Our research is on AI’s disruption of entry-level employment and what universities need to do in response. As a quick intro, give us a picture of what you have going on right now that ties to that question.

So yes, I think about this a lot. I am a bit of a unicorn in that I do have a very strong technical background, and I’ve now come into business and think about all this differently.

The question about job loss is one of the biggest questions students ask me. And these are online MBA students. So these are mid-career professionals, predominantly. They’re parents. They ask about their career, they ask about their children, what they should say. So it’s a lot. And I think about it a lot.

And I tell them, I don’t think you change the advice simply because AI and autonomy and robots are now here. I think that whatever made you special as a human is going to carry forward into this new world.

I recently wrote a thought piece on this called “The Economics of AI Anxiety,” now published on my site at innodative.com and by Gies College of Business. The basic premise is, it’s easy to sit here and say, well, AI can do that job, AI can do that job, AI can do that job. But the challenge is that we do not have infinite AI. We are limited, and we’ve seen this with OpenAI and Anthropic, they’re changing their pricing models. They don’t actually have a good way to value their product. They don’t even have the final product. We’re still using a chat. This is how we first introduced ChatGPT, what, three and a half years ago. So until we get more embedded AI that makes our devices more seamless, we’re going to have that issue. But regardless, we still hit this issue of inference costs. They don’t know how to price it.

Another big issue that people gloss over is, yes, AI can, let’s just take a hypothetical, AI can do every task better than a human. Right? I don’t think that’s true. But if we take that hypothetical, what do we have? Well, every job’s going to go away, because why would you pay a human to do that if an AI could do it better? But economists have taught us for two centuries this idea of comparative advantage. Just because somebody can do something better doesn’t mean society always has them do it, because they will focus on the high-value return they have, and that leads to advantage for others. This is why companies still trade all the time.

So I think there’s going to be comparative advantage with AI, particularly as we’re still trying to figure out how do we price it, how do we value it, what’s the product, as well as the whole ability to build data centers. We aren’t guaranteed we’re going to be able to build enough data centers on Earth. There’s already opposition. They’re talking about building in space. There’s a reason they’re talking about building in space, because they know we’re going to need a lot. There’s a lot more room up there than here.

This is a complex problem. You’re seeing a little bit of this with software development, where you had, whether it’s true or not, lots of reports of layoffs in the software industry because, well, AI coding. AI coding is going to take all these entry-level positions. But now people are realizing, that’s not really what it is. What we really need are smart young people who know how to work properly with AI and can have AI augment their skills. That’s really a thing that we need. We aren’t really doing that very well, either in academia or in industry apprenticeship. So that was a long-winded way to basically tell you, yeah, this is an important problem, but it’s also very complex.

I want to push back on something you said. You said that you would always give the same advice, basically follow your dreams, do what you’re passionate about. If someone said, my dream is to be a translator from Spanish to English, that’s my passion, would you say, follow your passion?

You’re right to push back on what I said, because I was paraphrasing. I think it is important to pursue your passions, because you should try to find purpose in life and purpose in work. But that doesn’t mean you do it blindly.

I would just take myself as an example. I pursue my passions. It doesn’t mean I’ve done the same thing. I haven’t done the same thing for more than five years in my career. Part of that is because I keep my eyes open, and realize that as I learn more about the world, as I learn more about my interests, my skills, what I really find value in, I change.

I thought I wanted to do cosmological theory. That’s when I started grad school. And then I got involved with data. And I’m like, well, this is really fascinating. And I actually like how machines think. So I did a lot of that. And then it was a lot of application of that. And then I started doing data science. It’s not just, oh, I’m 16, this is what I really find fascinating, I’m going to do this the rest of my life. I don’t think that.

But in your example about translating, maybe you’re really interested in this. Okay, but maybe that doesn’t mean you’re doing the translating. Maybe you’re running a company that does automated translating.

Or maybe you need to do real-time translating. There are lots of opportunities to be pursuing your passion that keep your eyes open to what the world enables.

The other thing you asked about was augmentation. I’ll just use a simple example. When AI first came out, my job as Chief Disruption Officer is to literally try to live in the future and see things that are going to potentially cause problems. I’m basically always thinking about not what’s here right now, but what is going to be important six, 12, 24 months out. And so I and some students, we built a version of ChatGPT probably six months before it was public. And one of the things we did with this was, I shared it with a senior leader and I said, I think you need to be aware of this. Put in a prompt and you’ll get an essay out. Obviously at the time, a lot of our courses, particularly our online courses, were very dependent on text essay-based assessments. Here’s a business plan, right? So he did it. And again, this is my team’s attempt at building one of these. He did it and he comes back and goes, well, it was probably a C-level essay that came back, but it did it in two minutes. And I’m like, this is our team. This is not some big company doing this. I told the senior leader, these things are coming, we need to be thinking about it.

So I don’t think we were quite as caught off foot as many universities.

Even so, most faculty members, when they first confronted this, were like, oh my gosh, students are just going to cheat on all my assessments. I instead took a different approach, and the college is now doing more of this. It was, okay, we can’t fight it. So we need to embrace it, and what does that mean?

It means we redo our assessments and say, you’re going to use Gen AI, and here’s how you’re going to use it, and you’re going to do this. And then the last part is, now you’re going to think about what you did and how it worked, and you’re going to write about that, and don’t use AI for that. And the only time we ever really paid attention to whether Gen AI was used improperly was that part. And we always came down hard on them. Like, what are you doing? You’re outsourcing your future, your brain, to AI. You can’t do that if you want to succeed. And we rarely see that anymore.

I’m augmented as a faculty member. The very first assignment in this course I’m teaching right now is basically doing basic data analytics with ChatGPT. You basically say, hey, can you generate a data set for this type of a business, and it generates it. Can you generate it and make it dirty, like have all kinds of errors? It will do it. Now we’re going to clean that data. So students are doing things they never would have done in my class, but they’re getting that hands-on experience of working with it to solve different problems that they normally wouldn’t be solving in this type of a class. That is what I mean by augmentation, that you’re using AI as a partner to solve problems that might be outside your comfort zone, might be outside your normal flow.

I see you talking about the augmentation thesis a lot. You have this verbatim from your essay: “AI adoption is less about job displacement and more about AI augmentation, where workers shift from routine tasks to strategy, oversight and problem solving.” I want you to defend the augmentation thesis. When we are at a point where most institutions, most governments cannot even formulate the responses they’re going to have for this. It seems like AI is already so much ahead of us that the institutional responses are not even clarified. Universities are either outright banning AI. How can you be so sure of your augmentation thesis when that seems to be the case?

Let me go back and answer something you said earlier. You mentioned your sales team, you can now do the work of six or seven people. You’ve been augmented. You are now able to do more work than you could before.

So one way to look at that is, oh my gosh, five people just lost their jobs. And I don’t think that’s how capitalism, how free markets work. It’s not a zero-sum game. Just because you’re able to do that work doesn’t mean the company couldn’t use two people to now do twice as much as the entire group did before.

So I think we’re going to see, and maybe this is the AI optimist view, and I’m not saying I’m fully in that camp, but this might be what they’re saying, that we’re going to get to this place where you’re able to do things you like, and everybody’s going to do more, and the costs are going to drop, and we’re going to have UBI and all that. That’s an argument some people will make.

But I do think there is some merit to that. Just because you don’t need as many people to do the same amount of work as before, it’s not zero-sum across all industries, across all jobs and economies. So there will be rough periods. I don’t mean to downplay that at all. There’s going to be rough periods as businesses, as individuals, as leaders figure out how we navigate this process. We are not to the point where people graduate college or high school and they’re already augmented and know how to do it. We aren’t there.

But to that point, we already have companies, and one of them is McKinsey, who come and recruit students, and part of their interview process is the students have to work with an AI tool. So companies are already expecting this augmentation from students. And we’re struggling. We don’t know your tools, so we can’t teach them how to use your tool, but we can make sure they know some basic AI skills that they can move between tools and different employers. So that’s the overall thesis.

But why am I so optimistic? I’m optimistic because I think the free market will require it. I already talked about the economics issue, that we’re not going to have cheap inference that allows you to do whatever you want. You’re going to focus on high-value targets first, and maybe over time as inference costs come down, that’ll change the equation. We also have many industries that are regulated, or that data has to be protected, or that people expect a human touch. So that’s part of the reason why I think this augmentation, even if it’s not universal, will be a dominant theme in the market. Because people who are augmented, you yourself demonstrated, are doing more than those who aren’t. And you’re still going to need that human aspect, both because of comparative advantage and cost, but also because, whether you’re in healthcare or education or finance, there are regulated industries where it’s just required, where people want to be able to talk to a person.

If you were to speak to a dean at a university that just bans AI, what would be the first hard step you would tell them to take?

I can’t speak to European institutions and cultures and policies and all of that. But if I was talking to a generic US institution that had that sort of policy, I would say, you need to get your resume polished up because your institution is going to become extinct.

Higher education is facing many pressures in the world right now. The US has a demographic cliff. It’s well-known. It’s been known it’s coming for over a decade, that the number of graduating high school seniors is dropping in the US. We’re already seeing a lot of smaller institutions facing very serious pressures, a lot of smaller liberal arts colleges. If they take a hit, and there are some, even just in Illinois, that I’ve seen anecdotal evidence have had 20, 30, 40% enrollment drops, that is a non-sustainable business model.

So if you think about those sorts of pressures, and on top of that, you have the pressure of students saying, you’re not preparing me for the jobs if you don’t allow AI, if you don’t allow me to learn how to use AI. And I want to be clear, when you say augmented, you get the images of chips in your head. That’s not what I’m talking about. It’s really, can I be augmented in what I’m doing so that AI helps me do more, augmenting the output. If you are not training students for that, they’ll vote with their feet. Employers will vote with their feet, and they’ll go elsewhere. If you’re already facing demographic challenges, why would you want to risk bigger demographic challenges?

Now, the flip side is, academics are notoriously resistant to change. “I’ve been teaching this subject for 20 years, how dare you come in and tell me I need to do it differently.” And so it is hard to change. At the University of Illinois, we have a faculty senate, and major curricular changes have to go through this whole process, and it can take years. That’s very difficult to be nimble when you have something like AI that is just evolving so fast.

But I think that’s the challenge we have, that we have to meet that. And if we don’t, we run the risk of being replaced, either by institutions that are, or by technology firms saying, screw it, we’ll just hire high school seniors and teach them what we need. We’ll give them a six-month training and then maybe a three- or four-year apprenticeship. Why would a student not be interested in that? They’re making good money, they’re getting practical skills, and they just don’t necessarily have a diploma at the end. Or maybe Google University gives them a diploma.

I have said this to my institutions, and usually from leaders I get, yeah, okay, we see this, we’re going to have to do something about it. But on the flip side, I’ve had a senior faculty member at another Big Ten institution tell me, “thank God, I’m retired in three years.”

You may look at Gies and be like, wow, they’re really at the forefront, they’re really doing all these things. We look at it differently. Wow, we are not doing enough. There are so many things we need to be doing. But universities are very decentralized, at least in the US. They tend to be very decentralized, and so you can’t from the top say, this professor, you’re going to do this, this professor, you’re going to do that. So we tend to provide incentives and allow people to take things they’re already interested in doing and see if they can’t do more.

When you see these sort of strategies and these sort of teams, can we find ways to encourage people to be more assertive about bringing augmentation into their courses? We don’t even know what it would mean to teach AI inside any specific functional discipline yet. And even if we did know what it would mean, it doesn’t mean we know what would be relevant in a year when our students would be graduating.

So you can’t just say, we’re going to shut down for six months, we’re going to think deeply about this, and then we’ll come out of it, we’re going to add the new curriculum and all that. Even if we did, we’d then have years of negotiations and arguments, and our alumni would get involved, and businesses. So we can’t really do that top-down push. Instead, we’re trying to use the many-flowers-bloom approach, and say, let’s see what people can do. And then can we bring them together in a way to have coherent strategies. Okay, seeing this is working, this is working, this is working, let’s bring those together, share this more broadly, and encourage people to start adopting best practices or ways to do this.

In your essay, you write that educators’ first concern around AI tends to be how to ensure academic integrity. But you say that by focusing on the how we educate, we risk missing the more important question of why and what we educate. From your vantage as Chief Disruption Officer, what are universities still getting wrong on the why and the what?

In my class, I have polls regularly in this live session we’re having. I was chastising the students because many of them were choosing an answer that basically was, “it depends.” You can’t, in your profession, you can’t bet on “it depends” for everything. So I’m tempted myself, I could give that answer, it depends.

But in reality, the big thing I say, and it goes back to the economics driver of higher education, I think too many faculty members forget that higher education is a business. Unless you have a monarch or some government that’s basically endowed sufficient funds that you’re set up and can do whatever you want, most of us are not set up that way. Certainly in the US, we are not. The Ivies, to an extent, are, but even they face financial pressures.

So we forget we’re businesses. And often, faculty are like, well, I’m here to do research, or I’m here to do X. But you won’t be there if the university doesn’t exist. So I keep coming back to this why. Why are we doing what we do? And if you go back to history, it’s to provide that education so that people can do what they want to do. And in Western liberal democracies, that tends to be, I’m going to get a job, I want to be able to provide for a family, I want to have those things, and ideally I want to have some purpose and contribute positively to society.

So if you’re going to do all these things, you can’t think, the world’s going to be different in two years or five years, but I’ll just keep doing what I’m doing. We need to think about how the world is going to be different. AI is here. It’s not going to go away. It’s only going to get better. You’re using the worst AI you’ll ever use, you’re using right now. It’s only going to get better. What do we, as an institution, think we should do? In my view, we need to embrace that. We do need to embrace AI. We need to embrace automation, robotics, a lot of these technologies that are coming. Think about how do we rethink what we do in this age, where all of these technologies are changing how society works, what we’re doing.

We won’t get it all right. You quoted Dario earlier. Dario is a brilliant computer scientist, but I don’t think he understands economics to the degree that his quotes might imply. His quotes are being taken out of context, and sometimes they’re used as marketing. But 50% of jobs, in practice, 50% of the jobs, that’s already been shown false, because he said it like a year ago. And while we’re not there right now, 50% of entry-level jobs would be such bad unemployment that our economies would be crashing. We’d be in real trouble.

So I think it’s always more nuanced. It’s always more in that middle of thinking about how these things relate.

*This interview has been lightly edited for clarity and readability. The interviewee reviewed and approved the transcript before publication.