If last year’s AI summit was all about businesses showing off the fruit of a year of generative experimentation, attendees at this year’s iteration of the Manhattan conference seemed more concerned with how to turn those shiny prototypes into real moneymakers.
Sessions at Informa’s gathering were preoccupied with “real-world scaling,” moving “beyond the pilot,” and “AI on a budget.” Multiple moderators tried to gauge attendees’ progress beyond experimentation with informal shows of hands.
“We wanted to talk today about how we are starting to overcome this ‘failure to launch’ challenge,” Traci Gusher, Americas AI and data leader at EY, said at the start of one representative session.
Yet despite potential roadblocks, interest in the ninth annual AI and business-themed event remained high; organizers reported their biggest year yet, with more than 5,000 registered attendees. New York Governor Kathy Hochul even stopped by to address the crowd on the second day in a surprise appearance.
The conversations at the summit reflected the larger scaling challenges that organizations are facing two years after ChatGPT’s release kicked off a massive hype wave in the business world. A Gartner study earlier this year predicted that 30% of companies’ proof-of-concept AI projects will be abandoned by 2025. The research firm has classified the tech as entering what it calls the “trough of disillusionment” in its famous hype cycle trajectory.
Jim Rowan, principal and head of AI at Deloitte, said on stage that around two-thirds of businesses have moved 30% or fewer generative AI experiments into full production, per the consultancy’s most recent survey data.
“That’s really created a challenge and a tension, and that tension is really about ROI,” Rowan said. “Why am I investing so much money in this technology if I can’t see the unlockable value creation?”
Keep up with the innovative tech transforming business
Tech Brew keeps business leaders up-to-date on the latest innovations, automation advances, policy shifts, and more, so they can make informed decisions about tech.
That’s not to say that no businesses have managed to fully commercialize generative AI, Rowan said. “We’re starting to see some frontier organizations that are really embracing AI start to unlock that capability.”
Priceline CEO Brett Keller talked about how the travel booking site developed a chatbot called Penny to serve as a guide and assistant in the reservation process. As the company thinks about scaling the tool, Keller said the next stage of the mission is adding more personalization, though he declined to share conversion rates so far.
“How do you now personalize this and make it much more relevant? In the travel industry, there’s a lot that we can do with that,” he said. “‘We know that you travel with a dog. We know that you love warm destinations.’ Those are the kinds of things that will help us to really put the best recommendations in front of you.”
Booking flights is an oft-cited example AI companies use to demonstrate the utility of agents, but Keller said this “ultimate agentic flow” is still four or five years off.
“Travel is too complicated to address all of the sophistication related to airline schedules and times and delays, and it’s just a complex thing to pull together,” Keller said.
Suman Giri, VP and global head of commercial data science and analytics solutions at Pfizer, said there are two broad categories of obstacles at play in holding back companies: those related to data infrastructure, organization, and compliance; and problems navigating the confusing landscape of vendors and consultants.
Gusher predicted “there’s going to be a dramatic shift in focus on data readiness” as well as “consolidation of some boutiques and AI startups,” she said.