 Welcome to The Get, the podcast for Enterprise Leaders, delivering timely insights for today's global economy and tomorrow's competitive advantage. I'm your host, Chris Kane, President of the Center for Global Enterprise. And today we sit down with two global technology leaders to discuss generative AI, artificial intelligence, and its impact on business and society. Tom Friedman, Pulitzer Prize winning columnist of the New York Times and author of Widely Regarded Technology Related Books, and Sam Palmasano, Chairman of the Center for Global Enterprise and former Chairman, President and CEO of IBM. Tom and Sam welcome, and thanks for being with us today to discuss what will change in a world where generative AI technology such as chat GBT becomes widely accessible and used by individuals and business. Tom, you recently wrote a very compelling column, our new Promintheum moment, arguing that humanity is on the verge of a new technological revolution powered by generative AI tools like chat GBT. You state that it will fundamentally change the way we live and work. There's been a lot of buzz about the impacts, both positive and negative, that this new era of innovation will pose. Economic disruptions, job displacement, ethical dilemmas are some of the negatives mentioned. But addressing some of the world's most pressing problems from climate change to healthcare to inequality are some of the positives mentioned. The excitement in expanding use of chat GBT has ignited a race among AI labs to develop and deploy even more powerful tools. As this is happening, many stakeholders such as government regulators, business leaders are now scrambling to understand the impact their app emergence of new AI tools will have on consumers, businesses, society and global economies. Regulators are considering new rules and guidelines for AI, while even some leaders in the tech community are calling for a time out to better understand their amplifications of generative AI. In this podcast, we will explore the implications of these new AI tools and the challenges and opportunities they present for business and others in society. Sam, given your deep experience with technology, perhaps we can begin with you, which is what is AI for our audience? How long has it been around? And why is so much attention now being given to generative AI? Chris, that's an excellent question and thank you for having me today. If you don't mind, I'll take you back more than a half a century. I know I'm old and I'll admit that, but let's go back in time and Tom will recall some of this as well. Fundamentally, it all started with people and technologists doing research, tried to simulate the human brain. And that went on its course for several years. However, it led to the thing called the cold winter of AI, which everything was set on pause at that point in time because it was extremely difficult to simulate the functioning of the brain, especially given the computing capacity at that point in time. I mean, if you think about it, you need acres of technology to simulate, which is in a shoebox called the human brain from a cognitive perspective. So that led to what I'll call massive data and statistical lookup, which we know AI today, artificial intelligence. It's really artificial or augmented intelligence, if you like. Having said that, a lot had to happen within the technology infrastructure to make this what it is today. You needed massive scaling and lower computing costs beyond what you knew with supercomputers, now it's large cloud infrastructures. You needed the internet. The internet was important because it drove large data sets and lots of different information, not just ones and zeros in table format and those sorts of things that were traditional in classical computing science at that point in time. And so all those things had to come together and you had to get the algorithms or the software tools to do the statistical analysis. Now, what's changed now, it's led to this thing, generative AI or truly large language models. So therefore, what you're talking about now is almost like human interaction in a way, if you could think about it that way. It was an analogy and it's not a great analogy. It's the browser to the internet is what this is now doing for AI. It makes it available to everyone now. You don't have to be a data scientist to take advantage of these technologies. That's why it's so transformative. Thank you. Tom, in your column you wrote, and I'm quoting here, this is a Promethean moment we've entered. One of those moments in history when certain new tools, ways of thinking or energy sources are introduced that are such a departure and advance on what existed before, you can't just change one thing. You have to change everything, end quote. Clearly you see the possibilities good and bad of this technology's use. What makes you so convinced that the technology can change everything? That's a big order, everything. And then Sam, are you as convinced as Tom? But Tom, please go ahead. Well, you know, I define a Promethean moment as a moment where you get the introduction of a new technology or way of thinking that does require you to change everything, how you govern, how you learn, how you teach, how you do commerce, how you fight wars, how you manufacture, and even how you commit crimes. And we know what these Promethean moments in history are, they're the printing press, scientific revolution, the agriculture revolution, the industrial revolution, and this moment. And I call this moment, this Promethean moment the age of acceleration, amplification, and democratization. Never have more people had access to more tools that amplified their physical or brain power at a steadily accelerating rate and were being democratized at the same time to more and more people. So let's start at a far edge of the spectrum. You now an illiterate Indian farmer with a smartphone. Will now be able to ask a generative AI app, anything from soil fertilizer needs to water, and get an answer, and this word is the interface is so important, as Sam can tell you, get an answer in his own Indian language. And it may not be Indian, it could be one of 22 Indian languages, and I guarantee you two years from now it'll be in one of 122 Indian languages. He'll be able to get that answer any way he wants it, or text by voice. And again, if he's illiterate, that would be very important. And that answer, once this gets going, will be the best answer that is possible. So for the first time, everyone everywhere will have access to the best of everything. And that is a Promethean moment. And that will change everything. Powerful, Tom, you wrote a book a couple of years ago called The World is Flat. What it listening to you, it sounds to me like you're saying life is flat, given the power of the generative AI capabilities that you just described, it's been individualized to come to the personal level. Well, the theme of The World is Flat was that the new, new thing, you used to have to be a country to act globally. Then you could be a company to act globally. Now you could act globally as an individual. And now you build that globally as an individual with access to the very best and any range of subjects. Yeah, it's quite exciting. Sam, thoughts? Oh, I agree with Tom. And the vision he paints is actually the most positive view of the outcome, and I think Tom would agree with that. And there's a lot of innovation along the way to get to where Tom is talking about this. And what do I mean by that? There are a lot of capabilities, just like the internet in its early stages that had to be built up to what it is today. I mean, it was basically you're passing scientific documents between research organizations, and then all the technologies came along. I won't bore you with the path of 30 years or so, but fundamentally all that came along. It's a similar thing here. The only difference is this one's moving out heck of a lot faster. I mean, a heck of a lot faster than the early days of the internet. Okay, I'd like to turn now to the focus on business and the implications for business. And you both have spent years assessing technology impacts on business and society. If generative AI tools have such a pervasive impact across industries and society, what's your advice to CEOs and government leaders about where to start for them to determine its relevance for good outcomes and risks for bad outcomes? Tom, why don't we start with you? The best place to start, Chris, is to start. And what I mean is whatever business you're in, whether you're selling hamburgers and french fries or software and hardware, the only way to learn about this is to dive in. That's what I'm trying to do. Dive in, play with it, and see where it takes you. There's no other way to learn. And one of the features of generative AI is that the engineering is the head of the science. This is both a good thing and a bad thing. James Vanneke from Google made this point to me. It's a bad thing because we don't know how these systems do what they do entirely. He gives the example of they cannot give, barred Google's generative AI system scraps of Bengali. And it can start writing poetry in Bengali. They actually don't quite know how it does that. So the engineering of it is actually head of the scientific understanding of it. Now, the upside of that is these things may be capable of doing so many more things positively than we fully understand. They equally may be, and certainly are capable of doing so many more bad things than we understand. So that's the moment we're at. And the only way we're gonna find out is to dive in. So whatever business you're in, you're a journalist or you're a manufacturer or you're selling hamburgers and french fries, I'd dive in. Sandbox strategy, get in the sandbox. Sam, thoughts on where CEO should start? I completely agree with Tom and I'd add one more thing to that point. Firstly, it's gonna become ubiquitous. So you have to assume it's gonna become ubiquitous. So Tom's right, you need to start learning now. Now, I'll answer from a CEO perspective versus a government official, since I've never been a government official, I have been a CEO. You need to establish a management system within your company on usage principles. Cause everybody's gonna use this. Remember, it's gonna become ubiquitous. Everyone will use it. So you need a structure and principles of usage guidelines. No different than chatbots and those things that are now on the internet. Everybody established usage guidelines for their enterprises. And then you have to let the people play with it as Tom's saying, cause they have to learn. And there's one other thing I would add to it. You have to establish a culture of trust. Back to Tom's points, it could be good or could be bad. And the engineering's ahead of the science, I totally agree with that. But having said that, you as an enterprise, whatever you're doing as a business, you have to be trusted by society and your customers. So therefore you have to build a culture within the organization that's principle-based that long-term builds trust as you use these technologies. So in a new space like this, building trust can be a difficult task cause people don't know what to hold you accountable for. And it seems to me that enterprises and government agencies will have to proclaim what their value system is relative to this new environment because otherwise I don't know how to hold them accountable. In other words, I don't know how trustworthy they are. Do you see that as a necessary first step? Tom and Sam about having organizations that are beginning using these powerful tools talk about how they're using them for why they're using them and what they will not use them for? Well, you know, this is a classic example of another point that I've been making for several years now, which is that you cannot govern the world we're in now from the left to the right. You have to govern it from the center out. And what that means is that you need actually complex adaptive coalitions to manage any number of challenges. And AI is one of them. So if we want to govern AI properly, what do you need? You need engineers. You need technologists. You also need moral philosophers. You need government regulators. So this is not like in the old way, I government regulate your business innovate. There was a binary system. This has to be much more like nature, like an ecosystem approach where you get all the stakeholders together, not just the stakeholders, people who can understand the stakes and define them and bring them under one tent in order to write the regulation. Because we have to be very careful. We don't want to over-regulate it and basically retard its potential, but you don't want to under-regulate it either. Now, one of the problems with generative AI, one of the challenges is that regulating it will not be one size fits all. Because if you're competing with China, you don't want to regulate it at all. If you're worried about aberrant outcomes, you want to regulate it a lot. If you're worried about ISIS getting it, you need arms control. So it's got a lot of different variations and that's why regulation has to be a merchant, can't be from the top down, and it can't be one size fits all. So you've jumped ahead, it's a perfect... My next question was going to be to you about your call for complex adaptive coalitions and you've already gone there and that's great. And I want to... You both will remember back in the late 90s when internet regulation was a topic among policy makers both in the US government and around the world. And the governments from around the world did exactly what you were just saying, Tom, which was to have a regulatory approach that gave space and time for the internet and its application and value to develop. From a policy standpoint, it was called forbearance. Governments were going to forbear command and control regulation to see what was going to happen. And it started to talk... Governments started to talk about behavior, right? Here are the behaviors that our society considers to be appropriate. But that was a very different time. We seem to be more cohesive and more together as a society and you talk about needing to come organically from the center and go up and go out. Today though, is that really possible? We're living in a world where we have so much separation and division in society. Can this approach to generative AI or generative AI in and of itself bring us together or just separate us further? And if there are historical examples, either one of you have that can point to the fact that the power of a tool like generative AI can bring us together, I think that would be really powerful to discuss and important to discuss because there seems to be such a lack of trust in society today. And yet if what is needed here to be the critical success factor is a foundation of trust, can we get there from here? Sam, why don't we start with you? Yeah, Chris, I think there are excellent analogies to what you've said. You need to let learning occur, right? If you don't let learning occur, you're not gonna get the benefits associated with the technology. You go back to the internet, yeah. People weren't quite sure and it started with putting up your information about your company, your products or your government websites and those sorts of things. No idea would be transformational as far as inclusive society where there's sharing of ideas where you could grow globally. I mean, Tom's example of a farmer in India, all those things came about because the world was connected. So one could argue that there's benefits to that and yes, standards did emerge over time and that was the role of the regulator, quite honestly, to establish some of those standards. I'll just say we've learned a lot about how to create these standards. There are organizations and they're nonprofit organizations, NGOs around the world that know how to do these things and it'd be great if you could convene those to Tom's point because it's multidisciplinary and bring those together. I'll give you an example of an institution that's trying to do this. I'm on the advisory board called the Human Center Institute for Artificial Intelligence at Stanford and it's multidisciplinary and it's people of all kinds of backgrounds both the philosophy department, the law department, obviously computer science and business, et cetera, et cetera. They're trying to convene and bring this together but that's just one example and that's not enough quite honestly to have this occur. I think to rely strictly on governments today to bring the world together, given the nature of the governments and how they are, that's gonna be very, very difficult to do but perhaps this kind of nonprofit, intellectually honest entities and they exist all over the world, coming together might address the points that Tom's making. The Tom is spot on, I completely agree and if not, there could be, I think that the negatives could outweigh the positives. I never would have said that back to the internet days but I really worry a lot that given the nature of some of our leaders today, at least in the public sector, the negatives could outweigh the positives. Tom, thoughts? I've been working on a book and I have a chapter on this question of regulation because what happens basically when technology goes so deep, you notice we added the adjective deep to everything. There's no global lexicographer that ordered that. We just did it intuitively because we sense that, oh, that's not a fake, that's a deep fake, Chris. Yeah, that's not just medicine, that's deep medicine. It's just research, that's deep research. Everything had to go deep because we sense technology was going so deep that we really didn't even know what it was. So that old relationship which Sam alluded to, I government regulate you business innovate, really breaks down when it gets that deep. I have this leaked email from these Boeing engineers and one of them's talking about his basic as FAA regulator and these are in the New York Times, a person leaked to me, there was a Times reporter there and the guy says basically my regulator is so clueless about what I do. Watching him watching me is like watching dogs watching television. Oh, that's a lot of where sort of regulation is today. Watching government regulators today, watch chat GPT is like watching dogs watching television and the speed and depth of it is just way, way too far. And so the old top down command and control system just not gonna work. And the speed that this is coming at because these are really powerful tools. This is not the internet, this is not just about reaching somebody. You can start wars with this, you can rob banks with this in a way we've never seen before. There's no time to waste and therefore obviously you want a coherent recovery structure working together and not two parties who spend every morning trying to subvert each other. And you can only hope that the system responds that way. Maybe we can talk a little bit about hype. Every new exciting technology has a hype cycle affiliated with it it seems. And the real value proposition is whether everyday impact versus specialized impact time you were just referring to deep. That connotes that there's a specialized impact that something has but everyday impact and I agree with you where the world starts to change, right? Where people change their behavior, their interests, their investments, their time because they see a benefit for doing so. But these new technologies have a hype cycle and that everyday impact takes a while to stick. Examples, right? The personal computer, the internet, smartphones, social media, FinTech, Bitcoin and cryptocurrencies are just a few, right? Of the technologies that broke onto the scene and there was a tremendous hype cycle affiliated with them. From a practical value, pervasiveness perspective, where do you think we are in the generative AI hype cycle? I think it's underhyped. I was going to say, Chris, if you look at traditional adoption cycles to the points that you've made, this thing is like definitely faster than the adoption cycle. It's up to millions and maybe billions of users already that are just sort of talking about it within a year. Large language models quite honestly have been in research for maybe a decade or so. I mean, I've been invested in companies for seven or eight years now that have been working on large language models. But the fact that it's coalesced around this chat GBT that was now actually a open source piece of technology that people could have access to, the adoption, it's well ahead of anything that's ever happened historically back to Tom's point. So therefore it's the hype, but it's also the adoption. Now, what's happened historically, and will happen here, because there's a ton of innovation going on and there's a ton of investors flowing to this space, there are gaps in the technology. As we know, you don't know the source of the answer as Tom alluded to, so therefore you don't know the accuracy of the answer. There really isn't a fact checker per se as we would hope to have as far as the accuracy of this answer because the text is so persuasive that therefore you conclude it must be right. It's almost human-like, but it could be completely wrong. And I think some of the inventors of this are free to, like Sam's admitted, this is completely wrong at times and we need to be able to deal with those sorts of things. So my point is those gaps will be filled, but I think they're going to be addressed and filled much quicker than they ever have in the past. So those questions about data accuracy, quality, profit and sandwich you've just addressed, right? I mean, so people use the tool, they get a result. At some point in time, there's going to be reflection or scrutiny about that result was either, to Tom's point earlier, a deep fake intentionally or it was just wrong. And then it seems to me that those moments have a capacity to constrain adoption. Tom, you talked about how democratized this capability is, is the reason you think it's underhyped is because it's in the hands of everybody and therefore its adoption is not gated by certain stakeholders where traditionally in economics and society, there were certain institutions that allowed certain things to come to market. And until they allowed it to be widely distributed, the adoption was constrained. Or in my off point, the distribution and impact aspect. Nothing you're right. I mean, it's been distributed so fast and therefore people are going to find its benefits and its problems much faster. So I think it will improve much faster. But, you know, my wife found the museum in Washington DC called Planet Word. We had a board meeting a couple of weeks ago, Craig Mundy from Microsoft did a briefing for us on chat GPT. He asked it to do a 400 word summary of Planet Word. It did it perfect. Then he asked it to do a 200 word. It did it perfectly. Then he asked to do it in Mandarin. It did it perfectly. Then in Arabic, then in Spanish, then in Shakespearean verse and then in Abyssinian verse. This is the Model T version did that. Yeah, right. And you're telling me this is hyped? Okay. This was the Model T. I shudder to think what the Ford Mustang version is gonna be like. You know, look in our lifetime, personal computer, internet, AI. I mean, this is like, you know, generative AI. This is one of those moments. You are here to Promethean moment when Gutenberg invented the printing press. Some priest turned to some monk and said, now that is really cool. I don't have to write this Bible out long-hand anymore. I can't actually stamp it out. You know, I mean, like we're here at that moment. Only that moment took about 200 years to spread scale and ultimately give us the reformation and the renaissance. This, I mean, hold on to your hats. This is coming so fast. And we have a Congress. Then knuckleheads up there can't balance a checkbook. And the idea that they are going to be able to regulate this alone is ludicrous. We have a lot coming down the pike right now. Yeah. And that monk said, we can sell more copies of the Bible too. Yeah. And they did. And a lot of them, you know, got into, because then people realized they could publish something other than a Bible. Yeah. And they had enough money to build a huge cathedrals. Right. So there's a lot of economic gain by the monks. Well, maybe we'll come back to the history of organized religions on another episode. All right. Well, look, thank you very much. This has been a great discussion. You know, you really feel like you're at the beginning of something transformative, Tom. Your column was great, very compelling. I recommend everybody who's listening today to read it as well as to think about the implications to your daily life and whatever business and vocation you're in about how this will change, what you have to do and what you can do. Before we close our episodes here on the get, we always like to take a minute and ask our guests whether there is one, in one word or one phrase, is there one emerging issue that you see on the horizon that business leaders need to put on their radar? And we call it our emerging critical issues moment. And so, Tom, let me go to you first. I know this has been a pretty provocative conversation, but something either embedded in this or beyond generated AI that you think business leaders have to put on their radar, what would it be? It's hard to be what we just talked about because it's gonna change your business, your customer's business, your supplier's business, your community's business. And if you aren't diving into it right now, you're too late. Okay, Sam? I would add to what Tom's saying, one word, Chris, trust, this is beyond ESG. Everybody wants to talk about ESG. This is well beyond ESG. If you're running any kind of enterprise that touches society anywhere in the world, you have to be trusted. Therefore, you need a value system and principles. So if you don't have one, get it created. You don't wanna be Disney in Florida. You can't go back and say, well, I used to believe this, now I believe this because of chat GBT or generative AI. So I think it's fundamental, it's simple, it's one word. Start the day every day saying, is my organization and am I trusted? Great, okay, thank you both very much. We'll come back to these on a later show. I'm sure that we will organize an episode around the concepts of trust and how trust is operationalized in today's very connected and distributed world. So thanks, Tom and Sam, for your time and your insights today. You've been listening to The Get sponsored by the Center for Global Enterprise, celebrating 10 years of convening global enterprise leaders around the most important business transformation issues.