Managing the entry-level jobs crisis

BlackRock CEO, Larry Fink, warned last month that the Class of 2026 could face "the highest unemployment rate in years, even without a recession." His concern centres on AI eliminating the entry-level roles that have long served as training grounds for college graduates. For business leaders, this creates an immediate challenge: if junior positions disappear, where do you find talent ready to contribute from day one?

 

The answer lies in understanding what AI actually is, and what it fundamentally can and cannot do.

 

 

The Category Error

Artificial intelligence conjures images of machines that think and reason. But what we call intelligence in many of these systems is actually prediction: the ability to sift and recombine existing knowledge at scale, operating on probability and pattern. That’s impressive. It’s why AI has become increasingly valuable. But it’s not truly comparable to human intelligence.

 

As a neuroscientist, I can say with confidence that the danger isn’t that AI will surpass us. It’s that we’ll mistake it for something it isn’t – and hire accordingly.

 

The way generative and agentic AI models work loosely echoes computational structures in our brains. But that resemblance tempts us into a category error: mistaking language for knowledge, and a sophisticated prediction engine for insight. As investors deploy the equivalent of trillions of pounds worldwide chasing AI’s promise, it is incumbent on business leaders to recognise what sets human intelligence apart, because as AI becomes ubiquitous, what distinguishes leading companies will be the people, not the machines.

 

 

Human Intelligence Is Social

AI is derivative by design. It analyses and engages words, diagrams and even equations that humans have produced, and it can, similarly, parse incoming visual or acoustic data (in some of the more sophisticated models). Human intelligence is different. It is social and expansive, coming into existence as we explore a multi-dimensional world of which language and abstract representations are only a part. It emerges when people argue in good faith, struggle through problems together, or find themselves changed by unexpected insights. It thrives on surprise, ambiguity and emotional resonance – the elements that spark genuine innovation.

 

Ask a chatbot to imitate Abraham Lincoln, and it can generate a convincing speech if trained on his style. But it cannot do what Lincoln did: create from his own mind the Gettysburg Address, with its power to inspire 160 years later.

 

 

The Way Forward for Students

Over the last few years, computer science graduates from leading national universities began struggling for jobs as tech companies cut entry-level hiring. Current students are noticing, recognising now that if AI can code, pure technical skills aren’t differentiating.

 

At Pomona, we’ve started to see greater interest in social sciences like public policy analysis, philosophy and international relations. It’s a sophisticated market response: these students understand their value lies in the very capabilities AI lacks.

 

A number of companies have taken notice as well.

 

Microsoft’s chief scientist recently told The Wall Street Journal that liberal arts education is now "really important" for Gen Z. McKinsey’s CEO admitted the firm is "looking more at liberal arts majors, whom we had deprioritised." These aren’t sentimental pivots – they’re strategic responses to workforce reality.

 

Justina Nixon-Saintil, Chief Impact Officer at IBM, offers the following to those with young children: “I’d lean into project-based learning as much as possible because that is where kids build the skills that will matter in any future job, like problem solving, collaboration, communication, adaptability and critical thinking. Learning the tools will matter too, but I’d treat that (at this age) as secondary.”

 

This is sage advice. After all, the people who invented large language models weren’t learning about AI tools in school or college. They were taught to think – broadly, deeply, critically. The next breakthrough will come from today’s graduates – but only if they’re trained to be creative, open-minded and collaborative.

 

 

What Business Leaders Should Look for

When a recent Pomona graduate interviewed at a consulting firm, the hiring manager said: "I assume you can use ChatGPT. What I need to know is: can you tell me when it’s wrong?"

 

That’s the question that matters. It requires judgment AI doesn’t have – the ability to see what’s missing, recognise when algorithms optimise for the wrong outcome and ask questions no one thought to programme.

 

The future belongs not to those who can prompt a model, but to those who have been taught to think critically and can:

• Define problems in new ways. AI performs best when problems are clearly framed with existing training data. It struggles with genuinely novel challenges.
• Imagine possibilities not encoded in the past. AI recombines what exists. Creating what’s genuinely new requires human thinking and judgment.
• Navigate collaboration across differences and negotiate conflicts successfully. When people from different disciplines and perspectives collide, friction generates innovation. AI can summarise but it cannot facilitate the human dynamics that produce breakthroughs.
• Make decisions under uncertainty. When there’s no template and competing principles conflict, human judgment becomes essential.
• Build organisations that serve human dignity. This isn’t about soft skills - it’s strategic thinking about long-term value creation beyond quarterly metrics.

 

A Hiring Strategy for the AI Era

If your process still prioritises technical credentials over these capabilities, you’re selecting for yesterday’s workforce.

 

The apprenticeship layer where graduates learned judgment over ten years? It’s gone. They need to arrive with those skills now. The antidote is the Liberal Arts. And it’s a road-tested antidote at that: at Pomona, 92% of our graduates secure employment or begin postgraduate education (often Ph.D programmes) within six months – a rate holding despite widespread entry-level contraction. That’s not because we train students to use AI better than anyone else. It’s because we train them for what AI can’t do, not just what it can.

 

 

The Strategic Imperative

Larry Fink’s warning isn’t about the economy. It’s about a talent pipeline that’s in some places is no longer fit for purpose. Business leaders face a choice: continue hiring for skills that worked in 2020 and struggle to find graduates who contribute or recognise what the changed landscape requires.

 

The graduates who’ll drive businesses forward aren’t necessarily those with the most technical credentials. They’re those who can do what AI cannot: frame new problems, make decisions under ambiguity, collaborate across differences, and apply judgment when there’s no training data.

 

These aren’t supplementary skills. When AI handles the technical work, they’re the primary capabilities that drive business value.

 

---

 

Dr G. Gabrielle Starr is President of Pomona College in Claremont, California, and holds faculty appointments in neuroscience and English. Her research focuses on how the brain processes aesthetic experience.

 

Main image courtesy of iStockPhoto.com and Hispanolistic