Lessons on AI Leadership

Michael Cato: Hello everyone and welcome to the Integrative CIO Podcast. I'm Michael Cato and I'm here with my colleague and co-host Cynthia Golden. We are so happy to have you join us for what we assure is going to be an enlightening and timely conversation. Today we're exploring leadership lessons and reflections from leading AI efforts in higher education institutions and we have two colleagues who we hope are going to be able to speak from different but related vantage points. We have Liv Yesbam, vice president and CIO at Denison University. Hi Liv. Hi

Liv Gjestvang: Michael, good to see you. Hi Cynthia.

Michael Cato: And we have Stan Waddell, vice president and CIO at Carnegie Mellon University. Welcome, Stan.

Stan Waddell: Thank you. Hello everybody. Great to be here.

Michael Cato: Well, we're glad to have you both here with us.

Cynthia Golden: So you are both leading AI strategy at very different kinds of institutions. Stan, you're at one of the world's preeminent research and computer science and engineering schools and leave at least from our conversations as I recall, you're at a residential liberal arts institution that has a very strong focus on teaching. So before we dig into everything we want to talk about today, I'd love to hear from each of you a brief description of your institutions or how your institution's identity shapes the lens through which you see AI. So what does AI mean for your mission in particular? Liv, let's start with you.

Liv Gjestvang: Great. So as a liberal arts college, I think we feel that we have a really important role to play in the way that students are learning both about the technology, but also really what I would say is at the heart of a liberal arts education, which is what does it mean to be human? How do we learn about and engage in citizenship and the construction of knowledge and then really exploring the social, cultural, ethical concerns that we face in our society. And so AI presents challenges in every single one of these areas. It presents huge opportunities in every one of these areas too. And so we look at really the relationship between the rapidly changing technology environment and the kinds of skills that our students and really I think I would say citizens of the world will need to have to be able to guide the ways that technology is being implemented, the kinds of implications it has for the economy, for the world of work, for education.

Liv Gjestvang: And that for us is I would say a very holistic part of our mission is focusing on preparing leaders to be able to lead through all of the new challenges and opportunities that are arising and also ensuring that as part of that, our students have deep understanding of and experience with AI and other emerging technologies so that they can then apply those larger thinking and collaboration skills to the world in which the technologies are playing out.

Cynthia Golden: That's an important foundation I think for our conversation today. Stan, what about at Carnegie Mellon?

Stan Waddell: Yeah. At Carnegie Mellon, ostensibly we're the birthplace of artificial intelligence. And when you think about that and how important artificial intelligence has been as an educational domain for us, but also as a research domain for us and the birth of Carnegie Mellon being a relatively new higher education institution around the century mark versus some of the longstanding higher education institutions. And we were formed really with a translational or a direct impact mission in mind. Carnegie Mellon was created to impact industry to create educational pathways for the children of skilled tradesmen and artisans within the Pittsburgh environment. And that impetus and creation really still exists in the mantras of how Carnegie Mellon operates today. So when we're looking at artificial intelligence, we're looking at it with how it's going to impact society, how it's going to be implemented in meaningful and significant ways. And we sort of have that historical perspective of the creation of the domain and how that's translated into where we are today.

Stan Waddell: And I watch our researchers and educators continue to learn and to advance the body of knowledge when it comes to artificial intelligence and then to put that into the hands of our students from an undergrad and graduate research perspective, but also into translating into real skills that they can Liverage to go out and be productive members of the workforce and productive members of society. Many of our students are starting companies in AI fields and they're going and being hired at top AI firms and they're taking that mindset from Carnegie Mellon out into industry and into the community and impacting the development of artificial intelligence in very meaningful ways as the technology rapidly unfolds.

Cynthia Golden: So, thank you for that important perspective.

Michael Cato: And it's really interesting because as you're both responding, one of the things that crosses my mind is for those in our audience who may not know it, you're both very active beyond your institutions, right? You're both active in active voices and helping shape our higher education thinks and pursues these issues around AI more broadly through convenings, national summits, peer networks, policy conversation, just to name knowing you both well, you're all over the place and have been around these issues for a while. So I'm curious, how do you think about those aspects of your external role? What are you trying to influence and how does that impact your work at your respective institution? And perhaps Stan, I'll start with you this time.

Stan Waddell: Awesome. Awesome. And sort of following on my original comments, again, there's a real desire at Carnegie Mellon to have this translational impact and transformative impact on society. And that bleeds into staff as well as our educational pursuits. So faculty have that mantra and staff do as well. And so we want and therefore I want to have that kind of impact and being at a place where the science is unfolding I think does give me a unique perspective around the power of the technology as well as some of the risk associated with the technology. And I think it's one of my responsibilities and one of the responsibilities of Carnegie Mellon to get out and share those perspectives. I've had a number of conversations with peers, a number of conversations with industry in various forums, whether they be the small forums or very large forums, sharing perspectives, hearing perspectives and bringing those things back.

Stan Waddell: And I think that that bidirectional communication will only make us all better and better equip us to handle this very disruptive technology set.

Michael Cato: Yeah, I appreciate that, Stan. Liv, what about you? How do you think about this work externally?

Liv Gjestvang: Yeah, I think we are, I think, in a very fortunate position in higher education in that we work in a field that is extremely collaborative. I think we are working across the country and across the globe to be able to engage and support learners to be able to have them ready to go out and lead in a society that needs really strong leaders and collaborators. And so for us, we've been very interested in convening conversations and bringing leaders initially the first two national summits that we held were for liberal arts colleges and it was great to connect with people with a very similar mission reaching pretty similar student bodies to be able to think about what are each of us learning? How can we accelerate our work by building some strategies together? The third summit that we held, we actually broadened the work or the colleague set that we were partnering with and had a really wide range of institutions.

Liv Gjestvang: So R1s, we had a Ohio State, Michigan case and then a collection of liberal arts colleges and other large institutions. And there we're able to learn about some of the ways that we and smaller institutions are able to go deep and connect in these very personal ways with one-on-one student faculty relationships and research with small classes, lots of interdisciplinary work and what was happening at a much larger scale at schools like Penn State and Michigan. And so I think I believe that we go further and faster when we are able to kind of elevate our work based on what we learn from each other. And I think doing that within higher ed allows us to have a stronger influence on the way that this field unfolds. I will just say, and we can maybe circle back to this, but I do think that as a field, I think there are opportunities for us as higher ed institutions to come together more closely and have our voice heard more clearly in places where decisions are being made.

Liv Gjestvang: So I think that's incredibly important and that as kind of regulatory and policy strategies are evolving or maybe need to evolve more quickly, I think we have to be a part of that conversation. And I'm part of a couple of evolving groups that are working on that now, so happy to share more about that if we have a chance.

Michael Cato: That sounds great. And it's interesting that I hear a theme across both of your responses of that bidirectional nature of that work so that you're sharing outside what's happening at your institution, but also bringing in ideas inside. And to your point, Lee, there might be opportunities to put that to work in service of the larger issues. Appreciate that.

Cynthia Golden: And I was really glad to hear what you had to say about that, about that sharing your perspective and hearing the perspective of others. Because I started out at Carnegie Mellon years and years ago and it was absolutely part of the culture that we were expected to get out there and share what was happening on our campuses on our campus. I mean, it was just part of the way we work. And so I'm glad to hear that that environment and that expectation is still there.

Stan Waddell: Yeah, definitely still there.

Cynthia Golden: So leading AI strategy means navigating some real tensions and it's things like the tension between innovation and safety or between the excited early adopters and then the people who are not so excited, the more cautious folks and things like academic freedom and institutional responsibility. So there's lots of tensions that I know work their way into everybody's work with AI. So what's the tension you find yourself returning to most often? And how are you thinking about those tensions right now? And Liv, I'll start with you.

Liv Gjestvang: Yeah, I mean, there's a really interesting one that I'm not entirely sure where to land on this because I think it's a really, really important question. And it's the on that is around what are the ways that we do dive in and use this technology in ways that support learning that accelerate people and really amplify people's students and faculty members' skills. I think there's a huge opportunity there. And we're also have been reading more and more about research that talks about some of the negative cognitive impacts of not wrestling deeply with questions or writing multiple drafts. And we know that those things are important both in terms of the development of the ways that we think and learn, but also I think a really critical sense of resilience and grit and tenacity that is really important. So for me, and I think this is the fact that there are so many questions in this space are part of what makes it exciting to me.

Liv Gjestvang: So how do we recognize the pressures in both of these spaces and how can we as a community use this as a moment to dive into what are the ways that we might want to expand or develop pedagogy so that it allows us to really deepen the critical thinking and drive and resilience that comes from working through hard problems, getting stuck, hitting a wall and coming back to it. And also figuring out how we do leverage technologies that can help us learn in really different ways. Like I said, you can develop so many skills so quickly go further. So I'm really interested in that and I think that comes up. I hear it from students, I hear it from faculty, I hear it from staff in terms of how they want to engage and the questions they have about where this technology is going and how we can ensure it's developing in ways that serve people and serve communities broadly though.

Cynthia Golden: Yeah, I think those are important questions. I've been having conversations about that with instructional designers as well, what's the role going to be there in the future? So Stan, what about at CMU?

Stan Waddell: Yeah, I mean, I think Liv and I would use a lot or I would use a lot of the same words that Leave just used. I think we have a very similar philosophy in this vein around the tensions. So maybe I'll take it into a different direction and talk a bit about the speed of adoption and disruption. I think that that is something that society is wrestling with and I worry that maybe we don't wrestle with it enough because the technologies are moving so quickly and the capabilities shift weekly and you watch successive rounds of models get released, you see the things that they can do start to increase and the mind has to go to a set of conclusions that point to some amount of work that human beings do needs to change and some of the work that human beings do maybe we don't need to do.

Stan Waddell: We don't drag carts and things like that around as a principle means of moving stuff anymore. And in many cases, we don't use one of the oldest technologies that there is paper to convey information in the way that we used to. And I think that large language models are going to shift that conveyance of information even further into that technical medium and it's going to shift a lot of the work that we do that provides marginal value. And what I mean by that is a lot of the things that we do on a day-to-day basis, moving files and things around and doing basic analysis and things of that nature, we really don't have to do that anymore. And I watch, again, on a weekly basis, automation and agentic workflows really start to impact people's work sites. And I will share that my own personal impression and my own workflows are nothing like they were three years ago.

Stan Waddell: I don't open apps to start writing email. I don't open apps to start writing documents. I do free form creation of knowledge or information. And I do think that that's one of the principle values that humans bring to a work chain is that we create knowledge and that's where I concentrate my workflow on now is how do I do the things that I do well and add value to the workstream and how do I let the computer do the things that it does well and free me to do more of that value creation that I do? But that is going to force us to change the way work unfolds and it's going to impact and maybe even eliminate some types of work. I think we as technology professionals have bought into and continue to talk about the line that work may not be, some jobs may not be eliminated, all jobs will change, some might not be eliminated.

Stan Waddell: And I think it is more of the some jobs, not all, but some jobs will be eliminated and most jobs will be changed.

Michael Cato: If I could pick up on the topic of tension and I really appreciate what you both shared already. One tension I've been observing and now wrestling with a lot more directly is the move from the seat license model that we've had for the first two and a half years or so of the generative AI wave to now a consumption or usage based model. And I'm curious if that shift is complicating or has implications for the work you're already doing, or am I the only one who hasn't figured this out yet? So Lee, I'll start with you first.

Liv Gjestvang: I mean, I'll just say I think that absolutely and I think the most direct place that we kind of dove in there was around our campus' use of LLM tools. And so in that space we ended up building out a multi-language model using OpenWeb UI on Amazon Web Services and have we pay for tokens for 17 different models across Anthropic and OpenAI that allow students, faculty and staff to be able to experiment with different models. They can do side-by-side comparisons and we're paying on a token basis as opposed to a user basis, which is just cost prohibitive for most institutions. And just as a side note there, I will say part of what I think we need to be really careful about is that these resources don't turn out to only be available to institutions or students at institutions who have extensive resources. We have got to be able to ensure that we're educating students in every type of institution to have access to these tools.

Liv Gjestvang: So this model was a great way for us to do an incredibly reduced cost model because we're paying for usage as opposed to the license-based model. And I will also say we're playing with how this works too. All of our incoming students this year are doing a foundational AI course focusing on ethics and kind of practice and use as part of their summer enrollment checklist before they come to campus. So all students will be doing that and then have access to those models. But that's one place where we've been able to change our consumption model. I think we're all watching very closely what's happening with the large enterprise tools that we use and how this will play out in that space.

Michael Cato: Really appreciate Lee. What about you, Stan?

Stan Waddell: Yeah. For us, we've been much more diligent about applying caps to consumption, especially on the ... And it's sad because it's the more capable models that consume more tokens and the tokens cost more per drink, but we really do have to be more vigilant and diligent about capping the use of those and helping people sort of understand more around what's the right tool for the right use case. So not to name any vendor's ecosystem, but as illustrative, maybe you don't need Opus if you're doing some simple document editing or document summarizing, maybe Haiku is fine and Haiku's going to be a much cheaper proposition for the university, going to be very capable, but there are scenarios where you're going to need to use the more expensive models and we're going to help you be mindful of that by giving you a little bit of a nudge that says you can only consume so much.

Stan Waddell: And then Leaf mentioned this as well, but I am encouraged by edge computing and what I mean by that is local language, large language models. Those models continue to become more and more capable and continue to be able to be run on less and less compute intensive environments. And so with a handful of GPUs, you can host some fairly capable models in your data center that your campus community can take advantage of at the cost of power and the upkeep of that hardware and it's not super intensive cost. So it's those kinds of things. And I think that's going to force us to think differently about when you use things, when you use certain models, when you don't, and how do you make more locally available models available to the campus?

Michael Cato: Yeah. And all of this is really helpful. It also suggests to me that part of what we're building now is a new portfolio of solutions, but the challenge with portfolios, it's harder to explain and there's some education involved in use this one for this one and this one for that one and here's how you can move your workflows across. I was also struck recently at Liv, it's similar at my institution, we have an AI, a chat platform, a gateway that we've had for a long time that you can move across models, but even there we've discovered that to your point, Stan, the use of specific models is driving up our cost six times this month than it was three months ago, just because not more people, but the same number of people having much more intensive exchanges with the models, which then drives up the cost.

Michael Cato: So we're realizing there's a lot more work to do there. Well, if we could shift gears a litle bit to our faculty and our instructors on the academic side as academic institutions, we know that not just our AI strategy, but many times strategies to pursue things that are new can sometimes be the place where it gets challenging for those changes to take hold or sometimes they can stall out. How are you all pursuing approaching AI in the context of the faculty relationships and maybe what's your honest read of where things were faculty sent and that stands right now broadly speaking or at your institution, depending on how you want to answer it first? Stan, I'll start with you.

Stan Waddell: Awesome. So I'll kind of start with Carnegie Bell is a little bit different beast again with the birthplace of artificial intelligence and technology is integral to pretty much all of our education domains. And so our faculty, I think, think a little bit differently about this set of tools. I think early on when organizations were struggling with academic integrity and whether or not students should be allowed to use generative AI. And there was some of that at Carnegie Mellon as well, don't get me wrong, but we weren't having the same types of arguments that other institutions were having. We were more focused on how do we put the tools into the students' hands, how do we help them understand the capability of the tools and how do we help them understand where they could be forestalling their opportunities for education by over reliance on the tools versus Liveraging the tools to get, again, more the less valuable work done so that they can concentrate on critical thinking and their own discovery.

Stan Waddell: And I think that that's been a through line for Carnegie Mellon as these technologies have continued to evolve. And again, we're at a place where we really are trying to educate the students on how best to use the tools and how not to let the tools have a very, very negative impact on them from their discovery perspective or not to have a very negative outcome and impact on society when they go out and they leverage these things in the professional arena. And so we are a little bit different in this respect. I also think that we're again, really interested in making sure that people have understandings around the capabilities of the tools and how they can impact Carnegie Mellon. So we see our faculty doing outreach, doing seminars, creating workshops, reaching out to industry, reaching in to Carnegie Mellon being parts of our educational pursuits around where the technology space is going and how to best use it and have really given us some great thoughts and ideas around how you build introspective programs and communities of practice at Carnegie Mellon.

Stan Waddell: And I think the outcomes have been very, very positive for us.

Michael Cato: That's awesome, Stan. Thank you. Lee, what about you?

Liv Gjestvang: Yeah, I think I will say that the advent of the kind of rapid drop on the scene in 2023 of AI tools I think has just through open a set of conversations that were met by, I would say, a mix of trepidation and at times enthusiasm from the faculty across our institution. It was really reshaping what does this mean about how we teach, how we assess. So I think what I've been really happy to see is just progression of these conversations over time. So what started for us with bringing ... I mean, we ran multiple, I think workshops and symposia with faculty with large, large numbers of faculty who came and were part of conversations early on in 2024. And then we started doing showcases that were featuring work, innovative work by faculty, students and staff. And I will say this has been a really interesting development for us.

Liv Gjestvang: I think it's the first time in my career that I have seen such open side-by-side learning of all three of those core populations on our campus where we're holding showcases where students are sharing work that they're doing alongside faculty members alongside staff or administrators. So that's been really fun to see. We did a lot of work early on with departmental listening sessions and just went to talk and learn about the ways that AI was impacting teaching and learning in different disciplines because it shows up really differently. And we did last fall a symposium for all faculty. The provost runs a symposium in the fall and that was focused on AI. And in this case, I will just say what I think this is an example of a model that has worked well for us. We had faculty showcase examples of the ways that they were using AI across disciplines.

Liv Gjestvang: So we had, I think about 10% of our faculty were showcasing their work at tables, a huge event in the gymnasium in the main facility there. And then we essentially put together a collection of those resources, hundred pages of examples across disciplines of what faculty ... Mostly this was faculty teaching faculty. So I would say there's a lot of curiosity and interest. It comes at a cost because we've all kind of mentioned the way the technology is changing so quickly, it's a huge lift. I think for us and our roles to stay on top of what these changes mean, it's a lot for faculty to try to stay on top of their discipline, how the technology shows up, and then how they can continue to integrate these changing models. But I think I've seen a lot of interest. The other thing, I mean, of course, in my context at Denison and I think across liberal arts colleges, we see faculty members who are thinking about in an intro writing class, AI may not have a place.

Liv Gjestvang: And so So how do you establish that and build those expectations out with students? How can that be communicated clearly? I'm seeing some faculty who are doing interesting things where they will assignment by assignment on a syllabus essentially describe what the learning outcomes will be for students who do use AI versus those who don't. And they can choose about the ways that they want to develop their thinking and skills. One other thing that I'll just add that I think has been really interesting to see is that in some cases where people are wanting to not have AI shape learning in a classroom and are going back to things like oral exams, I've heard a lot of enthusiasm for that. I talked to a faculty member in anthropology who said she wanted to really talk to students to make sure she could understand. She was able to connect with them around what they were learning and what they understood.

Liv Gjestvang: And so she did oral exams and she was a little overwhelmed. What would it mean to sit with students, all of these students and do those? And she just said it was really, for her, really meaningful and inspiring because she was connecting actually in a deeper way than she had prior to this kind of push for her to be really thoughtful around whether her assessments were measuring the learning she hoped to see.

Michael Cato: It's really powerful. Thank you.

Cynthia Golden: Yeah. One of the things I'm thinking about as you're both talking is what a wonderful opportunity this time that we're in gives us for true partnerships with IT, with the teaching center and with faculty members together. I think as you said, Liv, this is a heavy lift for people and I don't think faculty can always do it on their own. And that's where the partnership opportunities come in and I think they're really important. And we could probably talk about those all day, but just switch gears a little bit. Stan, at CMU, you're surrounded by people who are building this technology. So does proximity to the research make you more optimistic or more cautious about the future?

Stan Waddell: Honestly, a little bit of both. For some of the reasons I've shared earlier, but I would say in regards to the optimism, getting to watch the domains get pushed is incredibly enlightening for me. I'm sure the Carnegie Mellon community, but also very likely society because folks are continuously adding to the body of knowledge. Some of the things I'm most excited about are physical AI and autonomous science. And Carnegie Mellon's made of course a huge investment in robotics. That's one of the other things that we are known for along with our very, very fine College of Fine Arts, basically owning the Tony's and whatnot.

Stan Waddell: But we're known for artificial intelligence, we're known for physics, we're known for robotics. And so getting a chance to sort of see things like the National Robotics Center and our Flame Center for Artificial Intelligence and watching robots going, I mean, I've literally seen this in the time that I've been here, watching robots going from human beings programming activities and then watching the robots do those activities to actually watching robots go and become learning entities on their own, learning about their environments and how to interact with those environments. And you're seeing the real fruits of that starting to get commercialized now. And to be able to see that is just impressive. The new term is physical AI, not so much robotics, but also watching the edge expand in that respect. So this is the place where I think we get to be impactful is to help bring resources to the research community to help enable the work that they're doing and then to watch them utilize that in very, very meaningful ways.

Stan Waddell: One of the examples I've given this respect is, again, with the Flame Center, which is basically Carnegie Mellon's large language model research center, it's inter faculty collaborative with faculty from across Carnegie Mellon. And a couple of years ago they came to administration and said, "Hey, we really need some compute resources in the vein of GPUs that will let us compete across higher ed and maybe even compete a little bit with industry and the very, very large infrastructure that industry is able to field." And we were able to put 300 GPUs into that center in partnership with Google using their cloud GPU resources. The time to science was cut very, very short away from a year to provisioning to actually being able to sign the check and turn the stuff on the next week. And they are writing award-winning papers out of those resources. And again, they are pushing the boundaries on what's possible on some of the smaller models as well as making refinements into the way the larger models actually work.

Stan Waddell: And so they're improving those as well. It's just a super exciting time to be involved in technology and to be at a place like Carnegie Mellon for somebody like me that's a technology aficionado in and of myself, it's like a dream come true. But on the other side of that, I do worry about what the long-term impacts will be for society. And I think we've got to wrestle with those things around power consumption. We've got to wrestle with those things around the economic impacts as jobs are impacted and we've got to wrestle with bias and error handling for these large language models. They are not perfect. We all know this. We know they hallucinate and that's the big thing that we talk about, but they also have bias and we have to be careful that we're not letting them introduce bias into processes that are already biased and to amplify some of our worst impulses.

Cynthia Golden: So all important points. Haliv, given your perspective at Denison and your institution's identity, what about you? More optimistic or more cautious?

Liv Gjestvang: I mean, I'll kind of piggyback on some of what Stan said. I think there's a really interesting transition that is happening and just the depth of reach here. So we've taken our oldest academic building, doubled it in size and are opening this fall, a new center for data and analytics, computer science that is incredibly interdisciplinary. We're bringing in faculty who teach from across disciplines and have students and student projects. I think we were one of the first liberal arts colleges to launch a data analytics major and have students who are across disciplines working in data and analytics. The space has robotics labs, AI, VR labs. So a lot of interesting kind of new opportunities. And I think they're happening across institution types that are shifting, I think, some of the ways that we think and the combinations in which we think students need to learn.

Liv Gjestvang: We're a liberal arts college getting ready to launch a finance major. And some of these things, it's important to be able to sort of weave some of these threads together because they're just not distinct anymore and I think are continuing to come closer and closer together. So to me, there are some ways that the kind of emergence of these technologies into learning spaces are driving collaboration that they are elevating in a way that I would not have I think anticipate elevating these very core kind of human capabilities that are so important right now. So to be thinking about the ways that we know that today's leaders and tomorrow's leaders have to have the ability to analyze information, to make ethical judgments, to communicate clearly, to write, to engage in meaningful debate. The centeredness of those human qualities is exciting to me. And it's a place that I think we all know.

Liv Gjestvang: I know that these are critical. I am glad that in this moment in time that they are rising up as something that I think we are acknowledging as educational institutions and as a culture that we have to invest in. I think we need to do more because while we talk about it a lot, and we are investing in it in educational space, I think we need to ensure that those kinds of qualities are also driving both the developments and the reach of the large tools. It's a small number of companies that are driving a lot of this right now and just to ensure that these deeper, more sort of ethical, Stan referred to a lot of the challenges that people who are focusing on and engaging deeply with those human qualities are really at the center of conversations around how these tools evolve and continue to impact society.

Liv Gjestvang: So I can't say that I am ... I think I also carry both optimism and concern, and I just think that our continued deep engagement in these conversations, finding ways to show up and ask questions that need to be considered at the highest levels is really important.

Cynthia Golden: Thank you.

Michael Cato: Really appreciate that from both of you. So this is the Integrative CIO podcast, but we're always trying to be mindful that it's not just CIOs who listen to it. So for the members of our audience that are not in that senior most seat, but they're department heads, directors, aspiring leaders in their own right, what's a suggestion you might give on how they might be able to contribute meaningfully to their institution's AI journey? Liv, I'll start with you this time.

Liv Gjestvang: Yeah. So I think that individuals at any level, really in any department can find others to collaborate with. I think working collaboratively in this space is super helpful and it can literally ... I'll say we started early on with a staff collaborative with identified individuals from across every division of the college and that group just got together. It was 30 people initially who were working together to say, "These are problems I'm facing. These are solutions we're finding." And it was just kind of a brain trust. That evolved into a staff certification program where we actually ran a program for, I think about 20% of our staff participated in the first year to learn and build skills. But a lot of what we learned, having a program was phenomenal, but getting people in the room to work together, I think was at the heart of the impact of it.

Liv Gjestvang: So I would say it could be within your team. You could sit with your team of four people and say, let's find two hours and let's hack our way. Let's sit down with our computers and a problem that we face in our area and let's figure out what are some ways that we can try to use these technologies to accelerate our ability to solve problems, to offload some of the things that can be more automated. And so I think that type of partnership, there's a great book called The AI Driven Leader. Ethan Molack has a great book Co-Intelligence. Find a book to read with people. And I think building these connections with people either within or outside of your division, I'm seeing people do this with staff and students, staff and faculty groups, lots of different kinds of configurations. I think that's a great way to just start to learn what it is how you can use this and then have some insights that you can start to share out with people around you or above you who might be driving bigger initiatives.

Michael Cato: Stan, what about you?

Stan Waddell: Awesome. I love Leave's framing. I think I have some similar leanings, but what I would focus on is that yes, artificial intelligence is a science. And so the place where our researchers are engaging with the body of knowledge, that's scientific exploration. Where I think we differ and what I would sort of ask or try to impart on aspirational leaders is don't necessarily think of it as a science. As a practitioner, it's at this point much more of an art and we're literally limited by our imaginations and what we can vet out of the tools in regards to the output that they put that they bring out. So you don't want to get out in front of your skis in regards to being leveraging as a source of truth from an information standpoint, but you really do want to experiment with the tools to increase your understanding of what the art of possible is.

Stan Waddell: I think the big limiting factor that our campus communities have right now is that they want to engage with the tools in a very routine way like, "Tell me how to use it. " Well, I mean, I can tell you the basics, but if that's all you try to do, you're going to be very limited in how you'll interact with the tool sets. So what I tell my leaders is like, "Experimentation is keying here or is the ruler here. Figure out some proofs of concepts that you can engage with. I'm also going to monitor that performance. That's a performance metric for us and deliver some results. And a negative result is just as valuable or valid as a positive one. If you figure out something that is not great at doing, awesome. We won't try to go down that path as much, but then also be willing to share those experiences.

Stan Waddell: And one of the things I'm really proud about, Carnegie Mellon and our leadership team is that willingness to share. So we've got a couple of different things that we are doing with the campus community. One, we do an AI in day-to-day that's coming up not next week, but the first week of June where we do hands-on engagement with the technology. We also have speakers and whatnot, but we have folks showing what they're doing and the value that they're reaping from the tool sets. And that's leaders from throughout the technology organization. We have an AI Center of Excellence that our Business Innovation Office runs. It's a multi-day hands on engagement with campus. They can run about 12 people through that event for every session. They do about two a month and we're booked all the way through the end of the calendar and there's a waiting, Liz, that'll probably take us into the summer of next year and that's been the rolling thing about a six-month waiting list and we've been doing this for about a year and a half now.

Stan Waddell: Yesterday we had our Global Accessibility Awareness Day event and I was so pleased to see folks show up and demo how generative AI can be leveraged for people with disabilities to Livel the playing ground. People with visual challenges, being able to leverage generative AI to take pictures of environments and describe the environments, describe what's going on, read things that might not necessarily be accessible to them in the moment without special tools. Artificial intelligence is just doing that work. People with cognitive differences, leveraging artificial intelligence. And I'll share that. I mean, at this point, I've offloaded so much of my memory because as I get older, I can't remember as much, but I can take notes and artificial intelligence is such a great way to summarize notes and recall notes that I feel like I've got a memory buddy, but to see people demo things like that and help others understand where value can be derived, but also increase that imagination and creativity, "Oh, I never thought about that. I wonder if I could do this thing." We all have a role in doing that.

Cynthia Golden: And I have to say in my experience that when it comes to disruptive or new technologies that peer engagement and peer learning, there's almost nothing more powerful than that. It really can have a huge impact on the campus. I know we're reaching almost the end of our time, but we had just a couple more things to talk about. And so quickly, one of the things keeping with the CIO theme, we know the CIO role has changed a lot over the last years and how has leading through this AI moment specifically changed how you see the role itself? Has it changed your relationship to the president or to the academic leadership or to the board? Stan, why don't you comment and then maybe Live can jump in.

Stan Waddell: Awesome. Awesome. I would share that I actually have seen the role of the CIO shifting for many years. I think that shift was accelerated during COVID where people realize the strategic value of technology and how it impacts an organization's ability to reach its mission and imperatives. And generative AI, artificial intelligence, machine learning is the next sort of accelerator. So we're getting faster now and the number of conversations with leadership or with the campus community around what technology means and how resources can be acquired and put into their hands is accelerating almost in a logarithmic fashion. I mean, there's more work to be done than we can realistically accomplish. And so we're using generative AI to improve our efficiency so we can engage in more of these conversations and can drive and derive more value with the campus community, but we're not going back. We are not plumbers anymore.

Stan Waddell: We are strategic partners and as far out as I can sort of foresee it, there's going to be more work than we can do on behalf of the campus or in partnership with the campus, then we will realistically be able to.

Cynthia Golden: Liv, anything further on that?

Liv Gjestvang: Yeah. So I'm going to very quickly touch on two things. One is that we've had an opportunity to connect with much more closely with our State Department of Education. The State of Ohio passed a Senate bill last year that mandated every K-12 district to have an AI policy in place by July one of this year. And as we were talking to the district leaders, they were really looking for some guidance to help think about what does a good policy look like in a K-12 educational environment. And so we partnered with the state and ran have been running some convenings for leaders in K-12 to really think through what are their values, what drives the way their district works, and then how can they use that to translate into the ways they're integrating AI? And I will just say as a CIO, I spend a lot of time talking to and thinking about what's happening in industry and what do we need, what do our students need to be skilled at and what do they need to know to be successful when they leave college?

Liv Gjestvang: And for me, it's been really fun and interesting to be drawing this line in the other direction of what are K-12 learners coming with? How do we better understand that as higher ed institutions and then how does that translate into their lives after college and built this kind of stronger line across the education and work spectrum. So that's been really fun for me. The other thing I think ties to what Stan was talking about, which is just the breadth and expansion of work that we need to do has actually really pushed me to really actually deliver on something that I believe in, but I don't always get right, which is that we have to prioritize and we have to identify what's the most important work for us to be delivering as an institution in these wild times for higher education. Then how does my team identify what are the most important parts of that for us to do and then how do we really prioritize those most important things and figure out how we reduce or eliminate some of the less important things I just read a great newsletter from Rashad Tobakawala and he talks, I won't get into depth about it, but he foregrounds these three elements of empathy, curiosity, and generosity.

Liv Gjestvang: And we just ran our annual retreat with a focus on those three areas and thinking about how do we bring those qualities to our work and then how do we focus our workload so that we're able to actually deliver in a way that keeps us engaged and healthy and here for the long haul?

Cynthia Golden: I think that's an important way to wrap up this conversation because-

Michael Cato: Absolutely.

Cynthia Golden: Yeah. Thank you.

Michael Cato: That is powerful. Well, just want to say thank you, Liv. Thank you, Stan, for joining Cynthia and I on today's episode of the Integrative CIO Podcast.