WTF Summit Recap: Jessica Lessin’s Fireside Chat With Amazon’s Colleen Aubrey

As the artificial intelligence hype cycle wanes, business leaders are pushing to move beyond experimental approaches and find applications that yield tangible value for their business. But with an emerging technology like AI, it can be difficult to distinguish between groundbreaking new use cases and short-lived fads.

For Colleen Aubrey, senior vice president of AWS Solutions at Amazon, the answer lies in moving beyond the big picture and getting granular about business problems.

“It’s useful to get very pragmatic and very simple,” Aubrey said during a fireside chat with Jessica Lessin, founder and CEO of The Information. “What is the problem that we can solve today with a new technology that we could not solve before? And if we’re able to solve that problem, how does that change the way we work?”

During the chat, held during The Information’s Women in Tech, Media and Finance Summit, Aubrey discussed the current AI landscape, some powerful real-world applications, and why it’s nearly impossible to predict where the technology is headed in the coming years.

The State of Enterprise AI

Lessin asked Aubrey to characterize the AI adoption curve among Fortune 500 companies. Aubrey’s answer: “It depends, and it’s a mix.”

Aubrey noted that some companies are better positioned than others to deploy AI. “Using AI as a tool requires you to have good data—which requires you to have organized your data in a place that is accessible,” she said. “We have been going through this process of getting ourselves organized for many years, and there are some companies that are further along. The realization that we’re all in the business of data…I think AI has only accelerated that.”

AI in Action

Internally, Aubrey said, Amazon has saved 4,500 years of developer time by automating Java upgrades via Q Developer, its coding assistant. While such monumental time savings is only possible at a company with such a scale, Aubrey noted that the example illustrates the opportunity for organizations to eliminate low-value “toil,” free up their employees to focus on more fulfilling tasks, and achieve a clear, measurable return on investment using AI tools.

“If we can eliminate four and a half thousand years’ worth of developer time, is that useful—yes or no?” she said. “It’s a relatively easy answer. But I think there are many cases of that.”

Outside Amazon, Aubrey noted, companies like Autodesk and Pfizer are using AI to enable more sustainable aircraft designs and accelerate drug development, respectively.

Aubrey also pointed to how Amazon Connect, the company’s commercially available contact center solution, leverages AI to provide customer service representatives with just-in-time resources and suggestions for next steps. “What I like about the way the team has integrated AI there is [that] it just sort of happens,” she said. “If I’m an agent, I don’t have to learn a new process. Things are happening and toil is being eliminated, because AI is being used in a very simple way.”

Looking Ahead

Aubrey noted that AI is in the “super early” stages of its maturation, prompting Lessin to ask why leaders should be confident that the technology will eventually become essential.

Aubrey acknowledged that there are still many unknowns—including what business problems could be solved with artificial general intelligence—but she predicted that AI will dramatically transform the business world in the years to come. “When I think about the size of shift, I think about the industrial revolution,” she said. “There will be a lot of things that we [currently] spend energy and time doing that we could have machines doing on our behalf, which will be lower-value, repeatable tasks. Hopefully, we go down the path of applying ourselves to more interesting problems to solve. But it’s way too early for me to even see where the extent of that would be.”

Aubrey once again stressed the importance of businesses taking a pragmatic approach to this journey, saying that often businesses are focused too much on going “to the moon” with their ideas for AI, instead of focusing on the near-term problems. .

“I would like to see more…‘What are we going to do this month, next month, next quarter?’” she said. “‘What are those milestones that we are going to achieve?’ I would like to see more of that and less of the visioning.

“AI is a way to learn, iterate quickly, go deep, and then… be the catalyst for some big ideas.”