Inside Amazon’s AI cloud strategy

At WSJ Tech Live, Amazon Web Services chief executive Matt Garman addresses the perception that the cloud giant has been slow to adopt generative AI, and discusses cybersecurity, regulation and his company’s strict return-to-office policy.
At WSJ Tech Live, Amazon Web Services chief executive Matt Garman addresses the perception that the cloud giant has been slow to adopt generative AI, and discusses cybersecurity, regulation and his company’s strict return-to-office policy.

Summary

Matt Garman, CEO of Amazon Web Services, says the company didn’t get off to a slow start with generative AI. It was just more deliberate.

When Matt Garman took over as chief executive of Amazon Web Services in June, one challenge stood out among the rest. He is now responsible for directing the efforts of the cloud-services business—the main driver of Amazon.com’s profits—to harness the potential of artificial intelligence for its customers. In doing so, he has to compete with powerful rivals like Microsoft and Google parent Alphabet.

At the recent WSJ Tech Live conference, Garman talked with Emma Tucker, editor in chief of The Wall Street Journal, about AI and other significant challenges, including government regulation and the potential effects of Amazon’s decision to call an end to remote work for corporate staff in the new year.

Here are edited excerpts of that conversation.

A big bet

WSJ: Like a lot of big tech companies, Amazon is spending huge amounts of money in the AI boom. Do you worry that all this competitive spending is creating a bubble? Or do you think it’s leading to tangible results?

MATT GARMAN: I firmly believe that AI, and particularly generative AI, is a transformational technology that is going to change every single company, every single job, every single workflow out there.

WSJ: I think there’s a general perception that Amazon was off to a bit of a slow start with generative AI. What are you doing to play catch-up?

GARMAN: I think the whole world was amazed when ChatGPT first came out. So what you saw was a bunch of providers rush out and try to get something out to market quickly. We took a bit of a different approach.

We firmly believe that your data is the thing that ultimately differentiates one company versus the other as you go and actually apply these models inside your businesses. And so companies want to make sure that that was very secure, that they had control over how that data was accessed when they’re using it in an application. And then we also wanted to enable people to go build interesting models and interesting technologies.

And so we had this fundamental view that there wasn’t going to be one model that was going to be the rule, but there’s going to be a lot of different models, used in concert with each other to go build interesting applications. And that people are going to need a rich set of tools to deliver interesting business results. So we took a step back.

We basically said we’re going to go build that platform, so that our enterprise customers, our startup customers, our government customers can go build those applications and build what’s going to be unique to them, and not just get something to market quickly.

WSJ: Another big challenge I know you spend a lot of time thinking about—so much so that Amazon has bought its own small nuclear reactor—is energy. I think I read that these small modular reactors aren’t going to come on stream until 2036 at the latest. So what are you going to do in the meantime to meet the energy demands?

GARMAN: I think it’s 2030 hopefully, not 2036—we’re more optimistic they can deliver a little sooner than that. But they’re just part of the portfolio.

In the near term, there are projects in place that can meet all of the needs in the next couple of years. And many of those are renewable, by the way.

And we think that nuclear can be a big part of this. There’s a bunch of nuclear around the world that’s been shuttered because people didn’t need that source of power. And so the very first project we did was a partnership with Talend [a data-management company] in Pennsylvania to bring eventually almost a gigawatt of power back online that just wasn’t being used.

And some of these small modular reactors, they’re not going to solve anything in the 2020s, but in the 2030s they could be an excellent source of energy.

Too much regulation?

WSJ: Do you fear that too much regulation will hold back the development of these new technologies around AI, or do you think the U.S. has roughly got it right?

GARMAN: We’re happy to play with whatever rules the government wants.

What we advocate for is really thinking through not setting regulations or policies that are inadvertently going to lead us to the place that you’re trying to avoid. I think you just want to be super careful about that, because it’s very easy to construct a scenario where you actually give China the leg up that you’re trying to prevent by accidentally holding back the companies that are doing all this incredible innovation.

It’s a hard question. I don’t think we know the answer, and I think zero regulation is probably not the right answer either. We have to be thoughtful about it.

Back to the office

WSJ: In September CEO Andy Jassy ordered everybody back into the office five days a week, from January. You were speaking about this last week, and you were very clear—shape up or ship out was basically your message.

GARMAN: That’s not quite what I said.

WSJ: It was definitely the message. So have people shipped out, or has there been a lot of pushback?

GARMAN: Just to be clear what I actually said, it was more like, for us, we think that being in person is super important. We think that our teams are more innovative, and—particularly as we think about how do we want to disrupt and how we want to invent on behalf of our customers—we find that there is no substitution for doing that in person, just the creative energy and how fast you’re able to iterate. When you’re sitting there writing on a whiteboard, or you’re talking to people in the cubicle next to you, or you’re running into people that are in a different department but you see them at the coffee line or whatever it is—there is just that exchange doesn’t happen when you’re remote.

We tried three days a week first, and then what happened is Bob would come in on Monday and Sally would come in on Tuesday and we kind of didn’t accomplish what we wanted because everybody picked a different set of three days.

And so if it’s not for you, then that’s OK. You can go and find another company if you want to, but for us, that’s what we’ve decided is the best way to operate our company.

WSJ: So could you put a number on what percentage of the workforce you expect to churn in January or before then?

GARMAN: I don’t know. I’ve had plenty of people reach out to me and say, part of the reason I came to work at Amazon is because I want to work together with a bunch of smart people and be together. I am optimistic that most of our employees actually are excited about that in-person working.

Catch all the Corporate news and Updates on Live Mint. Download The Mint News App to get Daily Market Updates & Live Business News.
more

topics

MINT SPECIALS