AI Labor Tracker / Foundations
The sci-fi scorecard: what the futurists got right about work
We graded a century of predictions about machines and work, from Bellamy and Forster to the Jetsons and Her, against what the data shows in 2026.
29 hours
Average weekly hours per UK worker in 2023, against Keynes's predicted 15
The writers who imagined the machine age got the gadgets mostly right and the work mostly wrong. They kept predicting we’d work less. We kept finding more to do.
That’s the short version of the scorecard below. The longer version is more interesting, because the misses tell you as much as the hits.
Here’s how this works. I picked ten of the best-known predictions about technology and work, from an 1888 utopian novel to a 2013 film. For each one I took what the author or show actually said, checked it against data, and gave it a grade.
Grades are mine. The data isn’t.
One ground rule: I grade the prediction about work and daily life, not the plot.
The scorecard
| Work | Year | The prediction | What happened | Grade |
|---|---|---|---|---|
| Looking Backward, Edward Bellamy | 1888 | Everyone works from 21 to 45, then retires; people pay with a “credit card” | Credit cards are everywhere; full Social Security retirement age is 67 | C+ |
| “The Machine Stops,” E. M. Forster | 1909 | People live alone in cells, talk through screens, fear direct experience | In-person time with friends fell by two-thirds, 2003 to 2020 | B+ |
| “Economic Possibilities for our Grandchildren,” J. M. Keynes | 1930 | Living standards 4 to 8 times higher in 100 years; a 15-hour week | Living standards up 4.4 to 5.5 times; the week is about twice as long | B- |
| Player Piano, Kurt Vonnegut | 1952 | Automated factories need few workers; a managerial elite runs things | Factory output up, factory jobs down sharply; mass idleness didn’t follow | B+ |
| The Jetsons | 1962-63 | Two hours of work a day counts as a hard day | Private-sector workweek averages about 34 hours | D |
| Isaac Asimov, NYT | 1964 | Machines do routine work; boredom spreads; “work” becomes the most glorious word | Routine work automated; people still work; population forecast too low | B |
| Arthur C. Clarke, BBC | 1964 | Run a business from Tahiti as well as from London | About 35% of employed people work from home on days they work | A- |
| Star Trek: First Contact | 1996 | Money no longer exists; people work to better themselves | 811 million people still in extreme poverty; top 1% hold a third of US wealth | D |
| WALL-E | 2008 | Machines do everything; humans drift in hover chairs | Adult obesity hit 40%, now edging down with new drugs | C |
| Her | 2013 | People form emotional relationships with an AI | Nearly 3 in 4 US teens have tried AI companions | A |
Table 1. Scorecard of predictions about technology and work. Grades are the author’s. Sources for each row are cited in the sections below.
Now the details.
Looking Backward (1888): C+
Edward Bellamy’s hero falls asleep in 1887 and wakes up in Boston in the year 2000. Everyone serves in an “industrial army.” “The period of industrial service is twenty-four years, beginning at the close of the course of education at twenty-one and terminating at forty-five,” Dr. Leete explains in Chapter 6 of the Project Gutenberg edition.
Retirement at 45 didn’t happen. For anyone born in 1960 or later, full Social Security retirement age is 67, according to the Social Security Administration.
But Bellamy deserves real credit for one detail. In Chapter 9, citizens shop with “a credit card,” a piece of pasteboard loaded with their share of the national product. He used the term generations before anyone carried one.
The plastic part he nailed. The equal share of national income, not so much. Grade: C+.
The Machine Stops (1909): B+
This is the most unsettling story on the list, because it reads like a description of a phone.
E. M. Forster’s Vashti lives alone in a small hexagonal room underground, connected to everyone through screens and speaking tubes. “She knew several thousand people; in certain directions human intercourse had advanced enormously,” Forster writes in the story, available in The Eternal Moment collection. When she has to leave her room, she is “seized with the terrors of direct experience.”
The data is uncomfortable. Americans spent about 60 minutes a day with friends in person in 2003 and about 20 minutes in 2020, according to the Surgeon General’s 2023 advisory on loneliness, which drew on the American Time Use Survey. The 2020 figure is distorted by the pandemic, but the decline was well underway before it.
Forster also imagined people turning to the Machine for everything. About 61% of working-age Americans used generative AI in the third quarter of 2026, according to the Real-Time Population Survey published by the St. Louis Fed. But Forster’s people are sealed off from the physical world, and ours still commute, travel and crowd into stadiums, and the 2020 low owes a lot to the pandemic.
The mood is right; the extremity isn’t. Grade: B+.
Keynes (1930): B-
John Maynard Keynes made two predictions in his famous essay, and they split cleanly.
First, living standards. “I would predict that the standard of life in progressive countries one hundred years hence will be between four and eight times as high as it is to-day,” he wrote in “Economic Possibilities for our Grandchildren”.
Real GDP per person in the US was about 5.5 times its 1930 level by 2022, and in Keynes’s Britain about 4.4 times, according to the Maddison Project data compiled by Our World in Data. That’s inside his range, eight years early.
Keynes nailed the money
Multiple of 1930 real GDP per capita
Bar chart. Use the up and down arrow keys to move through 4 categories. A data table follows.
United States, GDP per capita 2022 vs 1930: Multiple of 1930 5.47
Figure 1. Real GDP per capita in 2022 as a multiple of 1930, US and UK, against Keynes's predicted range for 2030.
Source: Maddison Project Database 2023 via Our World in Data, accessed Nov 1, 2026; Keynes, 1930 Download CSV Methodology
Data table
| series | Multiple of 1930 |
|---|---|
| United States, GDP per capita 2022 vs 1930 | 5.47 |
| United Kingdom, GDP per capita 2022 vs 1930 | 4.43 |
| Keynes's prediction, low end | 4 |
| Keynes's prediction, high end | 8 |
Second, the work. “Three-hour shifts or a fifteen-hour week may put off the problem for a great while,” he wrote. That’s where it falls apart.
The average UK worker put in about 1,523 hours in 2023, roughly 29 hours a week spread across the year. The average American worked about 1,789 hours, around 34 a week, per Our World in Data. BLS payroll data puts the average private-sector workweek at 34.4 hours in September 2026, according to FRED.
The 15-hour week never showed up
Hours worked per worker per year
Line chart. Use the left and right arrow keys to move through 6 periods. A data table follows.
2023: United States 1,789, United Kingdom 1,523, 15-hour week 780
Figure 2. Average annual hours worked per worker, US and UK, selected years, against a 15-hour week (15 x 52 = 780 hours).
Source: Huberman and Minns (2007) and Penn World Table via Our World in Data, accessed Nov 1, 2026; reference line is our calculation Download CSV Methodology
Data table
| year | United States | United Kingdom | 15-hour week |
|---|---|---|---|
| 1929 | 2,316 | 2,257 | 780 |
| 1950 | 2,022 | 2,023 | 780 |
| 1980 | 1,840 | 1,654 | 780 |
| 2000 | 1,888 | 1,568 | 780 |
| 2019 | 1,797 | 1,545 | 780 |
| 2023 | 1,789 | 1,523 | 780 |
Hours did fall by about a third in Britain since 1929. Keynes’s number implied a cut of about two-thirds.
We took most of the gains as stuff, not leisure. Grade: A on wealth, D on hours, B- overall.
Player Piano (1952): B+
Kurt Vonnegut’s first novel imagines a society where machines run the factories and engineers run everything else. Ordinary people, the “reeks and wrecks,” are left without much purpose, and the hero, engineer Paul Proteus, eventually joins a revolt, as Kirkus Reviews summarized it when the book came out in 1952. Kirkus called it “a tough, tightly written book of a brave new world to come.”
The factory half of the prediction came true. US manufacturing employment peaked at 19.6 million in June 1979, per BLS data on FRED.
It stood at 12.7 million in September 2026. Since 1987, real manufacturing output rose about 63% while factory employment fell about 28%, according to BLS output data.
More stuff, fewer people
Index, January 1987 / Q1 1987 = 100
Line chart. Use the left and right arrow keys to move through 5 periods. A data table follows.
2026-04: Employment 72.1, Real output 163.3
Figure 3. US manufacturing employment and real manufacturing output, indexed to 1987 = 100. Output is quarterly; employment uses the first month of each quarter shown. Indexing is our calculation.
Source: U.S. Bureau of Labor Statistics via FRED, series MANEMP and OUTMS, accessed Nov 1, 2026 Download CSV Methodology
Data table
| date | Employment | Real output |
|---|---|---|
| 1987-01 | 100 | 100 |
| 2000-01 | 99 | 157.3 |
| 2007-01 | 80.2 | 168.7 |
| 2019-01 | 73.2 | 164.7 |
| 2026-04 | 72.1 | 163.3 |
What Vonnegut missed is that displaced workers mostly found other work. Unemployment was 4.2% in September, per BLS. The jobs moved to services, and not all of them were good ones.
But the mass idleness of Ilium didn’t arrive. Grade: B+, and it might get upgraded if AI does to offices what robots did to factories.
The Jetsons (1962-63): D
“Yesterday, I worked two full hours!” George Jetson complains in the episode “G.I. Jetson,” as Matt Novak documented for Smithsonian. His wife asks if his boss is running a sweatshop.
Novak notes the push-button future of leisure “was a mainstream assumption made by even the most conservative of prognosticators.” It was also wrong. Average weekly hours in private payrolls sit at 34.4, per FRED.
To be fair, the show was more realistic about money. George lives near paycheck to paycheck while a billionaire neighbor lives in a mansion, as Novak points out in another piece.
On inequality, the cartoon beat Star Trek. Grade: D, with partial credit for the billionaire.
Asimov’s World’s Fair (1964): B
Writing in The New York Times in 1964, Isaac Asimov imagined the world of 2014. He predicted that people would need to be trained as “machine tenders,” that high schoolers would learn computer languages, and that “mankind will suffer badly from the disease of boredom,” in his essay “Visit to the World’s Fair of 2014.”
His most quoted line: “In a society of enforced leisure, the most glorious single word in the vocabulary will have become work!”
The computer education part looks good. The enforced leisure part didn’t happen, at least not yet. And his population number was too low.
He expected 6.5 billion people in 2014. The actual figure was about 7.35 billion, per the World Bank.
But his core worry is the one this publication exists to track. If machines do the routine work, what gives people’s lives structure and dignity?
That question is more relevant in 2026 than it was in 1964. Grade: B.
Arthur C. Clarke (1964): A-
In the same year, Arthur C. Clarke told the BBC that “it will be possible, perhaps only 50 years from now, for a man to conduct his business from Tahiti or Bali just as well as he could from London,” as Fortune recounted. “Men will no longer commute. They will communicate.”
On days they worked, 35% of employed Americans did some or all of their work at home in 2025, according to the American Time Use Survey. People still commute, and cities did not empty out.
But the idea that knowledge work could be independent of distance is now ordinary. Grade: A-.
Star Trek: First Contact (1996): D
“The economics of the future are somewhat different,” Captain Picard says in the film. “The acquisition of wealth is no longer the driving force in our lives. We work to better ourselves and the rest of humanity,” according to a fan transcript of the film.
We are a long way from that. About 811 million people still live in extreme poverty, roughly one in ten, according to the World Bank’s 2026 Poverty, Prosperity and Planet report. In the US, the top 1% held 32.5% of household net worth in the second quarter of 2026, up from 22.8% in 1989, per the Federal Reserve’s distributional accounts.
Star Trek was always more aspiration than forecast, and it is set three centuries out. But the trend on wealth concentration is moving the wrong way. Grade: D, with time left on the clock.
WALL-E (2008): C
Pixar’s humans have spent generations on a cruise ship where machines do everything. Roger Ebert described them as obese “Hoverchair” people gliding around in comfy chairs, in his 2008 review of the film.
The obesity part got uncomfortably close. In 2021-23, 40.3% of US adults had obesity, according to the CDC. But here is a twist no screenwriter saw coming.
Self-reported obesity fell to 37.0% in 2025 from 39.9% in 2022, as the use of GLP-1 drugs for weight loss rose, according to Gallup. Technology made part of the problem and is now helping fix it. Grade: C, and the hover chair is still on back order.
Her (2013): A
In Spike Jonze’s film, a lonely writer falls in love with his operating system. In one of the film’s most memorable scenes, she reveals she is talking to 8,316 others at the same moment and is in love with 641 of them, according to the film’s quotes on IMDb.
Thirteen years later, 72% of US teens have used an AI companion and 52% use one at least a few times a month, according to a Common Sense Media survey. In a four-week randomized study of 981 people by MIT Media Lab and OpenAI, participants who used a chatbot more each day reported more loneliness and emotional dependence, and less socializing with people, according to the researchers. The study could not show that the chatbot caused the loneliness.
Her got both the product and the ambivalence right. Grade: A.
What the misses have in common
Look at the bad grades and a pattern jumps out. The Jetsons, Keynes on hours, Bellamy on retirement, Star Trek on money. They all assumed that when machines made us richer, we’d take the gains as free time and the end of scarcity.
That’s not what people did. We took the gains as more stuff, better health care, bigger houses and new kinds of work nobody had imagined.
Work kept its hold on us, partly for the money and partly because it gives people identity and a reason to get up. Asimov understood that better than anyone on this list.
The hits share a pattern too. Forster, Clarke and Jonze got the texture of daily life right: the screens, the distance, the strange intimacy with machines.
Those are predictions about human behavior, not about economics. Put simply, the futurists were better psychologists than economists.
That should make us humble about today’s predictions, including the ones on this site. We ran a similar test on the claim that AI is bigger than the Industrial Revolution in an earlier piece. The people forecasting either a jobless utopia or mass unemployment from AI are making the same kind of call Keynes made, and he was one of the smartest economists who ever lived.
How to read the next forecast
Read the old futurists, but read them for the questions, not the answers. Asimov’s question about dignity when machines do the routine work is the right one for 2026. Forster’s warning about isolation is worth taking seriously every time you pick up your phone.
And when someone tells you exactly how AI will change work by 2035, check their track record against this table. If they have never been wrong before, they have probably never made a forecast. Bet on adaptation, plan for disruption, and put your training hours into the skills machines can’t copy.
Sources
- Bellamy, E., Looking Backward: 2000-1887, Houghton Mifflin, 1888. Project Gutenberg eBook #624. https://www.gutenberg.org/ebooks/624
- Forster, E. M., “The Machine Stops” (1909), in The Eternal Moment and Other Stories, Project Gutenberg eBook #72890. https://www.gutenberg.org/ebooks/72890
- Keynes, J. M., “Economic Possibilities for our Grandchildren,” 1930. Marxists Internet Archive. https://www.marxists.org/reference/subject/economics/keynes/1930/our-grandchildren.htm
- “Player Piano” (review), Kirkus Reviews, Aug 1, 1952 issue. https://www.kirkusreviews.com/book-reviews/kurt-vonnegut/player-piano/
- Novak, M., “Automating Hard or Hardly Automating? George Jetson and the Manual Labor of Tomorrow,” Smithsonian Magazine, Feb 19, 2013. https://www.smithsonianmag.com/history/automating-hard-or-hardly-automating-george-jetson-and-the-manual-labor-of-tomorrow-20694353/
- Novak, M., “The Jetsons and the Future of the Middle Class,” Smithsonian Magazine, Jan 14, 2013. https://www.smithsonianmag.com/history/the-jetsons-and-the-future-of-the-middle-class-1397497/
- Asimov, I., “Visit to the World’s Fair of 2014,” The New York Times, Aug 16, 1964. https://archive.nytimes.com/www.nytimes.com/books/97/03/23/lifetimes/asi-v-fair.html
- “Arthur C. Clarke, ‘2001: A Space Odyssey’ author, correctly predicted modern living 60 years ago,” Fortune, May 29, 2024. https://fortune.com/2024/05/29/arthur-c-clarke-space-odyssey-remote-work-ai-predictions-bbc/
- “Star Trek: First Contact” transcript, Chakoteya.net. https://www.chakoteya.net/movies/movie8.html
- Ebert, R., “WALL-E” (review), RogerEbert.com, 2008. https://www.rogerebert.com/reviews/wall-e-2008
- “Her (2013) quotes,” IMDb. https://www.imdb.com/title/tt1798709/quotes/
- “Annual working hours per worker,” Our World in Data, accessed Nov 1, 2026. https://ourworldindata.org/grapher/annual-working-hours-per-worker
- “GDP per capita” (Maddison Project Database 2023), Our World in Data, accessed Nov 1, 2026. https://ourworldindata.org/grapher/gdp-per-capita-maddison-project-database
- “Average Weekly Hours of All Employees, Total Private (AWHAETP),” BLS via FRED, accessed Nov 1, 2026. https://fred.stlouisfed.org/series/AWHAETP
- “All Employees, Manufacturing (MANEMP),” BLS via FRED, accessed Nov 1, 2026. https://fred.stlouisfed.org/series/MANEMP
- “Manufacturing Sector: Real Sectoral Output for All Workers (OUTMS),” BLS via FRED, accessed Nov 1, 2026. https://fred.stlouisfed.org/series/OUTMS
- “The Employment Situation, September 2026,” U.S. Bureau of Labor Statistics, Oct 2026. https://www.bls.gov/news.release/empsit.nr0.htm
- “American Time Use Survey, 2025 Results,” U.S. Bureau of Labor Statistics, Jun 25, 2026. https://www.bls.gov/news.release/atus.htm
- “Our Epidemic of Loneliness and Isolation,” Office of the U.S. Surgeon General, 2023. https://www.hhs.gov/sites/default/files/surgeon-general-social-connection-advisory.pdf
- “Generative AI Usage Share, All (RPSGENAIUSAGESHAREALL),” Real-Time Population Survey via FRED, accessed Nov 1, 2026. https://fred.stlouisfed.org/series/RPSGENAIUSAGESHAREALL
- “Starting Your Retirement Benefits Early,” Social Security Administration, accessed Nov 1, 2026. https://www.ssa.gov/benefits/retirement/planner/agereduction.html
- “Population, total (SP.POP.TOTL),” World Bank, accessed Nov 1, 2026. https://data.worldbank.org/indicator/SP.POP.TOTL
- “Poverty, Prosperity, and Planet Report 2026,” World Bank, 2026. https://www.worldbank.org/en/publication/poverty-prosperity-and-planet
- “Share of Net Worth Held by the Top 1% (WFRBST01134),” Federal Reserve Board via FRED, accessed Nov 1, 2026. https://fred.stlouisfed.org/series/WFRBST01134
- “Obesity and Severe Obesity Prevalence in Adults: United States, August 2021-August 2023,” NCHS Data Brief No. 508, CDC, Sep 2024. https://www.cdc.gov/nchs/products/databriefs/db508.htm
- “Obesity Rate Declining in U.S.,” Gallup, 2025. https://news.gallup.com/poll/696599/obesity-rate-declining.aspx
- “Nearly 3 in 4 Teens Have Used AI Companions, New National Survey Finds,” Common Sense Media, Jul 2025. https://www.commonsensemedia.org/press-releases/nearly-3-in-4-teens-have-used-ai-companions-new-national-survey-finds
- Fang, C. M. et al., “How AI and Human Behaviors Shape Psychosocial Effects of Extended Chatbot Use: A Longitudinal Randomized Controlled Study,” MIT Media Lab and OpenAI, Mar 2025. https://arxiv.org/abs/2503.17473