AI and Jobs Part 2: Seven “T” Jobs for the Future
When Mark Twain read news of his death in the paper one day, he quipped: “The report of my death was an exaggeration.”
So too have been the rumors of the death of the job market for the past 600 years. At various times, people have predicted rampant unemployment as a result of the invention of the printing press, the cotton gin, the steam engine, the automobile, the plane and the Internet. And people were wrong every time: with each technology breakthrough, new jobs, doing things we could never before have imagined, sprang up.
That’s how tech optimists like Sam Altman, CEO of Open AI can predict both that AI may blow up 70% of jobs and create a better jobs future. At a Senate hearing last month, he said: “People will lose jobs. Many new jobs will be created. And I think better jobs. The definition of work changes.”
Others joining the tech optimist side include the World Economic Forum. In just the past couple of years, the WEF’s Future of Jobs report has gone from predicting AI will destroy more jobs than it creates (the 2023 report predicted a net loss of 14 million jobs) to predicting this year that AI will result in a net gain of 78 million jobs by 2030.
AI and automation mean a seesaw ride for all of us going forward (I asked deepai.org to create this image)
Pessimists are pretty good at adding up the specific kinds of jobs that are likely to go away in our AI future. We won’t need as many computer programmers (AI can increase coding speed by 20%-50%) or lawyers (up to an estimated 57% of lawyer’s work is automatable), plus salespeople, copywriters, accountants, designers and others -- as AI improves. And because people haven’t been able to spell out what the new jobs will be, tech pessimists are winning the war with worry words. Venture capitalist Chris Sacca put it this way on a recent edition of Tim Ferriss’s podcast: “We are super f*cked.”
After my most recent piece on AI, several of you were ticked off, and asked me to give you hope – to try to imagine which kinds of jobs seem most likely to continue or appear as we automate and AI-ify. I’ve come up with seven categories of human-performed jobs of the future.
Think of these as the 7 “T’s” that might beat back the double “A” assault of AI and automation.
Some “old” jobs will probably stick around – with changes
Trades:
I’ve written in the past about the challenges and opportunities finding folks to work in the skilled trades (think carpenters, electricians, mechanics, pipefitters, welders, plumbers). The advantage of these jobs going forward is that they are more resistant to efforts to scale them, instead requiring bespoke creative solutions to individual situations. As existing independent contractors, business owners and workers in these fields age out, McKinsey is projecting huge shortages in these fields: they estimate the US will need 740,000 new electricians and 426,000 new welders by 2032.
AI can create images of skilled tradespeople, but it can’t replace them yet (I asked deepai.org to create this image)
Many of the jobs are performed in the construction industry: an analysis by Associated Building Contractors notes that they will need 439,000 net new workers in 2025 and 499,000 in 2026 to meet demand. AI can continue to make inroads in the design process of construction, and there is room for efficiency in building through increased modularization, but most of the other trades appear to require humans for the forseeable future.
The construction sector continues to be very slow in adopting automation and AI (graphic from Associated Building Contractors).
Tool and Die:
There’s a race continuing between human employees and robots at places like tool and die manufacturing plants, but so far the humans are still winning. The latest figures show that there are about 400,000 manufacturing jobs open in the US and, if President Trump is successful in his “reshoring” efforts, there could be even more jobs available.
A recent survey by the National Association of Manufacturers makes clear companies want to “speed up digital transformation” (84.7% of manufacturers say they are trying) but demand for real human beings continues to grow. One interesting irony: as AI builds more and more giant data centers, the “chillers” needed to keep those centers at the right temperature will require human maintenance. One heating and air conditioning exec estimates between 400,000 and 500,000 technicians will be needed in the next decade.
Touch:
There’s still a future for those with high-touch skills – mental and spiritual “touch” folks like pastors, mental health counselors, divorce lawyers, social workers, and physical touch folks like masseurs, medical providers, and physical therapists. AI is already beginning to make inroads into the counseling field and automation is replacing some of the physical touch jobs, but I think it will take a number of years before a majority of us are willing to give up the deep empathy and responsive touch we can get from a responsive, sentient human being.
Tech
Tech employment is going through a similar transition. We are seeing high profile job cutbacks at high profile companies, and AI is undeniably cannibalizing the job market for entry-level programmers (down 8% over the past three years for 22-27-year-olds). But Dice’s 2025 Tech Salary Report finds tech unemployment sits at 2.9%; well below the national average. The key is having the right skills.
Chart from DICE on jobs requiring AI or Machine Learning skills: Since this chart came out, the numbers have continued to climb. As of May 2025, the percentage is 36%.
As of May 2025, some 36% of all job openings now require deep knowledge of AI skills, an increase of 98% over last year. The demand for computer-aided design skills is up 156% over last. The World Economic Forum's “Future of Jobs Report 2025” projects cybersecurity jobs to be the second-fastest growing job sector globally.
There will be a lot of new kinds of jobs
Tying together:
Predicting what new jobs will be created by AI is tricky, but a recent article in The New York Times Magazine by Robert Capps highlights three categories that seem promising.
The first set of jobs have to do with “tying together.” Folks in these jobs would do things like assessing where a company does and doesn’t use AI; evaluating which uses are working and which aren’t; determining how humans and AI are working together and ensuring that the company “voice,” “look” and “personality” are consistent, providing customers with a confidence that they can know what to expect from the company.
Trust:
Another new job opportunity Capps highlights is human jobs built around ensuring “trust.” Even as AI produces gobs of data and work product, companies are going to need real people that keep an eye on that product. We’ll need to have AI fact checkers (making sure someone is looking out for AI’s proclivity to “hallucinate,” or, more eloquently, “make sh*t up.”) Capps imagines jobs for people who serve as a company’s “trust directors”; “legal guarantors” to responsibility when AI screws up; “AI ethicists” to draw the line on what a company does and doesn’t do; and “escalation officers” who can provide a human brain or face when a customer doesn’t want to interact with a bot.
Taste
Over the past few years, more algorithms have attempted to take over “taste-making” in our society. Bots determine much of what I see and hear and read on my social media platforms and I expect that to continue. It’s hard for me to imagine a time when we won’t want human beings involved in the generation of ideas, the creation of art and music, or the brand and personality of our companies, but I think it is also likely the way we do those things is going to change.
AI will continue to get better at generating on-demand art and “creative” products, but we are still going to want humans involved (I asked deepai.org to create this image)
Capps predicts that in the future “writers” may spend less time on direct research, handing that off to AI, while spending more time polishing prose. “Product designers” might review options created by AI while still making the call on what goes out the door. Composers and screenwriters might be able to review a larger number of options for words and music before assembling the final product. The skillsets of all these people might broaden from individual creativity to an ability to write good AI prompts, then discern which results are most impressive.
If there is one conclusion I feel most confident about, it is that we are entering a period of our jobs future that is likely to demand more transition, more flexibility, more willingness to retrain or relearn more quickly than we have ever had to face.
Scribes had a long offramp as the printing press was catching hold; the Wells Fargo wagon phased out slowly as railroad lines were constructed; horse farriers had time to get ready as the automotive shoe dropped.
The speed of the AI transition is coming quicker than employers and employees and public policymakers are accustomed to. “I don’t think anyone knows exactly how fast this is going to go,” Sam Altman said at that Senate hearing. “But it feels like it could be pretty fast.”
We should start getting ready for those changes – pretty fast.
-Leslie
Part 3: What might an AI jobs policy approach look like?
Notes:
Mark Twain on death rumors (he apparently did not say “the rumors of my death have been greatly exaggerated.”) https://simple.wikipedia.org/wiki/Death_hoax#:~:text=Mark%20Twain%20is%20one%20of,death%20have%20been%20greatly%20exaggerated%22.
World Economic Forum’s Future of Jobs 2025 report: https://www.weforum.org/publications/the-future-of-jobs-report-2025/
Blue collar job shortages: https://www.nytimes.com/2025/06/23/business/factory-jobs-workers-trump.html#:~:text=The%20country%20is%20flooded%20with,built%20in%20the%20United%20States.
Construction worker shortages in 2025: https://www.abc.org/News-Media/News-Releases/abc-construction-industry-must-attract-439000-workers-in-2025
US National Association of Manufacturers 2ndQ 2025 survey: https://nam.org/wp-content/uploads/securepdfs/2025/05/Q2_2025_Writeup_FINAL.pdf
Shortages in the skilled trades: https://www.boneconnector.com/writings/work-construction-trades
McKinsey projections for skilled trade needs through 2032: https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/tradespeople-wanted-the-need-for-critical-trade-skills-in-the-us
Loss of manufacturing job applicants due to incarceration: https://www.boneconnector.com/writings/look-to-former-prisoners
Chris Sacca on Tim Ferriss podcast: https://www.facebook.com/reel/1017352013686116
BLS projected growth in computer science 2023-2033: https://www.bls.gov/ooh/computer-and-information-technology/
DICE analysis of tech jobs needs: https://www.dice.com/recruiting/ebooks/ai-jobs-hiring-guide/
Capps article: 22 new jobs in the AI future: https://www.nytimes.com/2025/06/17/magazine/ai-new-jobs.html