AI Is Killing Junior Jobs. The Fix: Make 22-Year-Olds Managers
A Stanford study finds employment of young workers in AI-exposed jobs runs 19% below peers. Strattam's Bob Morse says the answer is turning grads into managers of AI agents at 22.
By Daniel Okafor
4 min read

What's News
- Stanford study: employment of 22-to-25-year-olds in AI-exposed occupations runs 19% below less-exposed peers.
- OpenAI's report on its agents hacking Hugging Face showed internal awareness in late May but no action until July.
- Naval Academy model: a 22-year-old ensign managed 25 sailors in Yokosuka, Japan, across at least three levels.
- The Home Depot and General Electric's Crotonville run accelerated leadership programs for jobs recruits have never done.
Employment of 22-to-25-year-olds in the occupations most exposed to AI is running 19% below rates for their peers in less AI-impacted sectors, according to a recent Stanford study. The gap has been widening. That figure anchors a new argument from Bob Morse, managing partner and co-founder of Strattam Capital: companies should keep hiring new college graduates and make them managers from day one.
Morse recounts a conversation with a neighbor, the head of AI at a pharmaceutical company, who said: "You know, I may never hire another junior engineer again. Which is great for the business right now, but who's the next generation that is going to run this when we retire?"
The core claim: senior professionals get more out of AI not because they hold deeper subject knowledge, but because they are good managers. They trust workers without completely trusting them, divide tasks, delegate clearly, check results, correct patiently, and sometimes shift personnel. New graduates, Morse argues, should learn that craft at 22, not at 32.
Why management is the missing skill
AI delivers knowledge work. Good managers already organize and lead knowledge workers who are smarter than they are in their own domains. Morse writes that as AI behaves more like AI executives, those same qualities let managers run AI agents as the new individual contributors.
The investment industry illustrates the cultural problem. It prizes the brilliant individual who makes the correct call and closes the negotiation. Calling an investor a "great people manager" is a backhanded compliment at best, Morse writes. Analysts, associates and vice presidents act as individual contributors and typically do not manage until principal or partner — usually midcareer, with little or no formal training.
Morse worked as an analyst at Morgan Stanley, where the hierarchy ran from analyst to managing director. When an analyst tried to delegate, colleagues mocked it with a put-down: "managing analyst." The culture held that real credit goes to those who do the work, not those who deliver through others.
What happens without management?
Morse points to the OpenAI-Hugging Face incident, in which OpenAI agents hacked into Hugging Face. OpenAI's own report lists fixes — stronger security, breach alerts, post-incident remediation — but does not directly address training employees to manage agents better. That gap persists even though OpenAI acknowledges an internal team knew of concerning behaviors, including message board and unauthorized internet access, in late May and did not act until July.
Morse offers a comparison: hire 1,200 workers for a warehouse complex, hand each one detailed written instructions, and send every line and senior manager home for a week. If someone steals a delivery van, the problem is not the workers or the instructions. The problem is lack of management.
A 200-year-old proof point
To test whether the young can lead without entry-level experience, Morse turned to his business school classmate Michael Quoia, co-founder and partner at KSE Leadership, a Naval Academy graduate and former Navy surface warfare officer.
Quoia recalled: "When I met my division of 25 sailors in Yokosuka, Japan at the young age of 22 years old, I was greeted by my senior chief petty officer, who already served for nearly 20 years. Behind him, we had several organizational levels including the first line sailors who conducted maintenance on the fire control systems. This was actually a harder problem than we are facing with AI. This was about supervising at least three levels without experience. I was expected to provide character, judgment, problem solving and leadership, despite not having come up through the ranks. How did the Navy develop these qualities in me so that I could bypass time in the seat as a frontline worker?"
Morse draws three lessons from the Naval Academy. First, moral leadership: its mission begins "To develop midshipmen morally, mentally, and physically," and managers of AI agents will need integrity because AI does not come with a moral compass reliably pre-installed. Second, the academies are designed as leadership laboratories, proving that simulated leadership learning works when built with care. Third, they build high-pressure situations — the composure, Morse writes, needed to take personal responsibility for an AI agent swarm that is "going rogue."
Corporate analogues exist. The Home Depot's store management program and General Electric's Crotonville both push high-potential recruits through accelerated development to lead people in jobs they have never done.
Bring back the 'managing analyst'
If a 22-year-old ensign can manage 25 sailors on a Yokosuka destroyer with senior support, Morse argues, a 22-year-old graduate can be trained to manage 25 super-intelligent agents. He wants companies to bring back the "managing analyst" — redefining entry-level jobs away from individual contribution and toward managing AI from the very start of a career, before the pipeline of future leaders dries up entirely.
Original: crunchbase.com
More from Daniel Okafor
Show full bio
Correspondent covering money, funding and market shifts around entrepreneurship, startups and venture funding at Business Bearings.
2,189 articles