AI may eliminate most jobs within a generation. We already know what happens when people get abundant free time. The answer is more complicated than either side wants to admit.
For most of human history, the question "what would you do if you didn't have to work?" was a parlor game. It's becoming a policy question.
AI is accelerating toward a point where large categories of human labor become optional. Not every job, and not all at once, but enough that entire societies will face a question they've never had to answer at scale: what do people do with their time when survival no longer requires selling it?
The optimists paint a picture of liberation. Freed from the grind, humans would pursue art, philosophy, exploration, community. The pessimists point to despair. Without work, people lose purpose, structure, identity. They drink more, sleep more, and do less.
Both camps are guessing. But we have data. We've already run versions of this experiment: retirement communities, trust-fund inheritors, universal basic incomeCash payments made to individuals with no strings attached and no work requirement. Dozens of pilot programs have been conducted worldwide since the 1960s, with the largest modern studies running from 2017-2024. pilots, and entire populations that crossed the threshold from subsistence to surplus. The results don't confirm either story. They reveal something more interesting.
Every year, roughly four million Americans retire. Each one gains about 2,500 hours of unstructured time.
The research on what happens next is extensive, and it doesn't tell a simple story. A 2022 study in the Journal of GerontologyStudy on retirement and life values using a 3-year panel tracking retirees' priorities before and after leaving work. Found that what people value shifts significantly after retirement, with relationships and personal growth rising in importance. tracked retirees over three years and found that their values shifted after leaving work. Relationships and personal growth rose in importance. Career achievement and status fell. The people didn't just fill time differently. They reorganized what mattered to them.
But the reorganization doesn't happen automatically. Research from the National Institutes of HealthPMC review of retirement and sense of purpose, analyzing multiple longitudinal studies. Found that retirees who maintained a sense of purpose had significantly better physical and mental health outcomes, while those who lost purpose after leaving work experienced measurable decline. shows that retirees who maintained a sense of purpose had significantly better health outcomes, both physical and mental. Those who lost purpose after leaving work experienced measurable decline. The difference between the two groups wasn't wealth, health, or education. It was whether they found something to organize their days around.
Sense of purpose and social connection outweigh income and former career status by wide margins. The data suggests that what you did for a living matters far less than whether you replaced it with something meaningful.
The pattern is clear: money helps, health helps, but neither predicts satisfaction the way purpose does. Retirees who volunteer, mentor, create, or care for others report well-being levels that match or exceed their working years. Retirees who fill the hours with passive entertainment tend to report the opposite.
Retirement is a preview. A small-scale version of the experiment AI might run on everyone.
Between 2017 and 2024, more than 150 guaranteed income pilot programs ran across the United States alone.
The programs varied in size, duration, and payment amount, but they all tested the same core question: when you give people cash with no strings attached, what do they do?
The critics predicted couches. That people would stop working, stop trying, stop caring. The boosters predicted liberation. That people would start businesses, go back to school, become better parents. The data, as usual, landed somewhere more nuanced.
The OpenResearch studySam Altman's OpenResearch ran one of the largest U.S. basic income experiments, giving 1,000 participants $1,000/month for three years (2020-2023). Published results in 2024 through the National Bureau of Economic Research. Found modest reductions in work hours and mixed effects on well-being., funded by Sam Altman, gave 1,000 people $1,000 per month for three years. The participants worked about 1.3 fewer hours per week on average. Some left jobs entirely. But the reason they left matters: many shifted to caregiving, enrolled in training programs, or started small businesses. They didn't stop working. They changed what they worked on.
The Minneapolis pilotMinneapolis Guaranteed Income pilot (2022-2023), administered by the Federal Reserve Bank of Minneapolis. Gave $500/month to 200 low-income residents. After one year, participants showed improved food security, financial stability, and psychological well-being. found that $500 monthly payments improved food security, financial stability, and psychological wellness within a year. Participants used the money to pay down debt, buy groceries, and stabilize housing. Kenya's GiveDirectly programGiveDirectly's long-term UBI study in rural Kenya, the largest in the world, provided monthly payments to over 20,000 people across 197 villages. Preliminary results showed recipients invested in assets, started businesses, and improved nutrition., the world's largest UBI study, found that recipients invested in livestock, started small enterprises, and improved their nutrition.
The pattern across all these programs: when basic needs are covered, people redirect their energy. The work migrates from what the market demands to what the person values.
There's a population that has lived without work requirements for generations. They're called trust-fund kids, and the data on their well-being is complicated.
Research on inherited wealth and life satisfaction shows a split. People who inherit enough to never need employment report higher rates of both purpose and purposelessness than the general population. The distribution is bimodal: some thrive spectacularly, and some struggle in ways that money can't solve.
The ones who thrive tend to share a trait. They found something to care about that wasn't handed to them. Philanthropy, art, entrepreneurship, scholarship. The specific domain varied. The common thread was that they chose it voluntarily, pursued it with discipline, and tied their identity to the craft rather than the outcome.
The ones who struggled tended to share a different trait. They optimized for comfort. Travel, leisure, consumption. Experiences that felt like freedom in the short term but produced a specific kind of restlessness over years. Researchers who study this population call it affluence-related malaiseA cluster of psychological symptoms observed in individuals who have no financial need to work, including chronic boredom, identity diffusion, difficulty with motivation, and a paradoxical sense of purposelessness despite having abundant resources and freedom.: the paradox of having everything and finding it insufficient.
The split isn't about intelligence or character. It's about whether the person interpreted freedom as "freedom from" or "freedom to." The first framing removes constraints. The second creates new ones, voluntarily. The data is unambiguous about which framing produces better long-term outcomes.
I was part of the Stockton SEED program. Five hundred dollars a month, no strings. My wife thought I'd finally fix the garage. My boss thought I'd quit. My mother thought I'd waste it.
What I did was start coaching little league again. I'd stopped when the second shift started because the games were always at 4 PM and I was always at the warehouse. With the extra money covering gas and groceries, I could drop to four days. Four days still paid the bills. The fifth day, I was at the field.
The kids didn't care where the money came from. They just knew I was there. That year I slept better than I had in a decade. Not because of the cash. Because on Fridays, twelve kids expected me to show up, and I did.
Across every dataset, the same finding repeats: free time is neutral. What you do with it determines everything.
Retirees who find purpose thrive. Retirees who don't, decline. UBI recipients who shift their work toward caregiving or craft report greater satisfaction. Those who simply work less without replacing the structure report little change. Inheritors who choose a discipline flourish. Those who optimize for comfort erode.
The variable is always the same: voluntary obligation. A commitment you chose, that others depend on, that you show up for even when you don't feel like it.
This has direct implications for the AI transition. If automation eliminates jobs but societies fail to create pathways for voluntary obligation, the result won't be a leisure paradise. It will be a purpose crisis at a scale we've never seen. But if we build those pathways, if we design institutions around what the data says people actually need, the transition could produce the opposite: a population freed to do the hardest, most meaningful work of their lives, with none of it forced.
The question was never "what would humans do without work?" The question is "what would humans do if they got to choose their work?" And the evidence, from every corner of this research, says: they'd choose harder work than the market ever asked of them. They'd coach kids, build gardens, learn instruments, care for aging parents, mentor strangers, and create things no one is paying for.
They'd work. They'd just finally get to pick what on.
Retirement data draws from longitudinal studies published in the Journal of Gerontology (2022 panel study on retirement and life values), the National Institutes of Health (sense of purpose and health outcomes review), and the Social Security Administration's behavioral research on retirement decisions. UBI pilot data comes from the OpenResearch/NBER study (2024), the Federal Reserve Bank of Minneapolis guaranteed income evaluation (2024), and GiveDirectly's long-term UBI study in Kenya. The Guaranteed Income Pilots Dashboard tracks over 150 U.S. programs. Research on inherited wealth and well-being draws from clinical psychology literature on affluence-related identity and purpose, including the work of Jessie O'Neill and Thayer Willis on the psychological effects of wealth. The Stockton Economic Empowerment Demonstration (SEED) ran from 2019-2021 under Mayor Michael Tubbs.
Humans don't need jobs. They need purpose. They need structure they chose, people who count on them, and something worth getting better at.
If AI gives us that freedom, the question is whether we build the scaffolding for people to find their way. The data says they will. They just need the invitation.
Meaning in the Age of AI
