Designing the Future Campus: Systems, Strategy, and AI Integration

Rachel Cossar:
Hello everyone. Welcome to another episode of Conversations in the Future of Work. We’re kicking off a new season between here and the end of the year. I’m so excited. We have lots of amazing guests planned, and we’ll be digging into AI, the future of work, leadership development, and academia from a number of different angles and perspectives.

So for our first episode, I’m really excited to welcome a good friend with whom I’ve already shared a number of very fascinating conversations, James Genone. Welcome.

James Genone:
Thanks for having me, Rachel.

Rachel Cossar:
So excited to have you. Would you mind sharing a little bit about your background, your general entry point into AI technology, and what you’re working on these days?

James Genone:
Sure. So I’ve worked in higher education pretty much my entire career. I started out as a professor, and I taught courses in philosophy and cognitive science, including, interestingly, some courses about artificial intelligence.

I always wondered if I would live to see the day when it would become a bigger part of our lives, which, of course, it has.

About 10 or 11 years ago, I shifted out of my more academic work. First of all, I helped to start a new university, Minerva University, as one of the founding faculty members there. Then I moved into a leadership role there and ultimately shifted into working for the company that had incubated that university, which is called Minerva Project.

I built a partnerships business there that worked with universities globally on program innovation, technology innovation, and so on. So that was really the shift in my career into educational technology, strategy, and innovation.

Since then, over the past several years, I’ve worked very heavily in AI as a product leader and as a go-to-market leader in startups, and also in a stint at Northeastern University, which was how we met, as the Senior Vice Chancellor for Learning Strategy, where I created the university’s AI strategy for teaching and learning, brokered a big partnership with Anthropic, and rolled out Claude to the entire university.

That was a really interesting opportunity to understand, from the other side of the table, how AI was really being adopted and integrated into teaching and learning.

I’m now back in the private sector. I’m about to start a full-time role at Instructure, the maker of Canvas, to lead strategic partnerships and innovation there.

Rachel Cossar:
Awesome. Very full experience there, which means I’m already going to divert from our plan.

You’ve had this interesting dual experience in the entrepreneurial space, but then also working with institutions that historically take forever to make changes happen, just given the distributed nature of universities in particular.

Have you noticed that AI has really forced some of these larger institutions to adopt more of a nimble, entrepreneurial approach and mindset?

James Genone:
Nimble and entrepreneurial may be a little bit too generous, but I do think that the pace and frequency of strategic-level discussion, and how people are thinking about the potential size of the change that’s needed right now to their standard operating procedure, is definitely different than what we would have seen in years past.

Things still take a long time, I believe, to get through academic governance, to get through probably board review, and making bigger changes is still hard.

I actually think this is a feature of universities rather than a bug. They tend to be conservative by nature, but that’s made them incredibly stable over long, long periods of time. That’s something that I think is healthy for society.

But we’re also seeing, not just with AI but with digital technology more generally, a much bigger change to the way that we work than we’ve probably seen with previous technological innovations historically.

The amount of people who are in knowledge-work kinds of jobs whose work is impacted by technology, and now especially AI, is quite significant. So I think universities are being forced to think much more seriously about a faster pace of change than they ever would have in the past.

Rachel Cossar:
And do you feel like it has generally been the case within universities where students lead a lot of the innovation or pressure to change? Or is that a newer thing?

I feel like I keep hearing anecdotal evidence of students really getting their hands dirty with AI on their own accord, with whatever systems they can access. Has that dynamic changed as a result of some of the accessibility of these technologies recently?

James Genone:
I think it may be changing. I also think there are some other factors that lead into that, largely to do with demographics and some of the changes in the workforce and the kinds of programs that students are seeking out.

Over the past many decades, when higher education has largely been on a growth trajectory and many universities have been able to add seats and grow their enrollments, students, in some sense, had less power because they were competing for scarce seats and trying to go to the best university that they could.

Whereas now, there’s more competition, especially below the very most selective institutions, at the tier below that and then down from there.

We’re seeing students voting with their feet, essentially, and saying what kinds of institutions they want to go to. Some of this is pretty well documented.

One trend we’re seeing in particular is that institutions like Northeastern that are more focused on career preparation are doing exceedingly well. Others that have less of that reputation may struggle.

We’re also seeing institutions that are smaller and less able to absorb fluctuations in enrollment really struggling financially. That’s a big factor. We see universities closing and so on.

So I think that ability to vote with their feet does give students, en masse, a certain kind of power to influence the educational system. You’ll see more and more universities copying strategies that I think are working for some of the big universities right now.

Once students are enrolled, institutions differ in what kind of a role they give students in terms of helping to determine the direction of a university and their approach. I don’t think that they have a ton of influence, even at ones that do try to include their voices.

But as some of these institutions try to understand ways in which they may need to evolve, I think some of them are starting to realize that listening to students can be a tremendous help.

Rachel Cossar:
Yeah, that’s really interesting.

It’s always hard to tell what’s real and what’s just media, but I feel like recently there has been a lot of student-led movement that’s affected university policy. That’s even going beyond the tech, AI, or curriculum piece, but I guess it really just depends on the institution.

James Genone:
Yeah. Certainly in some of the more political topics that have come up in recent years, I think student voices have at least been something that universities have had to respond to and navigate.

I think AI is an interesting one there because you do hear a lot in the media about students being concerned about the use of AI and so on. On the other hand, you’ll see surveys and other things showing that 90 percent plus of students are using AI in one way or another.

So it’s a little bit hard to know how to understand that.

I know from my own engagement with students that they have quite a bit of frustration with the lack of coherent strategy and policy on the part of universities. I think that’s an ongoing pain point, and they’re definitely making their voices heard about that.

As you mentioned, I think others are concerned about sustainability issues and those sorts of things, and maybe just not having great information about what the university’s point of view is.

Rachel Cossar:
For sure. I think that was also something that came up, certainly at universities but in corporations and other organizations throughout COVID, right?

It was that unified and properly communicated stance on where we are and where we’re going in a world of such ambiguity. So it’s a challenge on both ends.

Let’s get into where we are today in the world and academia and the different options that are available to leaders in education.

More and more, we’re seeing different school districts outright banning AI. How can university leadership—provosts, deans, and academic operations—build coherent, campus-wide strategies, products, and partnerships that foster systematic AI literacy without arbitrarily banning, or fully releasing AI?

James Genone:
Yeah, it’s a great question, and it’s a complex one.

Where I start on this is that I believe an institution has to have a point of view. They have to have a North Star that can guide the formulation of the rest of the strategy and the policy and so on.

That’s what I think is lacking at a lot of institutions.

There’s a decent amount of activity. There are pilots, there are experiments, there are different kinds of documents issuing from different parts of administration about AI.

But the lack of coherence derives from not having a particularly clear vision about what integration with—or what their responsibility is with respect to engaging with—AI.

I think something that most of us can agree on, even if we may have different nuances to this, is that the role of universities is to prepare students for their futures.

Some will focus on the professional aspect of that, and others will say we have to prepare people to be good citizens and members of society.

Presumably, in either case, having a job for most people is going to be a part of that.

With the changes that AI and other kinds of technology are bringing about, one would think that the way students are being prepared would need to change.

So I think that immediately starts to focus on topics like AI literacy and being prepared to use artificial intelligence for people who are going to go into knowledge-work fields.

Even for people who aren’t, it’s still impacting other fields, including things like the trades, in ways that they’ll be able to engage productively with the technology and integrate it into their work as their industry is changing and evolving.

I believe that has to be the responsibility of educational institutions broadly, and universities in particular, to navigate because, in most cases, high school education won’t be enough.

High schools, K–12, are having their own challenges adopting and integrating AI. I think they’re more likely to ban or prevent its use, or at least not teach it or adopt it in a systematic way, than colleges and universities.

Then we also know that employers themselves are struggling to figure out how their businesses work with AI and what they can do to retrain their existing workforce.

I think they’re also going to be really challenged to address entry-level workers and recent college graduates in this regard.

So I think that means universities really have a responsibility to be able to help prepare students for this. That means making big changes to all of their different kinds of programs.

And that requires making, I think, a really important distinction that a lot of people aren’t making right now.

That’s the difference between preparing students to use AI in future jobs and to learn the foundational skills and concepts from their discipline when AI is available.

Those are two very, very different things.

When you’re trying to increase your productivity, for example, you want to use AI to do work for you, in many cases at a high level of quality that you can evaluate and approve or modify as needed.

But when you’re trying to use AI to learn, it’s exactly the opposite.

You often don’t want the AI producing the work. You want it quizzing you, critiquing your work, helping you understand things, and then helping you practice them further.

So that difference between learning with AI versus learning to work with AI is one that I don’t think is getting nearly enough discussion.

That’s what’s confusing quite a bit of the discourse around the adoption of AI in education.

Rachel Cossar:
For sure.

I feel like increasingly it’s harder for—I wouldn’t say people necessarily, but maybe certainly mainstream media, which is where a lot of people get their news—to hold any kind of nuance.

I think, to your point, there’s a difference between empowering people to learn how to effectively leverage AI to improve their skill base, to lead to deeper development of themselves as a social human being or whatever professional avenue they’re taking, versus using AI to produce results for you.

Certainly at Virtual Sapiens, we spend a lot of time thinking about that because there are lots of expressions of AI within the coaching and training industry that arguably could lead to atrophy.

If you think about an AI coach, if the AI is not really functioning as a classic coach but functioning more as an advice giver, then the person who’s using it is not actually pausing to reflect and do the thinking work themselves.

They’re being fed advice, and then they can just move forward on that without thinking critically about it all that much.

Versus using AI as that practice and feedback mechanism, which, by definition, will help that person more deeply or more fruitfully interact with the human being that they’re intended to actually have the conversation with.

From a product and learning-science perspective, how do we build AI systems that act as “cognitive scaffolding” to build human capacity rather than cause skill atrophy?

James Genone:
Yeah, it’s really tricky because there’s been a lot of concern, for example, about young people developing a dependence on AI companions and then not socializing as much with their peers.

I do think that’s a real risk that has to be taken seriously.

But if you talk to some of these young people, they will say things like, “Hey, I’ve never felt entirely safe talking about certain things with my peers or a therapist or my parents or what have you. This is a space where I can express things that I’ve never been able to express before, and it’s really helped me. In fact, it’s maybe helped me even feel more confident and comfortable interacting with other people in ways that I haven’t before.”

I think that also has to be taken really seriously.

So it’s not as simple as being able to make blanket statements about the detriments or opportunities that the technology affords. It’s really quite nuanced.

When we come back to learning, I think we have to ask ourselves a question: What’s our goal? What are we trying to accomplish?

In cases where actually engaging with another human is going to be the best way to build a certain skill, then that’s something that we have to make sure that we maintain.

In cases where we’re trying to get something like a benefit from high-volume practice—which is actually not possible because our friends, classmates, teachers, and TAs aren’t willing to let us practice our delivery or speech a hundred times—that’s where the technology is really positive because it’s allowing us to do something that we literally couldn’t do before.

Rachel Cossar:
Yeah, and you know what’s so interesting?

When we sell to university partners and our university partners incorporate the platform as a practice and feedback tool for students who have these assignments that are due, we see the highest number of retries and engagement from the student population.

It’s really amazing.

I was looking at some data this morning, and there were a few students who’d practiced in our Practice Mirror, which is a presentation-type practice tool that we have, over 50 times.

These are short, one- to two-ish-minute presentations.

To your point, I don’t think any peer would have been like, “Yeah, sure, let’s try the 50th time.”

But, for better or for worse, that is what that student needed, and they don’t feel judged by the AI in the same way that you would feel judged by a human counterpart.

So there’s no question in my mind that the safe space that AI can create—particularly if the AI tool has been designed properly, and from a privacy and security perspective things have to be watertight—providing that safe space to students is very compelling.

James Genone:
Yeah, couldn’t agree more.

Rachel Cossar:
And professionals, right? Not just students.

But we do see that very, very high level of retries with the student population.

In some of our previous conversations, you’d mentioned skill assessment in particular as an area where there’s also a lot of potential.

What kind of evolution do you think institutions need to lead when it comes to skills assessments and individual student development?

James Genone:
Yeah, well, this is where I’ll give the education system—and maybe colleges and universities in particular—a bit of a hard time.

We’ve said for a long time that teaching skills like critical thinking, creativity, leadership, and so on is part of what you go to university to gain.

But there’s not actually very much evidence that that’s actually occurring.

Certainly it’s not something that is systematically measured, and the few attempts to do that with tests of critical thinking and so on do not paint a very flattering picture of what’s going on in universities.

You see many results where there’s literally no gain in critical-thinking ability over four years on these measures.

I think there are a number of different reasons behind that.

They’re hard skills to measure. They’re even hard skills to actually define in ways that everyone will accept.

But there’s been enough work on that, I think, to at least have a solid starting point.

A lot of it is that the assessment opportunities are competing with traditional forms of assessment that are heavily focused on content—on memorizing content or being able to produce fairly standardized deliverables like essays, for example, or solve problem sets in different domains that are amenable to scalable grading.

That makes grading easier because grading is one of the most unpleasant activities that teachers have to deal with.

It’s also pretty well known that, with the exception of things like multiple choice, which are deterministic, we’re not very good at it. We’re not very consistent.

Intra- and inter-rater reliability among experts is not as high as you would want it to be.

So I think this is a place where artificial intelligence could really help us.

I actually think it would probably do a better job at traditional forms of assessment than humans often do.

But I also think it opens up the opportunity for these alternative forms of assessment.

Sometimes they’re called authentic assessment, where you’re assessing students on more realistic types of tasks that they might perform in a workplace, for example, in their field.

People talk about simulations and even experiential learning, when you’re in a real, live working setting with others and you can capture some of how the work gets done and then assess: Was that done well? Was it done to a standard?

The use of rubrics and really clear criteria, even if qualitative, can use these sorts of qualitative scoring approaches done by machines, maybe checked by humans until we’re comfortable with their reliability.

It would allow us to scale types of assessment that traditionally have been thought to be too hard to scale, particularly what are sometimes called the durable skills.

So everything from critical and creative thinking to communication skills, leadership, decision-making, and so on.

I think that has the potential to be a huge benefit because I think those skills are only going to become more important.

They’ve always been important, but they’re going to become even more important as many of the routinized forms of knowledge work that people have done in the past become automated.

There will be more space and opportunity for people to execute on ideas, make decisions, be creative, and so on.

But they need to have those skills, and they need to understand where their gaps are and where they can improve.

I think these AI systems will be able to help us with that.

Rachel Cossar:
Yeah, I think you bring up a really interesting point about the non-standardization of human evaluators and how AI could certainly level the playing field from the perspective of at least every student being evaluated against the same scoring rubric that has been explained in the same way and the same directions have been given to the AI.

I think that’s really, really interesting.

We’re having conversations with some firms who are in the leadership assessment space.

It’s interesting to think about the different stages—from current human-led assessments and human-led evaluation, to a hybrid human-in-the-loop, AI-powered option, so you can still vet, particularly depending on the outcome of the leadership assessment.

If it’s succession planning or recruiting, those are very high stakes.

Then ultimately, I guess the question is whether you could have fully AI-run, evaluated, decision-made assessments.

Again, that’s a slightly different category from skills assessments within universities, but I think some of the same questions apply.

It’s an interesting space.

James Genone:
Yeah. I’ll mention one thing that comes up quite a bit in that domain.

Especially if you think of automation of performance or automation of hiring processes and so on, these are, again, fraught areas where I think we don’t always have the highest confidence in humans’ reliability in doing these things.

Those processes are often not fair and are biased in various different ways that have been studied.

People are very anxious about outsourcing that to forms of systematic bias that you may see in AI models.

In particular, when large language models first came out, there was a lot of evidence that they were biased by their training data in very problematic ways.

I think that’s a real area we have to concern ourselves with.

However, there are straightforward technical solutions to many of those things, such as improving the training data or other measures you can put into place to countermand bias.

Over time, with those improvements, you may be able to make them better than human systems.

Now, I would argue that actually the human systems could also be improved if better processes were put into place, which some workplaces have done to mitigate against bias and other issues that we have in these forms of evaluation.

But whether we’re willing to bear the cost of doing those things and whether or not we can do it reliably is, I think, a real question.

You could see a point in the future where, especially if we do get the systems to work reliably, they end up being less biased than human systems.

That said, I would always argue for humans in the loop of these things to check on what’s going on.

There’s an important concept in AI development called evaluations, which is really evaluations of the models. These can also be automated, so you can have AI models checking each other’s work and really seeing if they’re meeting the standards and avoiding the biases that they’re required to.

That can then help the human monitors track those things.

So I think there’s quite a bit of innovation still to do in this area, but we may end up with a better overall system of evaluations across many different domains, from education to the workforce, when we get to the end of the road.

Rachel Cossar:
Yeah, totally.

So the last question is another tough one.

There’s so much more competition now, and access to education or information is just a totally different ballgame these days.

From your perspective across academia and edtech, what is the core, un-automatable value proposition that higher ed must double down on to justify its cost to prospective students?

James Genone:
Yeah, it’s a great question.

People talk about the cost of higher ed, and then they’re benchmarking that to private universities that have these very, very high tuition costs.

But there’s still incredible value in education in our public system, and people don’t necessarily always keep that in mind when they talk about tuition costs.

Although tuition costs have grown at public institutions as well, it’s nowhere near at the rate of private education. So that’s worth remembering.

The data also show that the value of a degree in employment is quite significant still.

So I don’t think that degrees are going anywhere.

In some ways, I think that the request to justify their value is misplaced.

I think the question is a little bit more: How do degrees need to evolve in order to keep providing value, given that we would anticipate that the needs of employers and the needs of students are changing and are going to continue to change?

To that end, I think a couple of things are true.

First of all, we need to align credentials in a much more granular way with the different kinds of employment that people seek.

A transcript with a GPA and a list of courses with a letter grade next to it is a very, very weak signal of what someone is prepared to do in their career.

To come back to what we were just speaking about, being able to assess people’s skills and capabilities, where their real strengths are, and so on would be a tremendously stronger way to help people get placed into careers.

I think if universities can do that and really be the validator of skills and of what people are capable of, that would be tremendous.

Of course, that also requires that they help people build those skills.

Evolving their pedagogy, evolving what they teach in order to make sure that the skills employers need are being cultivated, will be a part of that change as well.

If they can do those things, then the economics become more of a question of affordability rather than value.

It doesn’t matter if the value of the credential is strong in a career if you can’t afford it in the first place.

If you don’t have access to that because of your means, then it’s not serving society.

That really becomes, I think, a question of what our political systems are going to do, particularly in the domain of public education, to enhance the affordability of college.

In the last election cycle and before that, free college was a frequent topic of discussion. It’s kind of dropped off the radar, but I wouldn’t be surprised if that’s something—

Rachel Cossar:
Right.

Yeah, I think you make a fantastic point about the public university system and what is available to people there.

There’s such a branded, marketed draw to obviously get into the top private institutions. But the reality is that they’re competitive, which is part of the appeal. They have, I don’t know, five percent, three percent acceptance rates or whatever.

So I think it will also be interesting, from the perspective of some of the things we’ve discussed—skill assessment and skill development—versus the network access that some of these larger institutions are going to give, which is just different.

But at least from the perspective of what skills are we building, how are those skills being evaluated, and then how are they being matched to careers, that would be a pretty amazing thing to be able to refine and solve for.

Then people who do go through public education systems can feel like they can compete, at least to some degree, with people who go to some of the Ivy League private systems.

Something like a leveling of the playing field, so to speak, could be very interesting.

James Genone:
Yeah. This is getting really speculative now, but I think there are a number of different scenarios for how the education system overall could really shift.

Many have proposed or speculated that some of these very strong branded institutions, like the Harvards, MITs, and Stanfords of the world, could actually use their brand to provide broader access by essentially—whether it was through online or through acquisitions or something like that—really expanding the reach of the education that they provide.

Now, there’s a real question of what it would mean to get a Harvard education if you weren’t coming to Harvard Square and going into the fancy old brick buildings.

There are probably some good answers to that.

Obviously, there are tremendous assets from the faculty that Harvard has to draw on, but how they would scale those out to a larger student body and provide value for them is a really interesting question.

There’s also been a lot of speculation that some larger companies—the sort of hyperscale companies—might partner with, or themselves get into, the education business to help prepare the workforces that they think they need.

That would provide potentially a different kind of competition for the public systems at large scale.

So I think those are all really interesting scenarios, and what it would take for those to come to pass.

But in the meantime, hopefully they provide some incentive for the institutions as they are now to try to continue to evolve.

Rachel Cossar:
Yeah, absolutely.

Well, many, many great questions answered and many more questions for next time.

James, thank you so much. If people want to follow your thought leadership and your work, what’s the best way for them to connect with you?

James Genone:
Always a pleasure, Rachel.

Generally on LinkedIn. I haven’t been as active on LinkedIn in recent months just because of some of the work that I’ve been doing.

But as I transition into my role at Instructure, I expect to be a little bit more active there again.

You’re always welcome to connect with me there.

I should have a webpage up on the Instructure website relatively soon with an email, and people can get in touch with me that way as well.

Rachel Cossar:
Awesome. Great.

Thank you so much. Thanks again for all of our fantastic conversations and for sharing your insights with our community and audience today.

And to our community and audience, thank you for tuning in. We’ll see you next time.