A movement of technologists is predicting that by the early 2030s, AI and robotics collapse the cost of a millionaire's lifestyle to near-zero for the majority of humanity. The economics of demonetization are the easy part. The part they skip is what happens to status, power, and incumbents once the cost curve actually bends.
Peter Diamandis's August 16, 2026 newsletter argues that by 2031, four converging technology curves (AI tutors, humanoid robots, robotaxis, AI diagnostics) collapse the cost of a $1 million household's lifestyle to near-zero, and "the millionaire lifestyle doesn't trickle down. It collapses down."
The demonetization math might even be roughly right. What's missing is everything downstream of the price collapse: whether status stops being relative once everyone has the goods, whether the industries currently gatekeeping those goods (medicine, education, insurance) let go of that gate voluntarily, and whether owning the compute and robots that make things cheap is itself a new, more concentrated form of the power the theory claims disappears.
Diamandis runs Abundance360, co-founded XPRIZE, and has argued some version of this thesis since his 2012 book Abundance (with Steven Kotler). But the argument this newsletter makes — falling costs are the whole mechanism, and the mechanism runs on its own — shows up nearly word-for-word from two of the most influential people in AI.
"The millionaire lifestyle doesn't trickle down. It collapses down."
"We believe the measure of abundance is falling prices."
An "American Equity Fund" taxing high-value companies and land ~2.5%/yr, paid to every adult — his own estimate: ~$13,500/year within roughly a decade.
That third card matters more than it looks. Altman is the most rigorous of the three on this specific question, and his own proposal concedes the point Diamandis skips: demonetized production does not redistribute itself. Altman's plan requires a coercive, deliberate transfer — a tax on companies and land, collected and redistributed by policy. If the field's own most detailed abundance proposal needs a government-enforced wealth transfer to close the loop, "it collapses down" on its own is doing more rhetorical work than economic work.
(Worth a separate flag so it isn't conflated: Abundance, the 2025 book by Ezra Klein and Derek Thompson, is a different movement entirely — a policy argument that U.S. scarcity in housing and infrastructure is self-inflicted by permitting and regulation, not a technology thesis. Same word, unrelated claim.)
Status is a positional good by definition — it's rank relative to others, not an absolute inventory of stuff. If a robot chef and an AI physician become universal, they stop functioning as status markers and become baseline, the way a flush toilet or a smartphone already has. Diamandis's own closing line — "they will be kings, everyone will be kings" — is self-refuting under his own logic. A crown that everyone wears isn't a crown.
Will Storr's The Status Game (William Collins, 2021) makes the underlying case: status-seeking is an evolved, universal human drive that operates regardless of material conditions. It doesn't require scarcity to exist; it just needs something to rank people on. Remove scarcity in healthcare, education, and transport, and status competition doesn't end — it relocates to whatever's still scarce: land, attention, provably-human-made goods, access to actual humans instead of AI proxies. The abundance-movement writers describe the collapse of one hierarchy and quietly assume no new one replaces it.
Our own engine has fields for exactly the force this theory leaves out: laborResistance, consumerResistance, regDragProfile, pathLock. They exist because the people currently positioned at the top of a scarcity hierarchy have every incentive to slow the collapse of that hierarchy, and the tools to do it.
Concierge medicine's entire value proposition to a $1M household is that it's not universal. AMA-style credentialing, insurance underwriting, and state licensing boards are captured by the current winners of that scarcity regime — which is a documented reason our own model shows medicine and finance lagging hardest on AI-driven cost collapse, even where the underlying capability is ready. Diamandis's advice to policymakers — "accelerate the demonetization" — treats regulatory friction as an oversight to be corrected, not as the rational, self-interested resistance of people who lose status and income if it succeeds.
Even where demonetization genuinely happens, it doesn't obviously distribute power — it usually just relocates the toll booth. Online content collapsed to free; ownership of the ad-tech chokepoint that monetizes attention around that free content consolidated into a handful of platforms. Cheap computing didn't distribute chip fabrication; it concentrated it into two or three companies capable of the capital expenditure required to compete.
If humanoid robots and AI diagnostics get cheap the same way, the compute, the robot fleets, and the land under the solar farms powering them don't distribute by default — they consolidate onto whichever balance sheets can absorb that capex, the same pattern our engine tracks under customerConcentrationRisk, dataMoatDurability, and hyperscalerCapexBeta. A falling consumer price and a concentrating ownership structure aren't in tension. They're the normal shape of a demonetization cycle.
The abundance argument assumes that once everyone has resources, the behavior that scarcity produces — status anxiety, self-interest, entitlement — fades with it. The research on what wealth and social class actually do to behavior points the other way.
A 2012 study in the Proceedings of the National Academy of Sciences found that higher social class predicted increased unethical behavior across several independent experiments — including a higher rate of failing to yield to pedestrians at a crosswalk while driving, and, in a lab study, taking significantly more candy from a jar explicitly described as reserved for children when primed to feel upper-class (roughly double the amount taken by those not primed that way).
Piff, P. K., Stancato, D. M., Côté, S., Mendoza-Denton, R., & Keltner, D. (2012). Higher social class predicts increased unethical behavior. PNAS, 109(11), 4086–4091.
None of this says wealth makes people malicious. It says the correlation runs opposite to what the abundance thesis needs: more resources tracking with more self-interested behavior, not less, which is a problem for a theory whose entire promise is that giving everyone more resources makes hierarchy and self-interest fade on their own.
There's a second, subtler problem: getting the goods may not even deliver the satisfaction the theory is selling. The classic 1978 study comparing recent lottery winners, accident victims, and ordinary controls found winners' rated present happiness (4.00 on a 5-point scale) wasn't meaningfully higher than controls (3.82) — and, more tellingly, winners reported taking less pleasure from small everyday moments than either group, a contrast effect from having recently experienced something extraordinary. Getting the millionaire lifestyle for free doesn't obviously make anyone happier than they already are; it may just reset what counts as ordinary. (Brickman, Coates & Janoff-Bulman, Journal of Personality and Social Psychology, 1978.)
Scope it down first. "The majority of humanity" isn't really what's on the table in any of these theories. The $1M-household inventory Diamandis describes — concierge medicine, private school, a personal chef, a Tesla refreshed every 2-3 years — is a Western consumption pattern being demonetized for Western consumers. Even in the optimistic case, the curve is calibrated to markets that already have the electrical grid, the regulatory infrastructure, and the purchasing power to receive it first. Call it what it actually is: an uplift for the West, not for humanity.
Inside that narrower, more honest scope, there are two paths that would genuinely move the needle. Neither is permanent. Both come with a catch.
The fantasy version: some combination of neural intervention or low-addiction, low-side-effect pharmacology that directly lowers the fear response — fear of illness, of homelessness, of hunger, of pain — while preserving or improving focus and productivity. This targets the thing abundance theory gestures at but never actually names: most of what makes scarcity feel unbearable isn't the absence of goods, it's the anxiety of not having them. Relief from that anxiety, delivered at population scale, would be a bigger uplift than any amount of cheap robotics.
The realistic version: social change — a broadly applied safety net around healthcare, housing, and food security — possibly assisted by AI used collectively rather than as a private wealth-generation tool. Not AI making a $1M household's private chef cheaper. AI applied at the level of public systems: triage, benefits administration, disaster response, resource allocation. That's a different application of the same technology than anything in Diamandis's, Andreessen's, or Altman's essays — and it's the one that would actually touch the anxiety layer instead of the inventory layer.
Both are, at best, temporary relief, not a permanent state — the hedonic-adaptation research above applies here too. Whatever baseline of security a society reaches, people adjust to it and start worrying about the next thing down the list. And both run into the same wall the rest of this piece keeps hitting: neither gets built without the people who currently hold the levers — legislators, regulators, and the handful of AI/compute owners with the capital to deploy either one at scale — choosing to build it that way. Nothing about either path is inevitable, self-executing, or price-driven the way Diamandis's curve assumes. It's a decision, made by people who are not currently incentivized to make it.
The honest question isn't whether Diamandis, Andreessen, and Altman are lying. It's whether they're rationalizing — genuinely convincing themselves, not just their readers — that the harder problems (status competition, regulatory capture, ownership concentration) don't need addressing because the technology alone resolves them.
That's not resolvable from outside anyone's head. What is checkable is the incentive structure around the belief. All three run ventures whose value proposition depends on this exact narrative being true and unchallenged: Diamandis's paid communities and longevity ventures, Andreessen's venture fund betting on the AI buildout, Altman's company selling the AI that's supposed to do the collapsing. None of that makes the underlying claims false. It does mean the people most publicly certain that no deliberate redistribution or regulation is needed are also the people whose business models benefit most from nobody insisting on it.
Upton Sinclair put the mechanism plainly nearly a century ago: "It is difficult to get a man to understand something, when his salary depends on his not understanding it." (I, Candidate for Governor: And How I Got Licked, 1935 — often misattributed to H.L. Mencken, which is itself a reminder to check a quote before repeating it.)
This isn't an argument that AI won't get radically cheaper, or that living standards won't rise. Both are already happening and our own model prices that in. It's an argument against the specific claim that cost collapse alone resolves status, incumbency, and ownership concentration — three forces our engine models explicitly because history says they don't resolve on their own. Betting a portfolio on "abundance arrives and distributes itself" skips exactly the mechanics that determine which companies and industries capture the value when a cost curve bends: the ones controlling the compute, the data, and the regulatory relationship, not the population at large.
Our 30-industry AI-effects matrix scores every industry on exactly the forces this theory leaves out — regulatory drag, competitive dynamics, customer and supplier concentration, and adoption resistance — recalibrated continuously since February 2026, with every parameter change logged in public.
See the 30-industry matrixPart of the same honesty-first family as Hype vs. Reality and The Human Bottleneck — this page is the human-nature layer underneath both.