Why Your Degree Filter Might Be Costing You the Best Talent (And What AI Has to Do With It)

In my recent LinkedIn post, I shared some striking data from Peter Diamandis about the collapse in perceived value of university degrees. Job postings requiring degrees have dropped 6% since 2019 in the US, whilst college graduates now make up a third of the long-term unemployed.

But here’s the question I keep coming back to: why?

It’s not that education itself has become worthless. It’s that the gap between what’s taught and what’s needed has become a chasm. By the time a student graduates, the specific skills they learned in their first year are often obsolete. As one MIT administrator candidly admitted: “We can build a nuclear reactor on campus faster than we can change this curriculum.”

And nowhere is this more apparent than in the rapid evolution of AI and automation tools.

**The Real AI Revolution Isn’t What You Think**

Switch on the news and you’ll hear sweeping narratives about AI replacing jobs, transforming entire industries, or requiring massive company-wide digital transformation programmes. It’s all very zoomed out, very dramatic, and frankly, not particularly useful if you’re trying to run a business.

The real AI revolution is far more granular. It’s happening at the individual level, in specific tasks, for particular roles. It’s not about replacing your team with robots. It’s about equipping each person to do their job better, faster, and with less friction.

This is why the traditional degree is losing its shine. Universities teach broad theoretical frameworks. They prepare students for roles that existed three years ago. But the tools available today, the AI assistants that can draft proposals, analyse data, generate creative concepts, or automate repetitive admin, these didn’t exist when current graduates started their courses.

Meanwhile, someone who’s spent the last three years actually using these tools, adapting to updates, learning through doing, they’ve developed something far more valuable: the ability to learn rapidly and apply technology to real problems.

**Why Individual Adoption Beats Company-Wide Rollouts**

Here’s where most businesses get it wrong with AI. They think they need a grand strategy, a company-wide platform, an enterprise solution that transforms everything at once.

That’s not how this works.

AI tools are designed for individual workflows. ChatGPT, Claude, Notion AI, Jasper, Midjourney, these aren’t departmental solutions. They’re personal productivity multipliers. They work best when someone understands their specific role deeply enough to know exactly where the friction points are.

Take your marketing coordinator. They might spend two hours every week pulling together performance reports from three different platforms, formatting them, and writing summary insights. An AI tool could do this in five minutes, but only if that specific person knows which metrics matter, what format your leadership team expects, and what insights are actually valuable versus vanity metrics.

You can’t roll that out company-wide. You can’t train a whole department on “AI for marketing”. You need that individual to understand their task, recognise the opportunity, and learn the specific tool that solves their specific problem.

This is why hiring someone who can demonstrate they’ve already done this, who has a portfolio of problems they’ve solved, processes they’ve streamlined, results they’ve delivered, is infinitely more valuable than someone with a 2:1 in Business Management.

**Practical Examples: Person Plus AI**

Let me give you some concrete examples of what this looks like in practice.

Your sales team member who manually updates CRM records after every call. They could use AI transcription tools to automatically capture key points, action items, and next steps, then feed that directly into your CRM. That’s not a company-wide CRM overhaul. That’s one person using one tool to eliminate 30 minutes of admin per day.

Your HR coordinator who writes job descriptions, employee onboarding documents, and policy updates. AI writing assistants can draft these in minutes, maintaining your company tone and incorporating specific requirements. But only if that person understands what good looks like, what your culture requires, and how to prompt the tool effectively.

Your operations manager who processes expense claims and flags anomalies. AI can analyse patterns, spot duplicates, and even predict cash flow impacts. But they need to know which red flags matter, what your approval thresholds are, and how to interpret the outputs meaningfully.

Notice the pattern? It’s not AI replacing these people. It’s AI amplifying their expertise. The human judgement, the context, the understanding of what matters, that’s irreplaceable. The tedious, repetitive, time-consuming execution? That’s where AI excels.

**Training for Roles, Not Departments**

So how do you actually develop this capability in your team?

Forget generic “AI workshops” where someone shows PowerPoint slides about machine learning. That’s theatre, not training.

Instead, work with each team member to identify their three most time-consuming repetitive tasks. Not strategic thinking. Not relationship building. The stuff that makes them think “I wish this would just do itself.”

Then help them find and learn the specific tool that solves that specific problem. Give them time to experiment. Encourage them to break things. Let them discover what works through doing, not through theory.

This is hands-on, task-specific, individual development. It’s messy. It’s not uniform across the organisation. And it’s exactly what works.

Because here’s the thing: someone who learns this way develops something far more valuable than knowledge of a specific tool. They develop the confidence and capability to keep learning. When that tool updates, or when a better tool emerges, they’ll adapt. That’s the skill that matters.

**The Hiring Implications**

This brings us back to the original question: if you’re still using degrees as your primary filter, what are you actually filtering for?

You’re filtering for people who could navigate an academic system three years ago. You’re not filtering for people who can navigate rapid technological change today.

Look instead for evidence of adaptation. GitHub repositories. Side projects. Freelance work. YouTube channels teaching others. Anything that demonstrates they can identify a problem, learn a tool, and create value.

Ask candidates: “Tell me about a time you automated part of your workflow. What tool did you use? How did you learn it? What was the result?”

That question tells you more about their capability than their degree classification ever will.

**Final Thought**

The gap between education and application has never been wider. AI tools are evolving faster than any curriculum can keep pace with. The businesses that thrive won’t be the ones with the most impressive graduate recruitment programmes. They’ll be the ones who hire for adaptability and then empower individuals to solve their specific problems with the best tools available.

Stop filtering for credentials. Start filtering for capability.

Mark Bushkes