Cisco CEO talks AI and why the internet ‘needs’ the company

The AI conversation continues and the World Economic Forum in Davos, Switzerland was no exception. Cisco (CSCO) CEO Chuck Robbins joins Yahoo Finance Live to discuss the trajectory of AI and how it will impact the company.

Speaking to large language-based models in generative AI, Robbins believes that the companies that will stand out are those that are creating “value-based capabilities that help end-users take advantage of the data that they have.”

Robbins sees a wide range of possibilities for the incorporation of AI in networking equipment, “it’s still early, but I think there’s going to be a lot of opportunity.”

Known for being a company that makes acquisitions, Robbins notes that some company valuations are “insane,” but says the company has “gained incredible AI talent,” through recent acquisitions.

Looking ahead to the future, Robbins states that though the company has been building out its software business, it’s hardware business is still important too saying “the internet needs us to be a really good hardware company.”

The world of AI is constantly evolving, and Robbins states the company is ready to meet demands, “we’re now building a plan for how do we [Cisco] educate different aspects of the company of how they should be thinking about it.”

This interview is part of Yahoo Finance’s exclusive coverage from the World Economic Forum in Davos, Switzerland, where our team will speak to top decision-makers as well as preeminent leaders in business, finance, and politics about the world’s most pressing issues and priorities for the coming year.

Watch this full episode of Yahoo Finance Live here.

Editor’s note: This article was written by Eyek Ntekim

Video Transcript

[AUDIO LOGO]

BRIAN SOZZI: Lots of focus on all things AI here at this year’s World Economic Forum in Davos, Switzerland. Let’s keep that AI conversation going with Cisco CEO, Chuck Robbins. Chuck, always nice to get some time with you. What are we missing on AI? Tell us.

Story continues

CHUCK ROBBINS: Nothing.

[LAUGHS]

BRIAN SOZZI: Nothing, OK.

CHUCK ROBBINS: You have to know everything about it at this point with all the discussions this week. I think– look, I think it’s incredibly early, I mean, with AI right now. And everybody’s trying to speculate. But I think if you go back– think about when the iPhone launched. And we had no idea what applications were ultimately going to show up on that iPhone, and the things that we were able to do with it, and the use cases that came.

And I think we’re sort of in that same phase right now, particularly with enterprise use cases. I think our customers are really trying to– I think we know some that make a lot of sense like customer service and contact center and creating new interfaces to technology that are more language driven as opposed to typing on a command line.

Assistance that help you with different technologies where it helps walk you through what to be– how to be thinking about possible outcomes or possible causes of things that are going on. All that stuff, I think, we expect and we’re working on right now. But the big killer enterprise applications, I think, are yet to be determined.

BRIAN SOZZI: I was walking down the promenade this morning. And everyone has a window, and they all have these big TVs showing off demos of something. And I’m sure it’s backed up by large language models supporting something. All these companies can’t win, right, Chuck? There is going to be a shakeout, right?

CHUCK ROBBINS: Of course.

BRIAN SOZZI: What does that look like?

CHUCK ROBBINS: Well, I think if you look at the foundation– and I’m not an expert on this stuff. So the guys that you’re talking to are clearly probably more with the details. But what I believe is that the large language models that are largely being built off of available information– sucking information from the internet– over time, you’re going to get the information loaded, and they’re going to be effective. And they’ll be generally– I think even Sam Altman even said that it’ll go to zero and commoditize.

I think the ones that are going to win are going to be the ones that figure out how do I create value-based capabilities that help end users take advantage of the data that they have? And how do I help them actually turn that into something more valuable than it is today?

And I’ve met with a couple of companies that are doing some really interesting stuff right now that I think are thinking that way. And we’ll be partnering with a lot of them because of our enterprise presence. So I think that’s how it’s going to play out.

JULIE HYMAN: And you said that the way the AI might inform your business is that there’ll be more demand for networking equipment, because there’ll be more demand for compute because of AI, right? So–

CHUCK ROBBINS: That’s one way.

JULIE HYMAN: So, talk to me about that way specifically first. And then we can get into the other ways. Like where are we in that adoption cycle?

CHUCK ROBBINS: So how technical do you want to go? So–

JULIE HYMAN: Not too. Not too.

CHUCK ROBBINS: So let’s take NVIDIA as an example. Today, NVIDIA sells their GPUs, and they have a networking technology that they sell with it. It’s called InfiniBand. And most of the web scale players and others who are building these large language models want to use ethernet instead of InfiniBand. They want to use a standard technology.

There are some special features that we need in ethernet to make that viable. But we’re now running ethernet underneath the GPUs in pilot phase in a lot of the web scale players today. And so that’s a big opportunity for us, because as they build that out, and we get this technology underneath, then every time the GPUs go in, you’re going to see our technology underneath it.

Then the enterprise is really trying to figure out– you’re going to have– you’re going to need to build a stack of technology that allows customers to take small amounts of their data and put it at the edge of their enterprise, closer to where the customer is. And they’re going to want to run inference models. And so we’ll be– we’re working on designing what that might look like. In many cases, partnering with the GPU players with our technology.

So it’s still early, but I think there’s going to be a lot of opportunity there. And then you’ve got all the issues of security that have to be dealt with.

BRIAN SOZZI: Outside of– I know the Splunk deal, you’re still working through that. But Cisco, for decades, has been a major acquirer of very large businesses. I mean, it’s how the company has got to the point where it is today. I mean do you look at some of these companies on the promenade and think, wow, this could be a feeding frenzy over the next decade.

CHUCK ROBBINS: I just can’t afford any of them.

[LAUGHS]

BRIAN SOZZI: Well, the valuations too high.

CHUCK ROBBINS: Some of the valuations, I think, are insane. But I mean, that’s what the markets– the market is the market, right? So it’s– and part of the reason– I made a comment earlier this week about the valuations relative to their revenue, but the reality is is that to get started in these businesses required so much capital, that the valuations are just naturally going up because of the amount of capital that’s being infused to build out the underlying technology infrastructure. So I do understand how it’s happened. I just don’t know how it ends.

But no, I think you’re right. We’ve made– we’ve made several acquisitions in our security portfolio and in our collaboration portfolio that we have just gained incredible AI talent. So we’re pretty excited about that. So I think you will see us continue to do that.

JULIE HYMAN: One of the other things we’ve been talking about is companies have been around for a while that are undergoing some changes. The Splunk acquisition– and when that closes, that will be a change for you guys. You’re trying to diversify a little more into services, as I understand it.

CHUCK ROBBINS: Software, yeah.

JULIE HYMAN: Software. Away from the traditional business for you guys. So what does Cisco look like five years from now?

CHUCK ROBBINS: Well, we’ve been working on this transition– and when we close Splunk, I think our software business will be somewhere around $22 billion. And so we’ve made a lot of progress. And I don’t know if that makes us number five or six or something in the world.

But at the same time, we still have a pretty meaningful hardware business. And the internet needs us to be a really good hardware company, because all the technology that we need to continue to support all the increased traffic that we see around the world. So I think you’re just going to see both sides continue to increase.

I think as this AI thing explodes, my expectation is that our infrastructure equipment that goes with it will also go up. And we’ll continue to invest in our software assets and add more software assets to it where we can.

BRIAN SOZZI: We’ve talked to a lot of leaders here so far at the conference. And I think they know a little bit about AI. But for many of them, this is new. And it might be more new stuff coming at them tomorrow. As a leader, how are you staying informed about a technology that wasn’t around 20 years ago, and things might be different a week from now?

CHUCK ROBBINS: We have a– about three of my direct reports are pretty deep in it. And so they give us primers almost every week on what’s going on, what’s changing, how things are evolving. And so we just– we spend a lot of time with them educating the rest of the team.

And we’re building– we’re now building a plan for how do we educate different aspects of the company on how they should be thinking about it. I’ve asked every function inside the company to give me their plan for how generative AI is going to fundamentally change the productivity or whatever they do. And so everybody’s working through that as well. So it’s all educational as we go.

BRIAN SOZZI: Fascinating stuff. Cisco CEO Chuck Robbins, thanks for giving us time at Davos. Like always, we appreciate it.

CHUCK ROBBINS: Great to see you guys.