EY exec: If you think agentic AI is a challenge, you’re not ready for what’s coming

Companies struggling to keep up with the arrival of AI agents should buckle up: even more complicated technologies are quickly coming down the pike.

That includes physical AI, which includes robots — and which Nvidia pegs as a multibillion-dollar market —and quantum computing. Both are likely to disrupt a number of industries in the coming years. 

Companies can either adapt quickly and stay ahead of the curve or they will remain left behind, according to Joe Depa, chief innovation officer for Ernst & Young (EY).

Not surprisingly, the tax and advisory firm is taking a big interest in future looking technologies. The company has a history of embracing new technologies and advising clients on how to handle them.

Computerworld sat down recently with Depa to get his take on AI projects in the enterprise, the role of consultants and how companies can adapt to new technologies beyond AI agents. 

Many surveys show lots of AI proofs-of-concept, but very few get to production. What are you seeing on the ground right now? “The speed of technology evolution is accelerating. We’re moving from generative AI to agentic AI to physical AI, with quantum right behind it. ChatGPT was invented three years ago, and you’re seeing headlines around not getting value out of AI, but you’re also seeing clients get value.

“Some experimentation is standard for any technology life cycle, which is innovation theater in the beginning. But you’re now getting tangible use cases where AI is having impact.

“When we talk about agentic AI, we have use cases focused not just on productivity —particularly in back-office functions like finance, procurement, HR. Agentic AI is disrupting the software development life cycle. That area is ripe for agentic AI. Then physical AI is coming soon.”

Is AI adoption more like a sequential journey — getting one technology right before moving to the next — or is everything happening simultaneously? “The convergence of technology is happening all at once. You’ve got new processes being put in place while simultaneously replacing legacy infrastructure. You’ve got new technology, new talent being rolled into this convergence. Meanwhile, physical AI and quantum are coming quickly on top of agentic.

“Adaptability is the new job security. The ability to adapt is the most important skill for employees and the most important organizational differentiator. Organizations that can adapt quickly to new technology, redefining processes and training — that’s how they’ll differentiate. The ones that can’t will fall behind.”

With so many technologies emerging, how do you prepare for business changes that people can’t even yet anticipate? “It’s becoming not a technology issue as much as a business and process issue. The technology — whether AI, agentic AI, physical AI, or quantum — mostly exists to solve today’s problems. The issue is training, people, and adoption.

“Take healthcare. Robotic surgeries can be performed in some key categories at or better than human surgery. The robotic surgeon, if trained appropriately, isn’t tired, takes out the human emotional element, and performs surgery with laser-like precision.

“There’s a doctor shortage everywhere. Robotic surgeries remove some compression on the system and provide better health outcomes. But getting surgeons to adopt robotic surgery is a challenge. How do I train doctors on these robots? 

“In some cases, there’s resistance because they don’t think robots can do it as well, even if you show them data. The technology’s there, we know it works. But if I can’t get hospitals and doctors to adopt it, it doesn’t matter. It’s less of a technology challenge, more of a change management challenge.”

How do you educate customers? Do you focus on solving the data problem first, or are people rushing to get agents in place either way? “It’s a combination. Some industries, like financial services and healthcare [and] precision medicine — financial services has over-invested for decades in data and data quality for compliance reasons. They can use it for AI and quantum. Precision medicine is another category with high data quality.

“But without the right data, infrastructure, and sandbox, you’ll spread yourself too thin. You may try things, but it doesn’t get you value. Without a defined use case and focus area, you create innovation theater.

“Companies are getting focused on that first step: What use case am I trying to solve? If I can get specific around the use case or business outcome, then the next question is, ‘Do I have the right data’? 

“If I have the right data, let’s simulate and use this technology to produce the outcome. And if we have the right outcome, is it going to change behavior in your organization? What action are we taking? It’s the use case, the data, the simulation, and there has to be actual outcome or action to get out of innovation theater.”

What role do partners and consultants play in AI deployment? How does consulting fit in when AI is supposedly replacing consultants? “What’s starting to happen is an open innovation ecosystem. The world’s moving quickly, so you have to leverage alliance partners more closely. If you want to experiment with quantum, your best bet is not to build your own quantum computer. Partner with somebody.

“We pick a few partners we trust with similar business strategies. We create an open innovation ecosystem where we’re lockstep in how we go to market. That’s important because it provides the speed you need.

“When you ask about consulting, what people need is really smart people that understand technology, data and AI, that can help identify business problems and solve them using technology more efficiently.

“If consulting services changes, it’ll change to deliver services more effectively. It comes down to talent. Do you have people that know how to deploy AI and agentic AI? Do you have people that orchestrate across multi-vendor environments? Do you have people that understand regulatory risk compliance? 

“Consulting firms with the right talent are going to see great success. You’ll see more opportunity, but also bifurcation. Without that, those consulting services will go away.”Apple’s Creator Studio: Life after the App Store? – ComputerworldRead More