Friday, October 23, 2026
Supersonic Tsunami
History Lesson

In 1900, 41% of American Workers Farmed. The Real Lesson Is How Long It Took

The shift off the farm was the largest job upheaval in American history, and it unfolded over a century. The AI wave is running on a much shorter clock, and that is what we have to plan for.

By Bryan· 6 min read
An old farm field with a hand plow giving way to power lines and a data center under construction as electricians work on a pole

If you want to understand how a job market can be rebuilt from the ground up, start with the farm. In 1900, 41 percent of American workers were employed in agriculture. By 2000, that share was 1.9 percent. Measured by the share of workers whose way of making a living disappeared, nothing else in American history comes close.

We rarely describe it as a catastrophe, though. We call it modernization. I think the reason matters a great deal right now, because it says less about technology than about time.

What happened

The U.S. Department of Agriculture's account of the 20th-century transformation of American agriculture covers an entire century of change in how food was grown and who grew it. Farms produced more with far fewer hands. The work most Americans had done for generations stopped needing most Americans.

The people who left farming did not stay idle. They and their children went into factories, offices, shops, schools and hospitals, and later into industries no one in 1900 could have named. That is the first half of the pattern I write about on this site, and you can read more of it on our past waves page: a great wave removes an enormous category of work, and new categories appear that nobody predicted.

My friend in Ohio ran a 200-acre farm, mostly on his own, for almost 25 years after he retired from dairy farming. It was amazing to watch, other than the days he needed help unloading the straw and hay bales. I saw this firsthand.

Why the clock was the safety net

The second half of the pattern gets less attention. The farm transition took about a hundred years. A change that slow can happen largely between generations instead of inside a single career. A farmer's son could take a factory job. His daughter could finish school and work in an office. Nobody had to retrain at 52 because a whole occupation disappeared in a few years.

There was real pain. Families lost land, towns shrank, and many people spent their lives in work that paid less than they had hoped. But a century gave institutions, schools and families time to adjust. Slowness was the cushion.

This wave runs on a different clock

Now compare that with the predictions about AI. Elon Musk has called AI and robotics a "supersonic tsunami" that will eliminate desk jobs at a very rapid pace. Bill Gates, who is usually more measured, said this week that people should prepare for 20 years of "tumultuous" disruption. Even the cautious estimate is a fifth of the time the farm shift took. Analysts cited by Fortune estimate AI is already eliminating more than 10,000 jobs a month and that 9 percent of jobs are at risk.

Speed is not the only difference. Farm machinery replaced human effort in one sector. AI and robotics can be applied to almost any task, which is why I argue this time is different: for the first time, the supply of labor itself stops being the limit. If the change happens within a single working life, the generational cushion that protected the farm transition is gone. People will have to adapt inside their careers, and fast.

What the farm still teaches us

That is sobering, but the farm story still carries three hopeful lessons.

New work shows up in unexpected places. In 1900 nobody could have predicted the jobs their grandchildren would hold. Today the early signs are already odd and specific. The New York Times reports that data center construction has become a fast-growing market for paint makers. PG&E's chief executive calls AI "the best thing that could've ever happened to the grid". Building the machine world takes electricians, line workers, builders and maintenance crews.

The payoff arrives after the cutting stops. A Fortune analysis argues that AI won't appear in productivity numbers until companies move from cost-cutting to building. The farm era's gains came from what people did with freed-up labor. The same will be true now.

Where you live matters. The farm shift left many rural communities behind. It is encouraging that the Labor Department is putting nearly $82 million into rural job training aimed at high-growth and emerging industries. Programs like that are how a faster transition can still be a fair one.

What to do now

The farm families who did well tended to see the change coming and position the next step before it was forced on them. With a shorter clock, that has to happen within your own career. A few places to start:

  • Map your own job. List your weekly tasks and mark which ones are routine execution and which involve judgment, trust or relationships. Use AI tools on the first group now, so the second group becomes your job.
  • Look at the physical buildout. If you are considering a change, the trades behind energy and data centers are growing, and they are hard to automate away.
  • If you run a business, redeploy people before you cut them. The gains come from building, not just from trimming headcount.
  • Act as a citizen. Find out whether workforce and apprenticeship programs reach your area. Tell your representatives, whatever their party, that retraining should keep pace with the technology.

The farm transition shows that a country can rebuild almost its entire workforce and come out more prosperous. This time we have to do it in years instead of generations. That is hard, but it is possible, especially if we start now.

Sources

Keep Reading

All posts →