Why AI Is Turning New Graduates Into Tomorrow’s Managers

AI is changing the first rung of the career ladder, and the shift is forcing a major leadership question into the open: who will run organizations when today’s senior teams retire?
A neighbor who leads AI for a pharmaceutical company described the productivity his senior team is getting from AI, then paused with a warning: “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?” That question reaches far beyond engineering. It challenges the old path from entry-level contributor to experienced manager.
The First Career Step Is Under Pressure
College graduates typically begin with individual contributor knowledge work, the kind of work AI is becoming more capable of delivering. A recent Stanford study found that employment for 22 to 25 year olds in occupations most exposed to AI is running 19% below rates for peers in less AI-impacted sectors, and the gap has been widening.
That creates a difficult choice for organizations. They can reduce early-career hiring because AI handles more entry-level tasks, or they can rethink what new graduates are hired to do. The argument for the second path is direct: move graduates toward management at the very beginning of their careers instead of waiting for a long period of individual contribution.
Why are senior employees getting more from AI today? The advantage does not come mainly from greater subject knowledge. Good managers already know how to work with people whose skills exceed their own in specific areas, and those same habits can shape how teams work with AI.
Good management requires trust without complete trust. It means dividing tasks, delegating clearly, checking results, correcting and coaching with patience, and shifting personnel when a team needs a different mix of skills. The craft is organizing and leading knowledge workers who may be much smarter than the manager in their specialties.
AI agents are becoming those individual contributor knowledge workers. That makes management skills central to AI adoption, not separate from it.
Teaching the AI Generation to Lead
Andre Shojaie saw the shift during a meeting with a client when his son joined them and said he was already working with AI. The emerging generation will not meet these systems as a distant workplace tool. “Call them the AI generation: a generation whose habits of thought will have been shaped by daily contact with machines built to answer them.”
Education offers a clear example of how fast this change can move. Yash Mundada said that two years ago, the hardest part of teaching a student to conduct research was helping complete a literature review. Today, a model can produce a satisfactory literature review in less than a minute.
Mundada’s organization, Rise Research, helps match high school students with Ph.D.-level mentors for original research. At first, the organization treated AI use as cheating. It tried to prohibit the tool and detect its use, but after many months it made no progress. Detection is not reliable, and banning a tool that students will use throughout their careers would not be appropriate.
That experience points toward a different model for early-career development. Students and new graduates need practice setting goals, checking outputs, giving clear instructions, and improving results. Those are management actions, and AI makes them part of daily work much sooner.
AI Needs Onboarding, Managers Need Training
Every employee with a corporate license for Claude, ChatGPT, or Gemini now has access to a skilled data analyst, according to Pawel Rzeszucinski. “The analyst writes SQL in seconds and answers at 2 a.m. without hesitation.” But the analyst starts with no understanding of how the business works.
Rzeszucinski described a striking test from Anthropic. When the company pointed its own model at a data warehouse without supporting context, accuracy on internal analytics evaluations never exceeded 21%. Once curated context was added, accuracy rose above 95%. “Same model, same data, radically different answers.”
The lesson is clear: AI agents need onboarding just as people do. They need definitions, context, expectations, and checks. New managers who learn to provide that structure can guide both human workers and AI agents, turning raw capability into useful work.
But technical skill alone does not create effective people leadership. OpenAI and Anthropic are putting technologists and product operations people in charge of HR, raising the question of whether technical expertise translates into better management. Wharton professor of management Peter Cappelli warns, “Give the general counsel a heads-up, because ‘they’re going to be busy.’”
Cappelli says organizations may expect technologists to “sprinkle some pixie dust” and make workplace problems disappear. His research found that many executives are engineers by training, while companies have abandoned management development programs that once taught them how to lead. Engineers and technologists may approach problems through optimization, but applied to people, that mindset can prioritize cutting headcount over developing and supporting employees.
His advice is practical: technologists cannot become effective HR leaders without learning the fundamentals of people management, perhaps through a rotational assignment on an HR team before taking the top job. They can also help companies navigate the AI transition and build expertise in change management.
The same principle applies to new graduates. Organizations can teach management early, then let young leaders practice with human teams and AI agents. They can also make role changes a conversation: Kai Werner of TeamViewer says, “The most helpful first step is to have an open and honest conversation with your boss.”
That conversation can uncover why a person was chosen for a new role, what strengths a manager sees, what success should look like, and which skills need development. A transition may involve training, a gradual handoff, or expanded responsibilities while preserving some work the employee values.
The future of work will not be shaped only by smarter models. It will be shaped by the people who know how to direct them, question their results, and build teams around them. Organizations that teach management from day one may find the next generation is not missing after all—it is already learning how to lead.
Based on
- In The AI Era, Keep Hiring New College Grads — And Make Them Managers From Day One — news.crunchbase.com
- The AI Generation Is Arriving: Who Will Teach It To Lead? — forbes.com
- Four AI Lessons From Education That Business Leaders Can Apply — forbes.com
- The Importance Of AI Onboarding: AI Doesn’t Know Your Churn Definition — forbes.com
- Why engineers make risky CEOs, according to one Wharton professor | Fortune — fortune.com




