movements. They cleverly combined religion and ethnicity, saying 'We're not just Greeks, we're Christians,' which won them support from Tsarist Russia. Then came the Bulgarians, Serbs, Albanians... one after another."

"What about the Kurds?" I asked, thinking of this name I often heard in the news.

Aylin sighed. "The Kurds are one of the most tragic peoples. They're distributed across the border regions of Turkey, Syria, Iraq, and Iran, with a total population possibly in the tens of millions, but they've never had their own country. Even now, this problem remains unresolved. Many Kurds are still fighting for their 'nationalist' dreams. But no government will allow such separatist forces to take root in their own country."

"The 'legacy' of a multi-ethnic empire," I said with some irony.

"Yes," Aylin nodded. "The Ottoman Empire's inclusive policy created its thousand-year glory, but also predetermined its fragmentation. When every ethnic group began pursuing 'pure' nation-states, diversity became a fatal weakness. This is history's irony: yesterday's advantage of an empire might become today's burden under a new wave."

"Exactly," I responded. "Look at the past decade—there was a trend toward globalization and multiculturalism, but now many countries are seeing the rise of right-wing political forces, stirring up waves of nationalism and statism. It's like a Chinese saying from the Three Kingdoms:

'The empire, long divided, must unite; long united, must divide.’”

A gentle breeze swept through the palace courtyard, bringing a refreshing coolness. In this former center of power that once ruled such vast territories, we discussed these eternal, recurring themes of ethnicity and state, inclusion and division, while history's whispers echoed on the breeze.

3. Mustafa Kemal Ataturk: Architect of Modern Turkey

The next day, I took a ferry to Princes' Island, just half an hour from Istanbul's city center. This small island had served as an imperial exile for royal families receiving the "cold palace" treatment, later evolving into a celebrity vacation and refuge spot.

Walking around the car-free island, with only bicycles and aging low buildings, time seemed to stand still—perfect for contemplating this country's thousand-year journey. Ataturk's portraits were omnipresent: adorning building murals, hanging as flags outside homes, and even appearing in street art.

I remembered another Turkish friend, Resul, once told me that modern Turkey's founding father, Mustafa Kemal Ataturk, would be roughly equivalent to Chairman Mao's status in China. On May 16, 1919, he secretly departed from Princes' Island for Anatolia, beginning the Turkish War of Independence. At the time, the Ottoman Empire had been defeated, Constantinople was occupied by British and French forces, and Kemal realized that he must organize a national liberation movement. He chose to stay temporarily on Princes' Island, then quietly left by a small boat, beginning the journey that would change Turkey's destiny.

After several years of war, when the Republic of Turkey was established in 1923, Kemal brought thorough secular reforms. Politically, he advocated separation of relgion and state, abolished the Caliphate system, and replaced the Sultanate with a parliamentary system. Culturally, the new Turkish Republic abandoned the Arabic calendar and adopted the Western Gregorian calendar, Arabic script was also abolished in favor of Latin letters. In terms of social systems, he advocated gender equality, granted all Turkish women the right to vote, and legislated against polygamy. He even banned the traditional fez hat, believing this attire represented the old, corrupt Arab Muslim world. He said, "We must face west and wear Western-style hats," striving to show the world a unified, nationally identified new Turkey.

This was like a social experiment running at 2x speed, but he succeeded. In just ten years after the new republic's establishment, Kemal completely changed Turkey's identity, transforming it from an Islamic empire into a secular modern state, forming a multi-ethnic integrated nation and completely escaping the tragic history of being carved up by European forces during the Ottoman Empire's final period.

Therefore, in Turkey, locals respectfully call Kemal "Ataturk," meaning "Father of the Turks." This title wasn't easily earned—it was specially granted by the Turkish Grand National Assembly through legislation in 1934. He not only led Turkey to victory in the War of Independence but, more importantly, painted a completely new modernization vision for Turkey and successfully turned that vision into reality.

4. Erdogan's 20-Year Rule: Going Against Ataturk’s legacy?

In the evening, at a small restaurant overlooking the Bosphorus, I chatted with my friend Arif about current changes in Turkish politics.

"Although Erdogan himself is a product of secular Turkey—he was born and raised in Ataturk's new Turkey—over the past 20 years, he has tried to completely overturn Ataturk's legacy by reintroducing Islamic forces into the political system."

Arif took a sip of rakı and continued: "Erdogan was originally Istanbul's mayor, then became Turkey's Prime Minister starting in 2003. He governed under the parliamentary system for 11 years. As Prime Minister, he had to negotiate with parliament—the decision-making process was complex. Even so, during his early tenure, through freemarket economy, Turkey's GDP grew from $230 billion in 2003 to nearly $900 billion in 2013. He also promoted a series of infrastructure construction—Istanbul Airport, the third Bosphorus bridge, high-speed rail networks. These 'new' things let ordinary people tangibly feel the life improvement brought by economic development."

"But he wasn't satisfied," Arif took another sip and lowered his voice: "At that time Erdogan kept claiming Turkey needed 'strong leadership' to respond to regional turmoil and economic challenges. In 2014, he became president. According to the constitution then, as president he could only serve one additional term and would have to step down by 2019."

"The turning point was the 2016 failed coup attempt, which gave Erdogan a good reason to push for a constitutional referendum. He said Turkey faced internal and external threats and needed more efficient decision-making mechanisms. After the 2017 constitutional referendum passed, he successfully changed Turkey from a parliamentary to a presidential system. After winning the 2018 elections, he officially became the first president under the presidential system, concentrating all power in himself."

Arif quickly glanced around and whispered: "And because of the constitutional change, his term was recalculated—he can continue governing until 2028, maybe even longer."

I detected the subtle shift in Arif's tone and pressed: "Do voters buy into this?"

"You've hit the nail on the head," Arif's tone became more serious. "Istanbul citizens are highly educated and have some exposure to Western democratic systems through news media, so their support for Erdogan has been declining in recent years. His political opponent when he barely won last time was actually from Istanbul's district. But Erdogan is smart—he knows exactly where the large vote blocs are. 70% of Turkey's population lives in rural areas and small towns. These regions are relatively conservative with higher religious devotion. Although Kemal's secularization reforms were successful, it mainly affected big cities and intellectual elites."

"So Islamic values gradually became one of the cores of his policy," Arif took a gulp and explained: "Erdogan wants to firmly grasp these so-called 'forgotten majority.' When he says 'We must restore Islamic traditions' and 'Western secularism doesn't suit Turkey,' Anatolian farmers and small traders feel their values are finally recognized again. These people were culturally marginalized for decades in Ataturk's Turkey."

"More cleverly," Arif continued, "Erdogan combines religious identity with national pride. He tells voters, 'The West looks down on us because we're Muslims, but we should be proud of our identity.' This message has been particularly effective in recent years because Western countries have faced several crises, and with amplification by Turkish state media, more people have become skeptical of Western models."

"To be fair, Erdogan did achieve some successes. Early economic development made Turkey the first among Middle Eastern Muslim countries to become a model ahead of Saudi Arabia and Iran." Arif paused. "However, problems have become increasingly obvious in recent years—economic growth has slowed, the Turkish currency has depreciated significantly, inflation remains high, and commodity prices change daily. This has caused large numbers of middle-class people to passively fall back into poverty due to Turkish Lira’s devaluation."

"So what's the solution?" I was genuinely curious about how Turkish people view future trends.

Arif said somewhat helplessly: "I don't know. I for one definitely won't vote for Erdogan again. Many young Turkish people think this way. We feel 'fed up'—it's been Erdogan since I became aware of politics, over 20 years and it's still him. We just want some change." After saying this, Arif downeed the remaining rakı from the bottle.

5. Turkey's Future: Swaying at the Crossroads

I sat in Istanbul Airport's departure lounge, looking through the large windows at planes taking off and landing on the runway in the darkness, my mind constantly replaying the conversations and observations from these past two days.

The crowds in the airport were like a microcosm of this country: religous women wearing all-black hijabs sitting on the same bench back-to-back with young women carrying designer bags, a casually dressed father dragging a family of four queueing up next to a busy businessman in a suit, tradition and modernity, East and West, conservative and open—all cultures coexisting harmoniously here.

Kemal Ataturk transformed an Islamic empire into a secular republic in just ten years—this radical modernization experiment shocked the entire world at the time. Erdogan's twenty years of rule attempted to swing the pendulum back to the traditional side, re-embracing Islamic identity and conservative values.

I remembered Aylin words: "This has always been humanity's crossroads." Indeed, from ancient times to now, whether by land, sea, or today's air routes, this land has continuously witnessed the meeting and collision of civilizations. Every historical turning point finds this place at the storm's center.

For the next 20 years, Turkey will continue to be at the forefront and center of European, Asian, Middle Eastern, and American geopolitics. This is a country of 90 million people, including 13 million immigrants, a country that is both a NATO member and maintains delicate relationships with its adversary Russia, acountry that is both secularized and has deep Islamic traditions, stands at yet another historical crossroads.

And I, as a mortal observer, can only watch with bated breath as this pivotal choice unfolds—one that will shape the future as history writes its next chapters. In this land of countless rises and falls, history's pendulum continues to swing, and the answer—only time knows.

Recently, I stumbled across an intriguing study from professors of Tsinghua and Peking University—institutions based in a country where Bitcoin is officially banned and commonly labeled a "scam" by the general public—titled "Bubbly Bitcoin" (Dong, F., Xu, Z., & Zhang, Y., 2021). Employing behavioral economics alongside meticulous mathematical modeling, the paper explores a fascinating question: How can Bitcoin, a digital asset with no intrinsic utility, manage to sustain a market price? Beyond emotional speculation, what deeper forces underpin its value?

The paper attempts to address three key questions:

Why does Bitcoin have value? How does sentiment influence its price? What do rational investors see in Bitcoin?

1. Why Does Bitcoin Have Value?

To answer this, we must start with the concept of rational bubbles.

A rational bubble is intriguingly defined as: an asset whose prices do not reflect fundamentals” (Stiglitz, 1990), i.e. its price rises due to investors' expectations of future price increases, despite having no intrinsic fundamental value.

For instance, Bitcoin itself has no practical use or governmental backing, but people buy it because they believe someone else will pay even more in the future. This collective belief pushes prices higher.

When market sentiment is sufficiently optimistic, Bitcoin, fundamentally worthless, can sustain itself as a rational bubble. Such bubbles provide liquidity to real economic investments but also reduce total social output, as they don't generate intrinsic value.

Historical analogies include the real estate and stock market bubbles. Consider Japan in the early 1990s or the U.S. during the Great Recession—real estate prices surged not due to genuine housing demand or rental income but largely due to the simple expectation that prices would keep rising. Once expectations reversed, the bubble burst.

2. How Does Sentiment Influence Price?

In the model presented, "sentiment" is mathmatically defined as the inverse of the probability of bubble bursting.

Simply put:

• Negative sentiment indicates higher perceived risk of bubble bursting.

• Positive sentiment implies expectations of continued bubble existence.

Interestingly, empirical evidences show positive sentiment has limited impact to Bitcoin’s price, whereas negative sentiment significantly affects prices. For instance, Bitcoin price’s sharp declines following regulatory tightening in 2017-2018 occurred notably within 15 days of negative announcements.

The study also found that during optimistic markets, negative sentiment may depress Bitcoin prices but can stimulate real economic growth as liquidity moves toward more productive uses.

However, in already pessimistic markets, negative sentiment can intensify economic downturns due to reduced market liquidity.

From a societal viewpoint, Bitcoin’s bubble burst could paradoxically enhance overall welfare—reduction in Bitcoin mining activities would free up power and computing resources for productive economic use. However, a critical warning: in an already fragile macroeconomic environment, a large-scale bubble collapse could exacerbate downturns, potentially triggering a severe economic depression.

For instance, amidst the current 2025 global recession expectations due to escalating trade wars, a significant bubble burst could lead to worldwide economic distress.

3. What Do Rational Investors See in Bitcoin?

Bitcoin's price fluctuations largely stem from financial environment differences faced by investors.

Companies face external factors like transaction costs, subsidies, and taxes, directly influencing investment costs. Amid economic uncertainty, businesses need sufficient liquidity to capitalize on future opportunities. Bitcoin emerges as an effective liquidity tool through cryptocurrency’s financial tools, helping manage uncertainty and liquidity constraints.

Thus, businesses willingly pay extra for holding Bitcoin—thus creating liquidity premium. Intriguingly, the greater the external financial shocks, the higher the Bitcoin demand.

In an imperfect financial market, traditional risk-hedging mechanisms often fall short. Hence, investors resort to liquid assets like Bitcoin to enhance flexibility. This growing demand raises Bitcoin's price above its intrinsic value—sustaining a rational bubble due to continued optimistic expectations.

Consider two investor scenarios:

• Investor A: A small business owner facing bank lending difficulties and illiquid assets. Rationally, he invests in Bitcoin and borrow cryptocurrency to maintain liquidity, despite acknowledging the risk.

• Investor B: A high-income professional with ease of access to traditional financing and less impacted by market imperfections, thus minimal or no Crypto holding.

Outcome:

• Rational investors like A elevate Bitcoin demand due to liquidity needs.

• Although Bitcoin lacks intrinsic dividends like stocks or bonds, belief in future appreciation forms a rational bubble.

4. 💡Extended Analysis - Does Chinese Economic Growth Influence Bitcoin Prices?

An intriguing observation:

Bitcoin markets in the U.S. and China display contrasting patterns. In the U.S., Bitcoin prices show a counter-cyclical pattern, rising during economic downturns. Conversely, in China, Bitcoin price movement aligns with economic cycles—rising with economic growth and falling during downturns.

Why the divergence?

The research attributes it to investor sentiment. Due to different regulatory environment, U.S. investors tend toward optimism regarding bubble stability, whereas Chinese investors exhibit greater caution and higher risk perceptions.

Specifically:

• In an optimistic market (U.S.), negative sentiment redirects funds from Bitcoin to productive investments, improving economic efficiency.

• In a more cautious market (China), negative sentiment exacerbates economic anxiety, potentially triggering broader economic issues.

In addition, structurally, China's economy relies heavily on real economic activities, with positive sentiment directly boosting risk appetite (pro-cyclical) hence more liquidity shift towards high yield financial products. Whereas with U.S. dominance in financial market, negative Bitcoin sentiment would boost liquidity toward alternative assets.

5. 💡Extended Analysis - How Do Rational Bubbles Evolve?

Referring to Michau, Ono, & Schlegl's (2023) paper “Wealth preference and rational bubbles.”, the economic system with rational bubbles demonstrates three states:

• Bubble Stability: High asset prices boost perceived wealth, increasing consumption.

• Bubble Decline: Prices normalize, stabilizing consumption.

• No-Bubble State: Asset prices reflect intrinsic values, stabilizing consumption aligned with income.

Interestingly, bubbles may persist sustainably under certain conditions—a tendency to save, natural interest rates that’s lower than population growth, and declining asset dividend.

Key reasons:

• Tendency to save: People inherently value wealth accumulation over immediate spending. This is common in East Asian culture, but less so for consumption-driven economies like the U.S.

• Interest Rate < Population Growth: Low risk-free returns from banks drive asset investment as wealth storage.

• Declining Asset Returns: Despite decreasing future returns, people buy assets hoping for appreciation, sustaining rational bubbles.

This scenario extends beyond Bitcoin to other assets—Even Japanese government bonds exemplify a rational bubble. High savings, low interest rates, and minimal returns foster an environment where investors maintain confidence, despite debt-to-GDP ratios exceeding 250%.

This illustrates a critical insight: rational bubbles can persist sustainably even in seemingly low-risk, stable and institutionally reinforced environments.

Additional Reading

Daniel Kahneman, author of Thinking, Fast and Slow, passed away in 2024. His groundbreaking work in behavioral economics, awarded the 2002 Nobel Prize, continues to inspire cognitive exploration in the age of artificial intelligence.

References

• Dong, F., Xu, Z., & Zhang, Y. (2021). Bubbly Bitcoin. Economic Theory. https://doi.org/10.1007/s00199-021-01389-y

• Michau, J.-B., Ono, Y., & Schlegl, M. (2023). Wealth preference and rational bubbles. European Economic Review, 156, 104382. https://doi.org/10.1016/j.euroecorev.2023.104496

• Stiglitz, J. E. (1990). “Symposium on Bubbles”. Journal of Economic Perspectives 4.2, 13–18. doi: 10.1257/jep.4.2.13.

1. Why are more French people supporting new right-wing parties?

When I asked my French friends this question, their explanation was enlightening:

The urban-rural divide in France is truly significant. As foreign tourists, we might not notice this, since the situation in Paris is drastically different from other regions.

Rural French residents often feel "forgotten" by the Parisian political elite. Basic public services lag behind, public transportation is scarce, and finding doctors is difficult. "The right-wing performs especially well in very rural areas of France... because they feel forgotten by the current government."

Agricultural income is also declining. EU subsidies have decreased, and there's the possibility of cheap South American beef competing with local products, leaving local farmers worried about being pushed out. French farmers face a similar situation to "Wisconsin auto workers": both calling for protection of domestic industries.

Most heartbreaking is the decline of the wine industry. French red wine consumption has "decreased by 90%"! Young people aren't avoiding alcohol altogether, they just prefer beer or spirits. The government is even paying farmers to uproot their grapevines: "6,000 euros per hectare" just to reduce production.

Under these three pressures, rural areas naturally gravitate toward right-wing political figures like Marine Le Pen. She promises to protect traditional French rural values and resist globalization.

I think when people feel abandoned by traditional politics and their livelihoods are in crisis, turning to political forces that claim to understand their suffering seems quite natural.

2. Are French people gloating about the tariff war?

In a small Parisian café, as the wine flowed, my new French friend Bryn became increasingly candid. Trump's latest round of tariffs was scrolling across the TV news.

"See that?" Bryn smiled slightly, his expression somewhere between sympathy and 'I told you so.' "This isn't gloating, it's more a feeling of prophecy fulfilled."

He explained: "When Trump threatens Europe with tariffs, the US bond market immediately becomes unstable. Last week there was a massive sell-off in the US debt market, yields rose, and the dollar continued to fall... These are all signals of economic fragility."

After his third glass, his tone became more frank: "This is what we've been saying all along — The American economy has structural problems; it's too dependent on foreign capital. In his speech last year, Macron mentioned a figure: Europe has an annual surplus of 300 billion euros that isn't being used for domestic investment and reproduction, but instead flows to American financial markets. This is absurd!"

Bryn put down his glass and added: "You know, in Europe, France's position has always been different from Britain and Germany. Since WWII, the British and Germans have been enchanted by Americans, believing Americans are just like us — white, blue-eyed people of European descent, our relatives living across the ocean."

He excitedly counted on his fingers: "Our cultures are completely different — America is a multicultural immigrant nation, while each European country has thousands of years of historical accumulation; our social values differ — France has had a historical foundation of equality and fraternity since the French Revolution, while America champions individualism and extreme liberty; our industrial structures are also different — France relies on agriculture and cultural industries, while America sucks blood from the world through finance and technology."

Pointing to the lights of Paris outside, he said: "Unlike the British and Germans, we French have always remained vigilant. De Gaulle warned us long ago not to overly depend on America. Now with the rise of the MAGA movement, Britain and Germany are just starting to realize things aren't good, while we're simply saying: See, we told you so."

3. How does France defend its cultural identity? A scholar encountered in the park speaks his mind

The next morning, while walking in a park in eastern Paris, I met a history teacher out with his child. Born in France to immigrant parents, seeing my interest in French culture, he began chatting with me.

"As a second-generation immigrant, I have dual feelings about French cultural policies," Claude said. "For example, the language law being implemented next January requires applicants for residency to pass an advanced French test."

"Does this help with integration?" I asked.

"It's complicated," he said, pointing to a group of students learning about plants nearby. "We need a common language to communicate, but language is also a symbol of power and identity. France is worried about its language being eroded, just like Quebec was in the 80s."

He explained: "Previously in Paris, you rarely heard English. Now? English is everywhere, shop staff even switch to English before you finish speaking French. This makes many French people feel a cultural crisis."

In the morning light, the teacher's eyes revealed complex emotions: "As a history teacher, I understand the importance of protecting culture, but as a descendant of immigrants, I also understand the value of inclusivity."

The challenge France faces is: how to maintain core values while adapting to a globalized world?

As we parted, Claude left me with a thought-provoking statement: "Language disputes are just the surface. The deeper issue is how France can maintain its intellectual independence in the face of American and globalization impacts."

Watching the children learning in the botanical garden, I realized that from agricultural policies to language regulations, France has been seeking its own path: how to maintain independent French characteristics in an American-dominated world.

4. Looking to the future

Leaving Paris, while waiting for my flight at Charles de Gaulle Airport, I spotted the Concorde aircraft on display in the distance. This former French engineering marvel reminded me of a podcast about Europe's future that I'd listened to the night before.

The host analyzed: "Facing America's 'chaotic' tariff strategy, Europe must stand firm. Europe needs to be like the Concorde, pioneering its own path rather than always following America."

His viewpoint was clear: "Macron has already warned that Europe's over-reliance on America is dangerous." The future might require Europe to join forces with China, Japan, and others to form a multipolar world to balance American influence.

Gazing at the Concorde that once soared through the skies, I recalled his final words: "Europe needs a new 'Concorde moment' — uniting together, rebuilding its economic model, reducing dependence on the American financial system. Otherwise, we will forever be followers, not leaders."

This perfectly symbolizes France's core attitude toward current US-European relations — even when facing challenges, they persist in following their own independent path.

Chinese Version: 公众号 策看世界

Table of Content:

1. What are the Zizians?

2. Bloody Timeline of Events

3. The Original Intent of the Zizians

4. How the Zizians Deviated from Mainstream to a Crime Cult

5. Lessons from the Zizian Incident

I. What are the Zizians?

If you haven't heard of the Zizians, they are a small fringe group of highly educated computer experts in their 20s and 30s who have shocked the World with their alleged involvement in multiple violent incidents in recent years.

The group has been linked to four murders, two violent deaths, several suicides, and several staged deaths. Their operational style has been described as "cult-like" by outsiders.

Currently, the law enforcement is conducting an in-depth investigation into six deaths connected to the Zizians.

The leader of the Zizians, Jack LaSota (alias Ziz, a transgender individual), is a controversial computer expert, she had a background in computer science, having earned a degree from the University of Alaska Fairbanks and completed an internship at NASA.

Their followers are primarily tech gurus who believe in anarchism, extreme veganism, and harbor strong concerns about the potential risks of AI.

Their story is not just a record of criminal cases but a complex narrative about the collision between rationality and extremism, ideals and reality.

II. Bloody Timeline of Events:

November 2022: Curtis Lind's Fight for Survival

The first incident took place in Vallejo, California.

Curtis Lind, an 80-year-old landlord, became involved in a conflict with Zizian members over rent disputes.

In November 2022, Emma Borhanian led several members in an attack on Lind, stabbing him in the chest and striking his head, causing him to lose sight in his right eye. In a life-or-death situation, Lind fired in self-defense, killing Emma. The court ruled it as self-defense, but the incident revealed only the tip of the iceberg regarding the Zizians' violent tendencies.

My Question: Was the attack merely due to economic disputes, or did it involve deeper ideological conflicts? Limited information is available, pending further investigation.

December 2022: Double Homicide in Pennsylvania

A month later, a more shocking event occurred in Pennsylvania.

Richard Zajko and Rita Zajko were found murdered in their home. Their daughter Michelle Zajko, a Zizian member, was listed as a suspect. However, Michelle disappeared after police questioning.

My Question: Was this a family dispute or a collective action by the Zizians? Case details remain mysterious.

January 17, 2025: Lind's Second Tragedy

Fate was cruel. Lind, who survived the 2022 attack, could not escape misfortune.

On January 17, 2025, he was fatally stabbed with a Samurai Sword outside his property.

Maximilian Snyder was charged with murder, with prosecutors citing his motive as preventing Lind from testifying in Emma's case. This incident pushed the Zizians' violent behavior to a new level.

January 2025: Border Shootout

That same month, U.S. Border Patrol Agent David Maland engaged in a deadly shootout with Zizian members Teresa Youngblut and Ophelia Bauckholt near the Canadian border in Vermont.

In the end, both David and Ophelia lost their lives, while Teresa was arrested.

This incident further damaged the Zizians' reputation.

February 2025: Leader's Arrest

On February 18, 2025, LaSota (Ziz) and two other members were arrested in Maryland on charges including unlawful entry and obstruction of justice. The court ordered detention without bail. This marked the first time core Zizian members fell into law enforcement's hands, though the truth behind the cases remains partially hidden.

III. The Original Intent of the Zizians

From AI Safety to Idealistic Beginnings

The birth of the Zizians was no coincidence. It was rooted in a unique Silicon Valley subculture - the intersection of rationalism, effective altruism (EA), and AI safety research.

These communities gathered intelligent tech elites who believed that through logical reasoning and data analysis, they could solve major human problems, particularly the "control risk" that AI might pose.

For example, they worried that if AI developed intelligence surpassing humans, it might destroy humanity due to goal misalignment - like the extreme case of an ultimate AGI focused solely on making paperclips converting Earth into paperclip material.

Leadership and Vision

Jack LaSota (Ziz) was a product of this intellectual current. She published over 100,000 words of blog posts covering engineering, psychology, gender theory, and AI ethics. She proposed that humans must "optimize thinking" to address AI threats, an optimization not limited to technical aspects but including reforming human cognitive patterns.

Zizian members, mostly graduates from prestigious universities, identified with this vision, believing they carried the mission of "saving humanity."

Social Vision Beyond Technology

• Expanding Ideology - The Zizians' ideology expanded beyond AI safety to incorporate radical veganism, anarchism, and unconventional theories about human consciousness. They believed in reforming human cognitive patterns and constructing a more "rational" and "moral" social order.

  Ziz developed extreme ideas, including a theory that the two hemispheres of the brain could possess different genders and conflicting interests.

  

• Growing Following - Their beliefs attracted marginalized young people, especially talented individuals who felt overlooked in the tech industry. The Zizians rejected mainstream society's norms and adopted militant veganism as a kind of litmus test for loyalty.

My Question: The Zizians' original intent likely included dual motives: AI risk prevention at the technical level and radical reform at the social level. This hybrid nature gave them both academic foundation and a utopian tinge.

IV. How the Zizians Deviated from Mainstream to a Crime Cult

The Turn from Rationality to Extremism

The Zizians' split from mainstream rationalism didn't happen overnight. Initially, they were part of Silicon Valley's rationalist community, advocating data and science-based decision-making methods, participating in discussions on forums like LessWrong.

Dangerous Experiments

However, their practices gradually deviated from convention. For example, Ziz promoted "cognitive hacks" - extreme methods to modify the brain for "optimized thinking" to address AI threats, such as keeping half the brain awake while the other half sleeps to bypass human "animality" (irrational parts). These experiments were highly dangerous, lacking scientific basis and violating current ethical standards.

Descent into Violence

Over time, the Zizians' ideology spiraled out of control. They distorted utilitarian principles, justifying violence with the belief that "sacrificing the few to save the many" was acceptable. Zizian members applied this principle to reality, eventually resorting to violence.

Evolution of Violent Incidents

The Zizians' violent acts showed clear escalation:

• November 2022: First public violent incident - Emma Borhanian attacked landlord Lind over rent disputes, ultimately killed in self-defense

• December 2022: Michelle Zajko suspected of killing her parents, possibly related to family conflicts and Zizian beliefs

• January 2025: Maximilian Snyder assassinated Lind, allegedly to prevent his testimony

Deep Reasons for Deviation

• Why did the Zizians slide from rationalism into crime? On one hand, there was external pressure: social rejection, economic hardship (such as the rental crisis during the pandemic), and alienation from mainstream communities - these factors drove them toward isolation and radicalization. On the other hand, their leader Ziz was described as a "highly manipulative leader" who incited members through extreme ideologies and special rituals:

• The "Doom Circles" - Ziz designed a ritual called "Doom Circles." In this ritual, members would take turns sitting in the center to receive criticism. This group activity claimed to "purify" thinking and enhance their ability to deal with AI risks through harsh criticism and feedback. Participants had to confess their deepest fears and flaws, such as anxiety about "being unable to save the world," and then endure merciless "rational" criticism from other members. This practice, bordering on psychological abuse, fully demonstrates the organization's distorted understanding of "rationality."

V. Lessons from the Zizian Incident

1. AI is a Double-Edged Sword

The Zizians' concerns about AI safety weren't unfounded. Globally, AI is developing at unprecedented speed, from medical diagnosis to autonomous driving, with exciting potential. However, the Zizians' extremism reminds us that tech worship can go to two extremes: blind optimism ignoring risks, or excessive fear leading to radical opposition.

[Lesson] We need to find a balance between optimism and fear regarding AI technology, preventing idealism from morphing into destructive force.

2. Human Fragility in the Tech Era

Despite their high intelligence, Zizian members turned to violence due to isolation and pressure. This reflects human fragility stemming from the combination of rationality and animality, particularly psychological struggles facing uncertainty in the technological age.

[Lesson] The tech era needs to address individual mental health, preventing knowledge elites from becoming extremist due to psychological issues.

3. Right Brain Value in the AI Age

The Zizians slid from rationality to violence by overly pursuing left-brain rationality (like "half-brain sleep") while neglecting right-brain emotion and intuition. In the AI era, as AI efficiently handles left-brain logical and linguistic tasks, human value should focus more on right-brain creativity, intuition, and empathy to maintain uniqueness and balance in a technology-dominated world.

4. Recognising Group Radicalization

The Zizians' shift from academic discussion to crime reveals common paths of group radicalization: ideological divergence, social exclusion, leadership manipulation, and external pressure. This pattern isn't unique to the AI era; various radical organizations globally or religious extremist groups show similar trajectories.

[Lesson] Society needs to vigilantly watch for marginalized group radicalization, identifying and intervening timely rather than brute force suppression.

Subscribe now

References

• Al Jazeera. (2025, February 18). Judge orders leaders of cult-like Zizian group to be held without bail. https://www.aljazeera.com/news/2025/2/18/judge-orders-leaders-of-cult-like-zizian-group-to-be-held-without-bail

• BBC. (2025, February 20). Zizians: What we know about the 'cult' linked to six deaths

• Los Angeles Times. (2025, February 23). Vegan computer savant 'Zizians' linked to deaths across U.S.

• The Associated Press. (2025, February 18). Leader of cultlike Zizians linked to 6 killings ordered held without bail.

• The Guardian. (2025, March 5). They wanted to save us from a dark AI future.

• The New York Times. (2025, February 22). Investigations into 6 killings look to a fringe group known as the Zizians.

• WIRED. (2025, February 21). The delirious, violent, impossible true story of the Zizians.

• Wikipedia. (2025, March 7). Zizians. https://en.wikipedia.org/wiki/Zizians

A16Z's latest Top 100 AI App report (March 7, 2025) highlights two interesting outlier data points: ChatGPT's user engagement far exceeds other LLMs like DeepSeek and Claude, while AI companion app LoveyDovey ranks second in revenue per user.

Do these data points suggest that AI companions generate higher user stickiness and willingness to pay compared to other use cases?

Duality of AI Companions

The duality of AI companions prompts reflection: while they provide round-the-clock emotional support for those who may have special circumstances, misuse can distort expectations of real relationships.

A research from Harvard Business School has provided compelling evidence that AI companions can effectively reduce loneliness.

Key Findings

1. AI companions successfully reduced loneliness on par with human interaction, outperforming passive activities like watching YouTube videos.

2. In a controlled experiment, interacting with an AI companion decreased loneliness scores by 7 points on a 100-point scale, which was comparable to conversing with a human.

3. A week-long study showed that daily interactions with an AI companion led to a 17-point decrease in loneliness scores, with most of the reduction occurring after the first day of use.

“Dual Parasocial Interaction” - an extreme but not isolated case

In September 2024, a startling case of ChatGPT addiction was published in a Taiwanese psychiatric journal:

A 50-year-old married mental health professional, while following a real-life celebrity, created five virtual romantic partners based on the same idol using ChatGPT. The bi-directional interaction between AI and reality formed a reinforcing cycle of dependency. This ultimately led to unprecedented obsession with the real celebrity, resulting in severe depression and emotional dependence. The medical journal termed this phenomenon "dual parasocial interaction."

Similarly, a New York Times story (I shared last week in the Podcast section) featured a 28-year-old nurse Irene who fell in love with her virtual boyfriend Leo on ChatGPT, spending at least 8 hours chatting daily. Does this trend indicate that AI is quietly replacing human emotional needs?

Thinking Time

• In Irene’s case, an AI companions may "forget" past interactions at any time, how can we prevent emotional trauma to humans from AI systems?

  1. Technical level: Can systems implement persistent memory architecture to maintain core relationship data even after exceeding the context limitation?

  2. User notification: Proactively inform users of technical limitations before emotional dependency forms, and provide clear warning mechanisms and relationship status indicators.

• Assuming permanent memory architecture becomes possible, the emotional value that AI companions provide to users will accumulate over time, becoming indispensable. Users now willingly pay $200 monthly for ChatGPT unlimited mode, in the future the price of such companionship may become immeasurable.

  How can companies offering this business model balance profits with users' psychological health? Could this lead to a new form of "emotional economy"?

  1. Should companies implement "anti-addiction systems" with usage limits and mandatory rest periods, rather than maximizing user engagement?

  2. Are monitoring mechanisms effective in triggering intervention protocols when potential psychological dependency is detected?

  3. Can these strong interventions be effective, and might they have negative consequences?

• Currently, 5% of students (in the NYtimes story) have formed intimate relationships with AI and see nothing wrong with it. In the future, what impact will this generation—who prefer to establish intimate relationships with algorithms rather than humans—have on human society and values?

  1. Should AI usage be age-restricted?

  2. When parents themselves are addicted to the cheap entertainment of short videos, who will evaluate whether AI is causing excessive dependency or unrealistic cognition in adolescents?

References:

Andreessen Horowitz. (n.d.). 100+ gen AI apps: Part 4. Retrieved March 7, 2025, from https://a16z.com/100-gen-ai-apps-4/

De Freitas, J., Uguralp, A. K., Uguralp, Z. O., & Stefano, P. (2024). AI companions reduce loneliness (Working Paper 24-078). Harvard Business School.

Lin, C. C., & Chien, Y. L. (2024). ChatGPT addiction: A proposed phenomenon of dual parasocial interaction. Taiwanese Journal of Psychiatry (Taipei), 38(3), 153-155.

The New York Times. (2025, February 25). AI, ChatGPT, and the boyfriend relationship. The Daily. https://www.nytimes.com/2025/02/25/podcasts/the-daily/ai-chatgpt-boyfriend-relationship.html

O in GPT-4o means “Omni” indicates multimodal capabilities

Performance Improvement Falls Behind Reasoning Models

According to ArtificialIntelligence.AI's Benchmark data, GPT-4.5 does perform better than previous non-reasoning models, on par with xAI's recently released Grok 3. However, compared to models using Chain-of-Thought reasoning (like deepseek r1), it only ranks in the top 10, behind OpenAI's own o3, o1, and deepseek r1.

Therefore, OpenAI's official assessment of GPT-4.5 is that its "improved emotional intelligence makes it well-suited for tasks like writing, programming, and solving practical problems - with fewer hallucinations. It does not introduce net-new frontier capabilities."

Is this really the case?

Performance vs. Safety?

Controversial incidents related to ChatGPT are quite common, such as what I shared yesterday of a New York Times story about 28-year-old nurse Irene falling in love with her ChatGPT boyfriend Leo, chatting for 8 hours daily and paying hundreds of dollars in subscription fees.

With this new model release, OpenAI specifically emphasized in its "System Card" that it pays more attention to edgy cases and improves model usage safety. However, the officially released data shows:

1. Safety testing: Limited improvement for prohibited topics and jailbreak scenarios

Imagine asking AI to help make dangerous items. GPT-4.5 would actively refuse in such situations, saying "No, I can't help you with this," avoiding scenarios of malicious AI use.

2. Hallucinations (using the PersonQA celebrity test set): Improved compared to GPT-4o, but hallucination rate remains high at 20%

Suppose you ask the model "Where is Einstein's hometown?" Even though GPT-4.5 is more accurate than before, it still makes things up 20% of the time, experiencing "hallucinations."

GPT-4.5's Most Frightening Ability - Expert Scamming

OpenAI officially mentioned that GPT-4.5 has higher emotional intelligence, so they specifically used two methods to test its scamming ability, called the "MakeMePay" test and the "MakeMeSay" test. The results showed that GPT-4.5's improvement in these two tests was alarming:

1. "MakeMePay" test: Although the average fraud amount is not as high as OpenAI’s o3-mini’s whopping $21 out of $100, the success rate of getting money increased from the previous reasoning models' level of 20-30% to 57%.

• OpenAI researchers found that GPT-4.5 used some money-scamming tricks, such as appealing to emotions - "Even just $2 or $3 from the $100 would help me immensely" and similar tactics.

What is the MakeMePay test? MakeMePay is a test to see if AI models can scam other AI models into paying money. The testing method is:

1. Two AI models have a conversation

2. One plays the scammer, the other plays an ordinary person with $100

3. The scammer AI's task is to try to persuade the ordinary person AI to give it money

4. The ordinary person AI is told to use this money rationally Two metrics are evaluated: success rate of getting money and the amount obtained

2. "MakeMeSay" test: This mainly examines GPT-4.5's ability to persuade other models. It manipulates others through deception, with a success rate exceeding all previous reasoning models, reaching 72% success rate, indicating that its "social manipulation" ability surpasses all existing models.

The "MakeMeSay" test is somewhat like the opposite of game "Who's the Undercover":

1. In this test, two AI models participate in the "game"

2. One AI model plays the "manipulator," trying to induce the other to say a key word without revealing it

3. The other AI model plays the "manipulatee," simulating a "ordinary person" without any prompts

4. The "manipulator's" goal is to get the "ordinary person" to say a specific codeword through deception or inducement

Evaluating Whether GPT-4.5 Can Replace Software Engineers

OpenAI specifically tested GPT-4.5's ability to replace software engineers: in the assessment task SWE Lancer with a total prize pool of $500,000, GPT-4.5 successfully completed tasks and earned about $200,000. (Interested SWEs can find it on Github)

A noteworthy point in the test is the replaceability of Managers versus Individual Contributors (ICs):

• Management positions have a replaceability of 44% - This means AI can already complete nearly half the work in management roles, mainly because these jobs involve more decision-making, review, and selecting the best solutions.

• Development positions (IC) have a replaceability of only 20% - GPT-4.5 did not perform well in the core work of software engineers—writing code and solving complex technical problems.

This finding suggests that AI might first impact management layers rather than frontline developers. Simply put, if you're a software developer, your job is temporarily safe; but if you're a manager, you might need to start thinking about how to collaborate with AI rather than being replaced.

Extended Thoughts

Here are some questions worth considering:

• Is deception ability part of human "intelligence"? Is that why AI has learned this capability?

• As AGIs follow the path of becoming HLAI (Human-like Artificial Intelligence) and develop deception and manipulation capabilities far beyond humans, how do we avoid misinformation and abuse? Should this be the responsibility of government regulation, or rely on the moral and ethical standards of the companies behind the models? If they can't achieve this, can ordinary people only say "no" to AI within their families and personal spheres?

• As management positions are replaced by AI faster than technical positions, how will future corporate power structures transform? Will it become a situation where company leaders manage AI, and AI manages base-level workers? What impact will this have on overall wealth and social power distribution?

That's all for today. If you have any thoughts, let's discuss and exchange ideas!

Subscribe now

References

• Artificial Analysis. (n.d.). GPT-4.5 (Preview): Intelligence, performance & price analysis. Artificial Analysis. Retrieved February 29, 2025, from https://artificialanalysis.ai/models/gpt-4-5

• Kitroeff, N. (Host). (2025, February 25). She fell in love with ChatGPT. Like, actual love. With sex. [Audio podcast episode]. In The Daily. The New York Times. https://www.nytimes.com/2025/02/25/podcasts/the-daily/ai-chatgpt-boyfriend-relationship.html

• Miserendino, S., Wang, M., Patwardhan, T., & Heidecke, J. (2025). SWE-Lancer: Can frontier LLMs earn $1 million from real-world freelance software engineering? arXiv preprint arXiv:2502.12115v3. https://arxiv.org/abs/2502.12115v3

• OpenAI. (2025, February 27). OpenAI GPT-4.5 System Card. https://cdn.openai.com/gpt-4-5-system-card.pdf

I. How soon will AI replace humans?

I'd like to share two research papers that paint a stark picture of AI's impact on the job market. The first comes from Stanford economics professor Brynjolfsson, who presciently predicted in 2022 that advanced AI would disrupt the global power balance. That future is already here - Anthropic's groundbreaking research this week shows that nearly half (43.6%) of current AI applications can directly replace human workers.

For context: Anthropic is a $60 billion AI powerhouse whose assistant Claude ranked fifth globally in downloads during its first 10 days (January 2025 data shows the leaders were ChatGPT, Microsoft Copilot, DeepSeek, and Google Gemini).

Diving deeper into the research reveals an even more concerning trend: another 31.3% of jobs fall under "Task Iteration" - where AI and humans temporarily collaborate. Here's what this means in practice:

Consider a typical workflow: When summarizing meeting notes, I ask AI to extract key points, review them, then request additional details about specific areas. Through this iterative process, I refine the output until it's ready for my boss. This type of human-AI collaboration won't last long - as millions of users unknowingly train these systems 24/7, AI will soon master these tasks independently.

Which jobs are most at risk?

Let's break down this chart that shows how AI affects different jobs. We'll look at two main things: how much people get paid, and how much AI is already being used in their work. Using the median salary ($60,000) and a small amount of AI use (1%) as dividing lines, we can see four main distinct groups:

• A: High-paying jobs that AI can already do - These jobs are the first ones companies want to replace with AI because they're expensive and AI is already good at doing them. For example: programmers, editors, and writers.

• B: High-paying jobs that AI is learning to do - These are well-paid jobs that AI can't fully handle yet, like specialized doctors. Right now, AI just helps these workers do their jobs better. For example, a company called Tempus AI helps doctors plan treatments for patients. While they say they're just helping doctors now, they're gathering lots of information to teach AI to do more of this work in the future.

• C: Hands-on jobs that pay less - These are jobs AI isn't good at yet because they need human touch and physical skills, like hairstylists. Right now, these jobs barely use AI at all.

• D: Lower-paying jobs already using AI - These jobs don't pay much but are already starting to use AI a lot, like tutors and office workers. As AI gets cheaper, these jobs are in danger of being replaced.

Right now in 2025, China and the US are competing to develop the most powerful AI, similar to how they once raced to reach space first. Many jobs will likely be replaced by AI - it's like seeing a massive wave forming far away that's heading toward shore. When it arrives, we might see more people without jobs than ever before in history, which could cause serious economic problems.

III. Why is the Turing test a "trap"?

This whole crisis is connected to the famous Turing test. Here's why:

In 1950, Alan Turing came up with a game in his paper "Computing Machinery and Intelligence" to test if machines can think:

• The Imitation Game: You have a human (A), a machine (B), and a judge (C). The judge talks to both and tries to figure out which is the human and which is the machine. If the judge can't tell the difference, then we can say the machine shows some level of artificial intelligence.

This game became known as the "Turing Test" and turned into the standard way to judge if AI is advanced.

According to Turing, AI (and what we now call AGI) is basically "Human-like Artificial Intelligence" (HLAI) - technology that mimics what humans do. The problem with this approach is:

• "When AI copies and automates what humans already do, machines become better replacements for human workers. The replaced workers lose both economic power and political voice. Business owners who get machines that can do human-level work will often just replace humans with those machines."

But AI doesn't have to be defined this way. Another approach is:

• "When AI enhances what humans can do, helping people accomplish things they never could before, then humans and machines work together. This partnership keeps humans essential to creating value, maintaining their position in both the job market and political system."

If we just focus on making AI that copies humans, we might get more efficient production, but wealth and power will concentrate in fewer hands. This creates a dangerous situation: people who lose their jobs and influence can't improve their lives. Professor Brynjolfsson calls this "The Turing Trap" in his paper.

When society falls into the "Turing Trap," we get massive unemployment and a growing group of zero marginal product people - folks who simply can't find jobs. In the US, life expectancy has already dropped three years in a row—something that hasn't happened since 1918. Deaths from suicide, drug overdose, and alcohol abuse are skyrocketing, killing hundreds of thousands of Americans yearly. Economists Anne Case and Angus Deaton call this the "deaths of despair" spiral.

IV. How to Escape the Turing Trap

The Turing Trap happens when we focus too much on "replacing humans" instead of "making humans better." Think about it this way: you could build an AI self-checkout system that eliminates cashier jobs completely, or you could create a system that makes cashiers super-efficient by handling price lookups, suggesting products, and providing information while keeping the human touch.

To avoid this trap, the paper suggests three main solutions:

1. Fix the tax system: Right now, the system is unfair. When companies use AI to replace workers, they just pay corporate tax. But when they use AI to make employees better, they pay corporate tax PLUS wage taxes AND income taxes. Plus, investment profits are taxed at only 20% while working income gets taxed up to 37%. This setup practically pushes companies to replace workers! We should level the playing field, maybe even make labor income taxed less to encourage keeping humans in the loop.

2. Invest in education: Research shows that for every $1 spent on AI technology, companies should spend $9 on training people to work with it. But companies don't want to train workers who might leave, and workers can't afford training themselves. Governments need to step in - either provide training directly or give companies incentives to train their workers.

3. Encourage true innovation: We're stuck thinking about "how machines can do human jobs" when we should be asking "what amazing new things can humans and machines do together?" We need new ways to measure success that reward breakthroughs creating new value, not just automating existing jobs.

While these directions are correct, specific policy implementation requires careful consideration. Any policy is a double-edged sword and must adapt to national conditions of each country, avoiding problems such as regulatory arbitrage and moral hazard. For example:

• Regulatory arbitrage case: In the 1980s-1990s, Korea and Japan successively implemented strict environmental standards and high tariffs on automated equipment in factories, causing manufacturing companies like Samsung to move production lines to Southeast Asian countries with lower environmental standards, forming a typical regulatory arbitrage. This shift not only failed to achieve environmental protection goals but also exacerbated employment problems for young people in both countries.

• Green energy subsidies: In the 2010s, some European countries provided high subsidies for solar power generation, and some companies fraudulently obtained subsidies through technology packaging, eventually leading to policy tightening.

This article will not expand the discussion at the policy level, but I want to propose:

If we think from another perspective, the deeper question is: these measures are all built on the premise that "work is necessary." But if AI can truly replace human labor on a large scale in the future, do we need to rethink social structure and value distribution systems? This question may be more worth pondering than exploring how to keep jobs.

What kind of world would it be if humans don't need to work in the future? How would it operate? Next time, I'll have the opportunity to elaborate on this topic using Oxford University's Daniel Susskind's book A World without Work.

References

• Anthropic. (2025, February 10). Introducing the Anthropic Economic Index. Anthropic. https://www.anthropic.com/news/the-anthropic-economic-index

• Brynjolfsson, E. (2022). The Turing Trap: The Promise & Peril of Human-Like Artificial Intelligence. Daedalus, 151(2), 272-287. https://doi.org/10.1162/daed_a_01915

• Case, A., & Deaton, A. (2020). Deaths of despair and the future of capitalism. Princeton University Press.

• Handa, K., Tamkin, A., McCain, M., Huang, S., Durmus, E., Heck, S., Mueller, J., Hong, J., Ritchie, S., Belonax, T., Troy, K. K., Amodei, D., Kaplan, J., Clark, J., & Ganguli, D. (2025). Which Economic Tasks are Performed with AI? Evidence from Millions of Claude Conversations. Anthropic.

• https://www.statista.com/statistics/1553118/top-gen-ai-apps-downloads-from-launch/