Twentieth-century dreams, on a long enough timeline
Busted Bundles ft. remixes and samples from the cultural phenomenon "You did what you were Told, but the Deal didn't Hold"
I’ve been involved in many conversations1 of late about the cluster of arrangements that the twentieth century treated as the shape of normal adult life, and how many of them are quietly turning out to have been long-term illusions.2
Mark Fisher’s hauntology work is one of the more popular treatments of why this feels the way it does from the inside, the sense of futures that were promised and then withdrawn, lingering as a kind of cultural background radiation.3 Fisher, borrowing the phrase from Franco “Bifo” Berardi, called it the slow cancellation of the future, and located its onset in the cultural expectations “fabricated during the long period of modern civilization, reaching a peak after the Second World War.”
Several recent pieces, including Jasmine Sun Jasmine Sun’s expansion notes on her Times essay on AI and the labor question, have been adding fresh material to that thread. The throwaway line I want to develop here is one I left in response to a detail in hers, but the thinking was already running.
"Ah, one of many 20th century ‘dreams’ shown to be long term illusions. "
The dream in question is not “living with your parents is bad” or “career stability was a lie.” The dream is The Bundle. For maybe sixty years across (parts of) the industrialized world, a particular bundle was treated as the shape of normal adult life: you would leave the household you grew up in, find paid work outside the home, that work would pay enough to form a new household, the household would form on a roughly predictable schedule, the work would last long enough to fund retirement, and the institutions around all of this (employers, banks, schools, neighborhoods) would hold steady enough to make the sequence legible to the person inside it.
None of those elements is unreasonable on its own. The dream was that they came as a package, and that the package was durable. The package was the thing.
It is worth being precise about one part of this, because it is where most of the lived damage actually sits. The Bundle was not merely available; it was sold as permanent. The people who came of age inside the twentieth-century arrangement were taught, in the explicit curriculum of schools and the implicit curriculum of everything else, that The Bundle was the future and would continue to be the future.4
Go to college, the deal said, and the wage premium will pay for the cost of going. Work hard, and productivity gains will accrue to your compensation. Buy a house when you can, and the asset will appreciate. The trend lines will hold. The infrastructure will hold. The energy will hold. You can plan a life around the assumption that the world your parents lived in is the world you will live in, plus some upgrades.
This is what I mean, in shorthand, when I talk about this moment catching folks generationally underprepared: not that the people in it lacked skills, but that they were specifically developed to expect a future that did not arrive, in a register that made the non-arrival feel like personal failure rather than legible structural change.
The wage-productivity gap is one form of this. According to the Economic Policy Institute’s ongoing analysis, from the end of the Second World War until 1979 productivity and typical worker pay grew in lockstep; since 1979, productivity has grown roughly 2.7 times as much as the compensation of the typical worker. That gap is almost half a century old now and was once treated as a temporary anomaly by people who expected the post-war pattern to reassert itself. It did not reassert itself. College costs are another form.
Since 1980, the BLS index for college tuition and fees has risen roughly 1,200% while the overall CPI has risen 236%, a multi-decade compounding gap, financed almost entirely on the assumption that the wage premium of a degree would justify the price. The premium is not reliably there for the cohort paying it, and the financing structure still assumes it is. The cycles of resource panic are another form.
The 2000s carried genuine fear that the United States was past peak oil, until shale and horizontal drilling reversed the curve and U.S. production rose past its 1970 high; then hydrofracking was treated as either civilizational salvation or environmental catastrophe depending on the year, with the framing flipping every few years.5
These are not failures of forecasting, necessarily. They are signs of how unstable the underlying conditions actually were, and how much weight was being put on the assumption that the trend lines would hold.
What we are watching, in slow motion — or, in Douglas Rushkoff’s Present Shock sense, far too rapidly to process at all — is the package coming apart. Not in a single dramatic failure but in the quiet way packages come apart: each component still exists, but they no longer arrive together, in that order, on that schedule, for the people who were told to expect them.
Full-time children, the Chinese phenomenon Jasmine writes about, is one shape this takes. So is the gig economy. So is the rise of one-person companies. So is the trend of household formation slipping later by years, then decades. So is the particular flavor of professional drift that people in their thirties describe when they say they don’t know what they are doing anymore.
The reason to call these illusions rather than failures is that The Bundle was never a law of nature. It was the surface impression of a specific set of post-war conditions: cheap energy, demographic expansion, a particular configuration of capital and labor, an unusual degree of institutional stability, and a set of expectations that ran on the assumption that all of the above would continue compounding. The conditions changed. The Bundle dissolved. The expectations are still installed in most of us, which is part of what makes the dissolution disorienting rather than legible.
Even the trends that did hold for decades are starting to show their conditions. Moore’s Law is the cleanest example. The transistor count doubling every couple of years has, in modified form, held for almost sixty years, and the popular version of it — that computing gets cheaper forever — became one of the load-bearing assumptions of the late twentieth century.
Memory prices in 2026 are the counter-text. DRAM contract prices roughly doubled in the first quarter of this year, driven by AI infrastructure demand. Micron is exiting the consumer side of its memory business to serve hyperscaler customers; its CFO has said the company is “sold out for 2026” with new fabs not coming online until 2027 and 2028. IDC describes the situation as “the end of an era of cheap, abundant memory and storage.”
The popular trend held for as long as the conditions held, and a single large new demand source — AI infrastructure — was enough to reverse it inside a few quarters. The pattern was real. The pattern was also conditional. The conditions were doing more of the work than most of us understood.
The same pattern is being made in present-tense planning, not just retrospectively in old assumptions. Anton Leicht has a recent piece on the assumption — common in AI policy thinking outside Silicon Valley — that frontier-AI tokens will soon be abundant and broadly available, and that whole national strategies are being quietly built on top of that assumption. His argument is that the assumption is already breaking. The specifics of his case are about cybersecurity, compute, and U.S. government leverage; the structural move is the one I keep pointing at. A popular trend is being treated as a durable condition, and the people planning around it have not yet noticed how conditional the trend actually is.
There is real work to be done in noticing this clearly, separate from grieving it or celebrating it or shrugging at it.6 The grief response says the package should be restored. The celebration response says it was always confining and good riddance. The shrug response says nothing was ever guaranteed and you should have known. None of those responses helps you read what is actually happening, which is that a load-bearing arrangement is being replaced, in pieces, by something whose shape is not yet settled.
What is settling into place instead is partly visible in the symptoms: more relational economy, more informal work, more multi-generational household structures, more dependence on platforms for the connective tissue that used to come from employers or neighborhoods or extended family. Some of this is genuinely new. Some of it is older than The Bundle was. Multi-generational households were the norm for most of human history, before the post-war detached-house decades. The point is not whether the new shape is good or bad. The point is that it is real, it is forming whether we attend to it or not, and the people who will be best positioned in it are the ones who can read it clearly while it is still forming.
This is much of what I mean when I write about literacy in the body of work I’m developing. The arrangements we inherited were legible because they were stable enough to be named.7 The arrangements we are entering are illegible mostly because they have not stabilized yet, not because they are inherently obscure.8 The work of the next few years, for anyone trying to build inside them, is partly the work of seeing them sooner than the language for them arrives.
There are a lot of twentieth-century dreams. Most of them were dreams about how The Bundle was supposed to hold, and a few of them were dreams about how the trend lines would hold. The longer timeline is showing us, one piece at a time, that neither did.
A note of reference, which also impacted the subtitle for this piece, is yesterdays ekkolápto salon about computational meaning-making, language, and rap. Some of the finer points I may return to have to do with the ways that media is able to reference and ‘converse’ or add-on or remix itself, and how different forms of media, art, or expression are empowered or constrained to do this in different ways. See: YouTube, related Substack post.
The phrase in the piece’s title is from Chuck Palahniuk’s Fight Club (1996): “On a long enough timeline, the survival rate for everyone drops to zero.” It’s become a piece of internet-discourse shorthand for “given enough time, the pattern reveals itself.” I’m using it here to signal retrospective distance — we’re now far enough from the post-war period to see which of its arrangements were load-bearing and which were artifacts of temporary conditions.
Also consider the many variations of phrases like “20th century culture on high-speed internet”, and “Capitalist Realism”, perhaps. But in many of my writings in this project, I’m particularly disinterested in isolating capitalism as the Big Bad of it all. It feels like suggesting that if we go back in time and remove Hitler, we’d never have to deal with fascism. We need to get in the business of identifying viable alternatives, which I am aiming at, even if yes, I’m starting with common lament formats.
Not only were they taught, but those doing the teachings had tangible success (footnote 6); it was training and development reinforced by concrete realities, at the time. Even the great depression was not to get in the way of the technological and economic progress of the post-war world. But that compared to the cyclical “generational” set backs that the Millennials have experienced, juxtaposed to the promise and optimism of the 90s — no less the Gen Z / A lacking of such a boom-time — and we have a quite the difference of context. This is partly why elders’ advice can be incredibly difficult to translate well to youths.
One of my first “analysis and consulting” gigs was about energy, and, at the time (earlier 2010s), it many in the community were contending with a dominant narrative of continued economic and population growth. That trend and narrative has faced much more scrutiny in this past decade, while the broader topic of environmentalism and sustainability has centered (in the USA).
In the linked piece, I talk about what it’s like observing various folks go through these changes of faith, perspective, vision, or hope. There are also some JOPRO projects aiming to document elements of such experiences.
The aggregate version of this promise has a clean empirical signature. According to research by Raj Chetty and colleagues, about 90% of children born in 1940 grew up to earn more than their parents in inflation-adjusted terms; for children born in 1984, the figure is about 50%. The headline promise of The Bundle — that each generation would, on average, do better than the last — has been quietly halving over half a century. Chetty’s piece is from 2016, before any AI/LLM waves broke as they did in the 2020s.
"The old world is dying, and the new world struggles to be born: now is the time of monsters”, perhaps more true interpretation of its clinical approach: "The crisis consists precisely in the fact that the old is dying and the new cannot be born; in this interregnum a great variety of morbid phenomena appear”, from Antonio Gramsci. I would add that, some of the mutations will be what allows us to adapt, and some won’t; the messy middle is still the actual terrain we all have to walk. How biological!