Betting that generative artificial intelligence (genAI) is top of mind for every IT manager and corporate executive would be a smart move. An equally smart take would be to realize that most organizations still remain reticent to fully embrace the technology and its automation capabilities.
Given the AI hallucinations, output errors, organizational data fragmentation and a lack of skilled IT talent to manage it, corporate leaders have good cause to be pensive. Yet, there’s not likely a consultant worth their salt that would advise an organization to not explore AI’s efficiencies, cost-savings and production-enhancing capabilities.
Julie Teigland is a managing partner and the global vice chair of alliances and ecosystems at global professional services firm Ernst & Young (EY). In her previous role at the company, she led 150,000 employees in a business region that encompassed 98 countries, including Europe, the Middle East, India and Africa. In February, Teigland began her current duties — focusing on how businesses drive transformation amid new technologies, a key one of which is genAI.
As part of her responsibilities, Teigland is especially focused on technology consulting and strategy all across all EY business lines.
Computerworld spoke with her about why genAI and related technologies haven’t been more widely embraced and how organizations can find their AI sweet spots. The following are excerpts from that interview.
Julie Teigland, managing partner and the global vice chair of Alliances and Ecosystems at consultancy Ernst & Young.
Ernst & Young
Has there been a sea change around generative AI this year? “I think back last year to Davos [World Economic Forum meeting], I was listening to just the CEO chatter and talking to our clients, and I felt like every CEO had an AI pet project. They were all talking about their cool pilots and what they’re doing.
“And I think that’s changed. I think AI is going broader. It’s moving from much more experimental projects to being mainstream and being deployed across the industry, where people are looking at how to really leverage it and get widespread allocation, widespread deployment. At the same time, I really feel that AI has not been embraced.”
Why hasn’t it been fully embraced by organizations? “I think there’s three things that are getting in the way. The first is the skill set. The skill sets aren’t there. And with those skill sets not being there, it’s making it difficult to deploy.
“The second is the data. And, having your data in order enterprise-wide — and tools to be able to leverage that.
“And the third thing I’d mentioned is the infrastructure cost, which you know — it’s an issue. It’s a real issue in terms of making sure that companies are making those investments, and those are Capex investments at a time when the world is in such a volatile place. I think companies are at least considering those investments, pacing themselves and being certain before they deploy it.”
So, is the investment in AI technology worth it, or are the risks too big? “The upside is huge, so I don’t think risk is going to stop them [organizations], but I think it does give them pause.”
You mentioned skill sets, but what are the skill sets they’re seeking? Are they seeking the data scientists, data analysts, prompt engineers? Or are they actually seeking the AI experts at this time? “I think they’re seeking all of the above. I personally think data scientists are still in high demand. If you think about it, data is the foundation for any type of utilization for AI getting those data structures. The job opportunities in the AI field for AI scientists have gone up massively.
“There is a huge skills gap in data science in terms of the number of people that can do that well, and that is not changing. Everywhere else we can talk about what jobs are changing and where the future is. But AI scientists, data scientists, continue to be the top two in terms of what we’re looking for. I do think organizations are moving to partner more in terms of trying to leverage those skills gap….”
I’ve heard you use the term AI ecosystems. What does that mean? “I see companies going out and partnering with other companies in order to try to make sure that they’re employing everything from gig workers to leveraging hyperscalers, to partnering in doing joint projects to put things together. You’re seeing some of that. I would not say that’s the widespread key to success, but you do see that in the market.”
So, do you think more companies are successful when they don’t partner and instead deploy AI unilaterally across an organization? “I do think you’re going to need a long-term plan to deploy it unilaterally. And you need the three things that I mentioned. You need a data strategy. You need some level of infrastructure. And you need skilled people.
“Now, you could go hire those skilled people, but then you have to make sure that you have somebody afterwards to run it. And that’s a strategy you have to consider, too – that you’ll have to have somebody else help transform the function and then give it to somebody else to run it on your behalf. I personally think that’s the new definition of a captive, but that’s my personal perspective.”
How will AI impact software engineering jobs? We’re seeing a lot about “vibe coding” right now, where people use natural language prompts to tell genAI what they want out of code, and then it just goes and does it. Do you see it impacting the number of jobs? Or is it just going to allow more creativity for people doing that sort of work? “It is going to impact jobs in terms of what people do. I think you will see much less standard base code. But let’s be honest, when that code comes out, it still needs to be checked, reviewed and adapted. The code AI produces is not perfect, and I know that we’ve been experimenting at EY. So, there’s a quality aspect there, too.
“AI tools allow us to be much more efficient in coding, but it doesn’t mean that we can release the [developers] yet. Right now, you just allow people to explore what they can do with the technology.”
How is EY exploring AI internally? “We are leveraging as many AI tools as we can to create more efficiencies, and to explore uses — and to drink our own champagne. To be honest, I think that’s really important. How could I convince a client that they need to do an AI project when I can’t say we’ve tried it?”
Let’s talk ROI. Are businesses anywhere near finding ROI in AI deployments? And, if not, how are they going to get there? “I definitely think they are finding it now. I think we’re at the beginning stages of this, but we do see productivity gains. And when I look at our own business and how it’s leveraging some of the tools, there are productivity gains that allow our people to focus more on the creative aspect of their jobs and to go faster in certain base tasks, to improve the quality by having AI help pre-check work.
“The issue is how do we quantify that ROI and how is it changing the overall model? That said, there definitely are productivity gains to be had, and we’re starting to see those. I think the question is, are there enough to justify the investments that’s required, and those investments are the three that I mentioned before.
“The question people need to ask is, ‘Does it justify the hiring of new people with AI skills? Does it justify the infrastructure that’s required and enterprise-wide data access?’”
Are there some verticals that will find AI ROI faster than others? If so, what are they? “Absolutely. The more specific the case for the use of AI, the more easily you can calculate the ROI.
“Healthcare is going to be ripe for it. I’ve talked to a number of doctors who are leveraging the power of AI and just doing their documentation requirements, using it in patient booking systems, workflow management tools, supply chain analysis. There, there are clear productivity gains, and they will be different per sector.
“Are we also far enough along to see productivity gains in R&D and pharmaceuticals? Yes, we are. Is it the Holy Grail? Not yet, but we are seeing gains and that’s where I think it gets more interesting.
“Are we far enough along to have systems completely automated and we just work with AI and ask the little fancy box in front of us to print out the balance sheet and everything’s good? No, we’re a hell of a long way away from that.
“When I talk to some of the startups in the AI space, they’re convinced that we’re going to have that soon…, but I think it’s going to take a little bit longer than they think. I’m old enough to say I’ve seen this movie before [with the cloud, mobile and internet].”
Speaking of those technologies. Would you compare the impact of AI to that of the internet, the cloud, smartphones? What would you compare this tech to in terms of being a game changer? “Oh, it’s a game changer. I think it’s different than all of those. It’s going to revolutionize the way we think, especially if I look at my own business with professional services. I can remember when the internet came out, people said, ‘Oh, professional services is over.’
“Professional services sold transformation. We still sell insights and transformation. But that’s evolved, and we are evolving with it. That’s why I think this [AI] is another tool in the toolbox. It’s going to force us to evolve.
“With the Model T, Henry Ford created the assembly line and forced companies to rethink how they produce automobiles. That was revolutionary.
“I think AI is a fantastic tool. Is it the be all to end all? No. Is quantum computing the be all to end all? No. But with AI, quantum computing is going to be pretty powerful.
“I’ve been out looking at some of those early-stage quantum companies that are on the edge of coming through with some fantastic breakthroughs. That’s some pretty powerful stuff. If you can say what’s happening in a natural world and combine it with AI that can do the math and the algorithms faster, and we can and put those two together, oh my gosh, that’s going to be a wow moment.
“That’s going to change the world when it comes to medicine, creating new natural compounds and saving us from climate change and all kinds of things.
“AI today is going to change how we due process work. It’s going to change how we look at data and insights, and it’s going to make us more intelligent. But, I still think we got the name wrong. Why the hell is it called artificial intelligence? It feels like a name created by a guy who was watching a sci-fi movie because anybody else would say it’s augmented intelligence.
“It’s going to make us smarter and faster, but it’s not going to replace us – at least not yet.”
Why is quantum computing so important to this, and is the technology even there to be used for AI at this point? “Quantum computing is leveraging the answers and seeking the answers that we see in the natural world, and it’s far more specific and far more fine than we could ever do with our current compute power.
“When you combine the two — the ability to query and run analysis and simulation and predictive views in a probabilistic manner with AI next to the computer that can really understand and do the math using the laws of nature — that is powerful for what we need to do to understand the world we live in. And we do not have that technology today.
“We’ve already got really good test cases out there. I went out to see a company…in Paris and they have five [quantum computers] running. They are refining things as they go, but they’re up there running. They’ve proven it can work. They are really close to a breakthrough. I would say, realistically, it’s still 18 months away, though.
“Right now, quantum computing and AI is a race. There’s only five companies in the world that are really close to this. But it is coming and that is going to be really exciting.
“Now, is that going to worry an accounting firm like mine or the Big Four? No. Is it going to worry the business world? Probably not. Is it going to change how we do R&D? Absolutely. It’s going to change how we look at pharmaceuticals, biotechnology, chemical compounds, sustainability, and climate change. We will be able to solve some amazing problems. And, it will be necessary.
“Will everyone want a quantum computer? Hell no. Why? Because you don’t need it. But will some need it…and they will be immensely important. So, I’m super excited about how far the [quantum computing] world’s coming along.
“It changes how we work. It’s going to change the jobs we have. It doesn’t mean that there’s less jobs, but there’s going to be different jobs, and that’s where I get so excited.”