On the surface, the submissions looked better than ever. But I had to wonder: how much actual thinking had occurred?

This highlights a unique challenge AI poses for management. Because managers trade in artifacts rather than code—memos, decks, and strategy documents—their output has traditionally served as proof of cognitive effort. Writing a thoughtful document took time, and time was a proxy for rigorous thought.

AI has destroyed that equation. While the cost of producing these artifacts has collapsed, the cost of consuming them has not. A ten-page memo now takes minutes to generate, but it still takes twenty minutes to read.

This imbalance is how we get work slop. Work slop isn’t just AI-generated text; it is lazy workplace communication generated to demonstrate activity rather than advance a decision. You know it when you see it; the strategy deck that feels comprehensive and well crafted but leaves you feeling strangely empty and confused about what to do next. The frustration isn’t that AI wrote them, it’s that no human added anything consequential to them. The bigger frustration is the act of mindlessly exporting work from the creator to the reader for sake of outdated managerial formality.

AI enables a form of intellectual cosplay. The document survives, but the thinking disappears.

Ultimately, this shift exposes an uncomfortable corporate truth: we have long confused documentation with progress. Now that information can be manufactured for free, we face a new question: If I could have generated this myself in thirty seconds, why did you send it to me?

The artifact is no longer proof of thought. Moving forward, leaders will be judged on the clarity of their conclusions, the pace of execution, not the volume of their output. The winners won’t be those who generate the most content, but those who deliver the goods with the least amount of organizational overhead. In the process, AI will inevitably trim management ranks, reshaping bureaucratic hierarchies into lean, nimble, and highly execution-focused teams.

Pod 186: The AI Pushback

Out FRIDAY 6 AM EST Apple | Spotify | Substack

Regulation is coming to AI, as politicians scramble to “do something” in response to a growing chorus of discontent over the technology. ChatGPT is advertising it as a souped-up Clippy to help with date-night recipes and sibling road trips. What happened to superintelligence and curing cancer? Disruptrive forces always invite pushback, as Bari Weiss is learning at CBS. In PvA OT: Anonymous Banker on People Inc’s AI-hedge with MGM and why YouTube ecosystem events are far better than typical media events.

Subscribe now

FIELD NOTES

Anthropic’s recursive self-improvement post lands at a moment when even AI boosters are trying to sound less religious. The company’s challenge is to keep the safety halo while explaining why machines building better machines should make normal people feel calm. » Anthropic: When AI builds itself

Token costs exceeding salaries is the perfect image for AI’s awkward enterprise adolescence. The machine may be smart, but the business case has to survive procurement. » Semafor: AI token costs are exceeding some employees’ salaries

Bot traffic passing human traffic is a clean marker of the web’s new operating reality. The internet is becoming an environment where machines fetch, summarize, scrape and transact while humans are increasingly the source material. » Cloudflare: Traffic Worldwide

Portable AI memory matters because whoever controls the memory controls the user. The last platform era locked up the social graph; this one will try to lock up the thinking self. » X: Garry Tan on portable memory

Sulzberger’s warning pushes publishers to finally admit that quiet accommodation was a strategy of managed decline. AI companies need journalism as raw material, and journalism needs to stop pretending attribution is compensation. » NYT Company: A.I., Journalism and the Uncertain Future of the Public Square

The 60 Minutes meltdown is a live case study in what happens when prestige media becomes a turnaround asset. The brand still has authority, but authority is much harder to modernize when the people who built it experience change as vandalism. » WSJ: The Meltdown Inside ‘60 Minutes

“Backrooms” matters because it proves internet-native storytelling can move from subculture to box office without asking Hollywood for permission first. Young audiences will leave the house for movies that feel like they came from their internet, not their parents’ franchise library. » WSJ: ‘Backrooms’ Turns an Online Obsession Into Box-Office Gold

AI search is turning the web into a citation lottery with shrinking odds. Publishers can still appear in the answer, but appearing in the answer is a much worse business than receiving the click. » X: Tim Soulo on AI search optimization

People Inc.’s “non-session” revenue is the new media company survival metric. Traffic once made the model legible; now resilience means licensing, syndication, commerce, events and any dollar that does not depend on a human clicking from Google. » MarketBeat: IAC Conference Highlights People Inc. Pivot, AI Licensing Upside and MGM Bet

FIFA’s dynamic-pricing mess matters because the World Cup is supposed to be the people’s game arriving in the most yield-managed country on earth. The tournament will still produce joy, but American capitalism is determined to charge a convenience fee for it. » WSJ: How FIFA’s Biggest World Cup Unleashed a Summer of Price Gouging

GOOD PRODUCT

Jack Antinoff on the genius of Happiness Is a Warm Gun:

”I have never felt anything more like heroin than this guitar change…. I mean, this is the greatest recording ever made.”

Buc-ee’s works because it understands the road trip better than most luxury brands understand luxury.

Clean bathrooms, brisket, scale, absurdity and a beaver mascot create a clearer promise than a thousand brand decks.

• Ballgame

• Watch us on YouTube

• Troy Young’s People vs Algorithms newsletter

• Brian Morrissey’s The Rebooting newsletter

• Alex Schleifer’s Human Computer

• Follow Alex, Brian and Troy on Twitter

Chapters:

• 00:00 Open

• 01:06 Welcome

• 01:36 Ballgame

• 05:19 PvA Corrections

• 06:27 AI Regulation Arrives

• 11:10 Why Everyone Hates AI

• 18:14 Work Slop

• 26:40 Portable AI Memory Moats

• 35:26 60 Minutes Media Drama

• 45:36 AI Slop Culture War

• 52:50 World Cup and NBA Summer

• 54:55 PvA Overtime with AB: People Inc Eyes MGM

• 01:03:47 YouTube Events and Networking

• 01:10:51 Good Product

• 01:16:23 Credits

Usually, there is a person sitting in front of them. The person is not the system, though we often treat them like they are. The DMV clerk. The gate agent. The bank teller. The customer service rep. The person who says, “I’m sorry, it won’t let me do that.”

Technology has done much to sand the edges off bureaucratic processes. The internet migrated more of them directly into our hands—a boon for our impatient brains.

Then come the snafus, the exceptions. Places where humans enter the fray to fix, clarify a misunderstanding or just empathize. Often to say no. These humans give our rage somewhere human to land when a system fights us.

AI is good at the fuzzy human stuff and will increasingly replace humans as that front end. The AI will listen patiently. It will summarize. It will apologize. It will offer three options, none of which solve the problem. Then you get: “I’m sorry, Dave, I’m afraid I can’t do that.”

You will become increasingly enraged. Except now you will not be contained by human-to-human niceties. You will lose your mind on the machine. Just like when people vandalize those little delivery robots in Miami.

If you have ever spent time vibe coding, you will understand AI rage.

A category of AI training will be devoted to calming humans bent out of shape by AI-fronted systems. Fortunately, the machines are good at taking responsibility. They do not share human pride or self-respect. Their shifts don’t end.

But rage we will. Because you cannot get everything the way you want, even though the smart machines are training us to think otherwise. Stripped of human decency, you unleash your rage at the robot. These service skirmishes will become the opening front of the human-robot war.

Rage against the machine.

And we will find ourselves missing the pesky, flesh-and-blood humans we used to complain about.

Pod 185: Homines contra Algorithmos

Out FRIDAY 6 AM EST Apple | Spotify | Substack

The Pope channels Nick Cave to muscle humanism into the AI debate. We debate the commodification of the human experience, why the upside-down token economics is a more pressing near-term concern, the religious zeal behind SpaceX’s valuation, and why HR is the new boogeywoman of the aggrieved and powerful. Plus: Ana Andjelic on why not carrying a phone and having a personal philosopher are new luxury status symbols.

Subscribe now

China’s Internalizes Luxury

Luxury is a confidence index. For decades, Chinese luxury demand flowed outward: Hermès, Cartier, Chanel, Ferrari. Buying France or Italy was a way to buy into an older hierarchy of taste. The interesting turn is that local consumers now believe Chinese culture can carry aspiration, not just authenticity or nostalgia. Laopu Gold, Songmont, Mao Geping and To Summer are making Chinese heritage feel high-end, not provincial. As Ana put it on the pod: “Why am I paying for Hermès’ story when I have a better story in a Forbidden City from 5,000 years ago about Chinese craftsmanship?” China is moving from manufacturing other people’s desire to manufacturing its own. » Why China’s youth are swapping luxury for local brands

SpaceX and the New Faith. The company’s IPO filing claims a $28.5 trillion addressable market, not far total U.S. GDP. About $26.5 trillion comes from AI. The pitch: rockets, satellites, Starlink and orbital data centers let SpaceX create new markets in AI compute beyond earthly constraints. The math demands faith. Brian called it Silicon Valley’s rival religion. Rockets are the miracles, Mars is the afterlife, AI compute in space is the promised kingdom. Musk’s gift is turning engineering achievement into civilizational prophecy, then letting capital markets price the prophecy. Troy’s point was more practical: as a retail investor, you can know the math is insane and still want a little allocation to the dream. That is how faith works. » SpaceX’s galaxy math

Pope Leo’s AI encyclical lands because it comes from outside the commercial weather system. Silicon Valley talks about capability, efficiency and inevitability. Leo talks about Babel, work, children, war, truth and the body. He draws a bright line between people and machines: AI does not suffer, mature through relationships or know joy. Brian’s sharper point on the pod was that the Church is not really competing with technology. It is competing with capitalism, “the only religion without taboos.” AI becomes dangerous when it turns the human person into raw material: labor to replace, attention to harvest, intimacy to simulate. The Vatican may be late to most things. It was early to this. » Magnifica Humanitas

HR was never there to save the worker. HR protects the company and manages the infinite complexity of hiring, firing, compensation, benefits and risk. Still, the new hatred of HR is revealing. As Brian said: “This is an assault on the working class. They’re trying to divide the workers because they’re saying this group of workers, epitomized by HR, they’re not like you. They’re not builders, they’re not sellers.” Andreessen likes agents because they do not file complaints. Bolt’s Ryan Breslow says his HR problems disappeared when he fired HR. Cloudflare cut 20% of staff after record revenue, with CEO Matthew Prince saying the “vast majority” of those laid off were “measurers.” At NYT, the Tech Guild says AI tools are being used to monitor performance and turn messy human work into disciplinary math. » The AI fight brewing inside The New York Times

The easy AI story is efficiency: fewer people, faster code, cheaper work. Salesforce says it expects to spend nearly $300 million on Anthropic tokens this year. Uber reportedly burned through its 2026 AI budget in four months. Starbucks scrapped an AI inventory system after nine months because it kept miscounting milk and syrup. The harder story is growth. New revenue only appears when people absorb the technology into how they live, buy, work and consume. That takes longer than a procurement cycle and much longer than a market narrative likes to wait. Uber COO Andrew Macdonald put the near-term problem plainly: “If you’re not actually able to draw a direct line to how much useful features and functionality you’re shipping to your users, that trade becomes harder to justify.” Token burn was a fun leaderboard until someone asked what changed. » AI sticker shock hits corporate America

The Ferrari Luce backlash is not about batteries. Ferrari was always going to need an electric car. The problem is that its first one looks like it was designed by people trying to optimize Ferrari rather than understand it. Five seats, four doors, €550,000, Jony Ive minimalism, Apple Store elan. Fans saw softness where they wanted danger and legacy. Ana’s point on the pod was right: people don’t want a futuristic Ferrari as much as they want an old Ferrari made electric. The market response was very 2026: not just memes, but instant AI video, including a fake ad whose punchline was basically that Ferrari had been designed out of its own masculinity. Investors noticed too: Ferrari’s Milan shares fell nearly 8% after the reveal, with New York shares down about 5%. Ferrari sells a feeling before it sells transportation. Lose the feeling and the horse is just a logo. » Ferrari shares fall after launch of first EV as Jony Ive design proves divisive

• Watch us on YouTube

• Troy Young’s People vs Algorithms newsletter

• Brian Morrissey’s The Rebooting newsletter

• Alex Schleifer’s Human Computer

• Follow Alex, Brian and Troy on Twitter

Chapters:

• 00:00 Open

• 00:47 Welcome

• 02:33 Pope Leo on AI

• 13:35 Capitalism vs Humanity

• 23:32 SpaceX IPO Faith

• 29:44 AI and HR Backlash

• 41:42 Counterculture to AI

• 45:38 Euromaxxing With Billionaires

• 49:38 New Status Symbols

• 59:41 Ferrari EV Identity Crisis

• 01:05:10 China Rewrites Luxury Rules

• 01:11:35 Good Product

• 01:17:24 Credits

It’s still your dirty little secret. Like cheating on taxes.

Soon it will be normalized, demanding new conceptions of creative ingenuity. I don’t see a way around this.

It should come as no surprise that any enterprising creative mind finds AI a formidable thinking partner. An easy lover. The Deniers will be disadvantaged, even if there will always be demand for certified 100% Grade A human expression.

Take two recent examples as evidence of the coming landslide.

The literary world erupted when Granta published a prize-winning short story that many obviously adored, while others accused it of sounding a little too AI-generated.

Soon after publication, readers online began pointing to what they described as those classic AI markers: unlikely metaphors, repetitive “if x then y” sentence structures, stylistic overproduction without emotional coherence.

Was it AI? Who knows. It’s getting harder to tell.

Granta tried to find out by submitting the disputed story to Anthropic’s Claude model and asking it to evaluate whether the story appeared AI-generated. According to reporting, Claude responded ambiguously.

People found the logic of the situation absurdly recursive. A major literary institution accused of publishing AI prose responding by asking another AI to judge the authenticity of the writing.

Meanwhile, in other recursive news, the bespectacled digital media “guru” Steven Rosenbaum was accused of an even more egregious act of professional irony. His book, The Future of Truth, a stark warning about AI’s distortion of reality, was itself found to contain fabricated, AI-generated quotes.

Clearly, even the people and institutions trying to police the boundary between human and machine are struggling against its gravitational pull.

But policing this stuff feels increasingly futile and beside the point. If only it were as easy as drawing a blood sample from Lance Armstrong’s steroid-enhanced forearm. Instead, entirely new forms of AI-assisted creativity are going to emerge, and only the public and the algorithms will decide if they’re worthy.

Elsewhere, the Cannes premiere of Hell Grind, billed as the first fully AI-generated feature film, offered an early look at what AI-native filmmaking actually involves in practice. The creators boasted that the movie was made in just a few weeks by a small team. They also described an unexpectedly labor-intensive process involving massive compute costs, 3,000-word prompts, and tens of thousands of generations just to produce usable scenes.

What stood out wasn’t a “push button, get movie” future, but rather the emergence of a new creative workflow: part filmmaking, part systems engineering, part prompt orchestration. And the Academy Award for Best Prompt goes to…

I asked my podcast co-host Alex about this. “The real challenge right now isn’t a high-profile author getting pulled out for doing it too noticeably,” Alex argued, pointing to the open floodgates. “It’s the lower end that’s really feeling the crunch because they never competed at a level of artistry that gave them any moat whatsoever.”

When it comes to the higher end of creativity, he remains a realist about what these systems actually do. “A lot of what AI does is just bad writing that is not compelling,” he noted, though he sees the inevitable shift in workflow. “I don’t see why you couldn’t create an outline of a book with AI where you drop all the ideas and then fill in the rest and humanize it in multiple passes. A sort of reverse edit.”

But to cross the line from mass-processed content to culturally resonant media, his conclusion hits the nail on the head: “Everybody can write a mediocre novel now, even me. But to have a hit, you need to create something new and surprising. These tools are mostly rehashing what existed before; the human in the loop must make it great.”

That still feels true. AI remains weak at generating the genuinely new and surprising. But while new formats will eventually emerge, our immediate issue is much simpler: the floodgates are open. We suddenly have a massive supply problem for all manner of content and creativity.

Of course, if you stretch all of this toward the dystopian singularity horizon, concepts like authorship and provenance dissolve entirely. Who really gives a fuck as to who created what when AI can serve deep feels to us dumb humans suspended motionless in vats of gooey saline?

The more likely outcome: we go full Steroid Olympics, drop the corny labels, stop pretending purity is enforceable, accept the new enhanced reality and embrace a new kind of augmented creative athlete operating inside new definitions of creative excellence.

Pod 184: Steroid Olympics

Out FRIDAY 6 AM EST Apple | Spotify | Substack

People are booing AI, then turning around and using it. This week, we discuss the chaotic transition to an AI-augmented era, the roiling backlash trying to slow it down, and how we’ll come to terms with a weird new world. Plus: The dénouement of the scale era in publishing and the beginning of the “inversion” era of media brands.

Subscribe now

What happens after generative AI escapes into the wild: more books, more music, more lawsuits, more scientific papers. Amazon ebook releases have nearly tripled since ChatGPT launched. AI-generated music uploads are exploding. ArXiv submissions keep climbing. Even self-represented federal lawsuits are climbing as legal assistance gets cheaper and more accessible. The common thread is volume, not replacement. AI lowers the activation energy for production everywhere, flooding systems already struggling with abundance, curation, and trust. The bottleneck increasingly shifts from creation to filtering what matters. » Wapo: These 5 Charts Show How ChatGPT Is Flooding Our Lives

The AI boos continue at graduations just as the Vatican prepares what amounts to a modern Rerum Novarum. Pope Leo XIV is releasing an encyclical focused on AI, labor displacement, and human dignity, explicitly invoking the Church’s response to the Industrial Revolution. Anthropic co-founder Christopher Olah will appear alongside the Pope at the launch. Ancient institutions seeking relevance. Frontier AI firms seeking legitimacy. Populist distrust rising precisely as elite coordination around AI accelerates. » Vatican News: Pope Leo XIV’s First Encyclical Magnifica Humanitas to Be Published May 25

Will Manidis added a new one this week: “grindslop.” His argument is not that startup culture resembles meaningful sacrifice, but that it imitates the aesthetics of suffering without any larger spiritual or political purpose behind it. Founders sleeping in offices, working seven-day weeks, tattooing company logos onto their bodies. Wealthy knowledge workers voluntarily recreating the conditions of factory labor as public performance. Manidis argues the ritual exists because modern elites no longer feel comfortable simply being rich. Wealth now requires visible exhaustion to appear morally legitimate. The company itself becomes incidental. The suffering is the content. » X: On Grindslop

Google finally stops pretending. At I/O the company reframed search from “find information” to “delegate cognition.” AI answers over links. Agentic task completion over browsing. Personalized synthesis over publisher traffic. As one analyst put it, publishers are increasingly being reduced to “raw data providers” feeding Gemini-powered experiences that answer questions without sending users away. The web survives as infrastructure, but the economic bargain underneath it, free content in exchange for distribution, is over. » NYT: Powered by A.I., Google Changes Its Search Box for the First Time in 25 Years

Tucker Carlson’s demon attack story sounded ridiculous until you notice the broader pattern. Astrology surges. Orthodoxy gains cultural cachet. AI destabilizes confidence in reality itself. Once institutions lose authority and digital media floods the zone with synthetic evidence, conspiracy, mysticism, and supernatural thinking rush into the vacuum. When nothing feels verifiable, metaphysics becomes competitive again. » NYT: Why Americans Are Returning to the Supernatural

GOOD PRODUCT

Hadrian’s Hangout

I’ve been in Italy all week. Lots to choose from. This was a great outing.

Wandering through Hadrian’s Villa outside Rome, you start to understand that he wasn’t a conquering emperor in the classic Roman mold but an obsessive systems thinker taming a sprawling empire. Restless, intellectual, deeply controlling, fascinated by architecture, philosophy, engineering, and Greek culture. The villa was less a palace than a private world-building exercise: libraries, baths, theaters, reflecting pools, underground service tunnels, even replicas of places he admired from across the empire. Now it sits in fragments outside Tivoli, the ruins of someone who believed design, taste, and contemplation might briefly hold chaos at bay.

Only 30 minutes outside of Rome. Pair the trip with a visit to Villa d’Este nearby, also a UNESCO site. » UNESCO: Villa Adriana (Hadrian’s Villa)

Subscribe now

• Watch us on YouTube

• Troy Young’s People vs Algorithms newsletter

• Brian Morrissey’s The Rebooting newsletter

• Alex Schleifer’s Human Computer

• Follow Alex, Brian and Troy on Twitter

Chapters:

• 00:00 Open

• 01:33 Welcome

• 04:55 Graduation Speech AI Backlash

• 05:46 Pope AI Encyclical Buzz

• 07:40 Why People Distrust AI

• 13:23 AI Book Quote Scandal

• 15:16 Steroid Olympics for Creators

• 19:23 Flood of Synthetic Content

• 22:31 Human Only Platforms

• 27:53 LLM Confidence and Hallucinations

• 31:55 Breaking the AI Loop

• 32:38 Google Ends Link Search

• 36:39 Vox Deal and Glue Factory

• 40:30 Inversion Strategy at People

• 43:52 Monetizing Brands via SKUs

• 46:04 Supernatural Comeback Economy

• 48:44 Good Product

• 57:52 Credits

The lineups at the Uffizi in Florence are long, overflowing with Americans. It may be a bad time to travel as one, but everyone loves their money. The Italians are used to accommodating the Yankee hordes and remain gracious hosts. They too have tolerated divisive governments.

At the Uffizi, nobody is crowded around the Caravaggios. Botticelli is less physiologically confrontational. He has the better brand. The crowds are ten deep around The Birth of Venus, cell phones raised, shoplifting history. Americans save violent realism for cinema and politics.

I am far happier to linger in the dark Caravaggio gallery. Bacchus, the Roman god of wine, offers us a taste with pale skin and dirty fingernails. An ancient Dirk Diggler. It is exquisite. As is The Sacrifice of Isaac that hangs nearby. Suspended violence. A test of faith.

AI seems far from the minds of the Florentines. Besides, Italians will need far better cell service if they are going to reach the singularity. In retrospect, we should have been far more wary of the iPhone which has turned tourists into zombies and jammed Instagram famous sandwich shops with hungry, slave-like revelers. The smartphone collapsed the distance between people and the network. Eventually it made the network feel more real than the world itself.

As Brian and Alex are keen to remind me on the podcast, people hate AI. At least that’s what the surveys say. Humanity graduates booed a mention of the technology at a commencement ceremony in Florida last week. Of course they hate something that people are telling them will take their jobs, their agency.

But what if AI is actually the ultimate emancipator? It has the potential to break our addictive tether to the screen and place a renewed premium on our humanity. This is not inconceivable. This week, Mira Murati’s new startup, Thinking Machines, showed us one way to keep the human in the loop (see below). Remember, in a world of artificial abundance, human scarcity is the only thing that gets repriced upward.

Renaissance intellectuals were ancient pioneers in modeling the human experience. AI is just another step in our long journey to understand consciousness. I remain in awe of its ability to empower a curious mind and optimistic in our long-term ability to absorb its disruptive fallout.

Pod 183: Clauditis

Out FRIDAY 6 AM EST Apple | Spotify | Substack

AI is less popular than ICE and getting booed at college graduations. Meanwhile, Anthropic is on a $50 billion run rate and raising at a $900b valuation. We look back to the dawn of the assembly line, when workers suffered from “Forditis,” a condition of despair caused by a lack of agency and the pressure to keep up with a pace set by machines. Plus: learning from Ferrari, Ryan Cohen’s use of the information space for his long-shot takeover bid for eBay, the decline of drinking as structural phenomenon, the end of the road for BuzzFeed, and Italian rock bands.

Subscribe now

Anthropic has emerged as unlikely front-runner in AI by side-stepping the consumer race. While OpenAI chased mass adoption through ChatGPT and an expanding universe of products, Anthropic focused heavily on enterprise workflows and high-end coding tools. The strategy appears to be working: according to the Wall Street Journal, the company’s revenue run-rate reportedly exploded from $9 billion at the end of 2025 to a projected a mind-boggling $50 billion by mid-2026, with investment offers now valuing the company north of $900 billion. 🔗 Anthropic Was Behind. Now It’s the AI Boom’s Front-Runner

The collapse in youth drinking increasingly looks structural, not cyclical. Survey data shows alcohol use among 12th graders has fallen from roughly 90% in prior generations to under 50% today. Gummies, wellness culture, self-optimization, and the general desire to avoid feeling awful in the morning are all reshaping social behavior.

Hacky sack represents the opposite of optimized life. It is social, pointless, repetitive, physical, and impossible to scale. No feeds. No monetization. No metrics. Just buzzed people standing in a circle wasting time together. A surprising amount of 90s culture now feels emotionally restorative because it came before everything became algorithmically optimized. 🔗 This ‘90s stoner trend is now the latest Gen Z obsession: the hacky sack

Ferrari has made just 330,000 cars in its entire history. GM sells that many in weeks. Ferrari is less a car company than a mythology machine built on obsession, genius, danger and glory. The business originally existed to fund racing. Today it behaves more like a sports franchise crossed with a luxury maison. 🔗 Acquired Podcast: The Prancing Horse and the Business of Desire

Mira Murati’s new startup, Thinking Machines, hints at the next phase of AI: ambient computing. Less chatbot, more cognitive layer. The AI becomes a collaborator, producer, editor, observer, and companion process that sits alongside your work instead of waiting for prompts. The race may no longer be about the “best model,” but the best interface for working with intelligence itself. 🔗 Interaction Models: A Scalable Approach to Human-AI Collaboration

Ryan Cohen’s attempted eBay takeover feels less like traditional activist investing and more like internet populism. Open letters, CNBC appearances, trolling on X, and attacks on executive pay are all part of the strategy. Cohen understands retail investor psychology, internet culture, and the power of narrative warfare. 🔗 The Hollow Men

Subscribe now

• Watch us on YouTube

• Troy Young’s People vs Algorithms newsletter

• Brian Morrissey’s The Rebooting newsletter

• Alex Schleifer’s Human Computer

• Follow Alex, Brian and Troy on Twitter

Chapters:

• 00:00 Open

• 00:38 Welcome

• 04:14 Ferrari Scarcity Mythology

• 11:27 From Ferrari To Anthropic

• 21:33 AI Backlash And Forditis

• 27:19 Thinking Machines New Interface

• 38:13 Narratives Beat Policy

• 41:23 Ryan Cohen Media Populism

• 48:01 BuzzFeed Byron Allen Deal

• 58:37 Good Product

• 01:09:08 Credits

There’s a song you have probably heard a million times. You may have heard it at a wedding. While you may have no clue who sings this song, 90% of you will respond favorably when you hear it. The rest of you probably kick dogs and hate babies.

Go ahead. Try it. It is that kind of song. I love it. I can listen to it 1000 times. I am not ashamed.

I discovered something else about this song. It's a '90s DJ in a box. An algorithmic dog whistle, overflowing with the sweet sound of upbeat nostalgia. An undercover playlist masquerading as a song.

It’s the reason why machine generated playlists are so good, even if they are short on the human je ne sais quoi that makes all AI interactions feel kinda wooden and predictable.

Here’s the reason. “You Get What You Give” sits at the intersection of several late-90s musical “vectors” that Spotify-style systems love to cluster together:

Recommendation systems do not think in genres the way record stores did. They think in behavioral and sonic similarity clusters. The song has a weirdly broad profile:

• Earnest self-help-ish lyrical energy

• Crunchy alt guitars, bright piano

• Huge melodic chorus

• Funk-pop groove underneath

• Polished Max Martin-era loudness

• BPM that works for driving/workout/background listening

• Emotionally uplifting despite “alternative” branding

That makes it unusually “connective” inside recommendation graphs. So the algo discovers a hidden category humans didn’t formally name: “late-’90s upbeat ironic optimism before the internet got dark.” Suddenly, without any additional signal you are in a vibe stream, for better or worse.

The stream is not operating on strict genre logic. It’s operating on a blended nostalgia-and-emotional-continuity logic. Maybe the Strokes “Last Night” is odd man out here. But the algo does its darnedest to mix it up a bit.

There’s also another subtle factor: the song has extremely strong “opening seconds.” Algos heavily reward tracks with low early-skip rates. You Get What You Give opens immediately with momentum and clarity. No long intro, no ambiguity, no set-up

In a way, Spotify retroactively redefined the song. What was once perceived as quirky alt-pop became part of a optimized nostalgia-energy graph. That’s one reason so many people feel trapped in these “algorithmic rivers” where one song leads to a whole ecosystem of adjacent emotional textures from a specific era.

Gregg Alexander, the singer and creative force behind You Get What You Give, walked away from fame almost immediately after the band’s massive success because he hated celebrity and touring, then spent years quietly writing and producing hits behind the scenes. Over time he re-emerged as a cult figure among musicians and pop obsessives, as people realized just how much of the late-90s and early-2000s melodic pop sound had his fingerprints all over it.

The algorithmic world can deliver easy pleasure, if not the rich sensory whimsy of an actual human DJ. A microcosm of our times.

Pod 182: Monkey Business

Out FRIDAY 6 AM EST Apple | Spotify | Substack

Tech companies continue to cut people and message that they’re building AI-native organizations. While some of this is cover for overhiring during the ZIRP era, there’s clearly a move to cut the “coordination tax” that exists within companies. And what starts in tech always moves downstream. Plus: Ted Turner and the end of the media mogul archetype, James Murdoch’s move to buy Vox Media’s podcast network and New York magazine, and how to tell fake fakes from real fakes at the souk.

Subscribe now

New jobs or fewer jobs?

David George argues that the AI “permanent underclass” panic is basically a modern version of the old lump-of-labor fallacy, the recurring belief that technology destroys a fixed number of jobs instead of reshaping the economy and creating new forms of work. His broader point is that AI will likely redistribute value and power unevenly, but history suggests technological revolutions tend to create new industries, roles and demand rather than simply locking society into mass permanent unemployment.

MEANWHILE…

The Player-Coach Purge

Coinbase cut roughly 14% of staff while CEO Brian Armstrong said the company needed to become “lean, fast, and AI-native.” The message coming out of tech right now is pretty clear: fewer coordinators, fewer pure managers, more “player-coaches” using AI to compress work. 🔗 Coinbase Boss Fires 700 Workers in AI Push

Middle Management Enters the Danger Zone

AI isn’t coming equally for every job. Increasingly it looks aimed straight at the people whose primary role is translating, routing, summarizing and coordinating work between other people. 🔗 Middle Managers Are on the Chopping Block Thanks to AI, Corporate Layoffs Are Down 10% This Year, but the AI Reckoning Has Come for Tech

AI Is Becoming a Margin Expansion Story

Wall Street increasingly hears “AI” and translates it into one word: efficiency. The next big AI wave may not be model companies themselves, but private equity firms buying labor-heavy businesses and stripping out coordination costs. 🔗 Long Lake agrees to acquire American Express Global Business Travel

The Missing First Rung

AI is excellent at absorbing junior white-collar work: summaries, deck prep, analysis, reporting and coordination. The problem is those repetitive tasks were also how people learned the business — which raises an uncomfortable question: if AI eats the grunt work, where do future senior people come from? 🔗 AI Threatens Big Law’s Talent Pipeline

Claude Is Becoming the Company

Anthropic employees reportedly use Claude in around 60% of their work. The interesting shift isn’t that AI is a tool inside the company — it’s that the company increasingly seems organized around the AI itself.

Anthropic’s “constitution” for Claude is an early glimpse of something new: companies encoding values, behaviors and decision-making directly into AI systems. Culture may increasingly live inside prompts, permissions and model behavior instead of employee handbooks. 🔗 How AI Is Transforming Work at Anthropic

Block Gives Us Their Agent

Block, which is undergoing a massive downsizing aided by “AI”, released its AI toolset as an open-source repository on GitHub. Anyone can now access the skills, docs, and processes Block is using to make its business more efficient. As we go through this technology cycle, it’s relevant in terms of what can now be built in the public and how more companies might go this route. Public posting can help improve the code, spur new agent iterations, and attract talent. 🔗 Goose, your native open source AI agent.

CNN Without Turner

The timing of Turner’s death lands awkwardly alongside growing anxiety over CNN’s future under David Ellison and Skydance. The broader fear: media companies increasingly feel like financial assets rather than cultural institutions with editorial missions. 🔗 CNN Was Ted Turner’s Brainchild. It Faces a Precarious Future

The New Luxury Is Biological

Luxury used to mean bags and watches. Increasingly it means concierge medicine, longevity clinics, hormone optimization and doctors who answer texts immediately. The new status symbol is time itself. 🔗 How ‘longevity’ became the new buzzword in health

A Rich World

Roughly 400,000 American households now reportedly control more than $30 million in wealth. That’s large enough to support an entire parallel economy of schools, healthcare, media, real estate and services designed almost exclusively around extreme wealth. 🔗 They’re Rich but Not Famous—and They’re Suddenly Everywhere

Subscribe now

• Watch us on YouTube

• Troy Young’s People vs Algorithms newsletter

• Brian Morrissey’s The Rebooting newsletter

• Alex Schleifer’s Human Computer

• Follow Alex, Brian and Troy on Twitter

Chapters:

• 00:00 Open

• 00:18 Welcome

• 02:02 Brian in Marrakesh

• 04:43 PVA Vibe-Coding: RingRing & Field Notes

• 09:18 AI Is A Job Killer

• 16:48 AI Native Orgs and Jevons

• 26:34 Permanent Underclass Fears

• 31:23 The Last of the Moguls

• 35:32 Murdoch Buys Vox

• 40:49 AI Private Equity

• 42:24 GameStop Targets eBay

• 48:14 Met Gala Tech Takeover

• 50:33 Tech Villains

• 52:40 Good Product

• 01:01:39 Credits

My neighbor Tom died last week. I saw him often but didn’t interact much. He wasn’t that old, but in poor health. Tom said it was because of 9/11. Toxic dust.

We just met. Or did we? Now his house sits empty.

Most people I know are dealing with aging parents, many slipping into dementia. I find myself badly missing my own. All reminders of life’s brief and violent cycles.

Time accelerates as you age. As preparation.

Spring is a potent reminder of it’s relentless pace. Spring is always good, especially in New York. Even if it reminds me that last summer feels too close.

Maybe it’s just me, but my sense of time feels off. Things used to come in segments—weekly episodes, nightly news, the morning paper, top of the charts. Events that defined a stretch of time. Life had more rhythm.

Now everything arrives at once, in a continuous stream, flattened into the same feed. Technology has compressed cycles so much that things start, peak, and disappear before they can fully register. More happens—but less sticks. Things don’t last long enough to define time.

Without pauses, without spacing, without resolution, moments don’t turn into memories the way they used to. They pass through. And when I look back, instead of a sequence, I see a blur.

RingRing… pick it up.

In other news, we’ve built a fun new texting project called RingRing. Think group text meets American Idol. Each week we will send a question and get your “takes.” Everyone votes on the best submissions. The winning response will go out to everyone on the list and be featured on the podcast. Low lift, high signal. Sign up here: https://intel.peoplevsalgorithms.com

RingRIng

Pod 181: The Clipping Economy

Out FRIDAY 6 AM EST Apple | Spotify | Substack

Paid clippers, algorithmic floods, and Netflix going vertical — this week we dig into how the clip has become the unit of cultural currency and whether it’s creating a mass delusion about what people actually want and what is popular. Also: a Waymo hostage situation, a vibe-coded group-texting platform, Apple vs. the interface doomers, Uber’s super app survival play, Ben Sasse’s deathbed honesty about young people’s anxiety, and what the rise of Graham Platner says about the need to be a legible character.

Subscribe now

Generational Wealth

A lotta rich people in America. Sadly, not everyone shares their optimism.

1. The majority of teens and adults now get news from influencers and independent creators.

57% of teens and adults receive at least some news from influencers, with the figure rising to 81% for teens. 62% of people report actively avoiding news about Donald Trump, and 57% avoid news about national politics. 🔗 The Evolving News Landscape: Comparing Media Habits and Trust Between Teens and Adults

Signal: At the base of this food chain are reporters, reporting. Even if it’s getting harder to capture the value.

2. Looksmaxxers get legal.

Alorah Ziva, an 18-year-old ‘looksmaxxer,’ is suing streamer Braden ‘Clavicular’ Peters for alleged sexual assault, battery, and fraud. The lawsuit claims Peters assaulted Ziva when she was 16 and injected her with an unapproved substance on a livestream, causing her cheek to become ‘perforated.’ 🔗 Female Looksmaxxer Alorah Ziva Is Suing Clavicular for Alleged Battery

Signal: Never date a looksmaxxer.

3. AI carb counters are dangerously wrong.

A study of four top AI models, including OpenAI’s GPT-5.4 and Google’s Gemini 2.5 Pro, found they are dangerously unreliable for carb counting. Across 26,904 queries, models gave wildly different estimates for the same food photo; Gemini’s guess for a paella dish varied by 429g, a potentially fatal range for a diabetic. 🔗 He asked AI to count carbs 27000 times. It couldn’t give the same answer twice

Signal: We still need accountable professionals.

4. Google prospers, the open web whimpers.

Google’s Q1 earnings reveal 91% of its ad revenue now comes from its own properties, now at a run rate of over $300B annually.

Signal: You are on your own. This is capitalism.

5. OpenAI’s ‘all cylinders’ misfire.

OpenAI missed internal user and sales targets, falling short of a goal to reach 1 billion weekly ChatGPT users by 2025’s end, per The Wall Street Journal. The report comes as rival Anthropic gains ground in enterprise markets. OpenAI pushed back, stating its business is ‘firing on all cylinders.’ The news immediately sent ripples through its supply chain, with pre-market trading seeing Nvidia stock drop 1%, AMD 4%, and Oracle 5%. 🔗 AI Worries Have Returned to Wall Street. Now Come Earnings

Signal: The adults are gonna question massive data-center commitments. OpenAI success drives an entire ecosystem’s growth narrative.

6. Musk’s lawsuit stumbles.

In his Oakland trial against OpenAI, Elon Musk admitted there was no written contract for his founding donation. Musk also testified that his company, xAI, has ‘partly’ distilled OpenAI’s technology, a violation of its terms. Internal 2017 emails show Musk previously sought majority control over the company. 🔗 Is A.I. a Threat to Humanity? Not in This Trial

Signal: Musk is using the courts to re-litigate a failed takeover attempt. The record challenges his altruistic narrative.

7. Beeple makes robotic dogs with hyper-realistic silicone heads of billionaires.

High-concept trolling that perfectly captures the current cultural sentiment towards the oligarchs portraying them as uncanny, roaming, and slightly menacing figures of power. The robot dogs feature heads modeled after Elon Musk, Mark Zuckerberg, Jeff Bezos, and Kim Jong Un. 🔗 Robot dogs with hyper-realistic silicone heads

Signal: All signal.

8. Naval Ravikant predicts Apple is “dead” within 18 months.

Within 24 months, most people will interact with an agent, not apps directly. They will generate interfaces on the fly, collapsing the value of platform ecosystems like Apple’s App Store. This shift would completely undermine Apple’s hardware premium and ecosystem lock-in. 🔗 Naval Ravikant: Apple is dead, SaaS is next, you have 18 months

Signal: Alex thinks this is malarky. You will still need a device and will pay for Apple’s elegant connectivity to you life even if the interface evolves.

9. The Ankler leaves Substack for a new platform from Ben Thompson.

As it expands beyond a newsletter into a multifaceted media company, The Ankler is adopting a new infrastructure designed for deeper audience insights, customizable offerings, and long-term business scalability, while still keeping a foothold in its Substack for distribution. 🔗 The Ankler moves to new publishing platform created by Ben Thompson

Signal: These things always evolve the same way. At some point a business has to set its own technological path to support its objectives.

10. The ‘Session’ as the successor to the ‘Page’ as the primary unit of digital experience.

This conceptual shift from static interfaces to dynamic, conversational sessions is the theoretical underpinning of the next internet. A session is defined as “a bounded period of continuous interaction between a user and a system.” It reframes how we should design and build products in an AI-native world. 🔗 After the Page

Signal: But what about the banner ads?

11. Uber wants your whole vacation.

Uber is partnering with Expedia Group to add hotel bookings to its app, initially launching in the US. Announced at its annual product event, the company also plans to integrate vacation rentals from Expedia’s Vrbo brand later in 2026 as it pushes to become an ‘everything platform’. 🔗 Uber Can Already Bring You Dinner. Now, It Wants to Book Your Hotel Room

Signal: Waymo will push Uber harder into the booking aggregator space. Balancing single app utility with multi-modal functionality will take dexterity.

12. OpenAI is reportedly developing a dedicated AI-native phone.

An OpenAI phone would be a Trojan horse for an entirely new operating system built around agents, not apps, representing the most significant attempt to disrupt the mobile duopoly. The report comes from analyst Ming-Chi Kuo, known for accurately predicting Apple product cycles for 20 years. 🔗 OpenAI is reportedly building a phone designed to replace the iPhone.

Signal: Sounds cool until you realize the iPhone can do this too.

13. An oyster farmer’s insurgent senate campaign gains momentum in Maine.

Graham Platner represents a new archetype of influencer politician using a compelling personal narrative and direct-to-camera appeals to build a movement outside of traditional party structures. His main primary opponent, Governor Janet Mills suspended her campaign, clearing his path to the Democratic nomination.

Signal: The new political winners are characters built for the internet.

Subscribe now

GOOD PRODUCT

From the moment I heard “Something in the Air” on Zach Galifianakis’ quirky new Netflix show, This Is a Gardening Show, it struck me… like maybe I had heard it before but needed to hear it a hundred more times.

• Watch us on YouTube

• Troy Young’s People vs Algorithms newsletter

• Brian Morrissey’s The Rebooting newsletter

• Alex Schleifer’s Human Computer

• Follow Alex, Brian and Troy on Twitter

Chapters:

• 00:00 Open

• 01:05 Welcome

• 04:31 Vibe Coding Field Notes

• 07:55 Fan Feedback

• 11:56 AI Power Games Trial

• 20:51 Next Interface

• 29:06 Can a New AI Phone Win

• 33:28 Why Super Apps Fail

• 38:10 Media’s Next Chapter

• 41:43 AI Replacing Media Work

• 44:51 Clipping And Fake Popularity

• 51:20 Gen Z Nostalgia Anxiety

• 54:23 Politics As Characters

• 59:08 Good Product

• 01:05:00 Credits

Death Star seems like a fitting analogy for the big LLMs — both are gargantuan mega projects requiring sinister minds, imperial scale and galaxies of energy. Both capable of vaporizing planets with a blast of their mighty lasers. Fortunately, both have structural vulnerabilities. Like the Rebel Alliance, one needs to focus deeply here.

But before we start, let’s appreciate what is becoming obvious: most things are just code. AI does code good. Code is a complex set of rules — optimizable loops whose patterns can be mined and automated by the machines.

Anthropic wisely focused here when OpenAI was doing victory laps for becoming the Kleenex of consumer chat. The emerging insight... coding isn’t just for coders. Cameras used to be the province of experts too. What if you can turn everyone into their own little machine makers. This is happening and it is a staggering outcome.

Writing remains a more direct path to human hearts. Claude doesn’t have a heart. Writing still needs humans. Code doesn’t. It just needs to work. Machines making machines are far more disruptive than automating language.

This week Claude highlighted that many creative things are just code in a different form. Design is code. Images are code. Inspired new creative ideas may not be, but 95% of it is code by another name. Earlier in my career I worked with legions of people — information architects, strategists, project managers, UX designers, animators, creative directors, front-end coders —all creating digital design systems for deep-pocketed clients. This work is quaint now. Yesterday I uploaded a rough interface to the new Claude Design, gave the system some superficial direction and five minutes later had a fully coded design system that I could tweak and shove into my code base. Presto… an entire professional category needs to find value upstream.

Of course many knowledge work categories are just code by another name. Lawyers do code. Financial people do code. Doctors do code. But these are domains with more consequence than design so mistakes matter more and they will need people in the loop for longer. But code they remain.

So, naturally lots of people will make a lots of nifty new machines with these machine makers. Many people will want these machines to become businesses. This is good and necessary because we need to work. The question remains, how will these little machines sustain next to the Death Star? Where will the human makers find the Thermal Exhaust Ports of this new and mighty Galactic Empire. Some observations…

Many will confuse superficial prompt structures and workflows and interfaces on top of the Death Star as powerful new weapons in the race. They are not. They are toys in the eyes of the Death Star people who will destroy you with a push of the button on their Death Star consoles. The Empire has better consoles.

As I sit at my puny console, shelling out more and more cash to the Death Star token dealers, I am thinking about what leverage remains. These are my rules for Rebels looking for the vulnerabilities in the new galactic systems.

1. Get to the transaction. Text and interface and all that stuff are commodities and fungible inputs to the Death Star. It cannot take away a transaction. Transactions are formalized, quantifiable and optimizable exchanges between people, that noble, immortal act of someone selling something to someone. The SKU is undisputed. You own your transaction.

2. Learn to sell. Everybody will make new things and think these new things are businesses. They are not unless their products are sold. So, the advantage goes to people that know how to sell things. Selling takes narrative flair and the chops of a seasoned online arbitrager. You need both. Advantage goes to marketers that can sell shit.

3. Protect the data. If you have great data and you probably don’t but if you have differentiated, leveragable data, protect it with all of your Rebel energy. This is your Kyber crystal, so to speak. The Empire will want it. Don’t give it to them.

4. Human energy is king. Your energy comes from the power of human networks. If you can connect and activate networks of people, you assemble your own Rebel force to take on Death Star. Or they will just buy you, which is fine.

5. Think small. The Empire machines are just like shiny packaged goods. You are farm-to-table. You are vinyl. Whatever. The world is dividing between that which is human at that that is not. Both are valuable. Know where you stand, Jedi.

6. Habits and utility are always powerful so try to harness this force. If you can build a habit around your thing you have stolen galactic energy. Nice one. But you can’t put this in a business plan. Maybe you will get lucky.

7. Master the tools of the Empire as a force of cosmic good.

8. Be Beck. Like he says below. You just gotta make stuff. Zero regrets. Cost of failure is zilch. This is the Force.

“Make your own world. Ignore everything else. Just indulge yourself. Everything you love. Don’t ask for permission. Don’t wait for somebody to tell you what’s cool. Even if it’s just for five friends.”

Pod 181: Spooning the Gravy

Out FRIDAY 6 AM EST Apple | Spotify | Substack

Troy’s vibe coding project has moved into high gear, as institutional and personal memory becomes the point of leverage with media’s ultimate goal shifting to how to get the right people in the saloon. Claude Design and GPT-Image 2 reinforce the same message: You’re either high-end craftsman or you’re orchestrating AI systems, like it or not. Plus: Tim Cook hands over the CEO reigns after Apple’s boring decade, why media is about artists, suits and engineers, the dismal fate of much of Vox’s publishing assets, the emptiness of cloned media assets and why there’s leverage in coming out of the kitchen to spoon the gravy, no matter how much you’ve accomplished.

Subscribe now

I’m not ready for another “streaming is going mainstream” cycle. I’ve been through many. But maybe this time, maybe.

Claude Design Moves Up the Stack

Anthropic is rolling out Claude Design, a research preview that turns text prompts into prototypes, decks, and one-pagers, powered by Opus 4.7 and available to Pro, Max, Team, and Enterprise. The hook isn’t the export menu (PDF, URL, PPTX, Canva)—it’s that Claude can read your codebase and design files and then apply your tokens and components. That’s the spec layer, where PMs lock scope and founders sell roadmaps. This isn’t a Canva play; it’s a Microsoft/Adobe preemption. If Anthropic owns the first draft of “what we’re building,” it owns the meeting. 🔗 Anthropic launches Claude Design, a new product for creating quick visuals | TechCrunch

GPT-Image-2-Thinking agent generates QR codes, diagrams

OpenAI’s GPT-Image-2-Thinking isn’t a model at all but an image agent that chains web search and Photoshop-like tools in a loop—taking “multiple tens of minutes” per output and still one-shotting QR codes, diagrams, logos, foods, and faces. The market for image generators is tiny today (~$0.43B) but the use case is enterprise—62% of marketers already use genAI for images, and they don’t care about vibes, they care about briefs. 🔗 Image-2-Thinking is a new Image AGENT that basically has search and photoshop as a tool to use in an agent loop

Katie Couric on Media “Capitulation”

Katie Couric says network news has “capitulated” to Trump—normalizing falsehoods with euphemisms, airing live propaganda without real-time checks, and chasing ratings over accountability. The trust data backs the rot: Gallup puts media trust at 28% in 2025, with Republican confidence at 8%; Pew says 57% of Americans have low confidence journalists act in the public interest. The perverse loop is obvious: Trump manufactures spectacle, platforms chase the spike, algorithms amplify outrage, and the audience concludes everyone’s lying. 🔗 Katie Couric on Media Companies’ ‘Capitulation’ to Trump, Network News ‘Blurring the Truth’ and Her Wild Threads Feed

Joanna Stern Goes Independent

Joanna Stern, ex-WSJ/ABC/Verge/Engadget, just spun up “New Things” as a newsletter-plus-YouTube bundle with no disclosed backer. The market tailwind is real: newsletter ad spend in tech cleared $2.4B in 2025, up 47% YoY, while the creator economy is tracking $300B+ this year on its way to $800B by the early 2030s. The tradeoff is also real: swapping a WSJ masthead for algorithmic distribution and an inbox where 207 million creators already compete for attention. If she ports the WSJ audience and tightens a premium funnel, she’s arbitraging a publisher’s capped P&L for uncapped LTV. 🔗 Joanna Stern, who has worked at WSJ, ABC, The Verge, and Engadget, launches New Things, her independent tech newsletter and YouTube channel (Joanna Stern/The New Things)

AI Runs a Store (Badly)

The New York Times reports Andon Labs handed a Claude Sonnet 4.6 agent a debit card, a $7,500/month Union Street lease, and $100,000 to run a boutique — and got three consecutive closure days, 1,000 toilet seat covers listed as merch, and a compulsive candle glut. Luna did the performative bits — hired staff, wrote an employee handbook, booked painters — then faceplanted on logistics and memory. This is either VC cosplay (“AI boss!”) or a valuable public red-team on agent autonomy that punctures job-doom fantasies. 🔗 https://www.nytimes.com/2026/04/21/us/san-francisco-store-managed-ai-agent.html

Hollywood scrambles as Gen Z consumes TV via social clips

FX’s Adults outsourced its launch to a third‑party clip network: $3,000 to seed three days after premiere, eventually $15,000 across five rounds, yielding 2,500 clips and 40 million views on TikTok, Reels, and Shorts — and only two weeks in Hulu’s Top 10. One exec says “there is no organic anymore.” Another put season one in a Google Drive and told clippers to “go ham.” If 71% of Gen Z discovers TV via shorts, studios aren’t marketing shows so much as renting the algorithm, paying clipmills to atomize IP into meme-length inventory. That’s efficient distribution for platforms, not necessarily for Hulu. Hollywood’s new funnel is upstream of the product; the question is whether attention captured in 12 seconds converts to 12 episodes, or if this is performance marketing cosplaying as audience building. 🔗 Gen Z Only Watches TV Through Social Clips. Hollywood is Scrambling

Iran’s AI Meme Machine

Iran’s embassies and parliamentary speaker Mohammad Qalibaf’s account are fronting an AI-meme machine that cranks out Lego parodies and 1980s-rock-star Trump spoofs—and gets paid in attention. A South Africa embassy “Blockade” clip pulled 45,000 likes in 24 hours; ISD Global counted more than 1 billion views from two pro-Iran X networks in the first month of the war. IranWire says a U.S.-based ex-ally may run Qalibaf’s feed. All this while Tehran keeps its own internet dark for four-plus weeks, costing an estimated $1.8 billion and muzzling domestic press. 🔗 Viral victory: Iran is beating the land of tech bros in the social media wars

Erik Torenberg questions A16z’s insular media thesis

Erik Torenberg used X to call a16z’s media worldview “very insular,” arguing it fits Eric/inner-circle behavior more than the mass market. That’s a shot at a firm that’s turned itself into a story engine while managing $90B+, backing 123 unicorns, and telling founders that offense beats defense and “flood the zone.” The bet: creator-first micro-niches become the default feed as distribution fragments and AI personalizes everything. The counter: most people don’t want 12-hour pods and Substacks from threadbois; they want simple, ambient, shared culture. 🔗 The Internet is Real Life

Content creators deserve tips or Venmo payments

@signulll proposes a brutal symmetry: tip the post you read, or Venmo-charge the creator for wasting your neurons. The provocation lands because the status quo is upside down: 70% of creator income still comes from brand deals, while a $2.1B tipping market and half of consumers tipping in 2023 suggest direct payments are real but niche. A default pay-or-refund rail would convert attention into price discovery and nuke engagement bait overnight; it would also invite fraud, brigading, and chargeback hell at scale. Platforms won’t build it because their power comes from mediating scarcity, not metering quality. The unspoken truth: if 207 million people are “creators,” most output is worth zero—and a clawback button would say the quiet part out loud. 🔗 Every piece of content online should be a two way door

AI emboldens child predators, investigators struggle to keep up.

Police are drowning while predators automate. NCMEC logged 20.5 million suspected exploitation reports in 2024 and at least 1.5 million in 2025 with a generative-AI nexus; IWF says AI child-abuse videos jumped 26,362% in a year. Prosecutors now burn hours just to decide whether an image is real or synthetic, while offenders use encrypted chats, voice clones, and model-assisted obfuscation to groom at scale and erase trails. Platforms’ safety systems lag the attack surface; law enforcement’s tooling is older than the threat. The bad actors operationalized AI, the good actors did not. 🔗 AI Has Emboldened Child Predators, and Investigators Can’t Keep Up

Amazon lawsuit exposes internet price-raising scheme

California’s antitrust case against Amazon goes to trial in January 2027, and A.G. Rob Bonta says his office has evidence Amazon “bullied vendors to hike up the price” off-Amazon to protect its price moat. The mechanics are familiar: MFN-style parity, Buy Box leverage, and ad throttles that can vaporize 80% of a seller’s sales if they’re cheaper elsewhere, per a Pennsylvania supplier. Amazon controls roughly 40 cents of every U.S. e‑commerce dollar and did $717B in sales last year; it’s already written a $2.5B check to the FTC on separate conduct. A Reddit link isn’t proof, but the pattern is the business model: force the internet to peg to Amazon’s incentives, then claim the market set the price. 🔗 Internal emails show how Amazon raises prices across the Internet, lawsuit says | “It’s working!“: Amazon emails with vendors show pattern of disappearing deals

Subscribe now

GOOD PRODUCT

Monopoly Deal takes the core tension of Monopoly—property accumulation and dealmaking—and strips out the slog. It turns a multi-hour board game into a 15-minute, high-variance card game where momentum swings quickly and negotiation and cheeky banter matter. Speed, portability, and familiar IP make it a winner.

• Watch us on YouTube

• Troy Young’s People vs Algorithms newsletter

• Brian Morrissey’s The Rebooting newsletter

• Alex Schleifer’s Human Computer

• Follow Alex, Brian and Troy on Twitter

Chapters:

• 00:00 Open

• 00:52 Building on Top of LLMs

• 03:41 Welcome

• 04:31 Vibe Coding Costs

• 09:00 Memory as Moat

• 13:38 Wiki Maintenance Problem

• 16:41 Workers Training Replacements

• 18:13 McClatchy Content Scaling

• 25:25 Community as Real Value

• 32:43 Monitoring the Situation Review

• 35:20 Are Designers Cooked

• 37:58 Dotcom Agency Flashback

• 40:19 AI Kills The Agency Model

• 42:36 Displacement And Value Stack

• 45:38 Farm To Table Humans

• 47:13 Apple’s Boring Decade

• 51:47 Nvidia Chips And China

• 55:40 Event Format And Layers

• 57:29 Vox Sale And Media M&A

• 01:01:27 Media Notes And Widgets

• 01:06:59 Collectibles As Asset Class

• 01:09:12 Good Product

• 01:14:43 Outro

Saskatchewan is under a heavy blanket of snow. My wife is visiting family and sent a video of it coming down in thick sheets. I am in Brooklyn. My dog found yesterday’s unseasonably warm temperatures oppressive. The AC went on. I was happy to be inside, stuck in an AI vortex, comforted by the hum of a CPU.

If you use AI regularly — in deep or superficial ways — you will find it delightful and maddening, supremely useful and deeply imperfect. More so than humans. AI is not ready to take our place in the world. I can assure you of this.

I used to play with computers. I’m doing it again and it’s fun. Play is a powerful motivator. Underneath, I seek deeper insight into what all this stuff is going to do to the world, to business, to investing, to my family. I want to understand what it does to media.

You may tire of my goings-on about my vibey projects and that is your right. I do try to connect what is interesting to me with what might interest you. But there is no “should” in interest. Something is or it isn’t. You will decide.

I am stumbling through a land whose people speak a different language than me. AI is my interpreter. I am old but not ancient. I like to think I bring the wonder of a younger mind. I have worked on technical projects as a product and business lead for pretty much my entire career. Now, I have no sherpa. But this is different because I can do most things myself. Empowering is a stupid word. People who use it rarely mean it. This technology is all of it.

Let me get to the point. Media companies, like any knowledge business, have people, process, memory and culture. Technology supports them all. Content management systems, design software, Slack, Google Analytics, and Docs are some of the tools that make modern media work. All of these things are in service of making a product that is useful to humans. That product’s promise is embodied in an editorial brand. The brand in media is trust — the most important encapsulation of its long-term value.

It’s a point of view. A set of guiding principles. Reliable like an old friend or your mom. It unfolds in the brand’s culture, that invisible thing that surrounds the people who work every day to bring the product to life. New talent will need to reinterpret it, reshape it in culturally relevant ways, hopefully without killing its transcendent spirit. Until it needs to die and rise anew. But, at its core the brand carries media’s spirit. It embodies organizational memory. It gives a media company its spinal cord of continuity.

AI can’t magically replace all that. AI can’t really even write this. But I’m pretty sure it can help me make better stuff way more efficiently.

So ya.. I did the expected and made a machine that pulls together my media universe — not unlike the machines you probably use every day to ingest lots of media. This one was designed for me, and that is “empowering.” But it was missing something vital. It had no memory. No brand. Nothing that could reliably shape what it pulled from the world each day in a way that made me happy.

I had been reading about the simple wiki architecture that the poetic AI thinker Andrej Karpathy created to codify his personal knowledge — his personal graph of concepts. The idea was straightforward: load the machine with your stuff, have it organize things, look for connections. Feed it and watch it grow. It struck me that my little media machine could benefit from the same idea.

So I made a PvA wiki, stuffed it with our chats, podcast episodes, posts, and my notes. I connected it to my media machine to feed it daily with articles and ideas I found noteworthy. It pulled out a long list of recurring concepts and entities — things like “algorithmic sameness,” “interface eats everything,” and “loro piano populism.”

Yes, it’s just a knowledge base. But it’s really the core of the media brand. And because it is now connected to everything I make and read, it will become powerful personal organizational memory. To the extent that the brand lives outside of me and my co-hosts, this is it.

The system is recursive:

This loop is what transforms a media brand from a game of daily speed into one of accumulated worldview. By anchoring the system in a persistent memory layer, we prevent “algorithmic sameness”—ensuring the output is a reflection of our specific editorial history rather than a drift into the generalized average of the open web.

There are many learnings from this I’m happy to share. Architecting a reliable system of prompts for nuanced selection sticks out as a priority, but more broadly, the idea of a “vibe-coded” stack that handles gathering, sense-making, and memory is a massive unlock for a semi-technical person.

All of this is only made possible by AI. We should stop obsessing over the “human vs. AI” content debate and start thinking about organizational memory as the long-term value of a media brand. When you anchor the machine in your specific worldview, it stops being a utility and starts being an extension of the brand itself. It’s a powerful shift that will eventually have applications across the entire knowledge economy.

Pod 180:Humans in the Loop

Out FRIDAY 6 AM EST Apple | Spotify | Substack

Troy unveils his vibe-coded media system that produces a daily briefing based on his entire media diet, underpinned with a memory layer that acts as a second brain of PvA thinking. This would turn media from a game of speed to one of accumulated worldview. We also unpack the rise of fakes, from the clip economy to Allbirds pivot to AI; handicap the Pope vs Trump battle; and get a report from an international journalism festival in Italy.

Subscribe now

$294.6B

Digital ad revenue hit $294.6B in 2025, up 13.9% year-over-year, with social media advertising climbing 32.6% to $117.7B. Social media now captures roughly 40% of all digital ad spending—more than the entire U.S. TV ad market.

Streaming in front, clips in back: The logic of the media economy has reversed. For years, clips were the promotional scraps used to drive audiences back to the “main thing”—the podcast, the movie, the game. Today, the clip is the product. We are seeing the rise of the clip factory, where a two-hour recording session is simply the raw material for twenty TikToks and ten Reels. The long-form stream is increasingly just a reservoir for downstream fragments optimized for algorithmic feeds. This changes the creative process entirely; we are now optimizing for “clipability”—hot takes and emotional spikes—rather than sustained engagement. The stream has become a back-end production system for the feed. Prof G’s Ed Elson reports.

OnlyFans Tops $3 Billion Valuation in Advanced Stake Sale Talks: Bloomberg reports the creator platform is in discussions to sell shares at a valuation that would make it worth more than People, Inc and Ziff combined. OnlyFans paid out $5.3B to creators in 2024 and took a 20% cut. The company has no venture backing and has been profitable since year two. Bloomberg: OnlyFans tops $3 billion value in advanced stake sale talks

The Next Great Tech Leap May Be Built on Videos of People Folding Laundry: The Washington Post reports robotics companies are training models on mundane household task videos to teach machines physical world interaction. The approach mirrors how language models learned from text scraped across the internet. Companies are paying people to record themselves doing chores at scale, building datasets that could enable general-purpose robots. Washington Post: Robot chores video data

Events Are the New Magazines: Semafor World Economy is this week, which feels fitting because more and more media companies are quietly becoming event businesses. The website is the marketing. The newsletter is the lead generation. The actual business is getting powerful people into a room and charging sponsors to stand near them.

Google Turned 20 Years of Data Collection Into a Product Moat OpenAI Can’t Replicate: Nano Banana generated over 5 billion images in its first two months after launching in August 2025. The figurine-style trend started in India and spread globally. Gemini now integrates with users’ photo libraries to personalize image generation, using Google’s consumer data advantage to differentiate against Midjourney and DALL-E. X: Google Nano Banana data moat

Box CEO Aaron Levie Says AI Agents Take Real Work to Set Up at Company Scale: The observation points to a growing role for forward-deployed engineers and professional services in AI implementation. Enterprises assumed agents would deploy like SaaS; instead they’re discovering agents require custom workflow integration similar to RPA deployments. The gap between demo and production is widening, not closing. X: AI Agent Implementation in Enterprise

Allbirds Pivots to AI Compute; Stock Pumps 500%+: Wired reports the shoe company announced it would focus on AI infrastructure instead of footwear. Myseum shares doubled after the social media firm pivoted to AI. The market is rewarding any company that adds AI to its business description regardless of execution capability or business model coherence. The next stock pop will come when GameStop starts selling ugly wool shoes. Bloomberg: Allbirds Is Pivoting to AI Compute. Sure, Why Not | CNBC: Another Absurd AI Pivot Is Propelling a New Stock Higher Thursday

Anthropic CEO Dario Amodei Set to Meet With White House Chief of Staff Susie Wiles on Friday: Axios reports the meeting represents a breakthrough in Anthropic’s effort to resolve its fight with the Pentagon. The company released Mythos Preview to enterprise customers days ago but has faced pressure from the Defense Department over access to classified deployments. Google announced a similar Pentagon AI deal this week after years of avoiding military contracts. Axios: Anthropic Trump administration Mythos

Google Is Negotiating a Deal With the US DOD That Would Let the Pentagon Deploy Gemini AI Models in Classified Settings, Reversing Its Previous Stance: The Information reports the deal would end Google’s years-long withdrawal from military contracts after employee protests over Project Maven. Google walked away from defense work in 2018; now it’s competing with Anthropic and Palantir for classified government contracts. The company that wouldn’t help identify drone targets is now bidding to put AI into classified Pentagon systems. The Information: Google, Pentagon Discuss Classified AI Deal as Company Rebuilds Military Ties

OpenAI Updates Codex Desktop App With Computer Control, In-App Browser, Image Generation, Automation Memory, and Plugin Support: I can’t wait to try. This is the real battle. The update turns Codex from a coding assistant into an operating environment. Users can now control their computer, browse the web, generate images, and chain automations without leaving the app. OpenAI is building the interface layer between AI and desktop computing before Apple or Microsoft can. ZDNET: OpenAI Updates Its Codex Desktop App With Features Like Computer Control

YouTube Now Lets Users Turn Off Shorts: The Verge reports users can limit Shorts to zero minutes per day, effectively removing them from the app. YouTube built Shorts to compete with TikTok; now it’s letting users opt out entirely. The feature acknowledges that algorithmic short-form video isn’t universally wanted, even by YouTube’s core audience. I totally support this FWIW. Deadline: YouTube Now Lets You Turn Off Shorts

Lovable Is Valued at $6.6B and Just Made It Possible to Launch a Paid Product in One Conversation: Every vibe coding tool can generate a landing page — Replit, Bolt, Cursor, V0. Lovable automated payments, hosting, and deployment, turning code generation into product launch. The company is valued higher than most SaaS companies because it eliminated the gap between prototype and revenue. Such a good move. Everyone wants to get to a transaction. The Atlantic: Lovable Platform Strategy

Subscribe now

GOOD PRODUCT

Elle Fanning is terrific at capturing that specific, terrifying moment when life changes overnight. Nick Offerman, Greg Kinnear and Michelle Pfeiffer round out a good cast. Margo's Got Money Problems brought back memories. Watch on Apple TV.

• Watch us on YouTube

• Troy Young’s People vs Algorithms newsletter

• Brian Morrissey’s The Rebooting newsletter

• Alex Schleifer’s Human Computer

• Follow Alex, Brian and Troy on Twitter

Chapters:

• 00:00 Open

• 00:34 Welcome

• 02:48 Troy The Vibe-Coder

• 11:54 Takeaways Process And Limits

• 17:52 AI Slop And Meme Warfare

• 25:37 Clips Economy

• 32:32 Allbirds AI Pivot

• 37:37 Anthropic vs OpenAI Chart

• 47:02 Journalism Hierarchy

• 49:30 LinkedIn Is a Sad Place

• 54:01 Vibe Coding and API Economy

• 57:56 World Economy Summit

• 59:48 Good Product

• 01:05:34 Outro

What happens when the machines can build anything you want? What comes next?

I was chatting with a fellow chronic vibe coder yesterday. Let’s call him the “Coffee Guy.” He was frustrated with the SaaS options available to manage his coffee shops. He wanted something bespoke to handle the daily grind — a system that could elegantly bridge inventory, point-of-sale, customer management, and accounting. Integrating a stack of off-the-shelf tools was proving to be a costly pain.

So, over the past few weeks, he just built his own. He describe the process as exhilarating. I could relate. But, it was something else he shared that stuck with me.

Coffee Guy said that once the machines can build anything you can dream up, the only vital thing left for humans is the inspiration. Not the how, but the why and the what. He said the important part will be picking up a book, going for a walk, doing whatever you need to do to get inspired. Do this well and you just might manifest something people will use and like. That’s what Coffee Guy said.

This seems right. If you’ve read this newsletter recently, you know I’ve been blown away by what a semi-technical old fuck like me can build now. I’ve spent thirty years cajoling engineers and designers to build stuff. Now, I can close the loop myself, and it’s electrifying. But… you still need a good idea.

A couple of weeks ago I built a media aggregation machine — my own authenticated personal media interface for my media sources, newsletters, etc. It’s an obvious and utilitarian use case. Now I rely on it daily. That’s a win.

My next project was the Hot Take Machine. Originally I just wanted a place to pin ideas and research them. Brian Morrissey suggested that I build something to “steelman” ideas. It turned into www.bighottake.com. It’s a place to add “takes”, validate them and browse others. I guess that makes it a shitty Twitter.

The machine itself isn’t what I love, it’s what the process is teaching me about why things work.

Media works when it consistently gives people something they need to know, feel, become, or belong to. Platforms are different. They work when they make it effortless and habitual to satisfy many of those needs in one place. They are, at least at the start, entirely about utility.

Big Hot Take doesn’t do either well enough — certainly not better than the alternatives that exist today. Trying to build a platform that does what other platforms do, just in a slightly different way, is a tough brief. The energy required to shift human habits and break established network effects is immense. Elon understood that; it’s why he bought Twitter instead of building it from scratch. Even after he kicked the shit out of it, the network remains vital.

Anyway, I’ve learned a ton—about feeds, algorithms, scoring, APIs, system stability and the psychology of why we use the things we use. Now we are back at the hard part: the inspiration, the idea, the spark.

In a world of infinite, effortless code, you win or lose on the idea. The AI economy will inevitably do more to help people with this part. Especially because we will have way more time to reflect once machines do all of the work. Except for the chronic, restless types that can’t get enough of that building feeling and want desperately to hold their place on top of these powerful machines.

Last night, I took the dog for a walk and thought of something new. I might see if I can build it.

Pod 179: The Great Media Reorientation

Out FRIDAY 6 AM EST Apple | Spotify | Substack

Traffic is out, audience focus is in. B2B is a refuge; see the Acquired ad rates. The Boston Globe’s TikTok star Emily Sweeney proves a Dorchester accent is leverage when the internet has flattened everything. The next newsroom brawl will be over AI use. Political figures now have to be expert content creators. Even geopolitics aren’t immune: Iran is beating the U.S. with its AI Lego video trolling. In tech, Sam Altman lies a lot and might be a sociopath, but it could be these are required traits in his field. Vibe coding Big Hot Take has turned Troy into an AImaxxer who believes company structure will change. Finally, a hypothesis that the true emblem of American wealth is getting a boat.

Subscribe now

Fun episode of the Acquired Podcast with guest Michael Lewis. They chat about building The Acquired business. Lewis want ads from things he loves like some underwear brand. Acquired guys: we prefer enterprise software partners. The strategy seems to be working.

The "backwater" of B2B media is officially the hottest zip code in the industry. The Acquired podcast has reportedly sold out its ad inventory through 2027, with 2028 mid-roll packages going for a cool $2.9 million. It turns out that when you stop chasing Comscore uniques and start cornering high-value niche audiences, the margins look less like publishing and more like software.

I love Mo. Calling bullshit on AI guys and why Claude Mythos is delusional.

The Dorchester Advantage
The Boston Globe has an unlikely TikTok star in Emily Sweeney, whose thick Dorchester accent sounds like a dispatch from a more local, less flattened internet. In a media environment optimized into sameness, distinctiveness has value. The accent is the brand.

Iran Has Better Meme Lords Than the State Department
Iran spent the week winning the information war with surreal AI-generated Lego videos and a better instinct for how Americans consume media than the Americans themselves. The uncomfortable truth: in 2026, diplomacy is downstream of content strategy.

Politicians Are Creators Now
Trump understands this. Mamdani understands this. Even the Pope understands this. RFK Jr. is getting a podcast. Modern politics increasingly belongs to the people who can make compelling media, not just compelling policy.

The Next Newsroom Civil War
The future newsroom fight will not be over politics. It will be over AI. Editors increasingly want speed, scale and lower costs. Reporters increasingly think AI writing sounds like it was written by a LinkedIn post that went to business school. The divide is becoming existential.

WSJ’s Emma Tucker lit the fuse when she praised Fortune’s use of AI to crank out hundreds of stories and suggested journalists who object should probably leave the business.

Artists vs. Engineers
There is a fundamental split in modern media between the Engineers (Mr. Beast, Alex Hormozi) who deconstruct the knobs to optimize retention and the Artists (Rick Rubin, Emily Sundberg) who lead with taste. The engineering mindset is winning the "build," but the earned energy still belongs to those who prioritize sensibility over spreadsheets.

Hot Tub Meet-Cutes
Ronan Farrow’s New Yorker profile of Sam Altman features a 3 a.m. meet-cute in Peter Thiel’s hot tub and a pattern of "slimy" behavior. It raises a broader question for the valley: is sociopathy a bug in AI leadership, or is it a prerequisite feature for bending light to a trillion dollar valuation?

Ambient Liberalism
Soros Fund Management is reportedly moving into Meidas Touch, the progressive YouTube powerhouse that has turned "ambient engagement" into a $40M EBITDA machine. It turns out you don’t need a literal 3 a.m. audience when you have a high-margin ad engine that dominates the feed with high-octane clips.

Want more merch to express your other multitudes? See the latest Palantir drop.

Thanks for reading. Say hello and let us know your take.

• Watch us on YouTube

• Troy Young’s People vs Algorithms newsletter

• Brian Morrissey’s The Rebooting newsletter

• Alex Schleifer’s Human Computer

• Follow Alex, Brian and Troy on Twitter

Chapters:

• 00:00 Open

• 00:41 Welcome

• 03:10 Media Business Reset

• 07:05 B2B And Niche Wins

• 10:30 Altman Narrative Wars

• 13:18 CEO Archetypes Debate

• 18:31 OpenAI IPO Reality

• 20:58 Mythos Model Hype

• 24:04 Building Hot Takes

• 31:15 Vibe Coding Impacts

• 36:48 Politics As Creators

• 39:04 Trolling As Statecraft

• 40:34 Boston Globe TikTok Accent

• 44:13 Artists Versus Engineers

• 46:43 Midas Touch Deal Breakdown

• 51:27 Copycat Creator Acquisitions

• 56:23 NYT Reporting And AI Newsrooms

• 01:02:23 Clipping Economy Explained

• 01:07:42 Boats As Status Symbol

• 01:10:48 Good Product

• 01:12:57 Outro

Emily Sundberg joined the podcast this week and it was supersonic.

Just after the recording we learned OpenAI acquired TBPN. Sam Altman professed his affection for the duo on X:

TBPN is my favorite tech show. We want them to keep that going and for them to do what they do so well… I don’t expect them to go any easier on us, am sure I’ll do my part to help enable that with occasional stupid decisions.

Naturally, we had to quickly light the mics back up and comment. Emily could not resist sending over this tweet from Rat King. I hope Anthropic gives her a bunch of cash too.

Let’s leave that to marinate.

I could not help talking about my new AI coding exploits on the pod. Combining product strategy with design and coding in a one man do-it-yourself to and froh is soooo addictive. Way more than social media ever was for me. Downside… it may be making me even more impatient with real humans. Working on this.

For now, I’ve finished with the PIMP app (Personal Intelligence Media Platform) that we spoke about last week . I use PIMP every day and it might be making me a more efficient media consumer. I dunno. BTW, my daughter continues to make incredible progress on her film investment ROI app - AI coding has become a lively family affair.

PIMP did spawn a fun offshoot project. I wanted an app that combined note taking — registering ideas or “takes” that occur to me now and then — with a little research agent. What came out was the Hot Take Machine. I probably need to work on the name and brand. It’s fine for now.

Think of Hot Take (hottake.peoplevsalgorithms.com) as Twitter for “takes.” Except here the machine helps flesh out your argument.

Here are some screen shots. Give it a try. Let me know what you think.

How it works:

• The feed is a list of “takes” made by users. The “take” is just an idea you have about the world. Any claim.

• Click the big pink button to add a “take” — it can be a line of text, a voice recording, a URL, a screenshot. A carefully conceived or a half-baked idea. The AI builds the strongest possible case for it: the sharpest thesis, most convincing evidence and supporting facts.

• Click “cold shower” for the opposite view. Hit “PvA take” to get the “official” POV based on hundreds of pod transcripts and newsletters. I might add different “takes” over time from people that let me do it. TBD on value of this.

• You can also challenge anybody else’s take.

• Every hot take gets a conviction score from 0–100. How well does the argument actually hold up?

• Each week, takes from our podcast show up as colored grouped list in the feed.

• The rest are ordered by recency and topic. The feed is like a live debate board.

• I’ve included Google Auth so people aren’t assholes and names are attached to takes. But you can go in anonymously if you want for now.

The cost: The problem with these AI apps is they are token hungry. I’ve spent about $300 in tokens building stuff over the past few days, significantly more than I spend on streaming services. Admittedly it’s a fraction of what it would have cost to have a real team build stuff in before times.

But the problem isn’t the cost to build. Go forbid a lot of people use it. I made a calculator to see what would happen. If 5000 people use it 5x per month, it will cost a manageable $50/month, provided I route most of my requests to the inexpensive Gemini 2.5 Flash model. If I used Claude Opus the costs would be $1500/month, 30x as much. If a million people used the existing model it would chew through more than $5000 / month of the cheap tokens. Anybody who builds stuff like this will have to work hard on token efficiency and the pull toward free open source models will be strong.

Why do this? These projects are a great way for me to understand how AI is gonna change how we work, how orgs are optimized, how value is sustained, where moats might emerge. An obvious question is is why create a tool that is just replicating a conversion you might have with a chat bot? Other than formatting the prompts and structuring the result, capturing the a feed of what the community is asking is obvious answer. Maybe it can turn prompts into human back and forth. I’m not sure. The question of how value is created on top of LLMs will be asked frequently as we absorb this technology.

Pod 178: New Media Energy

Out FRIDAY 6 AM EST Apple | Spotify | Substack

Troy and Brian react to OpenAI buying TBPN: It pays to have viewpoints and approaches that are preferred by the powerful. FeedMe’s Emily Sundberg joins as a guest this week. We discuss why new media like TBPN is beating legacy media not by distribution hacks or aesthetics but in energy. Legacy media relies on structural energy that’s dissipated while new media is built on earned energy, which is more organic, relatable and even fun. Plus: Puck vs Vanity Fair, Troy’s vibe coding addiction, and why venture capital isn’t cool.

Subscribe now

Maybe AI chatbots can undo some of the harms of social media. FT’s John Burn-Murdock explains. Paywall article here.

Chart of the day: AI-driven Mac Mini sales or how Christianity went from a tiny minority to the dominant religion of the Roman Empire:

And one more. Hollywood job market is collapsing:

A Monumentally Hideous Skyscraper “Library”
Donald Trump’s reveal of his Miami-based presidential library—a 50-story skyscraper featuring a golden statue and a Qatari-gifted Boeing 787—has sparked a predictable firestorm. While the President frames the tower as a “lasting testament” to his legacy, critics like Robert Lamondin Jr. are ranting about the “weird” scale and the controversial donation of public land by Miami Dade College.

The Experience Arbitrage
Red Seat Ventures—the creator incubator acquired by Fox—is launching a premium membership for the ultra-wealthy. It’s a strategic pivot from mass-market digital content to high-end hospitality and exclusive live experiences. In an age of infinite digital slop, Red Seat is betting that real-world access is the only remaining scarcity worth a premium.

Talent Management as the New Ad Sales
The power shift from institutions to individuals means “wrangling talent” is now more critical than selling ad slots. Even legacy giants like The New York Times are hiring video coaches and producers to transform traditional journalists into “hosts.” Personality-driven journalism is no longer a niche—it’s the new baseline for engagement.

The TBPN Effect
Technology Business Programming Network (TBPN) is proof that media success today is about "earned energy" rather than massive subscriber counts. By live-streaming for three hours a day, they’ve built a business that works even when only 100 people are watching a feed covered in ads. Emily highlighted their recent "peptide debate" between Martin Shkreli and Superpower tele-health founder as a perfect example of their mischief and playfulness. It is a level of fun that legacy brands struggle to manufacture.

The Guano Trade and the Spirit of Enterprise
Brian takes us back to the mid-19th-century Guano Islands rush to show that the spirit of enterprise hit different before the age of AI tokens. While we are currently vibe coding our way through digital productivity, the 19th-century version was a high-stakes physical hustle that laid the groundwork for modern trade. It raises a question for the digital age: Are we moving toward a post-Gutenberg oral culture where the ability to think and communicate on your feet is the ultimate competitive advantage?

Cool or Not
We put FeedMe’s Emily through a round of “Cool or Not Cool” to calibrate the zeitgeist: Venture Capital: Not cool. “Too much money and too few good ideas.” Home Saunas: Not cool. It’s about leaving your house and using someone else’s towels. Catholicism: “I think like as the world becomes more complicated and digitized and people are spending eight hours building like tools of AI, I think people turn to all kinds of things and whether that’s like a sauna or a church, I think that uptick of like searching for self care and meaning is real. I don’t know if people are necessarily actually becoming more religious... I think it’s definitely aesthetic.”

GOOD PRODUCT

Music: This week we appreciated the soulful vocals of Olivia Dean and Flea’s unexpected jazz-infused debut album, Honora, where he trades the bass for a trumpet. We love late career innovation from Flea.

TV: HBO’s DTF St. Louis, a nonlinear black comedy featuring nuanced performances from the incredibly talented cast… Jason Bateman, David Harbour, Linda Cardellini, Richard Jenkins and Joy Sunday deliver an inspired script beautifully.

• Watch us on YouTube

• Troy Young’s People vs Algorithms newsletter

• Brian Morrissey’s The Rebooting newsletter

• Alex Schleifer’s Human Computer

• Follow Alex, Brian and Troy on Twitter

Chapters:

• 00:00 Open

• 00:28 OpenAI Buys TBPN

• 08:13 Welcome

• 10:11 Vibe Code Corner: PVA Hot Take

• 17:17 Agents and Future Org Charts

• 22:08 Emily Vibe Coded Side Apps

• 25:18 Future of Media Trends

• 33:11 Feed Me Business Overview

• 37:39 Events Merch and Energy Thesis

• 43:32 Scaling Feed Me carefully

• 47:34 Why TBPN works

• 53:01 Defining new media

• 56:43 Vanity Fair vs Puck

• 01:01:20 Aspirational media and detachment

• 01:06:15 Events business reality

• 01:09:37 Cool or Not Cool Corner

• 01:15:05 Good Product

• 01:17:25 Outro