The capabilities of artificial intelligence are constantly evolving.
Traditional analytical AI uses rules-based systems and statistical models to deliver use cases like spam filters, recommendation engines, and fraud detection.
Next, generative AI became mainstream in 2022 with the ability to create text, images, music, and code.
Agentic AI represents the next leap. This is AI that doesn’t just generate content but acts, adapts, and collaborates across systems.
For example, a manager might ask the agentic AI to prepare a summary of competitor activity and schedule a strategy meeting. The AI monitors news and social media, compiles insights, drafts a report, books the meeting, and sends out the invites, all autonomously.
On stage at the Annual Hospitality Conference in Manchester, Micheal McCartan, EMEA vice president, IDeaS, wondered whether hospitality leaders were even talking about agentic AI let alone experimenting with it.
Despite the buzz, agentic AI is still in its early stages. “It is an emerging concept. One year ago, no one was talking about it,” confirmed Jules Barker, associate partner, Real Estate, McKinsey & Company.
AI fatigue
He also acknowledged that business leaders are growing tired of hearing how AI is the solution to all their problems: “There is also a sense of AI fatigue and I’m very conscious of the phrase ‘AI is everywhere except the bottom line.’”
Most companies have only applied a thin veneer of AI to their day-to-day operations compared to what is possible, he added.
“I’ve heard many CEOs say: ‘We’re forward thinking in terms of AI. I’m getting all my executives to use Copilot as much as they can in their jobs. But that’s not really going to work. All you’re going to do is pay Microsoft slightly more each year and you’re going to write slightly worse emails slightly quicker.”
What should companies be doing instead? And, given the rapid pace of AI transformation, is the first wave of AI adoption wasted now?
Barker said: “No it really isn’t. It’s like planting a tree. The second-best time to plant a tree is now. The best time was ten years ago, and it’s the same with AI.”
A lot of what is difficult about digital and AI transformation in any business is not the technology, it’s the culture, the people, the processes, the data, the change management, he said.
Next wave
Work that companies have already undertaken related to data processing or identifying use cases is still valid, he said, and added: “Unfortunately the tech you’ve put in place may now have been superseded but it doesn’t matter because you’ve learned something incredibly important – how to get people to buy into the change.”
He advised business leaders to focus on the value of AI implementation within single domains rather than trying to do everything at once.
A vertical drill down into finance, asset management, or front-of-house, is much more likely to make a meaningful difference than a horizontal approach across the whole company.
“Pick an area of your business that is going to benefit the most from automation and then tackle multiple use cases within that domain and that will enable to you to build up the culture and change management muscle,” he said.
Then the managers and champions of AI change management can go from that area and replicate their learning and experience elsewhere.
Technology teams tend to operate on sprint cycles of two weeks and the scope of what they are developing is likely to change over time. CEOs need to stay in touch with what is happening at least on a fortnightly basis, to make sure it stays relevant in a rapidly changing world.
Barker commented: “Culture is the big thing here. It’s never just technology. The biggest risks are that the business doesn’t want it, or when the tech team ends up not delivering what the business wants.”
All quotes taken from the ‘Beyond the buzz: real-world tech that’s reshaping hospitality’ session at the Annual Hospitality Conference 2025 in Manchester. Moderated by Michael McCartan, EMEA vice president, IDeaS.