A Wharton AI Research Leader's Formula for Responsible AI

To make good decisions with data, you have to do a lot of thinking yourself without data first, annoying. Exactly what it is that you're trying to achieve. Hi, I'm Cindy Hausen, host of the Data Chief. If you've heard we talk about data and AI before, you know that I believe everyone, not just data teams should be able to get the insights they trust. That's why I'm proud to say that Thoughtspots sponsors this podcast. Thoughtspots' authentic analytics platform lets you simply ask a question in natural language and get clear, governed answers right when you need them. No fuss, no waiting. It's why companies like Cisco, Lyft, Hyatt, and Roach count on Thoughtspot. See what the future of analytics feels like at Thoughtspot.com. Welcome to the Data Chief. I'm your host Cindy Hausen. Today, we're joined by Stefano Puntoni, co-director of Wharton Human AI Research and professor at the Wharton School. Stefano is a leading researcher on how AI is changing decision making, work, and society. In this episode, we'll explore why data and AI insights so often fail to translate into action, and how psychological threat and change management shape the success of AI inside the organizations. Stefano, welcome to the Data Chief. Thank you for having me, Cindy. Great to be here. Yeah, and where in the world is here today, Stefano? I'm in Philadelphia. I'm at my office at the Wharton School. Philadelphia, just a little bit north of me, but I feel like I should ask you, what is your favorite town or city in Italy? In Italy, or I'd say Florence. I grew up in Tuscany, so I will be by. Yes, yeah, no, Tuscany is beautiful, beautiful. It is perhaps one of my favorite countries to go visit in the world. I recommend it to everybody for vacation. It's not, you know, Tuscany is not the best place maybe for AI professor, and that's I think, you know, the US is a lot more exciting that way, but in terms of like, you know, enjoying life and who the wine is great. Oh, for sure, the wine, the wine. I'm picturing a village in Montalcino atop a hill. Yes, you'll have to come visit me in Lewis, and we will open this huge bottle of Brunello. That would be fun. So, Stefano, you are a prolific writer, researcher, you're educating executives and youth on how AI is changing the world. I want to ask you about a point that really stuck with me, though, in one of your books, Decision Driven Analytics, and how sometimes more data does not lead to better decisions. Why is that? I think that we have this almost like instinctive reaction that when we see intelligent algorithms and amazing data systems, we take them almost as an excuse for, you know, thinking less. Because the machine is hard, then we can rely on them for insights. We don't need to think it's hard. I think that's a common fallacy, but I think it's the bad one. Because, you know, way the smarter the machines, the smarter we've got to be. And so, I advise everybody that I think to make good decisions with data, you have to do a lot of thinking yourself without data first, and knowing exactly what it is that you're trying to achieve. Framing the problem, clarifying alternative courses of action. How do you even know what success looks like here? And what are you doing in the first place? So, I think a lot of these questions are not very well thought through in a lot of data analytics programs and that ends up often being an issue because there is a chance in being created between the decision makers, the subject matter experts, the people are accountable for decisions like burn managers, product managers, people like that. And on the other hand, the technical function, the analyst, the AI engineers, and people maybe lack domain expertise, but they do have deep expertise on the tools. I think not having that good communication, not knowing why we're doing this except what we're looking for, then leads the analytics to be, you know, but useful. Yeah, so that's interesting. It's almost like you're saying the pre-work is more important or equally important as the insights themselves. Yeah, in a way, it's like, you know, you get what you put into it. So, if you thought very carefully about what you're doing and why, it's much more likely it's going to be helpful than if you were not. Now, to be clear, there's a lot of insight that can be gained using just data mining and the sproletorian analytics just to see what you find in there and people can learn from data. But a lot of the times that's not what you're trying to do, you're trying to know, for example, whether this ad is better than that ad or whether this target audience will appreciate this feature or whether it's, you know, whatever it might be. And so in that case, you have to have thoughts through exactly what are the options, how do we know which option is best and what kind of data do we need in order to answer that question. And that is very, very more of a conceptual challenge. It's not a data challenge. So it becomes a data challenge eventually, but I think first you need to have done a lot of thought. I think a lot too many companies, basically, you know, see the data as a value in itself, while data is just a means to an end. Yeah. And so is that almost, is that getting worse with AI? You've written and researched about cognitive offloading. Are we, are we offloading even more of that thinking to AI? Or how do you think we're doing here? Yeah. And I think I think it can actually, you know, the progress in technology, almost paradoxically, can exacerbate the problem in two ways. I think the first one is just like I mentioned before, as your machine gets smarter, smarter, we might rely on it. You mentioned this notion of cognitive offloading. I think something that many people worry about. And, you know, it's only natural. We have biological systems. So we are hardwired for energy conservation. If we give a chance to save energy and be lazy, we will do it. So you can count on that. And second, you know, the technology is getting so advanced and also complex that is going to take now a lot of technical know-how to fully understand what it is doing. And so you have this, you know, threat almost for a lot of the physio maker thing. I don't understand if I don't know what it is. I have the technical people taking care of it. And it's just not for me. And I think that's not the right, the right answer. I don't think the point will be for the decision maker to become an expert on the analytics. Everybody has their own competence and the should be division or labor. But they need to be involved so that we make sure that those analytics efforts are functional to whatever goal the company has. Because the people of the world, the media can't, will make the decisions of those who know best, you know, that they are the land or what it is supposed to be doing. Yeah, for sure. And I, as somebody who's worked at the intersection of data and analytics for 30 years now, I believe that the closer we put the data and the insights to the person who actually has to make the decision, the better the outcomes will be because they know what they're looking for. Or they have more hypotheses than, let's say, the order-taker data analyst, Ashboard developer. So if you think about how potentially game-changing conversational AI can be in bringing the data closer to these decision makers, you've discovered some interesting observations. Can you share that with us? Yeah, many, many things happy to talk about in this space. I think one, one thing you mentioned, I think it's interesting what you're saying here. I think conversational AI, it's a device that you can use to bring decision makers closer to the data. And I think that's a deep insight, if you think about it, because as the cloud gets more complicated, the statistical models can more sophisticated finding the data, managing the data across different systems, become more complicated. The decision makers naturally are relying more and more on analyst and technical people to support that decision making. But that means that they're this growing chasm I was just talking about before, between decision makers and the data. And what they I can do, some extent turn the analytics function is something that looks much more like a self-serve kind of function, where for at least a lot of things, I want to say everything, and certainly there's a lot of roads for analysts to contribute. But for a lot of things, you can now have a manager who has maybe a hunch or maybe just a curiosity and can explore it using data in a way that before maybe they would not have done it, because now they need to submit a query to the analytics department that someone has to basically respond, and they'll get back to you, and then they wait two days, and then you get your crosstab or your graph or table or whatever it is that you're trying to get. And now, you know, let me, I have an idea. I think that the basic discussion segment might actually be more sensitive to this than we've given it clear to or whatever. Then, you know, maybe you think, we've got the data because we're under campaign, you know, because we've got that vendor or whatever it might be. and then just go on conversational AI tool and pull that data and test that type of this right away instead of having a bit. So in a way you can say, and this is a more general point about innovation in organizations. I think one big role of AI in organizations is to lower the cost of action. But you're making it easier in many situations where you have a balance between cost and benefit and say, there is some uncertainty about the benefit of trying something or the likelihood that something will work. And so now sometimes it will be on the fence and say, is it worth my time, is it worth investment? Do I do it, do I not do it? Maybe they pay office on a BSB or maybe the likelihood of its exceeding is too low. And so you don't do it. And now you've got AI, so testing, analysis, even implementation is so much easier and faster that it dips over the balance to action. And I say, you know, let me try. Maybe it is low probability, but you do enough of them and then I hit the winner. So I think in terms of innovation, it's not just the data analytics function. I think everything, new product development, customer service, a lot of different things. And which adds as you work at the intersection of marketing as well, it's which ads will work, which audiences and personas do we target. How do we even think about segmentation? So you had an interesting point that it lowers the cost to action. I think it gets us beyond that analysis paralysis or just flying blind because you can't get to the data. So clearly leaders are excited about this, but sometimes workers are less excited. Tell us what the gap is here. Yeah, so that's something that we see in our data. We run every a Genai adoption tracker and we do see that Russell levels in the organizational hierarchy. You see marked differences. In the way that people feel about AI, both in general in terms of their own firms efforts. And this kind of divide has been demonstrated in other data sets. So I think it's pretty robust. And the pattern is this one. People at the top of the organization tend to be very bullish about AI. They see it as core competence to be developed. A massive accelerator of lots of things that we care about, including ultimately the bottom line. At the same time, people further down the organization feel threatened by this technology because it holds the promise of transforming organizational functions, affects the value skills very much. And that's a big deal here, invest a decade and lots of, you know, basically, blood and treasures to develop a set of skills. And now there are questions as to the value of those skills moving forward. And I think leaders are not helping with a lot of the times when leaders talk about AI to the press and, you know, court-learning announcements or, you know, press releases or whatever. Typically, the shape of that conversation is very much focused on productivity. And productivity is code word for, you know, headcount reduction in the years of a lot of workers. And so now, on the one hand, I hear this conversation inside where the leaders are sending us all these memories saying, "You've got to do this, you've got to do that, you've got to get on board with this, you've got to love it." And on the other hand, when they speak outside to investors, they're basically, basically, they tell them, "We're going to use AI to file lots of people." And then why should you be surprised? Then people inside are skeptical and, you know, look at this more with anxiety than with excitement. I think it's only normal. Yeah, for sure. And listen, mainstream media does not help here. I think about the Wall Street Journal article early in the year, "Use AI or be fired." Every week, we're seeing headlines about mass reductions. So it, I don't think these fears are misplaced. And yet at the same time, your research is showing that some people will outright sabotage adoption of AI for the good of the company because they go into self-preservation mode. So how should leaders and even workers, if I am, you know, a marketing campaign manager, how should I be rethinking my role with AI? I think you can think about it from both perspective, leadership and, you know, the same management of the workers who implement programs. There's a lot to be interesting, think to talk about here, but just to sketch a couple. From the point of view of the leaders, I think you've got to give very careful thought to how do you harvest the energy of the organization? You need basically vision and investment from the top. You know, the leaders need to know why we're doing AI. AI is not a strategy. AI is just a tool. So what we're trying to achieve, we need to make available the resources and, you know, whether those are technical support, compliance, and say, whatever it might be. And to make those tools available to people across the organization, and then it needs to bring basically energy and vision and communicate authentically about, you know, the path ahead, honestly, too, because obviously there might be also uncomfortable conversations to be had, but there could be transparent and be genuine. And on the other hand, that's not enough, because you also need energy from the bottom. The people at the center don't know how a particular job is going to be impacted potentially by the AI. Only the people doing that job and the people that write above them know that. And so how do we, either you bring in outside consultants and the level you or you're going to involve people in the organization to co-create a template or a formula to make this work. And nobody knows how to make it work. And the way that it works for one organization, I actually end up being quite different from the way it works in a next organization because of the differences in culture, scale, assets, capabilities, and, you know, whatever. And so you need somehow to develop AI problems that meet in the middle where you've got energy from the bottom and energy for the top and somehow we make it all work. If you have only energy from the bottom, leadership is engaged, they lack investment and vision. You're going to have lots of people who are experimenting to the AI because they'll see the value of making the jobs better or easier, but there's going to be a lot of fragmentation, a lot of liabilities and risk potentially where people are doing lots of stuff without, you know, kind of support and supervision. And that is probably not going to work very well for the organization overall. If you do only things at the top and you're like a baby shoveling down the throat of people, that's not going to work either because half of you people want the program to fail, it will fail, there's nothing you can do about that. So somehow you need to find a balance between the two. Yeah. And I think one of the things that most surprised me in your research is the realization that somebody is more resistant to adopting AI if it threatens their identity. And you gave a perfect example unwittingly, but I love my juror coffee maker, Swiss coffee maker because I'm not a barista. And so just pressing a button and having it filter the water make a perfect cup of espresso. I'm happy to offload that to a machine to AI. But if my identity is as a CDO whisperer or a trusted advisor or even a dashboard developer, and now you're going to have AI replace that part of my identity, it would be pretty soul crushing. So what are some of the techniques that you recommend that leaders address this identity equation of AI? Yeah. So I was just mentioning before the perspective on the top, how do I understand deployment so that I basically channel energy from both the floor and the ceiling kind of thing. And now you can turn it around and say, if I'm a worker, I see this technology coming up and the trajectory of the technology is still uncertain. We don't know where friends, but it's obviously still getting better every day. And so we should expect that to be the case for the time being. And so when you look at that trajectory, for many people that is a concern about what does it mean for me, what does it mean to be the kind of person? Will we even have the kind of personal organization in the future? And so I think basically as a worker, as a professional in you were mentioning now people in the data space, but I think this is a threat to lots of different types of people. If you are an advertising copywriter, if you are a manager, if you are an accountant, if you are a financial analyst, whatever, consultant, whatever, there's so many different professions that are going to be transformed by eye. I think many of them will not go away, they're just going to change. And so the question is, do you understand the direction of travel? And hiding your head under the sun is not going to make you go away. You know, you can think about it from the point of view of the kind of the stake and say, I think adoption of this technology, learning what it means from your job function, your job specs, and basically your career trajectory moving forward is going to be crucial to actually making progress. And so you know, get excited about it, because I think now it's a moment where many things are in play. And if you can be a leader in helping that change management process and information, you have amazing opportunities ahead of you, I think. On the side of the stick, so to speak instead, if you If you're just pretending this thing is not there, right over here. So you basically have to somehow learn to adapt. There's many things that people can do. One, we call it direct resolution where you tackle the challenge head-on, you basically upskill yourself so that you make yourself the best possible complement to this technology, which you don't want to be the substitute to the technology. Because if that is what you do, then there's no future. If you're a complement, actually that technology might be a multiplier of your productivity. And in many situations we know based on economic principles that if you have a productive asset, just got more productive, you want to use more of it, not best. And so there'd be more opportunities for you. And so things like, you know, do you know how to buy code? Do you know how to leverage these tools for whatever workflow? Can you actually think carefully about how you re-engineer completely a workflow around this and then lead that change? That's direct resolution. Then you have something we call fluid compensation. And that one basically means to say, okay, there are a bunch of tasks and skills involved in this job, in this space. And I can see that AI is basically making great strides towards taking over some of those. What I need to do is to pivot and figure out which ones are not being taken over by AI. And make sure that I'm as good as possible at those. So imagine a marketing manager, maybe in B2B. And they were spending a lot of time doing content marketing. And now they realize that a lot of SEO or writing content marketing kind of tasks are going to be automated. What do I do? Well, now the question shifts from writing the content to what content needs to be written. And so now you could desire, okay, I'm going to be able to strategy and say, my goal is going to be not that to create content, but that to work as trading as a swarm of agents. I'm going to do lots of stuff around content and be successful that way. So the two things can be related, but I think there'll be different in mindset. One is basically, okay, I'm going to be successful by using the tool and I learn the skills and anything to do that. The other one is to say, a figure out of space, it is still valuable moving forward in way AI is not going to be leading for whatever reason. And I'm going to just take that space for me. And so do that two different approaches, but I think work has to find their own way, whatever that may be. And that requires some thinking and progress implementation too. Yeah, and I think that first strategy that you described rather than seeing AI as replacing you, how can it make you better? And what is the reality there? Is it making you better or is it replacing you? And then gravitating towards the skills that are not easily replaceable. Technology is always having a combination of replacement and enhancement. It's going to make some skills more valuable and more productive and it's going to make some skills less valuable and maybe even obsolete. And that is fine, it's always a case, it's always been the case from whatever technology can think of. The problem is that really that one, the problem is that now we have a technology that is so advanced in its abilities to produce high order thinking and outputs that is carrying out a big chunk of potentially the opposite base of activity. And so finding what you must can contribute. I think there's lots that can contribute, but we don't have a blueprint, we are making it up to see if we don't know. We don't have a blueprint because it's so new and more disruptive and transformational than past innovation waves. You do have a framework though that you wrote about in one of your articles and we will link to several of your articles in the show notes. Tell us about the aware framework. Yeah, so that aware framework is an attempt to help leaders understand how to deal with the feelings of threat that you were mentioning before, especially identity threat. And what we do is to say, well, I mean, taking a step back, if you think about what kind of threat I can trigger in workers. One, we already discussed the threat to the value of skills, to once role and status in a company, maybe a threat to competence. Then you've got the threat to autonomy. I feel now put in, we call it an algorithmic cage where basically you feel that your range of actions is limited because you got to use this tool with very dictating, large chunk of what is to be done in a workflow. And then you have threats to relatedness where you might find that this technology entering the firm has some kind of alienating effects and I don't talk to a person anymore. I only talk to the AI, AI is being very strong at me in a way that makes me feel I know be valued by the companies or if you'll detach from my colleagues and my superiors. So they could be the stress to different elements of our wellbeing. If you are a leader, then what do you do? Well, you basically want to have a set of steps that enable you to monitor and anticipate and basically deal with this kind of potentially a very psychological reaction so that the work is experiencing as you're allowed to. And so the aware framework, every letter is basically a verb, like a line, watch and so forth where basically you say, okay, do I understand how my ID deployment is make them feel and think? Can I anticipate who in my organization is going to be experienced, especially strong feeling the threat? And why? And if they do experience that what kind of coping strategy, what kind of reactions is usually likely to trigger? And can I device programs that maximize the likelihood that those coping strategies will result in positive action rather than negative, maladaptive action, active sabotage on AI program, disengaging, discapism or some kind of negativity action, negative from both the standpoint of the individual worker, but also the organization. And so in that paper, it's in a Harvard Business Review, we basically sketched steps building of this work and technology and tried to provide some advice as to how you can pair a technical and deployment track with a human deployment track where you've got your CTOs, your CTOs, CIOs working with leadership to basically figure out the data and the cyber security and the vendor contracts and the financials that are on that and basically, you know, the privacy, whatever it might be, compliance and all of that. And that is basically absolutely central to making it work. And so you might hire some kind of tech outsourcing consultants or whatever. And basically get that done, but that's not enough. You need to completely complete the program by adding a more HR lead, maybe, or I don't know how this can work out in practice, maybe sitting within each function, I'm not sure. But basically a program that would consider those psychological issues we talked about and try to make sure that, you know, this is going to be seen by work as a win rather than loss. And if they see it as something is going to hurt them, they're going to react negatively to it in all kind of ways. If they see it as something that is exciting and potentially empowering, then I think it's going to double the light. You know, the energy and the power like we talked about before is how you liberate the energy of the organization. And maybe not everybody will get on board. Then you have to deal with that. But maybe relocating some roles and maybe some people just can't adapt to the new system where, you know, there would be some difficulties there, I'm sure. But basically I think a lot of people can be brought along with the right systems, I think. For sure. Thank you for that framework, Stefano. One of the other areas that you and I agree on is the importance of AI literacy. And I have said that in the board room or in the leadership team that AI literacy there, most determines the degree that organizations will be successful, impactful with it. You also teach a course on AI in society. So see it as a life skill. What do you think real AI literacy looks like? What can we do better here? I think it's a bit of a hard question because I don't know that I already know the answer right now. I think certainly there are a few things that we all ought to be doing. One is to just get acquainted with the technology, use it. I mean, there are many people who have maybe an account with, you know, such a PTO, a cloud or some other system like that. It is through a private account or through something made available by the company. A lot of them have a fair contact with Microsoft to have access to Copilot or whatever it might be. But the reality is that many people use it sometimes, but they haven't quite developed them up. So where the instinctive reaction is to see how that technology can help them do the job better and faster. And so I think for many people it's just change management. I mean, it's hard to do why people's brain. You've done a job for 10, 20 years and now you're asking to pose and think every time you're doing something, you can do it instinctively quickly and say, can I get help from this technology? It's not obvious. It's quite complex process of change. So but just practicing and getting it, inviting it to everything you do, this is what my colleague I'm Olick is always alive in people to say. It just has to the question, would he be able to do it? How well and how can I support it and how can work together? Be the shot and sometimes it will work amazingly well and you think, wow, I found a great way to do it better and save time. Other times you'll find not great and actually, you know, I want to keep doing it the way I was doing it, which is also fine. But in that case, go back to it six months later because it's moving very fast. So I think that I was testing months ago and I wasn't able to do very well. I would now to do very well. And so this is a moving target for sure. So that's one. You know, have the right mind set, just try it and use it and make sure you have these accounts and you just do it for lots of different things. Maybe even start from a hobby. see how you can create maybe Vibe Covanat to help this youth group or maybe develop something that helps you be a better musician, artist, cook or whatever. Somalia. Somalia, exactly, whatever it might be. And just see how that works and use it as an entry point. I think that's, I think. But then beyond that, I think we ought to think carefully also about the most specialized youth than just a general productivity tool. We'll all be benefiting when we do a deck, when we write an email, when we try to understand a complex report, can help all of us. But I think this technology will also have the ability to be developed into specialized applications that are relevant to a particular function. This could be something you do in a house through this kind of internal R&D and maybe lots of proper prototyping. Nowadays can be done with Vibe coding by people in the function that don't have technical skills just to see how this might work. And then to scale it and put in production might be a different game, but I think you can go quite far. And finding out all those ways in which you can help, engineering and new product development and finance and all the functions within a company, I think that's something that's going to have to rest with the people who know that function best. So that's an add-on track where you have to basically just think about this specialized application that might require a degree of investment, some testing, some development. Yeah, so trying it, I think it's important. You mentioned Stefano change management and a data point from an AWS survey in 2025, only 14% of organizations had a formal change management program. Now the prediction is that we will do better in 2026 that a full three quarters will have a change management program in 2026. I guess my first question is, do you agree with this prediction or too optimistic or what do you think? I don't know about the point estimate, but the direction of travel is the right one. Indeed, we'll see more and more companies realizing that those pilots and those efforts are not leading to the returns of the leaders we're hoping for because of all these resistance factors that we discussed. Actually, that's exactly what I was advocating just a minute ago, right, to pair this technical track with a human track and say, you can call it whatever change management is, basically what it is. And so whatever you start saying, I'm again, investments in Genai, you can just throw the money there and just hope it's kind of blossoms. You really have to think about carefully about deployment and change management. So I think it's such an obvious thing, but I think the problem is that the college is almost like taking all the oxygen from the room. There's so much attention and urgency around the tech itself that we often forget the people around it. And I think that's something that we are learning to correct because you do see that a lot of these programs don't work out if they're not embedded within the kind of change management program. I'm excited to see what 2026 is going to look like from the start point. Yeah, you say it's obvious, but I would challenge you on that. I think it may be obvious, but no one person owns it. It's not an HR problem. It's not a CEO problem. It's not a CDO problem. And so as one leader said to me, it's easier to replace 25 year old tech than to replace 25 year old mindsets. So I think we avoid the difficult or avoid when it is multifaceted. The fact that the problem is obvious doesn't mean that it's easily solved. Yes. Yeah, fair. All right. We've covered so much, Stefano. If you had one last bit of advice to data and AI leaders and CIOs on scaling AI responsibly, what would that be? Maybe one thing we haven't talked about, which you just made me think about. Is to think about how do you lead that change management project? And I think one thing that we see in our data that companies increasingly have a role called the chief AI officer. And in part half of the instances in our data, we find that this is a bit of a rebranding of an existing role, oftentimes CDO, CIO. And so imagine there, if you are a chief information officer, you add an A to the title and now everybody thinks you're cool. That's one thing. And so that's in a way it's more like a kind of branding. But I don't want this message as an important thing because I think for many knowledge dream organizations signaling that we are leading this is actually important for outside clients or whatever. But then you have also about half of these new roles altogether. And I think that it's an interesting idea that many companies, I don't say that they should do it, but I think they should consider whether that's the way to do it and maybe they come up with a different one. But I think it's maybe an opportunity to reflect a little bit on what's best in order to manage this. Would you then have a person who is physically tasked with being the board level account to be the person? So I now have someone in the board who's job is to make this work. And I think that is probably going to solve some of the issues that you were pointing out in the when you were talking about fragmentation and nobody doing their own thing and they don't amount to anything. So that's one issue this kind of accountability can help with. The other one is the CTO will have to take care of the technical stuff. And there's a lot to take care of and you can't expect the CTO to go to the functions and say this is how you know it can help marketing payroll. Whatever. That is something that is not you know it's not fair to expect the CTO to have to do that. You basically need to have someone else who is taking the technology from the center and bling thing out to the functions. And maybe the CAI role can be that person who supports every function in making it work in that function. And in that sense you can see the CAI role basically as a change management job. Perhaps it's not a technical person at all. It's just a person who really understands change management and has a good holistic view of the business. Yes. That's really one idea that I think a lot of companies will benefit from. No, I like this idea and the debate. I resist rebranding. I'm like, come on, everyone should have part of the AI. But maybe it is a necessary signal for right now. I think that's a really good point. Step on it. You're right. I mean, that could be a transitional thing. I mean, we see it for now. Education. We are adding a lot of AI courses. We even launched an AI major follow MBA program. We'll first be as cool to do that. And the internal discussion was a little bit about why should we have an AI major? Everything is going to be AI first in a way. And I completely agree with that. But just like with digital, maybe international business, the decade earlier, every way you would change organizations need to be spearheaded somehow by some people. So I think that an AI major makes sense today. CAI oak makes sense today. That doesn't mean we're still going to have it 10 years from now. So that may be just a conjecture. As we move down, I really understood how this technology is going to play out, then we don't need that anymore because everything is AI in some shape or form. But I think if I take a dedicated tool, I don't think it's going to be an 18-month thing, like a lot of people in tech predict. Yeah. I think it's moving fast, but the impact on society and roles will take longer. OK. So now you have an AI major as part of the MBA program, which I think is exactly where it belongs. As I said earlier, you make me want to go back to school. So Stefano, we've covered a lot of ground. Let's shift to a fun lightning round. What do you do outside of work when you're not in the world of AI? The usual thing. Look, I love food and lots of traveling, lots of spending time with my family. One thing that I do, and maybe that's kind of interesting to people, is to carve out time for a bit of money. I try to think whether it is a long run or some meditative moment or maybe just working in nature, something that can help my brain come up with ideas. And I think in a world where things are rushing so quickly, I think we'll all benefit from having moments for pause. Because I think those poses are important for well-being, but they're also important for ideation. For sure. So I'm picturing you jogging along the scookel river. I do. But I think you should come down to Rahobet along Gordon's path. I have been running over there once in a beautiful place. Yes. It is. How about someone famous or not who inspires you? Well, I'll say that there are many really interesting people in the social sciences now helping society navigate the AI transformation. We hear so much from the tech sector, obviously, and a lot of very exciting voices and important authority voices in that world. But I think I'd say pay attention to social scientists because this revolution is now as much a technical challenge as it is a social challenge, if it is a technical challenge. So colleagues like mine like Ida Mollek or Eric B. Olson, Stanford or a whole host of interesting people who are basically helping us understand what we can do with this. Yeah. And as you say that, you know, I was reflecting on some of the customers that I have the privilege of working with and they are. are different titles, but what I find interesting is that they do have a background in behavioral science or psychology. And it's almost like, have they met the magic of combining the tech with the human elements, why they've been able to be so successful in their orgs? So I do think that's important. All right, Stefano, I've asked all the questions. You get to choose the last one, depending on your mood in the moment. Either, what are you most grateful for? Maybe beyond the obvious of health and family? Or what is an accomplishment from the last week that you are most proud of? I may be take the first one because I think it's maybe more relevant to everybody. And I think there is a lot of fear when it comes to changing the world and some of it is very justified because the world is somewhat in turmoil, both in terms of geopolitics and democracy and stability. So there's a lot of processes that are really scary. At the same time, I think we are so lucky to be alive. I mean, it's an incredible time. I just play with this technology and I just can't believe what I can do. And so I understand the threat, I understand the fear. And there is a lot of reasons to be worried, but there is so much to be excited to. And I would like to balance a conversation and say, don't look this is something that you look from. Look at this is something you can gain enormous amounts from. What a beautiful thought, Stefano. Thank you so much for the work you do and for being on the Data Chief podcast. Thank you for having me, and there to be here. [MUSIC]