Ford’s 1.7M Skilled-Trades Warning: Jim Farley on AI
America needs 1.7 million skilled-trades workers every year through 2035. If your AI plan ignores the people who fix the machinery, it is a very expensive fantasy.
If your AI strategy is buying software licences while you cannot find people who can fix the machinery, you do not have an AI strategy. You have a very expensive fantasy.
Ford CEO Jim Farley has put a nasty number on the table: America is expected to need 1.7 million skilled-trades workers every year through 2035. Current training pathways are producing only 55 workers for every 100 required. That is not a talent shortage around the edges. That is a production constraint sitting in the middle of the economy, waiting to ruin somebody’s growth plan.
Farley’s point at Ford Pro Accelerate in Detroit on September 30 was more useful than the usual AI chest-beating from corporate leaders. AI will change work. Some white-collar jobs will be eliminated. But the businesses that make, install, repair and maintain physical things have a more immediate problem: they cannot find enough capable people to do the work now.
That should wake up every founder, investor and operator. The next competitive advantage is not merely who has the cleverest model. It is who can turn that model into output in the real world.
Jim Farley’s uncomfortable diagnosis
Farley said the dividing line between engineers and skilled-trades workers is becoming hard to see in modern manufacturing. That is exactly right.
A worker repairing robots, maintaining automated systems, handling fibre infrastructure or diagnosing a failure on a digitally connected production line is not doing the old version of a trade. They are operating at the intersection of software, machinery, safety and practical judgment. Calling that person “blue collar” as though the work is simple is a good way to ensure your business never hires enough of them.
Ford says it employs more than 10,000 skilled-trades workers, around 20% of its 56,000 UAW workforce. Farley’s argument is that AI will act more like a companion for people doing physical, technical work than a replacement for them in the first wave of disruption.
He made the contrast bluntly. Routine work in finance spreadsheets, call centres and entry-level programming is likely to be changed or eliminated faster. Work involving machines, fault diagnosis, physical dexterity and responsibility for safe outcomes will be changed too, but not casually removed from the payroll.
This is where plenty of executives get it backwards. They see labour as a cost line and AI as a magic eraser. In reality, the scarce worker who keeps a plant, fleet, job site, hospital system or data centre operating is an asset with compounding value. The more technology you put around them, the more valuable their judgment becomes.
Ford’s answer is bigger than a recruitment campaign
Farley is not pretending Ford can solve this alone. Ford, BlackRock, Google and Carhartt launched the Alliance for America’s Skilled Trades in July, and the alliance added 14 companies around the September 30 gathering. Its research examined 124 skilled-trades occupations and projected the 1.7 million annual openings through 2035.
Ford and its partners are also putting smaller but practical pieces on the board. Ford Philanthropy and JPMorganChase announced $550,000 in joint support for Focus: HOPE, the Detroit nonprofit that provides education and job training. The Ad Council is developing a national campaign aimed at helping younger people understand pathways into skilled trades.
Good. More of that, please.
But let’s not get carried away and call a coalition a solution. The American corporate world is brilliant at announcing alliances, panels and initiatives. It is less brilliant at doing the boring work: paying trainees properly, making apprenticeships easy to enter, giving people clear promotion paths, funding equipment, and treating technical careers as a route to wealth rather than a consolation prize for people who skipped university.
The real test is not how many logos appear on the alliance website. It is whether a 19-year-old can see the path from training to a well-paid job, build serious capability in three years, and continue progressing for the next decade.
If that path is murky, young people will choose something else. They are not stupid. They are responding to incentives.
The management failure hiding inside the AI debate
The fashionable story is that companies are about to become radically lean because AI will make every office worker twice as productive. Maybe. But that is not a workforce strategy. It is a spreadsheet prediction wearing a hoodie.
The smarter question is: where does work actually get stuck?
For many businesses, it is not in drafting emails, making slide decks or summarising meetings. It is at the handoff between the digital system and the physical world. A technician cannot be conjured by a chatbot when a vehicle is down. A construction project does not progress because an AI writes a better status update. A factory does not increase throughput because someone installed another productivity tool while the maintenance team is short-staffed.
Farley has been talking about what he calls the “essential economy”: the people who keep physical systems functioning. It is an unfashionable phrase, but the operating logic is sound. Every shiny bit of the AI economy rests on people who build power infrastructure, install networks, service vehicles, manufacture equipment and keep buildings operational.
Even data centres, the physical backbone of the AI boom, need electricians, HVAC technicians, construction crews, fibre specialists and maintenance teams. You can spend billions on compute, but it is still housed in a building that somebody has to wire, cool and repair.
That means labour scarcity in the trades is not a social issue sitting in a separate policy bucket. It is a commercial risk. It can delay revenue, inflate capital expenditure, reduce service quality and stop expansion altogether.
The overlooked angle: this is also a wage and retention problem
Here is the bit business leaders do not love saying out loud: shortages are not always caused by a lack of people. Sometimes they are caused by an offer that is not good enough.
If a company claims it cannot find technicians, ask a few rude but useful questions. What is the starting pay? What does a competent worker earn in year three? Do they get proper tools? How much unpaid travel or overtime is built into the role? Can they become a supervisor, specialist or business owner? Are they trained on the technology they will actually use?
If the answers are weak, the company does not have a pipeline problem. It has a proposition problem.
The 55-for-100 training gap is serious, but training alone will not close it. Operators must make the work financially attractive and professionally respected. That means managers need to stop treating frontline technical talent as interchangeable while paying consultants a fortune to explain why retention is difficult.
There is also an opportunity here for ambitious people. White-collar workers have been sold a story that a desk, a degree and a browser tab equal safety. Farley’s comments should put a dent in that story. The safer career may be one that combines technical capability, commercial sense and the ability to solve real problems where the consequences are visible.
That does not mean everyone should become a mechanic or electrician. It means people should stop confusing prestige with durability.
What this means for you
If you run a business, do these four things this week.
First, map your bottlenecks by task, not job title. Identify the five activities that would stop revenue, production or customer service if the right person did not show up tomorrow. You will probably find that several are physical, technical and badly underappreciated.
Second, measure your capability gap honestly. Do not ask HR for headcount. Ask operations how many qualified people are required to hit the growth plan, how long they take to become productive, and where the work breaks when they are unavailable.
Third, build your own talent engine. Partner with trade schools, community colleges and training providers. Offer paid apprenticeships. Give people equipment, mentors and a visible wage progression. If you need the talent badly enough, you should be prepared to help create it.
Fourth, use AI to make skilled people stronger before using it to make junior people cheaper. Give technicians better diagnostics, faster access to manuals, smarter scheduling and useful knowledge tools. The payoff is not just lower cost. It is faster learning, less downtime and more output from scarce people.
Farley’s 1.7 million-worker figure is not merely Ford’s problem, or America’s problem. It is a warning to every operator drunk on the idea that software will remove the messy bits of business.
It will not. The winners will be the companies that pair technology with people who can actually make the world work.