An elevator that always seems to be there when you need it. A dentist who rarely, if ever, misses scarcely noticeable cavities. A travel website that books an entire ideal vacation with minimal effort on your part. These are just some of the barely perceptible ways that artificial intelligence is helping to make people’s lives easier and allowing companies to both increase efficiency and innovate.
The above examples were among many discussed at Fortune’s virtual roundtable discussion, Leading With A.I., on Wednesday, with C-suite leaders from a wide range of sectors detailing their companies’ use of this technology, now and in the future. Much of the talk was spurred by a recent Accenture survey of executives on A.I., which showed that companies who invested in this area were much more likely to experience higher revenue and share prices.
Of the 1,600-plus companies surveyed, only 12% were considered to have achieved “A.I. Maturity,” which Accenture defines as having “mastered A.I.-related capabilities in the right combination to achieve high performance for customers, shareholders, and employees.” They expect that number to increase to 27% by 2024, mainly because these businesses have seen, on average, 50% higher growth in revenue over firms that are slower to develop their artificial intelligence capabilities.
Many of the executives participating in the roundtable emphasized the ways that A.I. is allowing them to go beyond efficiency-improving and time-saving tasks to predict and avoid issues that could negatively impact their customers. Visa, for example, is continually updating its deep-learning systems to flag fraudulent transactions, preventing a staggering $26 billion in worldwide losses per year while also reducing the amount of mistakenly declined payments by 30%. According to Visa’s senior director and head of government advisory, Michael Nunes, this goes beyond just standard credit card use.
“We’re constantly applying that to new technologies and new payment technologies,” Nunes said. “As ‘buy now, pay later’ gains traction, as we see new technologies emerging around different payment vehicles like cryptocurrencies, there are opportunities to use the A.I. that we’re leveraging for payments. We leverage it for our internal cybersecurity. That’s really important within the four walls of Visa, making sure that our network is safe and secure. We’re using A.I. to track and identify threats to the system, so I view it as something to protect the external environment to reduce fraud but also helping to keep Visa safe internally.”
Otis Elevator is using A.I. to identify problems before they occur in addition to more traditional uses, like optimizing the schedules of its repair technicians. Otis collects data on which floors use its elevators at which times to determine where to park the car when it’s not in use, and the company measures things like door speed and vibrations that could indicate a looming mechanical issue. Beyond that, Otis is looking to implement A.I. to integrate a number of external factors that could affect its ability to build, install, repair, and improve its products.
The company just started a pilot proof of concept with a startup in the Bay Area to take a closer look at its supply-chain orders, determining “Who are the vendors we typically buy from? Where are the shipments coming from? Is it coming by ocean, is it coming by train, is it coming by truck?” And Otis is using A.I. to look at geopolitical issues, such as examining what might be happening in a given country with freight and weather.
“It’s pretty exciting for a 170-year-old industrial company,” said Neil Green, executive vice president and chief digital officer for Otis Elevator.
That kind of forward-thinking approach is what Booking.com hopes to have down the line, according to CEO Glenn Fogel. The idea, he says, is to give customers the experience of dealing with an old-school human travel agent, who curated trips and took care of any issues after one phone call. To illustrate the kind of A.I. use he wants to see, he gave an example of a flight that’s canceled two hours before it’s due to take off.
“That happens, there’s no automatic rebooking. There is no adjustment for the car that was supposed to be picked up at the airport. There’s nothing being done in terms of if you’re arriving at the hotel late and having that room not canceled. There’s nothing about if you had a dinner reservation. None of this stuff is being done at all,” Fogel said. “What should happen is, these systems all talk together, all the data should be there, and it should be able to come up with a solution that is optimized for the traveler.”
A large impediment to making the most of A.I., many roundtable participants noted, is a reluctance to rely on the technology. According to Accenture, the health care industry has one of the lowest rates of maturity, a situation that Stryker CEO Kevin Lobo chalks up to “conservatism” on the part of doctors and hospitals. He gave two examples of A.I. that’s helping doctors, with one being an app that can determine if new mothers need blood transfusions by scanning a picture of a sponge in the delivery room.
“It’s really not to replace the health care professional,” Lobo said. “It’s to help them do their jobs more effectively.”
“Let’s face it, doctors don’t want to experiment on patients,” agreed Bruce Lieberthal, chief innovation officer at Henry Schein. “One thing we should add to the list of resistance factors is doctors. They’re businesspeople too, so they want to know that there is solid ROI. We don’t go into any customer to sell anything where maybe their first two or three questions isn’t, ‘Is this reimbursable?’ So, the payers have to keep up with this as well.”
The hesitancy also exists in the legal industry, according to Colleen Nihill, the chief knowledge management and practice services officer at Morgan, Lewis & Bockius. Her firm uses A.I. to pore over massive amounts of documents and to monitor the news around their clients to determine what public sentiments might affect them. The key to increasing the firm’s A.I. maturity, she said, is to have an internal team that can explain the value of the tech to skeptical attorneys.
“One of the things that I think is really interesting is there is an idea of ‘A.I couldn’t possibly help.’ They’re basically thinking that it should absolutely spit out exactly how we should practice law and thinking of it as a complete replacement for an attorney versus a tool,” Nihill said. “So we have tried to instruct the attorney that A.I. is only as good as them leveraging their subject matter expertise. What we’ve noticed is they need individuals that can speak both the language of law, as well as the concepts of machine learning, in order to accurately and adequately deploy that.”
The acceptance and implementation is just a matter of time for all companies, though. As Will Breetz, Salesforce’s senior vice president of product management, said, A.I. is a “must-have, not a would-like-to-have.” Even the possibility of a recession will not stop the progress, and it’s up to each business to determine how it will use the technology to innovate or be left lagging.
“We’re seeing a transformation from companies not just wanting to do what they’re currently doing but to do new things they could never do before. Companies still need, whether it’s a recession or you’re in a growth period, to serve customers, employees, to hit your growth targets,” Breetz said. “I don’t think that A.I. investments are going to slow down at all. Whether there’s headwinds or tailwinds, that could determine what companies choose to do with A.I. We do 170 billion A.I.-powered predictions or recommendations a day, and that’s doubled in the last 18 months. The growth is just unstoppable.”
