������ý����

Maria Kuntz: Today's LBIdea is the more you use AI the more you'll understand how it works.

Our guest today is Jeremiah Contreras, an assistant teaching professor and the Kala and Shiv Khatri Endowed Faculty Scholar at the Leeds School of Business. In addition to being an award-winning educator, Jeremiah has been at the forefront of incorporating new AI technologies into business school education across various disciplines.

Thanks so much for being here today, Jeremiah.

Jeremiah Contreras: Thanks, Maria. It's great to be here.

Kuntz: I'd love to start with how you first developed an interest in artificial intelligence. When and why did it spark your curiosity?

Contreras: Yeah, I guess you could say I was a fairly late bloomer to AI, in general. Although, I started playing with GPT in November of 2022. This was shortly before the official release. They had a playground environment, and so I was playing

��around in OpenAI's playground environment. And then on November 30th, a couple of weeks later of 2022, OpenAI released their ChatGPT version, a branded version.

Kuntz:��So, I want to pause you right there. You said you were a late bloomer, a late adopter, but you were in their sandbox. So, I think you're an early adopter compared to most folks. Tell me about that. How did you get in the pre-release sandbox?

Contreras: Yeah. So, I guess if you think of AI as starting, maybe, back in the ‘50s and people that have been involved for 20 or 30 years, so, I guess, from the generative AI, the release of ChatGPT, I was definitely an early adopter there.

And so, I would say, you know, November 2022 is when it came out. And I will say, when it first came out, my first reaction was fear, you know, thinking about how students could use it, thinking about the different ways that it might, kind of, destroy our ability to think, or using it as a crutch. And I guess, shortly after, as I started using it a little bit more, within a couple of weeks, I really saw this potential of what it could do, how it could revolutionize business, how it could really change education. And so, I really started to embrace what this new technology could potentially do.

Kuntz: I mean, absolutely. It's changing everything so fast. I just saw a headline today or yesterday that said students realize faculty are using AI and they're frustrated. I was like, “Oh, the tables are turning so quickly.”��

Contreras: Yeah, no, absolutely. And I'll say, to address the student frustration, I completely understand that, you know, full disclosure is such an important aspect. At the beginning of my class and in my syllabus, I really talk about how I use AI in the classroom. If I ever have AI help me create a reading or some kind of materials for the students, I put at the bottom that it was created with the assistance of an AI tool, whatever tool I've used.

You know, I think that disclosure and being upfront with it is important, not just from the student side, that is important, but also from the faculty side.

Kuntz: Yeah, I think it normalizes it. It also means it's easier to talk about ethical use and fair use. I gave a presentation and I used AI to help with the slides, and I said, “Well, I'm presenting my AI and I used AI to help.” So, you know, just that transparency.

Contreras: Yeah. And, you know, I think, from a student's perspective, a lot of times, they do see it as a shortcut. So, they see it as short-cutting the learning process. And I think, if we can show them how it can be used for efficiencies, right, creating a slide deck in seconds, what that can do to someone in business, you know, when you have a role. Now, it doesn't mean the content can just be created by AI, but if we give it the substance and we give it information and then it can help us make a solid product faster, that's just such a more efficient way to do business. And so, I think it does, kind of, model what it might look like in the business world.

Kuntz: Yeah. So, actually, I want to go a little bit more general, just what is Leeds doing around AI right now?

Contreras: Absolutely. So, it was, within the past year, Dean Khatri has kicked off the Leeds AI Initiative. And so, over the past year, we've really been spending time thinking about, “Okay, how do we incorporate AI into the classroom?” And one idea was to create an AI class that students could take, either mandatory or optional. But the other thought was, well, AI isn't necessarily just a topic. It's really embedded in everything we do and it's becoming more and more so. And so, what we decided to do is we have 15 core business classes, and so we made sure that we're adding AI into every single core class that all of our freshmen and sophomores take.

So, a big undertaking. A lot of courses are, about 50 faculty, and so a lot of people to manage. And so, we took five of those classes and we have a very coordinated approach. We have a class that talks about prompting, a class that, you know, incorporates ethics. It's our ethics core class. We talk about ethical AI. We have a business communications class, so we talk about AI used in communication.

So, we take a couple of those. And then every other class, like, marketing or accounting or finance, they may just have a little section that talks about how's it used in our industry.

Kuntz: I mean, I love this. The students going through this program are really just being immersed in that sort of thinking, that it's here to stay. Let's figure it out. Let's use it well. Let's use it right. And they're going to be, really, I think, ahead of the curve when they get into the work world.

Contreras: Absolutely. I really think Leeds has an opportunity to become a leader in this space, among business schools and even among other colleges and, kind of, have a roadmap for how AI can be incorporated into the classroom and into education.

Talking with employers, I spend a lot of time talking to firms and talking to companies, and that really informs a lot of what we do in the classroom. What are they looking for from students? What is the value add? There's a saying that AI isn't necessarily going to replace people, but people who understand how to use AI effectively will replace those who don't.

Kuntz: So, you've talked about using AI effectively. You know, what does it mean to engage AI well or to use AI effectively?

Contreras:��So, I would say that it's treating it as a partner more than a doer. And so, it's a collaborative effort. There's a saying that AI could be reversed and you could call it IA or “intelligence augmented.” And so, when we use it to augment our intelligence, not replace our intelligence, I think that's when it's the best partner.

And so, from a student perspective, it's hard to understand where that line is. If I'm learning a topic, where do I trust the AI? Is it going to lie to me and give me false information and what we call hallucinate? And so, it can be really challenging from a student perspective. So, that is one of the hard parts about learning how to use AI. But that's one of our roles, is to really show them what that looks like and the model, how do we use it as a thought partner.

Kuntz: I guess, like, so much of higher education, it's not black and white, right? It's a critical thinking exercise of, how do I use that? Where is the line of proper, correct, or ethical use versus not ethical?

Contreras: Absolutely. Critical thinking is, I think, that's probably my biggest fear, is that, if there's over reliance on AI, and students starting in middle school or high school are going to start using these tools, and if they're not really using them in a proper way as a partner, if they're using it to replace the activities and do assignments for them, they're going to lose the ability to think critically. And so, our job in higher ed is going to be to either retrain or, hopefully, even help K through 12 figure out how to address it in the classroom so that they can come to us with the ability to think critically, ask good questions, and really be able to go through the process of learning.

Kuntz: Yeah, I think these years ahead of us are going to be really interesting, because now, you know, you have students entering the university who have spent enough years in the K through 12 environment. I'm thinking about high schoolers. I have a few of them at my house. And they're getting mixed messages, like, don't use it. Just don't use it at all. It's all wrong. It's all bad. And then, I don't think that's actually the right way. I think it's too far to one end of the spectrum. And how we use it well is really the area of learning and growth. And it's going to be gray.

Contreras: I think that's absolutely right. There's a time and place for all sorts of tools. There's a time and place to learn math without a calculator. And then there's a time to introduce that calculator. And so, I think we really have to rethink, how does AI fit into that?

Kuntz: I love that analogy.

Contreras: Mm-hmm. What does it truly mean to learn something? What is the amount of knowledge that I need to be able to apply a topic? Like, you know, math, we learn addition. Do I have to know what 32 plus 64 is? Or do I have to know the basics of how to add and then I can use a tool? Do I have to know when to add and what it's doing? And how complex do we need to get? So, using AI as a supplement is important to know how to do properly.

Kuntz: Yeah, I love that. Well, let's move on. What do you see as some of the biggest misconceptions about AI in the classroom, either from students or faculty, or, maybe staff who you're working with?

Contreras: I think that there's a misconception that everyone's using it to cheat, for one, or everyone's using it to take a shortcut. But that almost becomes a self-fulfilling prophecy. So, I think, when students think that other students are using it to get a shortcut, they may be tempted to try to use it as a shortcut.

I think we do see this split between some students who are just staying away from it, partly because they don't know when to use it. And so, I think there's also just a mis… I don't know if it's a misconception, but just a misunderstanding of when it's appropriate to use. And so, from a faculty perspective, I think the misconception may be that it's okay to not address it. Students really need that clear guidance. They need to know, is it okay to use? If so, how? And if not, just let me know. And then, most students are going to follow those rules. But I think the lack of clarity, the ambiguity just leads to uncertainty.

I think a truth to that, though, is that a lot of students, because they don't know how to use it effectively, are going to misuse it. And maybe not intentionally, but, you know, the use of it, if we're not showing them and modeling how to use it properly, it will be used inappropriately.

Kuntz: This strikes me as so much a conversation about trust, right? Like, when we know how to use it. And we build trust by having clear guidelines, right? So, the faculty are saying, “This is what I expect. This is how I encourage you to use it.” And the students understand that. And so, it'll increase trust, which would, I think, accelerate learning in the classroom.

And also, I'm realizing faculty have to, now, learn and really be ready to teach and have conversations about a whole new topic that's affecting every classroom beyond what their area of research expertise is.

Contreras: Absolutely. There is, I think, a misconception and a fear of AI from a faculty standpoint. There's this idea that it's incredibly technical. And I think, really, it's just about communication with a tool that speaks our language. And so, when we break through that barrier of fear and just encourage people to play with the tool, I think that, the more people use it, the more they understand it. And so, absolutely, I think faculty have to overcome that initial just misunderstanding of and not wanting to address it.

Kuntz:��So, earlier you mentioned AI's been used in the 1950s. And just for our listeners, some of them might be early adopters and have been using AI for a long time, but maybe some of the listeners haven't. So, I was hoping you could just share a little bit about when and where do we already see artificial intelligence integrated into tools and technologies that predate ChatGPT and Claude and Perplexity.

Contreras: Yeah, AI has been around for a long time. And I would say, Google Maps, Apple Maps, whenever we put in to go somewhere, you know, the traffic algorithms and the way that it's analyzing real-time data, that's a form of AI. Self-driving autonomous cars is another form of AI. If anyone's ever gotten a notification from their bank about possible fraud, that's AI.

Kuntz: Oh, yeah.

Contreras: It's probably important to, kind of, distinguish different types of AI. So, before these tools like ChatGPT, this is called generative AI. Before those tools, you had things like machine learning. And so, those were a little more predictable. And that could detect fraud. So, from your bank, it would look, you could give it a lot of data and it could find these patterns of what fraud looked like. And so, then it could see, if it saw those patterns in your credit card, it would flag it as a potential fraud.

And so, that was machine learning. It was predictable. It was repeatable. You could trust a little bit more, rely on the information that was coming out. With generative AI, like these tools like ChatGPT, they are what we call probabilistic. They're different. So, if you put in a chat, it's going to give you a different response every time. So, they're a little more creative. They're not as predictable.

But AI has been around for a long time. This new version of AI, I think a good way to think about it also is, we used to have to learn to speak computer. We had to program to talk to computers and to write code and to do all of this. And what generative AI did is we actually taught computers to speak our language.

Kuntz: Our language, oh.

Contreras:��And so, now, we can communicate with, you know, I could build an app and I don't know a single line of code, but I can tell the system what I want it to do and then it can translate that into code. And so, it's just really changed how we interact with computers.

Kuntz: Wow. That just gave me a whole new idea for folks on my team who work with HTML. I mean, I've used HTML and I'm like, “Oh, I never even thought about trying to work with it to develop code when it's being tricky, even though I know folks who do it.” So, hey, I got a new idea.

Contreras: It's amazing. You can copy and paste code in that you're working on. It gives you fixes. And yeah, it's pretty impressive.

Kuntz: Great. Okay. So, you said you talked to a lot of companies, industry partners to see what's happening in their fields and what they need from students in the workforce of tomorrow. So, how do you see AI being used in those companies and workplaces?

Contreras:��So, companies right now are really having to rethink and reimagine how business runs. And so, I think generative AI has a time and place in certain aspects of business. It's constantly improving, so that's going to change. But I think generative AI also makes AI like machine learning. That's a little different type of AI. It makes that more accessible for companies.

And so, what this new AI era is doing is businesses are having to really rethink, where can I use it? I think one obvious place is increasing productivity. So, when it's used as a thought partner for people or to create PowerPoint slides as an example, when it's used to make you more efficient, you still own as an individual, as a worker, as an employee, I would still own the judgment. I would still own the product and the output, but I have an assistant or I have a partner who can help me become that much more efficient.

So, in that case, it's definitely changing how people do their jobs. It's getting to the point where they're creating these tools that are called agentic tools. Right now, if we use a chat, we ask it a question, it responds. We have to ask it another question, it responds. An agent is something that you could give it a task and it could complete the task A to Z several steps, several iterations. And so, it could start to take these jobs that maybe have less judgment and require less thought.

And so, I think it's really just changing. And so, a couple of things are happening. I think companies are seeing, “How can I use the tools that we have today?” And that's very promising. And they're seeing a lot of efficiencies. But you also have to consider to have that competitive edge as the tools progress, where am I going to insert them?

Kuntz: Yeah, “What's next? What more can we do?” Not just, “Great, we could do this today.”

Contreras: Yeah.

Kuntz:��So, also, we need people who think about that, right? We need folks who are business students and strategists who really understand what it takes to run a business, who then can apply thinking about AI and AI tools and technology to become strategists for the business.

Contreras: Absolutely. You know, education, oftentimes, is a bit lagging when it comes to incorporating technology. I think we really have an opportunity to help become part of the leadership, as far as predicting and foreseeing what can happen and almost being an incubator space to test some of these business ideas and help students grow these skills so that, when they enter the workforce, they're able to help these companies grow.

Companies I talk to right now, they really want students to come and try to solve problems. How can we incorporate AI into our business? People know that they should be thinking of AI, but when we have our jobs and our day-to-day, try adding on another full-time job and learning about AI and, how do we fit it in.

So, they want students to come in and think about, “What can we do with these tools?” And so, we have an opportunity in education to really help our students get to that next level.

Kuntz: It's the next C-Suite role, the Chief AI Officer.

Contreras: Absolutely. Right.

Kuntz: There's going to be a whole new set of careers in this field. Well, what advice do you have for people who are maybe just starting to explore using AI?

Contreras: Yeah, you know, I talk to people all the time, actually, that are just getting into AI or maybe haven't even heard of all of the tools. And I think the best way to think about it is just to play. If you think about animals and kids and humans in general, we learn so much by playing. Playing with blocks as a kid, you learn so many different things.

And so, I would highly recommend there are a lot of tools that have free versions out there. And so, OpenAI’s ChatGPT has a free version. We have Google has Gemini and NotebookLM. Anthropic has a tool called Claude. So, there's a lot of tools. And so, getting a free license and then just playing. I've used some sort of generative AI tool every day since the November 30th, 2022 release. That's a lot of days.

Kuntz: Yeah.

Contreras: And so, I use it for everything — cooking recipes, parenting advice, relationship stuff, how to rethink a class. And so, I think the more you use it, the more you get familiar with how it works.

Kuntz: Yeah, I'll tell you a story. I used it recently to create a training plan to train for a 10K. And it was really great until it wasn't. And it really had me overdo things one weekend, and so, now, I'm nursing an injury. So, I really learned through play, right? It was fun, but I also learned what was so obvious in some ways that every output does need a good editor. And I trusted it a little bit too much. And when I spoke to some friends who are more experienced runners, they were like, “Yeah, you probably shouldn't have done back to back long days.” And I was like, “Okay, duly noted.”

But it was great for integrating other things I was doing, right? I said, “Oh, but I do yoga on these days, so adjust it this way.” And I had worked with it, but I missed something that I just… I didn't know I wasn't an expert in. And I didn't tell enough information about me, apparently.

Contreras: Right. And trusting that spidey sense, right? Something that, if we, kind of, think, “Well, does that seem right,” maybe validate it somewhere else, double-check it, absolutely. But it's hard to know when to do.

Kuntz: Yeah. I'll listen to my spidey sense next time, for sure. So, looking to the future of AI, can you tell us a bit about the Colorado AI Act?

Contreras: Yeah, the Colorado AI Act was recently introduced. We were one of the first states to actually incorporate an AI law. It doesn't go into effect until February of 2026. And so, they're still working out some of the details. But as it's written, it addresses something called algorithmic bias. And so, it addresses when systems make a decision. And so, they define it as, if you're using an AI system that's a high-risk AI system, which is a system that either substantially makes or makes a decision that affects someone's wellbeing or their opportunity, then there have to be certain parameters around it. So, any user of AI system that really affects someone's wellbeing or their opportunities, an example would be, if it's used to make hiring decisions or for a college for admissions, or if I were to use it for grading, that could affect someone's opportunity.

Kuntz: What about healthcare or insurance? Are those industries… would those fit into that category?

Contreras: Absolutely. If it's making a decision on healthcare, so it's not that it can't be used. The law really talks about it. There has to be a process around disclosure when it's being used. There has to be a process for someone to come back and ask for validation or verification of how something was made. And companies have to take on the ownership of documenting and just testing to make sure that the system is working the way they want it to work.

Kuntz: Well, so hypothetically, you're a consumer and you find out that this decision which is affecting you, and you're not pleased with the decision, you could go back to the company and say, “Well, you disclosed to me that AI made it, and I'd like a human, a person, a real intelligence to review the decision.” Is that, sort of, where it's leaning toward?

Contreras: I think there's some ambiguity of what it would actually look like. That's, maybe, an initial interpretation. I think the way it's written, the state attorney general is the only one who could come forward to a company and ask them to remedy this situation. And so, I think we're going to learn a lot as we get a little bit closer. They're still trying to fine-tune some of the aspects. And there are people lobbying different sides of this to make the law weaker or to just change it in some regard. So, I think, as we get a little bit closer, we're going to start to see what this really looks like in action.

Kuntz: This is going to be an interesting year. I mean, we're in May to February. So, we've got nine months or so to see what happens.

Contreras: And in addition, there are other states around the country that are also adopting laws. So, we do have a handful of states. I'm not sure if it's 7 or 10 at this point in time. But other states have different types of laws, some that address data integrity and information, others that also may address algorithmic bias. And so, states are taking a different approach to how they're going to regulate AI. So, I think we're going to have this landscape where we really are going to have to figure out what is the best way to balance both, not stifling innovation, allowing innovation to be created, but also to protect individuals and to ensure that it's being used responsibly?

Kuntz: Really fascinating. So much there. I mean, what else? Is there anything else you see on the horizon when it comes to generative AI?

Contreras: You know, in general, I think that there's a lot of talk of something called artificial general intelligence. And that's when AI gets to a point where it can act as competently, use judgment, and make decisions just like you or I. It can work, you know, at a Ph.D. level and really make decisions. And then you get to artificial super intelligence where it starts to surpass where we are.

And so, there's a fear that it can get out of hand. So, I think, making sure that we have that relationship and that, together with society, that we're approaching it responsibly and that we're in the know. I think it's important for people to be engaged with it, to learn a little bit more about it. I think that's really important.

But I do think, on the positive side, it really has the potential to level the playing field when it comes to education. If people have access to these tools that couldn't access a tutor before, but now they have access to something that can help them learn a topic, people can learn things that they weren't able to learn before.

When it comes to medicine, it can be used in replacement or in addition to a doctor. In third world countries right now, people are actually able to get medical advice. That's much better than having no advice.

And so, I think it's really going to revolutionize, it has the potential to level the playing field when it comes to equality and access to…

Kuntz: Access to information.

Contreras: Yes, absolutely.

Kuntz: Wow. I mean, it really is like the internet in that regard, right? I mean, you mentioned developing countries. And I think, over the last few decades, the conversation about access to broadband and then Wi-Fi and how, once you have wifi and you have affordable phones, now people have access to the internet. And now it's access to real-time information beyond what you can Google — really specific tailored information.

Contreras: Absolutely. I mean, it's really just a game-changer. And while it does have the potential to level the playing field, it also has the potential to make a wider divide between areas. I think, you know, ensuring that there is some level of access is important. And that's what I'm wanting to do in education and make sure that we do provide tools to our students and give them access. So, it's not just people who can afford it.

Kuntz: Well, this has been such a fascinating, really great conversation. I have a lot more I would love to keep talking to you about, but we probably should wrap it up for our listeners. Are there any resources you recommend for people who are interested in learning more about AI or staying updated as it evolves?

Contreras: You know, there are some YouTube creators that I do follow that are pretty reputable. There are a lot that are, you have to, kind of, weed through. Google has some really good resources and videos. There are some LinkedIn videos as well. And there are a few websites that share some tips and tricks. And I'd be happy to share some of those with you if you can share that with your listeners.

Kuntz: Yeah, we'll put those in the show notes. That's great. I'm sure folks would like a vetted resource. And Jeremiah, if people would like to follow up with you, what's the best way that they can contact you?

Contreras: Absolutely. I would say, find me on LinkedIn, Jeremiah Contreras at ������ý����. I can also be found on the CU website.

Kuntz: All right. We'll put your links in the show notes. Okay. It's going to be super fascinating to see what comes next when we're talking about AI. And Jeremiah, I can't thank you enough for sharing your insights and expertise with us today.

Contreras: Maria, thank you so much. This has been fun.

Kuntz: Thank you again for listening to Leeds Business Insights. Make sure you're one of the first to hear every episode by subscribing to the show, wherever you get your podcasts. Leeds Business Insights Podcast is a production of the Leeds School of Business and is produced by University fm. We'll see you next time!