An Accenture report found a perception gap between workers and C-suite executives around AI, with 60% of workers anxious about being replaced by AI and only one-third of executives feel job displacement is a major worry for their workers.
Accenture (ACN) Chief Technology Officer Paul Daugherty joins Yahoo Finance for the latest installment of its AI Revolution special to help discuss the data and the ways workplace AI may effectively be incorporated into many business models to benefit workers.
Daugherty explains that the advent of AI is like nothing before, and so businesses will need to revolutionize how they approach work: "We believe what's happening, what needs to happen, is companies need to re-invent work itself and how it happens. That then changes the role that people have. So think about how to develop those new skills — think back to the internet when it came out. Before the internet, we didn't have web designers, eCommerce experts, search engine optimizers. We didn't have eBay, Etsy, entrepreneurs building businesses around the internet. That's the new creation we're going to see. We've identified 12 new jobs that didn't exist a year ago that we're building and hired within our company as we grow our AI talent from 40,000 to 80,000."
JOSH LIPTON: Switching gears as AI dominates markets and earnings calls. New research shows 95% of workers say they see the value of working with it. Yet Accenture's data also shows over half feared job loss and stress. This concern is underscored by recent layoffs at Microsoft following the lead of Google, Amazon, and others. Now we have the author of that report, Paul Daugherty, joining us for insights into navigating the AI revolution and how businesses can invest in the future.
Paul, it is good to see you. I actually wanted to start maybe more high level, Paul. There's been this boom of interest in AI. Certainly, investors are excited about it. Companies talk about the time and effort and money they're putting into this tech poll. But I'm interested, as you survey the landscape right now, how do you see companies putting that technology to work right now, Paul? What are the use cases that you see out there that get your attention?
PAUL DAUGHERTY: Yeah. Thanks for having me on the program. And it's a great question. I think we're in the midst of a shift in terms of how companies are viewing AI, now generative AI, especially. I look at 2023 as the year of education and experimentation. After all, it was only 14 months or so ago when ChatGPT burst on the scene. And we've done over 700 generative AI projects with clients. And that's a wide array across industries around the world.
And what we're seeing is a shift to more companies move from that experimentation to scale. And we really see 2024 as the year when executives are looking to see the results, see the impact, and see the value that can drive at scale. So some of the use cases that I'm excited about that we're working with companies on are areas like drug discovery and life sciences, where you can pair a scientist with generative AI models that understand chemistry and biology and can dramatically accelerate the drug discovery process. Very exciting.
All the way to examples like the customer service, where we're applying generative AI today to help customer service agents respond to queries faster, more productivity, give better results to customers, and really dramatically increase the satisfaction that customers are getting. So kind of a win-win all around. And the interesting thing from some of those examples is the workers enjoy the work better, because it removes some of the drudgery. It allows them to really focus on relationships and interaction and the things they like to do.
JULIE HYMAN: Paul, as Josh alluded to, we've also seen some redeployment of resources, right? We have heard about some layoffs this year. Accenture itself did some layoffs last year as well as companies trying to figure out where to put the resources. What strikes me, though, is I've heard a lot of reassurances about how AI is going to make workers jobs more productive, easier. But the folks who are being let go are not being retrained to do AI jobs. They're just being let go. And then other AI people are being hired is the way I'm seeing it. Is that what's going on here?
PAUL DAUGHERTY: We're really early in the cycle. And that's what the report that we released showed is that workers are optimistic. 95% of the workers surveyed do believe generative AI can help their careers. And they're excited to learn more about it. 60% of them have anxiety about whether the company's going to bring them along with it.
Then you look at the company side of it, it's a different story, where the c-suite executives that we surveyed don't see the broad release of people. They see the need to bring the people along. So we call it the trust gap that's forming. And there's a need for leaders to be better painting the vision of what they're doing and maybe be a little bit more transparent and then companies to invest in the learning platforms to bring the workers along.
So while workers want-- companies and leaders want to bring the workers along, only about 5% of companies are investing in those learning platforms and education at scale. That number needs to go a lot higher to address the need for workers we're going to have. The reality is you're not going to be able to get rid of today's workers and hire tomorrow's workers because the skills are changing so fast that the companies that win are going to be those that understand how to train and develop the learning program to bring the workers along with them and add the new skills that they need in this new generative world, the prompt engineering and other skills that are going to be table stakes for any job in the future.
JOSH LIPTON: What is it, Paul, about AI that helps companies here, more than the tech they already have in place?
PAUL DAUGHERTY: Yeah. It's important to step back and realize that. I've been doing this for over three decades, and I've seen every kind of every one of these waves come through. There hasn't been anything like generative AI over that period of time. It is different, and it has a different level of impact.
The reason is, one, the speed-- 14 months to the scale of impact we're seeing. A second is this human-like nature of the technology. It's the first technology that's like looking in the mirror in terms of what it can do, in terms of creating content and other capabilities. And as a result of that, it changes everything that people do. So the job of the CEO, the c-suite, the manufacturing worker, the frontline workers, everybody's jobs are going to change as a result of this. And that's what's really different about it.
And therefore, we believe what's happening and what needs to happen is companies need to reinvent work itself and how it happens. And that then changes the role that people have. So it'll be about how do you develop those new skills. Think back to the internet when it came out. Before the internet, we didn't have web designers, e-commerce experts, search engine optimizers. We didn't have eBay, Etsy, entrepreneurs building new businesses around the internet. That's the new creation.
We're going to see we've identified 12 new jobs that didn't exist a year ago that we're building and hiring within our company as we grow our AI talent from 40,000 to 80,000. And-- and that's just the tip of the iceberg as you look at enabling everyone to be able to participate in this generative AI-driven business world an economy.
JULIE HYMAN: Paul, finally I want to ask you about what you train the models on, right? There's been some copyright questions, right, about the data that some of the models are pulling from. We saw that play out for example with the Hollywood strikes. So how do you approach that and make sure that you have the rights to the data that you're training stuff on?
PAUL DAUGHERTY: Yeah. You know, we announced a $3 billion investment last year in data and AI to build the capabilities that companies will need as they move through this journey. And one of them is exactly what you said-- how do you choose the right models and train the models right.
There's over 600 models, different foundation models out there to choose from. So it's a bewildering array of choices that companies face. And there's strategic and investment and lots of implications in which models. And to your point, you have to understand how the models are trained. There's decisions on proprietary models versus open source models. And that's one thing we're working through a lot of companies on is, how do you make those choices, and then what degree of customization do you need to drive your business?
Initially, companies were thinking, well, I'll just use the model and I'll do what's called retrieval augmented generation or prompt engineering to get the right results out of the model. And companies are now realizing that to drive really strategic results. They need to invest more in using some of their own data to fine-tune or train the models to really drive that strategic impact. And that's-- with the investment, we're making in different industries and different solutions. That's how we're trying to help companies move through that and do it in a safe way to avoid some of the risks that you have with, as you mentioned, with intellectual property and other considerations.
JULIE HYMAN: Paul, it's great to catch up with you as always. Thanks for your time.