Author: KYLA SCANLON
Compiled by: Deep Tide TechFlow
Good morning, greetings from Washington, D.C.! This article is a bit long, so it might get truncated by your email. I've been on the road again for work recently, with destinations including Michigan, Kentucky, and Washington, D.C. While going through security, I saw a woman ahead of me coughing with her mouth wide open like a baby. I stared at her, first surprised by her "carefree" attitude, then feeling a deep sense of fear.
Most people are very friendly. But living in a society means dealing with different internal norms of others. Some people cough with their mouths open—that's the reality. I have a theory: they might believe that collective comfort is not their responsibility, perhaps because they lack a sense of belonging in the public sphere. This is a form of social drift, becoming increasingly evident in public spaces (like staring at a phone at a 90-degree angle and bumping into walls, or standing in the middle of pedestrian traffic and blocking the way).
But I think there are many similarities between these open-mouthed coughers and the ongoing economic downturn we're seeing. If you lose trust in the systems around you, why adhere to collective norms? Hard work doesn't seem to pay off, so why not "take a gamble"? Institutions are lying! But that YouTuber who makes the thumbnail isn't lying—his thumbnail features a big-mouthed emoji pointing at a bowl of pasta, asking some "big questions." We no longer trust each other. As Jordan Schwartz, the student chair of the Harvard Public Opinion Project, said:
"Gen Z is on a path that could threaten the future stability of American democracy and society. This is a five-alarm crisis, and we must act immediately if we hope to restore young people's confidence in politics, America, and each other."
The project he leads—the Harvard Youth Poll—surveyed over 2,000 Americans aged 18 to 29 about trust, politics, and artificial intelligence. When asked if they believe the United States is a healthy democracy, respondents showed clear partisan divisions, but the anxiety was evident.
Trust between groups is also collapsing. Only 35% of young Americans believe that people with different viewpoints want the country to move in a good direction. 50% see mainstream media as a threat. And only 30% believe their economic situation will be better than their parents'.
So, from this survey, we can identify three key concerns:
- Concerns about democracy
- Concerns about the economy
- Concerns about each other
I believe that without understanding how we discuss the economy, we cannot truly understand the economy itself. Here, we are dealing with the compounded effects of three factors: (1) the post-pandemic adjustment process; (2) smartphone-induced "micro-solipsism"; and (3) younger generations witnessing objectively unfriendly behavior in politics being rewarded instead. People (understandably) are experiencing a kind of "cognitive drift," some even calling it "medieval peasant brain," which is related to the constant flood of information on the internet (like people putting potatoes in their socks to "detox").
We are trapped in a compound crisis—the interaction of economic deterioration and cognitive overload creates a recursive trap, where each side makes the other worse and destroys the resources needed to break free.
- Economic pressures (like Baumol's cost disease, housing issues, weak labor markets) weaken our ability to think clearly, making us more vulnerable to scams, poor decisions, and exploitative markets, which in turn worsen economic pressures.
- Economic pressure + information overload erode trust in institutions.
- Loss of trust makes coordination impossible, problems remain unsolved, and unsolved problems deepen the crisis.
Currently, we are trying to understand the economy in a context where social and cognitive environments are changing faster than traditional economic indicators. This is the backdrop of the "Vibecession."
Note: Given that both Paul Krugman and Scott Alexander have recently revisited this concept, it's meaningful to reexamine what "Vibecession" meant in the past and how it has evolved today.
"Vibecession": Past and Present
I first proposed the concept of "Vibecession" in July 2022. At that time, inflation was falling (but still painfully high), the labor market was recovering, and the economy was growing. Back then, AI hadn't taken over all the focus, there were no tariff barriers, and large-scale infrastructure investments were underway. From the data perspective, everything seemed to be improving.
While the 2000s and 2010s had many problems (really, a lot!), the mood hadn't completely collapsed. Nowadays, there's even a trend of "nostalgiacore" on TikTok, where teenagers fabricate a beautiful fantasy of "2012"—they yearn for infinity scarves, third-wave boutique coffee shops, and the era when Instagram was just for sharing pictures of daisies in fields, not the hyper-competitive algorithmic battleground it is today.
There was still a glimmer of hope back then (this was also the core idea of Obama's campaign!), and people had "better" expectations for the future of the internet. Although the internet already had some issues, it wasn't yet profiting from stoking anger as it does now.
Some claim that "the entire past decade was a Vibecession," but sentiment data doesn't support this. In fact, the break in sentiment is clear and sharp.
The chart below roughly shows when the Vibecession began, illustrating the divergence between sentiment and economic data. After the pandemic shock, real disposable income recovered and continued to grow, returning to its normal trend. However, public sentiment never recovered. It slid into recession-like territory (or even lower) and has stayed there, even as economic fundamentals stabilized.
I think part of the reason is the accumulation effect. The chaos of the pandemic isn't over yet, prices of various goods are still unstable, stores are understaffed, teachers and students are exhausted, the public information dissemination system has collapsed, and institutions seem fragile. The daily frictions of life have increased in countless small ways. The surge in housing prices during the pandemic never came down. As the Fed started raising interest rates, mortgages "locked" people in place. Rents skyrocketed, and the path to adult life—moving out, renting, saving, buying a home—has been broken for many. If you didn't buy a home before 2020, you might never be able to.
But as Dan Davies wrote, the "Vibecession" might not have a specific trigger. "Vibes are like a supercooled liquid, just waiting for a random shock to undergo a phase change." And the pandemic was that shock.
The Vibecession came early. Now, the economic data matches people's mood, or at least is closer than before. We are now in a low-hiring environment, with persistent inflation and extremely strange trade policies. When the NBER defines a recession, it looks at three aspects:
- Depth: How severe is the economic downturn?
- Diffusion: How widespread is the pain?
- Duration: How long has this been going on?
If we look at the decline in consumer sentiment, it (roughly speaking) meets the definition of a recession—it's long-lasting, widespread, and sentiment levels are approaching historical lows. Kevin Gordon of Schwab calls it a "Vibepression"—extremely depressed sentiment, while GDP is propped up by AI-related investments. Can an economy booming due to AI data center construction make ordinary people happy? The answer is clearly no!
But why do we have this deep sense of gloom?
Part One: Economic Deterioration
A few weeks ago, Michael Green published an article claiming that "$140,000 is the new poverty line," pointing out that almost no one can afford the cost of participating in society nowadays. This article sparked heated discussion on the internet. Subsequently, Tyler Cowen, Jeremy Horpedahl, and others published rebuttals. However, as John Burn Murdoch wrote, the reaction to the article itself is very interesting.
Most people strongly agreed with the article's viewpoint (many rebuttals were countered with "Who cares if the data is accurate, the vibe is right!"). The article was republished by media outlets like More Perfect Union and The Free Press. People from both the left and the right read it and said, "Yes, this is why everything feels so bad. This is poverty. My economic pain is finally confirmed by data. What a relief."
Being "seen" in an analysis is a relief. Paul Krugman, in his series on the "Vibecession," pointed out that the following three key concepts are not well captured by traditional economic data:
- Economic participation: Can you afford the cost of participating in society?
- Security: Are you one bad tooth away from bankruptcy?
- Fairness: Are you being cheated?
People need to feel that they can afford a house, children, or a car, that they won't go bankrupt from a medical bill, and that others aren't cheating them. And the answers to these questions are becoming increasingly uncertain.
Regarding the first point—the Fed cut interest rates yesterday, sparking much controversy and division. Their dual mandate—price stability and maximum employment—is under increasing pressure. Inflation hasn't fallen to the 2% target (the bond market is very worried about this). The labor market is weakening, and inequality is worsening.
There is very real economic pain; participating in society has become more difficult. Young people once voted for Trump to improve the economy, but now they are turning against him. According to the Yale Youth Fall 2025 Poll, 18- to 29-year-olds show strong dissatisfaction with President Trump's handling of the economy.
John Burn Murdoch points out that we are facing "Baumol's Cost Disease."
The same productivity growth that drives down the prices of tradable goods causes the cost of face-to-face services to膨胀 rapidly. Industries like healthcare and education, which require intensive face-to-face labor, have slow (or even non-existent) productivity growth and must raise wages to attract labor that might otherwise choose higher-paying jobs in more productive industries. The result is that even if people consume exactly the same types of goods and services, as the country's standard of living rises, they find that more and more of their spending goes towards basic services.
Prosperity can actually make life more expensive.
Source: The Financial Times. Paul Starr, in The American Prospect, documented the collapse of cultural affordability under Baumol's Cost Disease, noting that "public elementary and secondary schools, public libraries, low-tuition land-grant colleges, and the mass media of the 20th century—including free radio and television" were once free or at least heavily subsidized. However, support for arts and education is now being cut.
In reality, this means the core elements of middle-class life—housing, healthcare, childcare, education, retirement—all belong to the Baumol sectors. The costs in these areas are rising faster than wages. Even if you "do everything right," you can still feel like you're falling behind.
In the 20th century, we somewhat solved the Baumol cost problem by socializing or heavily subsidizing these areas, like public schools, public libraries, low-tuition state universities, and public hospitals. Through policy, we made these expensive, low-productivity areas cheaper. However, now, at the worst possible time, we are privatizing (or undermining, or bureaucratizing) these areas. We are asking families to bear costs that were once socialized. Is it any wonder the middle class feels squeezed?
Of course, it only gets more complicated. AI will make non-Baumol sectors超高效 productive. Software development, data analysis, and anything computer-based will become abundant and cheap, meaning the productivity gap between scalable and non-scalable fields will become a huge chasm.
The second point—the government shut down this year over healthcare issues. The average annual health insurance cost for a family of four is $27,000. Insurance premiums are expected to rise 10%-20% next year. Many are one bad tooth away from bankruptcy.
The third point—we are rapidly moving towards a "quid-pro-quo" economic model. The US, once a beacon of democracy, is now trading land with Russia, requiring tourists to provide five years of social media information, threatening the independence of institutions including the Fed, and ignoring antitrust laws to support media control. When you read such news and see such headlines, it's natural to feel very bad.
Therefore, for many people, especially young people trying to build a life, the economic fundamentals have indeed worsened. But economic pressure alone doesn't fully explain this deep anxiety. This is where the cognitive layer comes in.
Part Two: Cognitive Overload
These problems aren't new, right? The US has been sliding towards a more difficult equilibrium for years. People have long experienced high housing costs, tight job markets, and Baumol's cost disease. However, the difference now is that these pressures are falling on a public that is already cognitively and socially overwhelmed.
For most of human history, literacy was scarce, but attention was abundant. Apart from work, most of the time people were in a state we now call "boredom." Today, the opposite is true—literacy is declining, attention has become a commodity, and people's cognitive load is completely overloaded. Jean Twenge wrote in The New York Times in an article titled "The Screen That Ate Your Child’s Education":
"In a study published in October in The Journal of Adolescence, I found that countries where students spent more time using electronic devices for entertainment during school days saw significantly larger declines in standardized test scores in math, reading, and science than countries where usage was lower."
And Brady Brickner-Wood, in "The Curious Notoriety of Performative Reading," wrote:
"Americans spend 40% less time reading for leisure than they did two decades ago, and 40% of fourth graders lack basic reading comprehension skills...... Meanwhile, universities are partnering with companies like OpenAI to introduce chatbots into student curricula, while humanities departments are being cut back."
If you don't trust any source of information, you won't trust economic data either. We conducted a huge experiment—giving people unlimited access to millions of things that could make them lose control—and the answer is, no, really not, it's like cooking an entire population into an egg.
The loss of education and deep reading has various knock-on effects: weak basic skills, declining media literacy, and more importantly, the collapse of trust. David Bauder's research on teen news consumption shows that "about half of the teens surveyed believe that journalists give special treatment to advertisers and fabricate details like quotes."
AI only makes this more complicated. Greg Ip's article in The Wall Street Journal, "The Most Joyless Tech Revolution Ever: AI Is Making Us Rich and Unhappy," sums it up perfectly. Almost two-thirds of people are uneasy about AI, and only 40% trust the AI industry to do the right thing. We have all this technology, but we neither trust each other nor feel very good.
Source: Greg Ip, The Wall Street Journal (WSJ)
So, when we talk about negative sentiment, there is indeed a "computerized" quality running through it.
We are collectively plagued by the "Bullshit Asymmetry Principle": We find that debunking lies is indeed 10 times harder than creating them. This leads to marketing and product strategies like "ragebait"—which is also a great way to raise a lot of venture capital money?
Disinformation has become an effective way to accumulate wealth: if you lie to a lot of people and make them angry at you, Twitter will pay you a lot of money. People abroad are also using this "money printer"—which makes logical sense!—and polluting American politics in a way that should probably be illegal?
Many are also "skimming off the top" at various points, cheating to gain an advantage, as Krugman mentioned with the "scam" problem. Every adult can feel their attention draining, their thinking flattening, their world filled with noise, no neutrality, and no institutions truly existing to protect them. Bro, your brain is being sold—as your attention drains, so does your cognitive ability, depth, and sense of certainty.
Confidence, optimism, and long-term thinking all require "mental space." If the information environment is chaotic, the emotional environment will be too. And if attention is the infrastructure of democracy, then that infrastructure is broken.
We are witnessing the consequences of outsourcing human learning to screens. Now, we might see what happens when we outsource humanity itself to AI. When you can't trust any source of information, you can't trust economic data either. When attention is fragmented and thinking is flattened, people become more susceptible to the next stage: extraction.
Part Three: The Extractive Economy
While the cognitive world is collapsing, the maintenance of the physical world isn't great either. The huge friction between decaying things in the physical realm (like bridges, schools, labor markets) and hyper-optimized things in the digital realm (like large language models, algorithms, and various operations in the advertising field) is becoming increasingly obvious.
In this newsletter, I'm quite harsh on AI—to be clear, I think AI is a tool that can indeed bring major breakthroughs to science—but AI itself is creating a complete downward spiral. In a documentary, Demis Hassabis's discussion of AI is very important, and as Linus Torvalds said in a recent interview:
"I very much believe in the potential of AI, but I'm not optimistic about the things surrounding AI. I feel the market and marketing are sick. This will lead to a crash."
Today, people become billionaires by expanding data centers, which drives up electricity costs and brings blackout risks. These data centers occupy huge physical space, but their impact is almost invisible to ordinary people, the only perceptible thing being rising electricity bills.
The AI race is a race for energy, not computing power. As the Financial Times wrote:
"In the global superpower competition, AI could be slowed for decades by outdated grid infrastructure and failure to provide sufficient power capacity."
Source: Financial Times
The Financial Times also reported that OpenAI's partners are already $100 billion in debt to build AI computing capacity. This is worrying because debt is where problems become dangerous. The dot-com bubble was mainly an equity crash, meaning there was no complex web of debt relationships. But once debt is involved, things can get very tricky very quickly.
The US also decided to sell some of Nvidia's top chips to China in exchange for soybeans and a 25% kickback. As the US Department of Justice said:
"The country that controls these chips will control AI technology; the country that controls AI technology will control the future."
So, you know, that's fine. The Financial Times reported that the US is losing the AI race, "Many US companies, including Airbnb, have become avid fans of the 'fast and cheap' Qwen." They also raised the question: "Can the West catch up with China?"
The chart below is very important—when we talk about the future, we often think of the US as the number one global superpower, but China is investing in the key factors needed for AI success—energy. The US is choosing the opposite direction.
Source: Phenomenal World
Barclays estimates that by 2025, more than half of US GDP growth will come from AI-related investments. People have realized that we are betting the economy on something that doesn't really promise much, like "Hey :) this thing will take your job :) now it can do art too :) might make some people very rich, but your electricity bill (which is already changing voter attitudes) will go up. Also, China might win. Also, suicide is against the terms of service."
Almost all young people are very worried that AI will take their jobs. MIT's "Iceberg Index" estimates that about 12% of US wages come from work that AI can do cheaper today, but only 2% of jobs have actually been automated so far. The capability exists, it just hasn't been fully activated yet.
How do you trust a system that doesn't seem to care what happens to your future?
Source: Harvard Youth Opinion Poll
This might partly explain why nearly 40% of Americans over 50 think the economy is "getting better," while most Americans aged 18 to 49 think it's "getting worse." These are two completely different economic worlds. Older people are largely insulated from AI and housing shocks, while young people are facing the direct impact of these threats.
Source: Civiqs
Adam Millsap wrote an interesting article about "Total Boomer Luxury Communism." This concept roughly refers to the older generation "hoarding opportunities and resources, while young people struggle to buy homes and support the generous social security and Medicare benefits that the wealthiest boomers expect." This intergenerational tension will only intensify with medically assisted longevity technology and resource shortages.
So, what should people do? AI is taking jobs, policy is increasingly designed around the elderly population, everything looks uncertain. How to move forward into the future?
Gambling?
Tarek Mansour, co-CEO of Kalshi, recently said: "Our long-term vision is to financialize everything and turn any difference of opinion into a tradable asset."
Financialize everything??? Every disagreement, every uncertainty, every future outcome—all become bets?? This is an extreme extension of Marx's logic of "commodity fetishism." When every interaction becomes a transaction, every opinion becomes a tradable asset, it becomes extremely difficult to form solidarity.
Gambling has become one of the few activities that can provide immediate returns, even life-changing ones. Living near a casino increases the likelihood of becoming a problem gambler. And when you live on your phone, and the casino comes directly to you, I guess you can imagine what happens. But the truth is, no one really wants this life. As the chart below shows, this is the truly depressing thing about the "casino economy"—no one wants it.
Attention is monetized, engagement is optimized, risk is financialized, everything looks like a scam. As Whitney Curry Wimbish wrote in The American Prospect, and Emily Stewart in Business Insider, layers of middlemen are extracting value, with almost no real regulation or protection. Some might say, "Well, obviously the market has expressed its preference, and that preference is for people to spin the roulette wheel." But I don't know how to respond to that.
When the labor market tightens, upward mobility stalls, wealth concentrates at the top and becomes increasingly out of reach, gambling seems like a rational response. In this structure, people lose a sense of purpose and meaning (Victor Frankl, save us), and that's when problems start to breed.
Reduced cognitive bandwidth + ubiquitous extractive systems = rational economic paranoia. People feel they are being cheated because they often are. When traditional paths become unpredictable, people turn to "narrative ladders"—like online communities, aesthetic categories, etc. These become ways to understand uncertainty. The "$140,000 poverty line" debate erupted in this context. People's reactions are about identity and experience, and whether their worldview can be understood by others. These reactions aren't all rational, but they exist.
When values diverge and common ground weakens, collective solutions become structurally impossible. Even if there is broad consensus—no one really wants a "casino economy"—we still can't coordinate to stop it because we can't agree on what "stopping it" should look like or who should have the power to achieve it.
For the past 70 years, the US was built on a simple contract: deliver growth, and people will tolerate everything else. But this year, after 40 weeks on the road, the most consistent feedback I've heard from people of different ages, regions, and income levels is that the basic trajectory of life no longer makes sense. These are just scattered anecdotes, but it's the most common thing people stop to talk to me about—their worries. They are not only worried about their financial situation but also anxious about the entire future.
Part Four: Trust
We are in such a complex crisis: economic pressure reduces cognitive bandwidth, reduced bandwidth fuels extractive behavior, and extractive behavior worsens economic pressure. Pressure and overload together erode trust, loss of trust hinders collaboration, and failed collaboration leaves problems unsolved, and unsolved problems further deepen the crisis.
This is not a problem that can be solved with a single policy lever. When the trap exists at the intersection of multiple areas like housing, AI regulation, and media literacy, you can't just "fix housing" or "regulate AI" or "improve media literacy." Economic deterioration and cognitive collapse are mutually reinforcing; they have already destroyed the institutional capacity and social trust needed to solve these problems.
This sounds bad! But breaking this cycle doesn't require solving all problems at once. We need to find the most actionable links and recognize that improving one part weakens the trap's effect elsewhere.
Directly reduce economic pressure—make Baumol sectors (like education, healthcare, etc.) affordable again (I know, it's that simple, right?). If people have more breathing room economically, their cognitive bandwidth increases. We all know this. More bandwidth means less vulnerability to extraction and fraud. Less vulnerability means better decision-making, and better decision-making means less economic pressure.
Directly regulate extractive behavior—ban or strictly restrict business models that profit from confusion and cognitive overload. Kalshi wants to financialize everything? We can say "no"! Prediction markets on elections can be banned. It's all about incentives. We can regulate physical casinos, and we can certainly regulate digital ones.
Make the benefits of AI clear—Currently, people's experience with AI is "your electricity bill goes up, and eventually it will take your job." If AI is to drive growth, that growth needs to manifest in tangible benefits in ordinary people's lives—like lower healthcare costs through diagnostic tools, cheaper goods, and more time.
None of this is easy, or even easy to imagine. But it's not hopeless. You don't need to solve all problems at once. This requires economic affordability (heard the "affordability journey" has just begun), national governance capacity, some friction, and an understanding of "humanity" in a world full of technology. And one seemingly simple but actually arduous task—rooting out crony capitalism and building some kind of common sense of reality. On my cross-country flight, the internet went down, just a few hours after that person coughing into the air. I was frantically typing, doing all sorts of "very important work," like writing this newsletter. We were all typing in the dark, emailing, busy on Slack. When the internet disconnected, we started opening the plane's window shades. Outside was one of the most beautiful sunsets I've ever seen. There's something to think about there too.
Finally, I really liked this quote from a recent interview with Kahlil Joseph:
"There's a famous story about Jimi Hendrix tuning his music for transistor radios because that's what the soldiers listened on, so he tuned his work for FM radio. He wasn't thinking about someone listening to his music on a thousand-dollar sound system. That always impressed me—to meet people where they are."
Thank you.
Notes:
The UMich Survey of Consumers switched from random-digit-dialing to online surveys in mid-2024.
Netflix and Paramount are vying for Warner Brothers. According to the WSJ, "David Ellison assured Trump administration officials that if he acquired Warner, he would completely overhaul CNN, which happens to be a frequent target of President Trump's criticism."
Reflections on the impact of the internet are starting to gain strong traction. Róisín Lanigan wrote an article titled "The Next Status Symbol is an Offline Childhood," the title itself is very direct. More and more articles like this are appearing, like P.E. Moskowitz's "The Internet is Destroying Our Memory and History," exploring the gains and losses brought by the internet.
Cracks are showing in financial markets. Applied Digital had difficulty selling bonds and had to offer a high yield of 10% to attract buyers. They provide data center services for CoreWeave, which in turn provides data center services for Nvidia and OpenAI.
Kalshi is facing a nationwide class-action lawsuit, with plaintiffs alleging it operates an "illegal sports betting platform." The lawsuit claims Kalshi "deceived" customers into thinking they were betting against other consumers when they were actually betting against the house (i.e., Kalshi). Additionally, a Nevada court ruled that Kalshi is not exempt from state gaming regulations, posing a major challenge to its business model. A Bloomberg analysis showed Kalshi's fees are so high that in most cases, users might be better off using FanDuel directly.
