# Is Gen Z Cooked? What I Took From Ed Elson's Round With The Compound Source: Realpha Blog (blog.getrealpha.com) Original article and charts: https://blog.getrealpha.com/en/blog/compound-2026-10-02-is-gen-z-completely-cooked-with-ed-elson/ > On October 2, 2026, The Compound and Friends hosted Ed Elson, co-host of Prof G Markets, for an argument that ran from whether AI is melting kids' brains to houses costing seven times annual income. These are listening notes and educational commentary — no investment advice, no buy or sell calls on any stock. Published: 2026-10-02 Locale: en Tags: GenZ, AIBubble, GenerationalDivide, InvestorPsychology, EducationGap ![A late-night university corridor; a young person sits on the floor in the foreground lit by a phone screen, while a doorway far down the hall spills warm light and the blur of a crowd](/covers/compound-2026-10-02-is-gen-z-completely-cooked-with-ed-elson-cover.png) > A young man's heart should reach for the clouds — > who thinks of him sitting in the cold, sighing? > > —— Li He, "Song of Offered Wine" (Tang dynasty, c. 9th century; my translation) On October 2, 2026, The Compound and Friends hosted Ed Elson, co-host of Scott Galloway's Prof G Markets, where he fronts four of five weekly episodes. The central argument: whether AI and phones have wrecked young people. Ed's numbers were houses at more than seven times annual income (three times, he said, when his grandparents were his age), one in five Gen Zers reporting zero friends, and 118 days a year spent looking at a screen; Josh Brown pushed back that the high schoolers he sees do college-level workloads and face admissions far harder than his own cohort did. Both were right, because they were describing the two legs of a K — so what this episode actually gives you is a method for handling "what I see contradicts the statistics," not a verdict on whether Gen Z is doomed. ![Two bars of very different heights side by side: the short left bar is the roughly three times annual income that houses cost in our grandparents' day, and the right bar, more than twice as tall, is today's seven times.](/figures/house-price-three-to-seven-en.svg) ## The awkward phone call The story starts while Ed was still in school. He studied classics at Princeton, became a hard fan of Scott Galloway's podcast, and one day, mid-episode, realized the guest was his old roommate's mother. He froze, then started working the connection. What followed was a phone call he describes as awkward — pitching himself to his own hero. The result: after graduation he came on as an intern, helping with books, research, and decks. I stopped the tape there. Without that call, someone else is co-hosting the Webby-winning business podcast today. He doesn't dress the story up. He tried to get connected, and he made an awkward call. ## His boss told him to go read the news After a stretch of intern work, Scott told him: I'm starting a podcast, and I need someone to read the news. Ed said okay. And that was the job — he wrote up what had happened the previous day, read it out, reacted a little. No point of view, no segment of his own, just reading. Then, he says, he started noticing he had opinions and wanted to voice them. The show grew, the roles changed, and now he hosts four of five weekly episodes with Scott joining for one of them to argue with him. Josh made an observation I think is exactly right: on that show, Scott Galloway spends long stretches sitting quietly and listening, which nobody expects, because it's Scott Galloway. Ed's own reading is that Scott bet on a young person, when plenty of people ask why this kid gets to have an opinion at all. ## He says you have to look at the numbers Josh steers into AI and education, and this is where it gets good. Ed's position: when Jensen Huang said in an interview that young people don't need long division, that was too casual. He cites OECD testing — students using AI daily score lower in science than students who don't, because AI hands you the answer and you skip the painful part where learning happens. And the gap between the top 10% and bottom 10% of American students is the widest in the developed world (only Luxembourg is worse, he says, adding "I don't know what's going on in Luxembourg"). Josh takes the other side of the whole thing. He has kids aged 17 and 20, and what he sees is brains in fine shape: high school workloads that look like college, schools that waved his generation through on a 70s–80s GPA now out of reach, the University of Florida possibly running a lower acceptance rate than Princeton. He also got a laugh out of me: elementary schools don't teach cursive anymore, good — somebody sent him a letter in cursive and he threw it out, because he can't read it. Midway through, Ed gives ground: you're right that standards have risen and the top kids keep getting stronger. Then he turns it back — the bottom is sliding, and the scores say so. Which quietly changes the subject of the fight, from "is this generation good or bad" to "the gap is widening." ![A single line splits into two branches: the upper one rises to the right while the lower one falls to the right, with two observers each watching only one branch.](/figures/k-shape-two-legs-en.svg) ## Those young people on the street in Austin The best stretch comes later. Michael Batnick says he was in Austin the day before and got hit by it: so many young, good-looking people who look like they're doing well, and every time he talks to young people he can't believe the distance between the statistics and what his own eyes report. Ed's answer carries in one sentence: you weren't in Austin, you were in a self-selected sample. The people who show up to that event are the ones who already got their lives on track; you don't see the ones in the basement at home, and that share is rising. ![A large crowd of dots on the left, a slanted sieve with a gap in the middle holding most of them back, and only a few dots passing through to be seen by an eye on the right.](/figures/self-selected-sample-sieve-en.svg) Josh then offers a sharper version. Every August his Instagram feed fills with sorority dance videos from southern schools — he has a college-aged daughter, he has no interest in this, and still they arrive one after another: healthy young people, huge smiles, clearly dozens of hours available for choreography, hair, makeup, costumes. Whose kids are those, he asks. Ed answers: doctors, lawyers, executives, real estate owners, people who made money in the market. Josh says I know that's 10% of the population, and we shouldn't treat that 10% as the problem. Ed says I don't want to drag the top down, I want us to take seriously that the bottom is falling. Nobody wins that exchange, and I came out of it knowing something: in one generation, two people each saw half, and both halves were real. ## Why the AI lab financials set them off The same logic lands on the market, and here one number is worth keeping. Ed's framing: this isn't a blanket AI bubble, the bubble sits in pockets. Nvidia trades at 16 times forward earnings, a consumer-staples valuation. Meanwhile, in the private market, a company with $4.5B in revenue and $42B in operating losses carries a $2 trillion valuation, and the people raising the money say don't worry about the financials, the total addressable market for AI is $28 trillion — larger than the GDP of China or of Europe. ![Four boxes nested from large to small: the outermost is the 28 trillion the market imagines, and the innermost revenue box is so small it is almost invisible.](/figures/ai-valuation-nested-scales-en.svg) Josh's rebuttal is blunt: profitability is the wrong yardstick at this stage. These companies are racing for scale, for whoever reaches three billion users first; if Bezos had chased profits in 1998, nobody would have heard of Amazon again. Ed's reply is the line I wrote down: I understand what they have to do, I'm only asking that we talk honestly about the risk in that strategy — this is a $2 trillion venture bet, venture bets fail, so stop pretending these companies are a settled fact of the economy. He also flags something I hadn't weighed: what unsettled him in the leaked financials wasn't the losses but the customer concentration — two customers making up a quarter of revenue. Losses are investment. Concentration is structure. ![A single bar standing for all revenue, with the leftmost quarter marked as two customers and the remaining three quarters as every other customer.](/figures/customer-concentration-quarter-en.svg) ## When your eyes and the statistics disagree This is the hole I fall into most often in my own investing, so I want to pull it out on its own. You walk through a store and the crowd spills out the door; you get home and read a report saying the industry is shrinking. Which one do you believe? Most of us go with our eyes, because they're vivid. But the Austin exchange says something uncomfortable: your eyes are showing you a filtered sample, and you don't know the filter. Here's how I handle it now. When what I see fights what the data says, I don't pick a side — I ask one question first: how was my sample selected? Those people in Austin were selected by "decided to come to this event." Those dance videos were selected by an algorithm feeding me what it thinks I'll watch. My good feeling about a holding was selected by the handful of accounts I happen to follow. Ask that question and you usually find your eyes and the statistic were answering different questions, one about the top, one about the average. Keep both numbers, filed separately. ![A bell curve where the average the statistics talk about sits in the middle, while the people you have seen with your own eyes fall in the shaded right tail.](/figures/your-sample-sits-in-the-tail-en.svg) It works on financials too. A company grows revenue 40%; you see 40%. Ask who contributed that 40%, and you may surface exactly the concentration Ed was pointing at. ## So why do investors keep funding the losses Here's the other question I get asked and have chewed on for years. You see a company burn billions and raise again, and the instinct is either these people are insane or they know something you don't. The Josh–Ed disagreement surfaces a third answer: they're playing a different game, so they're holding a different ruler. ![The same block in the middle stands for one company, with a ruler on each side whose scales differ completely, one reading failing and the other reading leading.](/figures/two-rulers-same-company-en.svg) Nobody on that cap table is asking for 2028 profits, because asking would work against their own investment. They want scale, scale converts to a winner-take-most position, and the profits get wrung out of the user base later. The logic is sound on its own terms — the Amazon comparison is fair. And Ed's counter stands just as well: the same logic was applied to a long list of companies that went under, in a stretch where the Nasdaq lost about 70%. What I take away is a different move. Don't argue about whether the ruler is right — ask where you'd see it first if the ruler is wrong. Ed names two observable places: customer concentration, and the price public markets will actually pay. The show mentions SB Energy walking the roadshow, showing the materials to Wall Street, then pausing the deal itself. Public markets are pickier than private ones, and they say so out loud. Those are moving signals, not beliefs. As for how private valuations get set, Ed's description is plain: they go to Silicon Valley, those people don't really look at the financials, they look at the technology and at the founder, they say this is an amazing founder, and they put him on their podcast. That's the pricing mechanism. ## One thing to take with you What this episode actually taught me: when your eyes fight the statistics, neither side is usually lying — your sample has a filter on it you didn't notice. Those people in Austin, the videos in your feed, the handful of friends around you who are doing fine, were all picked by some rule and delivered to you, and you've never seen the rule. Here's something I've tried, and I'd invite you to try it. Pick one belief of the form "everyone around me is doing X" — everyone upgraded their house, everyone's anxious about work, everyone's kid is in tutoring. On paper, write down the three to five people your "everyone" actually refers to, then one line each on how they entered your life: same company, same school, same community, same algorithm. When you're done you'll be looking at the shape of your filter. If you want to go one step further, find someone who isn't on that list and ask them the same question.