If the hype around generative AI is indeed losing any steam, you wouldn’t know from the lengthy, meandering check-in and lunch lines at the annual AI Summit in New York this week.
Tech execs, academics, and experts descended on the Javits Center for what attendees said was an especially hectic iteration of the conference focused on how businesses are putting the latest AI breakthroughs to use.
The event offered a look into how companies like FedEx, Johnson & Johnson, and Fossil are grappling with large language models (LLMs) and the inherent risks.
After a year in which it seemed like there was a big new enterprise AI tool debut every other week, many companies are still in the process of figuring out how the tech best fits their business. “Like a lot of companies right now, we’re kind of trying to figure [it] out,” FedEx CTO Adam Smith said in an onstage interview.
FedEx’s experimentation with generative AI covers three broad areas, Smith said: Finding the right partnerships, making “core processes” more efficient, and supporting the company’s developers with features like coding assistance.
The shipping and logistics company has experimented with everything from using LLMs to improve delivery estimates on orders to using a form of generative AI to design more efficient ways of stacking boxes through partner Dexterity AI, FreightWaves reported.
“If leaders don’t have small teams working in the generative AI space right now with very focused outcomes, they’re behind,” Smith said. “Because the reality is it’s evolving, and it’s evolving fast.”
Damian Fernandez-Lamela, global VP of data science and analytics at watch and accessories company Fossil Group, discussed how the company is experimenting with using generative AI for watch design, while acknowledging the complexities of navigating potential intellectual property issues that have made AI product design challenging.
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.
“Now, our designers have the ability to utilize this broad inspiration tool, where they will say, ‘Hey, I want a watch with two sub-dials with three hands…and immediately the system can provide a breadth of different ideas, a breadth of different concepts that they can iterate on,” he said onstage.
“We're not using the outcome of these models directly; we're using them as inspiration,” he added. “Then a human designer will take the ideas, and they will come up with the actual final design that ensures the ownership rights because if it is directly produced by the AI, there is an issue of ownership of that IP.”
In the healthcare space, Pablo Damasceno, principal data scientist at Johnson & Johnson, discussed how his company is navigating AI in medical settings. The company is currently looking to AI to help with everything from drug discovery to analyzing surgery results.
Damasceno detailed ways to root out bias in these kinds of tools: “You have to backtrack and say, OK, what’s going on here? Why is it not working?...And then you go in and acquire more data…So why is there bias here? What’s the cause and what can we do about it?”
Dan Diasio, global artificial intelligence leader at EY Consulting, told Tech Brew that this has been a year of experimentation for companies, but they will likely zero in on what works next year.
“My guess is that in 2024, it’s going to shift more toward a lot smaller, a lot fewer initiatives, but those initiatives will help the companies build strategic differentiation,” he said.