Product Field Work All writing
WritingWork after autonomous AI

After the engine starts

Once a company’s AI runs on its own, improving the product, the funnel and the support queue around the clock, most of today’s roles won’t be competing with a tool. They’ll be competing with something that runs a thousand experiments an hour. What happens to the people is the question I keep coming back to.

The honest starting point

I build with these models every day. I’ve watched work that took a team two weeks of planning, specs, handoffs and testing collapse into minutes of an agent’s time. When leadership can ask a system directly for what used to take a department months, the layers in between start to look like cost, not capability.

So I’ll say the uncomfortable part plainly: a lot of jobs, as people know them today, are going away. Not all of them, not all at once, and the data so far is quieter than the headlines. But the direction is clear, and waiting for it to be obvious in the unemployment rate means waiting too long.

What the data shows today

The door is closing before anyone is pushed out

The first effect isn’t mass layoffs. It’s that the entry-level door is narrowing in the jobs AI does best. Stanford’s payroll research found that workers aged 22 to 25 in the most AI-exposed jobs saw a 16% relative drop in employment, while older workers in the same jobs didn’t. The Dallas Fed traced it mostly to fewer people being hired in, not more people being let go.

Layoffs are catching up. Companies cited AI in about 116,000 announced U.S. job cuts through August 2026, roughly 22% of all cuts, up from about 55,000 and 5% in all of 2025.

Young developers are the canaries

Change in employment since late 2022, software developers at the same firms, by age.

Stanford HAI 2026 AI Index, from ADP payroll data, as reported in arXiv 2605.01160. Older-worker growth shown at the midpoint of the reported 6–12% range.

116,175U.S. job cuts attributed to AI in announcements through August 2026, the top stated reason this year
~8,000jobs Meta cut in May, about 10%, after evaluating some teams for cuts as deep as 60%
0.1 ptthe most AI could explain of the rise in U.S. unemployment so far, per the Dallas Fed. The signal is real, the aggregate still small

Both things are true at once. The aggregate numbers are still small, and a growing number of companies are telling investors, out loud, that AI is why they need fewer people. The second trend tends to lead the first.

Why this time feels different

The last escape routes are automating too

History is the best argument for calm. About 60% of the jobs Americans held in 2018 didn’t exist as job titles in 1940. New work keeps appearing, and it usually does. But the same research found automation erased work about twice as fast after 1980 as before, and its authors are clear that there’s no law saying new work arrives one-for-one.

What’s different now is breadth. The old advice was: if a machine takes your desk job, move toward physical work or care. Physical work is automating in parallel. Waymo went from 50,000 paid rides a week in May 2024 to about 500,000 in early 2026, with a goal of a million by year end. Humanoid robots are in warehouse and factory pilots, and a single Chinese factory can now build thousands of them a year.

Where work is exposed, and how that moves

My assessment, informed by BLS projections and exposure research. Toggle the horizon to see physical automation catch up.

Horizontal: share of the job’s tasks current AI and robotics can do. Vertical: how much the job depends on physical presence, dexterity, trust or care. Dot size: rough U.S. employment. Projections for 2034 from BLS, July 2026.

What looks most protected: work where the value is a person being there and being trusted. Nursing, home health and personal care, skilled trades in messy one-off environments like old houses and field repair, teaching young children, therapy, and roles that carry legal accountability. BLS projects healthcare and social assistance to add about 2 million jobs by 2034, the most of any sector. It’s not glamorous, and much of it pays less than the desk jobs it would replace. That gap is the real problem.

The shallow company

Fewer layers, fewer people, more tokens

When the engine runs, the org chart flattens from the middle. Agents take the execution work first, then the coordination work that existed to manage execution. What’s left is a thin layer of people who set direction, own the risk, and pay for the compute. The budget line that grows is tokens.

How a 1,000-person company might thin out

An illustration, not a forecast. Drag to see the org compress as agents take on more of the work.

1,000people
7management layers
$0share of the operating budget spent on AI instead of salaries

Public companies feel the most pressure to follow that curve all the way down, because investors reward it. Some leaders will keep people on out of loyalty for a while. I don’t think that holds once it’s obvious there’s nothing productive for those roles to do. Kindness without work isn’t a plan for anyone.

The question isn’t whether companies get smaller. It’s whether the gains stay with the few people who own the engine, or reach the people it replaced.

Two people, two years

What this looks like from the inside

Two composite people, not real individuals, in two of the most exposed jobs in tech. Step through their next two years, and what I’d tell each of them to do now.

What we could do about it

The options on the table

None of these is new, and each has real costs. Several are already being proposed in Congress and by the AI labs themselves. Open each to see what it does, what it would cost, and what the evidence says.

Where I land

  1. Measure it first. Require large employers to report AI-related layoffs, hiring freezes and retraining, so policy follows data instead of anecdotes. That idea already has bipartisan sponsors.
  2. Protect the income, not the job. Wage insurance and portable benefits help people take the next job even when it pays less, instead of freezing them in place.
  3. Keep the entry door open. Pay companies to train juniors alongside agents. If nobody learns the work, nobody will be able to supervise the machines doing it.
  4. Share the upside. Some form of broad ownership in AI’s gains, whether a public stake or an AI dividend, is worth serious study before the gap becomes a crisis.

These are my views on a contested question. People I respect disagree, especially on costs and on how fast any of this arrives.

What companies can do now

Before the next round of cuts

If you run a company

  1. Redeploy before you release. Meta moved about 7,000 people into new AI teams alongside its cuts; it can be done at scale.
  2. Say what AI changed. Report where it replaced work and where it created new work, before a law makes you.
  3. Keep hiring juniors, in smaller numbers, paired with agents. Your future reviewers and operators come from there.
  4. Put part of the savings into severance, retraining and equity for the people whose work funded them.

If you work in a company

  1. Move toward the judgment layer: owning outcomes, reviewing what agents produce, and being accountable when it’s wrong.
  2. Learn to direct and audit agents in your own field. That skill is scarce right now and will pay for a while.
  3. Build a second path near people or the physical world, where trust and presence still matter.
  4. Save more than feels necessary. A longer runway is the most practical protection there is.
What I’m least sure of

Speed. The aggregate data still says “early.” Adoption inside most companies is uneven, and a large Danish study found AI reshaping tasks without yet changing employment or earnings. Tokens may get expensive enough, for frontier work running around the clock, that some human work stays cheaper for longer than I expect. And new work does tend to appear in places nobody predicted.

But I’d rather prepare for the version where the engine works, and be pleasantly wrong, than the other way around.

Sources
  1. MIT Technology Review, May 2026, summarizing Stanford Digital Economy Lab (16% relative decline for ages 22–25 in the most exposed jobs) and Anthropic’s March 2026 labor report.
  2. Federal Reserve Bank of Dallas, January 2026: decline driven by fewer entries; at most a 0.1-point effect on aggregate unemployment.
  3. arXiv 2605.01160, citing the Stanford HAI 2026 AI Index: developers aged 22–25 down nearly 20% since late 2022; older developers up 6–12%.
  4. Challenger, Gray & Christmas, August 2026, and March 2026: AI cited in 116,175 cuts in 2026 through August, versus 54,836 in 2025.
  5. Fortune, May 2026 and TheStreet, citing Reuters: Meta’s ~8,000 cuts, and teams evaluated for cuts of up to 60%.
  6. Yahoo Finance, May 2026: about 7,000 Meta workers redirected into new AI teams.
  7. MIT News on Autor et al., “New Frontiers,” QJE 2024: about 60% of 2018 jobs didn’t exist in 1940; automation eroded work twice as fast after 1980.
  8. Yahoo Finance / TheStreet, 2026: Salesforce cut nearly half of its customer support organization as AI agents took over.
  9. The Rundown, April 2026: humanoid pilots in warehouses and factories; a Guangdong factory able to build up to 10,000 humanoids a year.
  10. TechCrunch, March 2026: Waymo at 500,000 paid rides a week, up from 50,000 in May 2024.
  11. BLS, July 2026: software developers projected +15.8% to 2034; customer service representatives projected to decline.
  12. Sens. Warner and Hawley, AI-Related Job Impacts Clarity Act: quarterly reporting of AI-related layoffs, hiring and retraining.
  13. Built In, July 2026, summarizing Anthropic’s and OpenAI’s policy proposals, including wage insurance and public equity in AI gains.
  14. Stanford SIEPR, July 2026: firm-level adoption uneven; a Danish study finds task changes without employment effects yet.