Andreas Welsch: 0:21

Today, we'll talk about how you can go beyond quick win use cases for generative AI. And who better to talk about it than someone who's an expert at doing just that. Randy Bean. Hey, Randy, thank you so much for joining.

Randy Bean: 0:32

Hi, Andreas. Nice to be here.

Andreas Welsch: 0:35

Awesome. Why don't you tell our audience a little bit about yourself, who you are and what you do?

Randy Bean: 0:39

Yeah I've been at it a long time. So I spent four plus decades and basically the data space. Now the data and AI space, worked with a major bank and at that time I was trained as a COBOL and Assembler programmer and I was working with all this data. And I said to the executives of the organization, I said, what do you do with all this data that you have? And they said the regulators make us hold on to it for eight years and then we're allowed to destroy it. And I thought what? And it was really from that time forward that I got interested in data. I spent some time with the database marketing company, a pioneer in that space, now known as CRM, ran their North American financial services, database marketing practice, went up to Silicon Valley in the internet era and did two venture backed startups. And then in 2001, I launched my own firm, New Vantage Partners, which I ran for 21 years until sold it to a Paris based global consultancy, WaveStone, in December of 2021. And along the way, I've done a lot of writing and speaking. When big data raised data into prominence and the C suite and with boards, I started writing for the Wall Street Journal for two years. Have continued that with Forbes and then am a regular contributor to Harvard Business Review and MIT Sloane Management Review, and also wrote a book during COVID, fail Fast, learn Faster Lessons in Data-Driven Leadership and in Age of Disruption, Big Data and AI.

Andreas Welsch: 2:11

I know that there's hardly a weak that goes by where I don't see one of your posts or articles in my feed. So I'm super excited to have you on and learn from you today.

Randy Bean: 2:20

Thank you.

Andreas Welsch: 2:21

Randy, should we play a little game to kick things off?

Randy Bean: 2:24

We can try.

Andreas Welsch: 2:26

All right, perfect. If AI were a month, what would it be? I'll give you 60 seconds to come up with an answer. If AI were a month, what would it be?

Randy Bean: 2:38

That's funny. I was gonna say December, but I have no reason or rationale for that.

Andreas Welsch: 2:45

Do you think it might be because people are already talking up the next AI winter and unmet hopes and dreams?

Randy Bean: 2:53

Thank you for answering that for me.

Andreas Welsch: 2:55

I don't know. Let's see.

Randy Bean: 2:58

Actually there's a reason. It's at the end of the year and that always represents the start of a new year. And with AI, when people say, what's happening with AI, I just say, wait until two weeks from now or three weeks from now or the next month, because it'll be something new, most likely.

Andreas Welsch: 3:17

Exactly. Yeah. And you never know what's that perfect storm that's brewing there. So perfect. Thank you for answering that on the fly. Now you already said that in the introduction, you've been working in this data and AI space for more than 35, 40 years. And I think now with gen AI there's a lot of excitement still that I see, which is also exciting for, me because there's so many different opportunities and things you can do and places you can go with these technologies. But I also feel that a lot of those use cases are more like, quick wins, low hanging fruit, whatever you want to call them. Generally, I'm curious to see what's next, what's the next big thing once we figure this out, how to write pithy copy, how to create social posts, right? All the different places where gen AI is being added. So maybe more broadly, I'm curious, what have you seen across your career, more than 30, 40 years, advising Fortune 1000 companies in data and AI? What's changed and what stayed the same?

Randy Bean: 4:22

I think the most important thing is to maintain perspective. Because the same questions that people asked me 40 years ago, how can we take this data and gain insights from this data? How can we learn are fundamentally the same questions that people are asking today. So in a sense, things have changed exponentially in terms of computing power, in terms of the volumes of information. But things haven't really changed from a human perspective in terms of many of the problems that organizations are trying to solve. And as a result of that, because I spent most of my time advising Fortune 1000 senior executives, you most of those business leaders, a lot of what I try to do is help them keep things in perspective. So I'm often asked what do you think about this? What do you think about that? And I say, I have no idea. And they say, what do you mean you have no idea? Aren't you the expert? And I say, first of all, expert's a very tricky word to use. I don't consider myself an expert, but then nobody really is an expert because things evolve so quickly. And so I guess my point in that is that you can't chase every rabbit down the hole. And for me, working with Fortune 1000 organizations, the test I've always used, I call it the critical mass test. In other words, if one organization asks me about something or two or three or four, it's it starts to get on my radar screen a little bit, but it's not until 50 organizations are asking me about it that I say, now this is something I should seriously take a look at and develop a point of view and perspective on. So I think that's one of the things that I'd say, and these days I'm attending a lot of AI events, a lot sponsored by universities and there's just so much out there that's coming so quickly that I think it is important to step back a little bit and say what are the core issues that we're trying to solve? How can AI be used for the betterment of humanity, keeping that in mind. So I could go on and on, but those would be a few comments.

Andreas Welsch: 6:34

Yeah. That's, exciting to hear, that the fundamental questions and challenges seem like they're the same, right? And it's just a new technology, but, I feel in a way, it's also a bit concerning, right? That we, over the last 40 years, haven't been able to drive as much change and, widespread awareness of what are the things you should be asking? How, do you get your things in order, right? So, you prepare for whatever the next hype and the next topic is.

Randy Bean: 7:02

I have all these lessons learned from my career. And at one point early on, maybe I was 30 at the time, CIO for a top three bank, he said to me, he said Randy nothing ever happens in less than a decade in this industry. And I thought, wow, what a amazingly cynical attitude. But I've come to find that is maybe it's not always a decade and it depends what the thing is. But yeah, things take time. People don't like to change. Organizationals don't like to change. And even if you think of the internet and digital adoption, which is roughly 25 to 30 years old. Now, the digitally native companies immediately Amazon, Google, et cetera, they started operating from day one. But if you look at the Fortune 1000, 90 percent of the Fortune 1000 by number are legacy companies. And those companies adopted the internet digital capabilities over a very long period of time, a couple decades. As a matter of fact, several large organizations said to me during COVID, we've done more in terms of executing on our digital strategy within the past year than we did in the previous 20 years. So it tells you something about the pace of adoption and also You know, if you're part of that 90%, you're not really competing with Amazon and Google, you're competing with the other insurance companies, the other banks, the other manufacturers, etc. So you don't need to be at the leading edge of AI or machine learning adoption, you just have to be where the others are and maybe a step or two ahead of them.

Andreas Welsch: 8:47

I think those are very grounding words and experiences. Now, looking at trends like AI, gen AI, and again, knowing that there are some fundamental things that haven't really changed, and probably some of them are in people's minds, some are in, in, the environments and systems where we all work with, what would you say are some of the greatest challenges to success in data and, AI and to moving to that next level of exciting value, high value use cases?

Randy Bean: 9:16

Yeah. I could answer that a few different ways, but what I think I'll do now is there's a survey that I've been conducting for the past 12 years, and it's come to be not by design, I guess just by attrition and survival, the longest running benchmark of Fortune 1000 data and analytics and AI leaders, so have been conducting this for a dozen years. And I'll just share some of the data with you. This year, some of the questions we asked are investments in generative AI a top organizational priority? 62.3% said yes. So that's pretty good. Are you increasing investment in generative AI? 89.6% said yes. Sounds very good. Do you believe that AI, generative AI, has the potential to be the most transformative technology in a generation? 64.2% said yes, so again, pretty positive. Do you believe there's needs for safeguards and guardrails for governing generative AI? 99% said yes. I still say I'm not sure what the other 1% were thinking, but there you go. Are the safeguards and guardrails in place at your organization today? 62.9% said yes. A fair amount of work to be done. Is the talent in place to responsibly implement generative AI? And 50.5% said yes and 49.5% said no. And just a few other pieces of data that we can talk a little bit about more, about if you're interested in. We asked about the state of generative AI implementation efforts. So 60.4% said they were at the experimentation and testing stage. 24.5% they were at implemented and limited production. 6.6% said they were at the planning and design stage. Only 4.7% said implemented in production at scale and 3.8% said not in use at all. And then just the two other things we asked what was the primary business opportunity created by gen AI and 49.1 said achieving exponential productivity gains, not unexpectedly. 23.6% said liberate knowledge worker from knowledge workers from mundane tasks. And 22.6% said improved customer service and experience. And then on the negative side, we said one of the primary risks. 44.3% cited a spread of misinformation or disinformation. 23.6% said ethical bias and 5.7% said job loss and job displacement. That's some data to chew on, and like most things you can view it as half full or half empty, depending upon the lens that you see it from.

Andreas Welsch: 12:14

Thank you for sharing those details, and especially what you mentioned. It's been such a long running benchmark across Fortune 1000. The one thing that I think stands out for me is it's not super surprising, I think, given where we are in the industry, but I think what makes it really tangible is the breakdown of we're experimenting with this, we are, putting this into pilots, we've rolled it out, and it's at scale. I think That's a very very narrow funnel towards the end. But it's still good to see because I feel last year there was a lot of talk, a lot of hype, everybody got excited about these things, let's try it out, let's see what we can do. And I'm curious to see how it shifts over the next one or two years. Assuming that we'll see more moving towards this, we're adopting it, we're adopting it at scale phase. And then also the other point that you mentioned, where are the opportunities? Customer service, knowledge work, productivity. Just, looking at how many emails we send each other in a large corporation on a given day or messages or where information is scattered across different, whatever, SharePoints, portals, documents. I think there's a huge opportunity already to begin with, right? And that's just the tip of the iceberg that we're seeing above the waterline.

Randy Bean: 13:36

Yeah I think this is very much a work in progress, and we're all both witnesses and participants at the same time. For example, generative AI, I guess was introduced in the fall of 2022. So as we got into 2023, I had a lot of people asking me what do you think? And I was like first of all, I'm just trying to understand what it is, but there were I characterize it as two sets of people. There was a lot of normal people too. There was people on the one hand that said, Oh, isn't this incredible? AI will be able to do magical thinking and all this type of stuff. And I'm like, Ooh. And then on the other side, there were people coming up to me and saying, wow you're evil. You're associated with artificial intelligence and artificial intelligence is going to destroy mankind and put everybody out of a job. How do you sleep at night? And I was like, Oh back, off. And so it was still a work in progress. I was trying to factor all these things in. And then I had the opportunity in October of 2023 to attend the Wall Journal Tech Live event in Laguna Beach. And it was about 200 folks and interestingly, Sam Altman was there as the keynote speaker, Bernard Koestler, Fei Li from the Stanford AI Institute, Mustafa Suleiman, who's now the Microsoft AI head. And listening to these folks I came away with a big takeaway and that was that it doesn't matter whether you love AI or you hate AI, it's inevitable. And it's going to be one of the most transformative technologies in at least a generation, if not several generations. Now, I was at an event at Harvard Business School last week, and they had some really exciting and brilliant panelists, but I was telling somebody yesterday, I don't know how many times, because I made a check every time I heard this, distinguished PhDs, top leaders, C suite leaders in industry, and they were saying, AI is the biggest thing in the history of mankind. And I heard that probably at least four different times. And I just thought back to that point about perspective really? Maybe, there is a 2% chance or maybe there's a 6% chance, but who really knows? But isn't it early and isn't it limited information? And have you actually ever read history or thought about like, how most of the things that the environment came to be? So it is this, Exuberance that is healthy, but there's also sometimes an over exuberance that clouds things and sometimes when you have over exuberance, it leads to backlashes, skepticism and things of that kind. So I think all of these things are going to come into play and that's why I think it's important to try to keep a level head and one's feet on the ground and try to put things into perspective as much as one possibly can.

Andreas Welsch: 16:53

I think that's a great way to frame it, right? There is indeed so much excitement and still so much hype in the industry. I feel just look at yesterday's announcements of GPT-4o and multimodal models and all the new things coming out. But I also feel putting them into perspective of, Hey, what can we actually do with it today? And what are the things where this delivers real tangible business value? And it's not just a shiny object or some some mirage that they were chasing because it's cool or because now there is tech, I think definitely approaching that with the right mindset and starting from a business point of view also makes sure it helps you make sure that, there are fewer lost opportunities, fewer money pits, and the opposite on that generating a return. just talked already about one example, right? People being overconfident that Gen AI is the latest thing, right? It's bigger, or at least as big as the invention of fire, discovery of fire there was another statement, I think, from Bill Gates a couple months ago. And to your point, I think that optimism that excitement is healthy because it helps us move forward. It helps us think differently, approach problems in a different way. We have this technology, this capability that will help us in many different ways. But we also I feel I'll still figuring out what can we actually use it for and where is it useful and where should we invest? So we're not over investing in things that we can solve with statistical models, for example, right? I'm wondering what are the challenges or opportunities that you see organizations facing when they use gen AI?

Randy Bean: 18:29

Yeah. Yeah, let me speak to your previous point. Yeah I've moved much more into the camp of being a techno optimist, but a techno optimist with certain constraints and so forth. And I'd also qualify that by saying that, for example, in my family, I'm absolutely the last person to adopt any new technology, which I don't think is necessarily a bad thing. In business, I felt it's made me a proxy for the mainstream. In other words, until I start to see tangible benefits, tangible value, and it becomes, prevalent enough I I don't, really go there you have a limited amount of time, and unless you're just somebody that plays with all the latest technology you have, other things to do. But where I'm really interested in these days in particular is the use of AI in terms of enhancing medicine and healthcare and human, human life. And I was at an event last September where some doctors from Harvard Medical School got up and they said that for certain classes of illnesses, the AI that they were using was predicting or diagnosing patients with 90% better accuracy than their physicians. Most extraordinarily when communicating and writing and so forth. The Art of AI capabilities appeared more compassionate 80% of the time. And so I've been exploring that ever since then. My wife was in healthcare and I texted her at the time. I said, is this possible? And she says, I'm sure. No doubt about it, having been there and lived that. And then I had a chief data and AI officer for Mayo Clinic on a panel a few weeks after that. I asked him the same question about these numbers. And he said, yeah, it's completely consistent with what we're saying. And then I've been doing some work recently with Washington University in St. Louis, and they're finding the exact same thing. So I think that in the context of healthcare and medicine, there's a lot that can be done in terms of more efficient diagnosis, of patient diseases. Treatment plans and all of those things can contribute to the quality of health and longevity. So I think that's very exciting. Now, don't get me started on social media.

Andreas Welsch: 21:02

Yes, I got it. Now that brings up another question, right? So if there is a lot more transformation, a lot more AI driven transformation and opportunities, who do you see leads those organizations, those large AI data driven transformations? We've talked about the Chief Data Officer a couple of years ago, right? They've, I dunno come and gone. Chief Data and Analytics Officer, and now we've got the Chief AI Officer as a new title, maybe a bit inflationary lately, if you look across LinkedIn and in the industry. But is that Chief AI Officer something that now every company needs to have to spearhead this or are there different leaders with different skill sets that are need?

Randy Bean: 21:48

Yeah. That actually happens to be my sweet spot. Like other thing, other, some of the other questions, I'm learning like everybody else. But, because I've been involved in helping Fortune 1000, have been involved in helping Fortune 1000 companies for 25, 30 years in terms of how they leverage data and analytics and AI. As it so happens, the first Chief Data Officer for a major bank was hired out of my firm in 2008, 2009 to go to Citigroup. And so we started working with a range of organizations to help them define and scope what the nature of that role was, and we put in interim Chief Data Officers at a number of the organizations. And I've really seen the role evolve from a defensive role to a more offensive role and bringing in the analytics and so forth. And so now there's the question of where does AI sit? And we actually asked that question in the survey this year. I'll try to dig it up. So right here. We said is analytics part of the CDO function at your organization? 78.8% said yes. And we said, is generative AI part of the CDO function at your organization? 61.7% said yes. Then we asked the opinion. We said, should generative AI be part of the CDO function at your organization? 79.4% said yes. However, in December, I participated in an event at the Institute for Experiential AI at Northeastern University in Boston, and it was divided into two days. The first day was the Chief Data Officer Day, and the second day was the first ever Chief AI Officer Day. Now, not being able to stay out of trouble, I turned to the organizers and I said, chief first Chief AI Officer Day? And they said, yeah. And I said, I thought that was in 1962. And they like, they thought they like misheard me or whatever. They just looked at me and said, okay, let's start the event. Then they thought I was like, what the hell is he talking about? In the past couple of weeks, I've had a chat with a few people that have been involved with AI for many years at both times, I said back in the 1960s, this and that was happening. I said, yeah I tried to make that point. in December, and it went completely over people's head, but on the first day of the CDO event, I hosted and moderated the CDO keynote panel, and most of that day was focused around all the challenges cultural challenges, etc. that the CDOs face, and the short tenures, and all the reasons why the job has been frustrating for many. But then the Chief AI Officer Day was like a greenfield day. The first day the room was packed, the second day the room was standing room only, and it was like out to the street. And I said to the Chief Data Officers, I said you should all just call yourself Chief AI Officers, and for two years everybody be off your back, they'd be bowing down to you and so forth. So to your question, do you need a Chief AI Officer separate from the Chief Data Officer. Who knows? It depends. It depends upon your organization, what's industry you're in, the culture of your company, where you place yourself in terms of an early adopter or fast follower. I don't think there's any single pattern, but it's, it'll be interesting to see how it plays out because some 30-35% of organizations now are actively looking for Chief AI Officers.

Andreas Welsch: 25:25

And I think the US government is as well, right? 400 Chief AI Officers to be hired?

Randy Bean: 25:30

Yeah so organizations have these mandates, but in the early days of the chief data officer role, people would ask me about it and I said, it's a check the box thing. So the organizations the regulators say, are you data driven? Yep, we hired a chief data officer. So now you have these things where. You announced that all the government agencies have to have a Chief AI Officer. It's a lot of it's for show and you could argue that it's good foresight to have somebody that's a central person that's thinking about AI, in the various government agencies. But private industry isn't moving at that same pace. A private industry is usually much faster at doing things and making things happen than the government. I, say great in terms of raising awareness of AI, but I can't imagine that, if taxpayers look at productivity at any point, it might be something that becomes called into question.

Andreas Welsch: 26:40

That's an interesting take. Thank you for sharing that. But it makes, sense, right? Put AI front and center also in line with the executive orders and everything, but time will tell. Now, Randy, we're getting close to the end of the show, and I was wondering if you can summarize the key three takeaways for our audience today before we wrap up.

Randy Bean: 27:02

Yeah I think that, as I've mentioned throughout, the importance of perspective and, realism that organizations adopt technologies and change their processes but I think that all organizations, everyone should be doing as much as they can to keep abreast of how AI is evolving, at the Wall Street Journal Tech Live event, Sam Altman talked a lot about artificial general intelligence, supposedly meaning when a machine can do any cognitive task better than a human. That's really the kind of holy grail. That's what Mustafa Suleiman talks about in his book, The Coming Age. There's some great books to read like that, and Feifei Li's book. So keep abreast of things, but also maintain a proper balance and look for the right opportunities where you can use AI in your organization.

Andreas Welsch: 28:03

Thank you so much for joining us today and for sharing your expertise with us.

Randy Bean: 28:08

It's my pleasure, Andreas.

Andreas Welsch: 28:09

Wonderful. And for those of you in the audience, thank you for joining us and for learning with us.