Hey everyone, hey Paul, hey Jim, how are you doing? Doing great, looking forward to the conversation. Likewise, very exciting topic here today. Hello again to all of our attendees. If you don't know who I am, I'm Rett, the founder and CEO of Lit Video Books. A little bit about Lit: we make learning from best-selling books as engaging and entertaining as watching a movie or a documentary. If you want more info, just stay tuned at the end of the chat. For now, let's settle in for what I think is going to be a really exciting conversation on AI, the future of work, and how humans and machines can thrive together. We're going to save 15 minutes or so for Q&A at the end, and we'll add the Slido link into the chat so you can add questions and upvote what you want to hear answered by Paul and Jim at the end. Paul and Jim, before we dive in, I'd love to share a bit about both of your backgrounds. Paul, as Accenture senior technology adviser, you bring nearly 40 years of experience helping organizations navigate emerging technologies. Previously, as the chief technology and innovation officer at Accenture, you helped shape their technology strategy and have guided some of the world's largest companies through digital transformation and how to implement AI. Also, as the co-author of "Human + Machine," you explore how AI isn't here to replace us but here to help us work smarter and drive innovation. Jim, you are the global managing director of information technology and business research at Accenture Research, where you study how emerging technologies impact workers and organizations. You've led research at Babson Executive Education and Bain & Company, contributed to Harvard Business Review and the Wall Street Journal, and also co-authored "Human + Machine" with Paul. Really excited to have you both here today and talk about all things AI. So to get started, Paul, could you tell us a little bit about your experience with AI and how you've seen companies utilize it? When did you start on your AI journey and start advising companies, and what are some of the common challenges you see that they're facing right now?
Yeah, we could probably spend the whole podcast on that question. In terms of AI, I've been working, studying, and researching AI since my university days in the mid-80s with some of the pioneering AI researchers that were active then. So it's always been something of deep interest to me. My primary focus across my career has been the intersection of people and technology—how technology impacts people and what we do, and how people influence the development of technology. Nowhere has that been more fascinating than with the evolution of AI. In the 90s, I was working with the AI at the time, which we referred to as expert systems, and applying them in business. That moved on to other forms of technology, machine learning and deep learning, and the generative AI of today. Over the course of that, in my roles at Accenture, I founded our artificial intelligence business about 13 years ago when we set that up formally. I've had many other roles related to advancing our AI business and working with many clients and many AI pioneers and big AI companies around what's happening with AI. One of the most fun parts of it is what we're talking about today: collaborating with Jim, who is my great partner here on looking at what the real impact is, this human-plus-machine idea. Jim and I have been working on researching that since well before the book and continue to do so with some new ideas for the future. I'm really excited for the conversation today.
Thanks Paul. And Jim, how about you? What kind of patterns have you noticed with companies and how they're incorporating AI? How did you kind of get your start working with AI?
Well, Paul and I, going back a number of years, found that we had a common interest around human and machine interaction. There was a book on the topic by Shoshana Zuboff that we talked about some time ago that kind of got this conversation going. At Accenture, I lead a lot of our cross-industry research initiatives on AI's impact, working with our clients, academic institutions, and our research teams. I also sit on the board of a couple AI startups, so I'm hearing all about AI in discussions with business leaders and seeing trends emerge in our large-scale studies and economic analysis. In this work, I consistently hear a few really basic questions around AI from executives. First, most business leaders have a basic concern about how to change the business around AI—what's the vision for what a changed organization should look like? A second pattern is how do you separate what is real from what is hype or even science fiction? AI has been discussed in our culture for almost 100 years, and there have been a lot of Hollywood movies on the topic, so we have biases around what AI can do and worries about it. When we first started doing this research and writing the book, we really tried to separate hype from reality. That leads to the third thing we wrote the book for: to give executives a real, grounded view on how to get started on their AI journey, not just a five-year view, but what you can start to do today or Monday morning to begin to evolve your career or reinvent your company. Those are three patterns we see and address in the book. I'm looking forward to talking about those today with you.
One thing I'd add to what Jim just said is that with a lot of the companies, executives, and people I'm talking to, first there's trying to understand the technology because it's changing so fast. You hear these mind-boggling new announcements continuously. Then there's the value and risks. Businesses are really trying to understand where's the value, how is it going to impact costs, growth, or whatever you're trying to optimize, and what are the risks in my business: how does it impact people, jobs, responsible AI, and a lot of the things we're going to talk about. I think that's why right now, almost every company believes they need to be doing something with AI. Most companies are doing some level of proof of concept, but only a small percentage—less than 20%, depending on how you measure it—are really doing scaled AI in big parts of their business. That's the hurdle companies are trying to get past.
It's really interesting. I think this is just such a timely conversation. You're right that the rapid pace of AI developments every day and every week—it seems like there's something new to keep track of. Really excited for the conversation. You know, we're transforming industries with AI at such an incredible pace, and it honestly feels like we're just scratching the surface. I want to introduce everyone to the "Human + Machine" video book now and show a clip explaining how we got here with AI and what's to come.
We have to back up and recognize that we're living in an era of exponential technology innovation. We can interact in a more human and more natural way with the technology. "Can you assist me in a dangerous task as a robot?" "I'm designed to help with difficult or repetitive tasks." The first generation was simply standardized processes. This came in with manufacturing, Henry Ford and the like, instituting discipline and standardized mechanisms in the way people worked over a century ago. With the rise of computerization and digitalization, we entered a second stage of automation of processes and the way work was done. Now, with AI and in particular with generative AI, the technology has a new and more unique humanlike quality to it, leading to this third wave of business processes where we work with the technology more like we would work with a colleague. This is leading to new capabilities in the way we as people do our jobs and work, and this is the real unlock of new capability that we're talking about in "Human + Machine."
Really exciting, that man having a full conversation with the robot. Paul, maybe you could kick us off here. How do you see AI evolving beyond a lot of the generative models that are gaining popularity today, like ChatGPT? What would you say is the next frontier in AI?
I think one thing we have to keep in mind is how fast AI, and generative AI in particular, is moving. We've talked about already on this webcast that AI has been around for a long time. Jim talked a bit about the history. Things like predictive AI have been around for a few decades and used in business. Recommending the right book for you as you're shopping, mapping applications—that's AI that's been very common in our lives. Then there's diagnostic AI, doing things like facial recognition we use to log into our phones and apps, and many other examples. So AI is already part of our lives today. Generative AI, to qualify what we're talking about, is what's new, made famous to the world through ChatGPT from OpenAI, just a little over two years and a few months ago. Generative AI is moving so fast. You have to look at where it's going. Just think about what's happened within recent months. We had some really big announcements at the Consumer Electronics Show about World Foundation Models, not large language models which work on language like ChatGPT, but world foundation models that understand the physical world of properties and objects. That's one of the big breakthroughs coming. When we have mature world foundation models that don't just scan everything on the internet but understand the world around us, that's a big breakthrough because we're running out of data to train the models of today. World foundation models will help AI do what humans do—learn by observing and interacting with the real world. World foundation models combined with large language models are what's going to enable that breakthrough. That was talked about a lot by Jensen Wong of Nvidia in his keynote at CES, as one example. Then we had the Stargate announcement of $500 billion in data centers, and other announcements about massive compute capability. The important thing is that there will be a tremendous need for new compute capacity for the use of the models, not just training them. As more of us use these models—that's called inference—that's going to require tremendous compute capability. Stay tuned and follow all the activity and investment going on in that front. As a result, amazing news like three different nuclear power plants in the United States being restarted just to power AI data centers, because of the increase in use we're going to see going forward. DeepSeek was really big news that many of you listening I'm sure heard about. DeepSeek wasn't an advance in AI; it wasn't better AI than existing models like Claude from Anthropic or ChatGPT from OpenAI. What it was was a better engineered AI. That's something to watch going forward: engineering advances that make AI better, faster, cheaper. That's just the law of technology—things become better, faster, cheaper. DeepSeek showed how engineering and algorithmic advances could be used to train AI models 50 times less expensively and have them operate faster. Those are some things to pay attention to. The concerns about cost, climate impact, and such need to be dealt with, but a lot of advanced technology will help solve some of those problems. The final thing I'd say is generative AI isn't the last exciting form of AI to come. I talked about predictive, diagnostic, and generative. Symbolic AI is something Jim and I have done a lot of work on as well. Generative AI is about predicting the next word, based on the Transformer architecture—it's a predictive statistical technology. Symbolic AI is modeling real things. Combining symbolic AI with generative AI is where the next set of breakthroughs will come from. If I'm thinking about the laws of gravity and chemical properties, I don't want to learn those and have inaccurate predictions. I want to embed physical laws and knowledge into models. That's where symbolic AI will come in, along with generative AI, and that's where the next set of breakthroughs, I believe, will come.
