James Kotecki (00:07):
This is CES Tech Talk. I'm James Kotecki bringing you one of the many conversations I recorded live at CES 2025 in the C Space Studio. You can find all our C Space Studio interviews on the CES YouTube channel, but we've selected this one especially to bring directly to your podcast feed. Enjoy.
(00:28):
Hey, I am James Kotecki and this is the C Space Studio at CES 2025. This is day one of our two day interview series here in the C Space Studio with leaders from advertising, marketing, media, branding, really talking about the future of human connection. I'm going to kick things off with a conversation about, what else, AI, with Mr. Scott Likens, who is the US and Global Chief AI Engineering Officer, very cool title, PwC.
Scott Likens (00:56):
Thank you.
James Kotecki (00:57):
Welcome to the C Space Studio.
Scott Likens (00:58):
Good morning. Thanks for having us.
James Kotecki (00:59):
So of course we're going to talk about AI. Why wouldn't we lead off our conversations with AI? But first can you just give me the quick snapshot for folks who might not be familiar or just heard the brand, what does PwC do overall?
Scott Likens (01:09):
So of course we're a professional services firm, a wide variety of services from assurance to tax to advisory services. And maybe something people don't know is actually deep in AI engineering and AI services.
James Kotecki (01:20):
Yeah. Well, of course, AI is going to impact all of the things that you mentioned and it's really impacting business across the board. I understand PwC has this report called the 2025 AI Business Predictions, and what can we expect from AI this year? What does that prediction report say?
Scott Likens (01:34):
So this is a report we've been doing for several years and we try and predict out the next kind of 12 months. As you know, in the world of AI, five years, forget about it. So this year we really focused on the transition we're in. Last year was kind of mayhem with the inventions that were happening. The enterprises had enormous energy around AI but they were trying to figure out what to do with it. So this year we looked at what's going to be tangible? Things like responsible AI, what did it mean? We believe responsible AI could be the key to ROI. Doing this responsibly means we're going to get a return maybe faster than if we did it irresponsibly.
James Kotecki (02:06):
I like that phrase, I'm going to ask other people about.
Scott Likens (02:08):
Absolutely.
James Kotecki (02:09):
I'm going to repeat that here in the C Space Studio, but everyone's talking about AI transformation, as you mentioned. I've hosted these shows for a number of years and obviously the AI hype has gotten bigger and bigger over the last few years.
Scott Likens (02:18):
It has.
James Kotecki (02:19):
But then there's always this question of what is the actual impact? So where are we actually seeing business impact? Beyond the hype, where are people actually using this to get value?
Scott Likens (02:28):
So I think if you look at the last 12 months, there was a lot of experimentation. It was focused on efficiencies. So we were trying to learn how to use this within the workforce. I think the difference with AI now, because AI has been around a long time, is it's very much intrinsic throughout the business. Depending on your industry, there's been different areas of opportunity. The efficiencies are great, but it's really about reinventing parts of the business. So you think about research and development, we think we could save probably 50% of the time you spend on research and development, whether you're bringing a new product to market, could we do that faster? We're seeing that in drug development. Can we optimize differently and integrate that throughout our organization? We look at every function in the business and we see opportunities and within each sector those opportunities are bigger or smaller. So it's pretty diverse.
James Kotecki (03:11):
We should probably pause here and say, as you mentioned, AI has been here for a long time. What most people are hyped about is generative AI. That's maybe the general understanding of... Even there's some conflation, I think-
Scott Likens (03:20):
Agreed.
James Kotecki (03:21):
... between people saying AI and meaning just generative AI. When you're talking about these business impacts, are you seeing generative AI as a big part of that, or is it really just the overall growth of AI?
Scott Likens (03:31):
So I think gen AI is the opener. Everyone wants to talk about that, and there's no doubt the invention a few years ago, about five, but the influence that Gen AI has had on the conversations at the executive level has brought along 70 plus years of AI research and development. So we like good old AI, machine learning and all the things we've been doing, but gen AI has made it easier and more ubiquitous. It's become more intrinsic. Everyone could use it. It's not just a data science tool anymore. That's the difference with gen AI. So I think gen AI is leading the charge, it's generating the investment, but it's also a big piece of the value.
James Kotecki (04:03):
So let's talk now about the next step, maybe when AI is out there on its own, let's talk AI agents. What does that mean in your context and what are you seeing that's actually on the front lines?
Scott Likens (04:11):
Agents is a hard term to define because people think about them differently. But I look at it as a way to decompose tasks and let the machines actually think or reason instead of having a hard-coded program or a human doing these things. So I hate the term RPA of old, this is really-
James Kotecki (04:27):
Robotic process automation.
Scott Likens (04:29):
Robotic process automation. This is actually letting LLMs and natural language be an interface to really intelligent AI systems, and they could be traditional AI systems fronted by generative AI. So having that autonomy and that reasoning built into systems giving us much more flexibility.
James Kotecki (04:44):
But then let's talk about the responsibility on the other side of that. When we talk about an agent, something that seems to have... I'm not going to get into terms like free will, but something that seems to be out there operating quasi-independently from a human, at least in the moment-to-moment of what it's out there doing, how do you then think about responsibility on the other side of that?
Scott Likens (05:01):
And that's where I start with responsible AI as part of the design. We have to embed the guardrails in the beginning. They're still using the AI-
James Kotecki (05:12):
The use of language is difficult here.
Scott Likens (05:13):
Correct.
James Kotecki (05:13):
Because we want to use words like thinking and free will, but it doesn't quite apply.
Scott Likens (05:16):
That's not truly what it is. But having the guardrails in place, and I think this era of AI, we've been talking about doing responsibility from the beginning versus before I think it was kind of added in the end. So I think infusing those responsible AI aspects as we design these systems, everyone wants to have trust at the end, whether it's an endpoint consumer or your workforce or your other business partners, we have to provide that trust and RAI is the way to do that.
James Kotecki (05:38):
So if I'm a business leader hearing now about agents and agentic AI, what are some things that I can be expecting that I might be able to use this for in the coming years?
Scott Likens (05:47):
That's actually one of our predictions this year. You may be able to double your workforce keeping the same head count because these agents could provide aspects of the tasks that people do today and to give the super power to be able to do much more of those tasks. So you're seeing it in customer service, you're seeing it internally in research and development, as I mentioned, in IT generating software or building technology itself. Agents are really good at those things already, and as the large language models get more intelligent, they can do more reasoning, more semi-thinking or whatever term we want to use. We start to infuse that into bigger parts of the process, not just task level. So those multi-agent systems. So think about your workforce expanding without actually increasing your head count.
James Kotecki (06:25):
And in some cases they can roam around the World Wide Web and go to websites and make decisions or click on things and do things like that.
Scott Likens (06:33):
You're seeing that, yeah, that agent computer interface, being able to actually control the computer, the promise of what robotic process automation was, but looking at the web, having more real-time information, you're seeing models by specific vendors saying, "We have up-to-date information." So yeah, that intelligence at your fingertips, that's just a super power for your workforce.
James Kotecki (06:52):
But people still need to be, I assume, observant of what's going on, and there still needs to be a human involved.
Scott Likens (06:56):
Humans in the loop.
James Kotecki (06:57):
Yeah.
Scott Likens (06:57):
We have to have humans in the loop. The actual judgment, the actual final reasoning should be humans. These are accelerators. These are going to take large parts of our tests and make them faster, but humans should be deciding.
James Kotecki (07:08):
Another trend that we've seen come up in past years here in the C Space Studio, blockchain, Web3, not everybody wants to talk about that anymore, at least not to the same degree-
Scott Likens (07:17):
I do.
James Kotecki (07:18):
But yeah, I want to ask you, where are we in that hype cycle and then where are we in the reality of it?
Scott Likens (07:21):
Well, I am a huge fan of the blockchain architecture, and I'd say over the past few years, it's been silently growing to a point where it's actually becoming very ubiquitous in certain areas of financial systems. So while we think about the crypto side of it, the infrastructure of blockchains being infused across a lot of systems. And if you think about the promise of maybe Web3, I think we're almost at Web4 where we now can control value and identity and interchange mathematically, which is great. Adding and infusing AI agents. So we have autonomy and decision-making and reasoning, we start to really get to some of the promise that we had years ago.
James Kotecki (07:57):
Speaking of promising technologies, I also want to ask you about quantum computing because look, you're doing it all over there at PwC, and quantum computing is definitely a hot topic here at CES 2025.
Scott Likens (08:05):
It is.
James Kotecki (08:06):
What are you seeing there and what are you doing in reality?
Scott Likens (08:08):
I'd say the past few years we've seen some really amazing, maybe not breakthroughs, there's been some headlines, but there's been much more investment. I think we're finally taking it seriously here in the US I know globally there's different strategies, but we're starting to really see some of the theories be proven. So to me, what quantum gives us is the ability to process much more permutation of data all at once, without getting into the quantum mechanics of it. So these mean bigger problem sets, bigger optimizations, more ability to go after really hard problems, material sciences, drug development, climate, things that are hard for us to solve on classical systems, quantum systems may give us some of those breakthroughs. Is it here today? Everyone will argue yes or no. But you see the investment, you see the increase in patents, the increase in headlines, the increase in startups. That's why it's exciting to me. We see it coming.
James Kotecki (08:57):
How do you keep all this straight? We've got this convergence, at least just here on this stage we're talking about three major trends, Web3, quantum and AI. If these three things just develop separately, if they all converge, if they all explode into some kind of dramatically new version of the world, how are you keeping all of this straight and thinking about how to actually make tangible recommendations to your clients?
Scott Likens (09:19):
I have a really brilliant team, and I rely on them to think in the horizons one, two, and three. One we can do today. Two is right in front of us, and three is those big moon shots. And for us, it's about bringing practicality to it. We're working with enterprises that want to know, "What can I do today? But what do I need to prepare for?" Quantum is an example. How do you build a workforce that understands quantum mechanics and the mathematics behind quantum AI? That's a whole different ballgame.
James Kotecki (09:42):
What's your favorite Horizon3 notion right now?
Scott Likens (09:46):
So I'm a huge fan of what's going on in space tech. Obviously not core to our business, but I just see that as pushing the pace of innovation from material sciences to rocket engineering to AI, it's really covering the gamut. So I think that is pushing the pace.
James Kotecki (10:00):
Is there a question that you think I should ask other people here in the C Space Studio?
Scott Likens (10:03):
Why are they going so slow? I think it's a really challenging time. So how are they adapting to the pace of change? I think that's the difference. The time scale has shrunk to days or weeks versus months or years.
James Kotecki (10:15):
Why are they moving so slow? Or how are they hanging on for dear life, right?
Scott Likens (10:18):
How are they hanging on? Yeah.
James Kotecki (10:18):
Scott Likens, PwC, thank you so much for joining us today.
Scott Likens (10:20):
Thank you.
James Kotecki (10:22):
And thank you so much for joining us here in the C Space studio at CES 2025. We are going to be live all day. Stay with us. I'm James Kotecki.
(10:31):
Well, I hope you enjoyed that conversation from CES 2025. That is our show for now, but there's always more tech to talk about. So if you're on YouTube, please subscribe and leave a comment. If you're listening on Spotify, Apple Podcasts, iHeart Media, or wherever you get your podcasts, hit that follow button and let's give the algorithms what they want. You can get even more CES at ces.tech, that's C-E-S dot T-E-C-H. Our show produced by Nicole Vidovich and Paige Morris. Our C Space Studio episodes are produced and edited by Cramer. I'm James Kotecki talking tech on CES Tech Talk.