I spend a lot of time thinking about the future of work. Lately, there is one question I can’t get out of my head, and I’m worried that we (society) are not planning for the gap:  If AI is doing the work that used to teach people how to work, where will our next generation of experts come from? 

We are spending a lot of time talking about AI adoption. How do we use it? What can we automate? How much time can we save? What jobs will change? Those are important questions. But I think we are missing a much bigger one: How do people get the experience they need to become really good at what humans do best? 

People don’t suddenly wake up with great judgment, critical thinking, creativity, contextual intelligence, or the ability to recognize when something isn’t quite right. They develop those things. Through experience, failure, adaptation, and shifting. Even when we think about “intuition” – that depends on experiences and the life lessons through, well, life. 

The experience gap 

Years ago, I attended a conference at the University of Waterloo where I heard about their entrepreneurial co-op program. One thing that stuck with me was that students were not allowed to use a business startup for their first co-op experience. They needed to experience the world of work first. 

Why? 

Because without that exposure, the problems students knew were largely education problems. So, unsurprisingly, they created businesses solving problems they had experienced as students. I remember laughing when the Co-op program leader said, “and all we ended up with was a bunch of ed-tech companies”. So they explained that the goal was to give them more exposure to workplaces, industries, customers, systems, and people, and suddenly they have a much bigger universe of problems to solve. 

I have thought about that story a lot lately. Innovation requires imagination, but imagination requires inputs. Experience gives us those inputs. 

We need experiences to encounter problems we didn’t know existed. We need to see how organizations actually work. We need to watch other people make decisions. We need to get something wrong and figure out why. We need difficult conversations and strange situations and customers who don’t behave the way the textbook said they would. 

Experience gives humans context. I worry that we’re starting to create an experience gap.  

The disappearing apprenticeship 

Think about a lot of traditional junior work. Research. First drafts. Basic analysis. Preparing documents. Taking notes. Creating presentations. Writing routine communications. Pulling information together. AI is increasingly capable of doing parts of that work, often much faster than a junior employee. 

From a business perspective, I understand the appeal. But here’s the paradox. That work wasn’t only producing an output; it was producing a human with experience. Some of it was boring. Some of it probably should be automated. I am certainly not advocating keeping meaningless work around just because previous generations had to do it. However, we have to recognize what else disappears when the task disappears. 

Where does someone learn to recognize a bad analysis if they have never struggled through an analysis? Where do they develop judgment? How do they learn which questions to ask? How do they recognize when AI has given them something that looks perfectly reasonable but is completely wrong? 

We could become so efficient at removing junior work that we accidentally remove the apprenticeship that develops senior talent. That’s the disappearing apprenticeship paradox. This creates a pipeline problem 

There is a very practical workforce problem here. If employers decide they don’t need as many junior people because AI can perform more junior-level tasks, what happens five or ten years from now? Where do the experienced people come from? You can’t hire someone with ten years of experience if nobody gave them years one through nine. 

I think we’re particularly vulnerable to this in Canada because we have built a strong culture around paid work-integrated learning —co-op, particularly here in Canada. This is something we hold in very high regard: paying students for the work they do. There is a real possibility that employers will move away from the excellence of this system toward unpaid internships. This is not something we, as a society, should support happening.  

These experiences matter both for the learner AND the employer. This is how the reciprocity of employment works – the human brings intelligence, knowledge, strengths, ideas, effort, and their individuality to a job, and the employer (should) benefit from all of those things.  
 
Relying on a wage subsidy to ensure employers are continuing to engage with early talent is not the answer long term. If you’re a small business and someone gives you $5,000 toward hiring a student, that’s great. You still have to find the other $5,000. And you need to figure out what that student should actually do. You need to onboard them. Supervise them. Give feedback. Answer questions. Create meaningful work. Spend time and attention developing another human being. 

For a small organization already stretched for capacity, that’s a very different problem than simply paying a wage. If we believe early career experiences matter, we need to start designing better ways to help organizations provide them. 

AI isn’t the source of human innovation 

There’s another assumption I think we need to challenge. We talk about AI and innovation almost interchangeably. AI can absolutely accelerate innovation. It can help us analyze information, make connections, test possibilities, and execute ideas much faster. 

But where do the interesting questions come from? Where does curiosity come from? Where does the moment happen when someone says, “Wait. Why do we do it like this?” Often, it comes from experience. 
From seeing something. 
From talking to someone. 
From understanding a customer. 
From being frustrated by a system. 
From knowing enough about a situation to recognize that something could be better. 

I see this even in something as simple as social media. I don’t particularly want an AI social media assistant generating endless content for me. That’s how we end up swimming in AI slop that all feels vaguely the same. What I want is a human to have an idea. I want them to notice something interesting, connect it to an experience, bring their perspective, and then have incredible technology available to help them turn that idea into something. 

That’s a very different relationship with AI. Human insight. Technology-enabled execution. Human judgment. The technology gives us capacity. Humans decide what that capacity is for. 

This is the Human Advantage 

I think we’re spending too much time asking, “What can AI do? Organizations need to start asking a different question: “What are our humans uniquely good at, and how can technology make them better at it?” 

Judgment. Curiosity. Relationships. Context. Courage. Creativity. Communication. Ethical decision-making. Critical thinking. These aren’t “soft skills.” They are “human skills,” and they are some of the most important capabilities in an AI-enabled workplace. 

But again, people don’t develop them automatically. Organizations have to create environments where people can practice them. Which means this isn’t just an early-talent issue. It’s a workforce development issue.ducation has a role here  

Experiential learning and work-integrated learning have been growing in importance for years. I think AI makes them even more important. Higher education is educating a dramatically larger and more diverse proportion of society than universities were originally designed to serve. While institutions understandably resist the idea that their sole purpose is preparing people for jobs, we also have to listen to why students pursue education. 

Careers matter. So education can’t simply be about transferring knowledge anymore, especially when knowledge itself has become incredibly accessible. Students need opportunities to use knowledge; to question it, apply it, communicate it, work with other people, use technology, recognize when technology is wrong, and make decisions when the answer isn’t obvious. That is where experiential learning becomes essential. 

We aren’t just giving students something to put on a résumé. We are giving them experiences from which judgment can develop. 

Employers have a responsibility 

We also need to stop pretending that people should arrive at work completely developed. Employers have a role in developing people, not just students – all their people.  

Leaders need development. Managers need development. New employees need development. Experienced employees adapting to AI need development. This might be one of the greatest opportunities AI has created. If technology gives us back five hours of someone’s week, what are we going to do with those five hours? Fill them with five more hours of tasks? Or invest some of that capacity back into people? 

Coaching.  
Mentoring. 
Experimenting. 
Talking with customers. 
Building relationships. 
Learning. 
Solving problems. 
Developing the next person. 

Maybe we need to start thinking about a human-development dividend alongside the productivity dividend we expect from AI. It’s ironic that we spend enormous amounts of money and attention developing artificial intelligence while investing remarkably little in developing human intelligence. 

We need to design both at the same time

I don’t believe the answer is to slow down AI adoption. I don’t believe we should preserve junior tasks simply because someone once learned from doing them. 

The structure of early-career work is changing. Evidence about what that will mean for employment is still developing, and we should be careful not to claim that AI has already caused mass graduate unemployment. But we don’t need to wait for that evidence to become overwhelming before asking better questions. 

If technology changes the developmental experiences people used to have, what experiences will replace them? 
If we automate junior work, how will we develop junior talent? 
If AI gives experienced employees more capacity, how will we use that capacity to develop human capability? 
If organizations want innovation, how will they make sure their people continue having the experiences that give them something worth thinking about? 

The organizations that thrive in the next era of work won’t be the ones that resist AI. 
 
I don’t think they’ll simply be the ones that automate the most. I think they will be the organizations that get exceptionally good at understanding their Human Advantage and designing technology around it. 

Develop the humans. Give them experiences. Let them build judgment. Help them understand what they’re uniquely good at. Then give them extraordinary technology and see what they can do with it. 

If AI does the work that once taught people how to become experts, we need to decide, intentionally, how we’ll develop the next generation of experts. I think we need to figure this out sooner rather than later.  

I’ve recently heard that people’s life satisfaction is strongly correlated with two choices: whom they marry and what they do for a living. While I’m not looking at the first part of that happiness equation (I will reserve that for another expert), the second part is in mass disruption. As a workforce ecosystem, we need intentional, practical solutions to ensure we don’t have a purposeless workforce struggling to develop their strengths and intelligence and apply their knowledge to something meaningful.