The spreadsheet class is getting automated first, and the guy running 10,000 union tradespeople just said the quiet part out loud.
The Summary
- Ford CEO Jim Farley told reporters that finance workers, call center staff, and entry-level programmers face job elimination in AI's "first inning" while skilled trades workers use AI as a "companion"
- Ford employs over 10,000 skilled-trades workers (20% of its 56,000 UAW workforce) who increasingly repair robots and maintain automated machinery
- The automation gap reveals which jobs require physical presence and which were always just pattern-matching in Excel
The Signal
Ford CEO Jim Farley drew a hard line between knowledge work and vocational work at Ford's Detroit conference last week. Spreadsheet builders are in trouble. Robot fixers are not. His framing matters because Ford is deep in the automation game, racing to remake its EV lineup while investing heavily in robotics.
The company needs humans to maintain the machines that replaced other humans. More than 10,000 skilled-trades workers at Ford now spend their days repairing robots, working with software, and keeping automated production lines running. That is 20% of Ford's entire UAW workforce. They are not getting automated. They are becoming machine whisperers.
"If you work in finance doing spreadsheets, or you're in a call center, or you're an entry-level programmer, those jobs are definitely going to be changed, eliminated."
Meanwhile, the jobs everyone thought were safe are the ones AI is eating first:
- Finance analysts building spreadsheets
- Call center workers following scripts
- Entry-level programmers writing boilerplate code
Farley called this the "first inning" of AI, which means he expects more rounds. But the pattern he is seeing tracks with what is happening across industries. AI agents can generate reports, write code, and handle customer service conversations. They cannot yet crawl under a stamping press to swap out a servo motor or troubleshoot a welding robot that is throwing errors.
The line between manufacturing engineers and hourly skilled-trades workers is blurring at Ford. The people on the floor are doing work that used to require a four-year degree. They are running diagnostics, reading error logs, interfacing with software systems that control six-figure machines. Farley positioned AI as a companion to skilled trades, not a replacement.
This is the Web4 inversion nobody predicted five years ago:
- White-collar "knowledge work" turns out to be highly compressible information tasks
- Blue-collar trades work requires embodied intelligence AI does not have yet
- The degree premium is collapsing for roles that were always just data transformation
The Implication
If you are building spreadsheets for a living, you have maybe 18 months to figure out what part of your job AI cannot do. The answer is probably not "build better spreadsheets." It is more likely "understand the physical business those spreadsheets describe." The people who will win in the agent economy are the ones who can move between the digital layer and the physical world.
For companies, this is a talent rebalancing problem. You need fewer analysts and more people who can maintain the infrastructure that runs the analysts' AI agents. Trade schools and apprenticeship programs are suddenly more strategic than another cohort of finance BAs.