How This Dot-Com Era CEO Is Navigating The AI Boom

0:00This episode is supported by Anthropic, the team behind Claude. Every business wants to move faster. Few can afford to be wrong. That's why nine of the Fortune 10 run on Claude for the work they want to get right. Safety and accuracy are not just line items for Anthropic. That's their whole mission. See what's next at Claude.ai.

0:23Welcome to Tech News Briefing. It's Monday, October 19th. I'm Danny Lewis for The Wall Street Journal. More than 40 years after Michael Dell founded his namesake direct-to-consumer PC company in his dorm room, Dell Technologies has pivoted to rapidly building the powerful supercomputers AI data centers rely on. But even as spooked investors flee some of the big tech companies that soared to new heights on the back of the AI boom, Dell says his company is ready to weather a storm. WSJ tech columnist Christopher Mims recently spoke with Dell about the company's strategy in the AI era and how it's prepared for a possible downturn.

0:58And for today's episode, we're bringing some of their conversation to you.

1:28typical US households, or maybe two or 300 if it's like a really high end. And this is the brains of these new AI data centers. If you think about what's happening now, we're sort of going through this shift from, we had like 60 years of calculating and computing. And that's what people did with computers. And now we have these models that are creating a kind of intelligence, Artificial intelligence or super intelligence, depending. Pick your phraseology, right? But basically, there's a lot of demand for that, whether it's in hospitals, improving their clinical workflows or manufacturing, improving their output and all sorts of other uses.

2:13I mean, it's kind of universal that people want to use their data more effectively to create better outcomes in whatever they're doing. Right. So AI is a way for them to get more about it. Right. And it can show up in many different forms, right? It's like in your PC, might be in your car, right? It might be in your refrigerator. It might be in your data center. It might be in the cloud, right? It shows up in many different places. And when you're building a large scale infrastructure, either to train the models or to serve up all the answers. Right. So I go on ChatGPT, I'm talking to the cloud.

2:54Exactly. Yeah. I want to anchor this in some very concrete things that you've built. So for example, Dell did half of the servers in the Colossus One AI supercomputer for SpaceX. And you did it really fast, right? I think that was a record fast build. It was like six weeks or something. Elon is probably the most demanding customer. No, he's definitely the most demanding customer there is. What do you learn from building that data center in six weeks? Which, you know, normally data centers take six months, right? Well, so, you know, Dell first entered the server business in 1989. Wow. When I was 24 years old.

3:34It didn't actually work the first time. Okay. It started working around the mid-90s. And so we've actually been in this business a long time. So it shouldn't be too surprising that we're also number one in AI infrastructure. But to your point, these systems are incredibly complicated. It's not just, you know, a rack with a bunch of chips in it, right? No, it's full of plumbing now. Well, it's, you know, millions of miles of fiber optic cable and engineering the liquid cooling to be superpower efficient, the storage, the networking, the deployment, you know, teams that come in and make these things all work and then keep them running all the time.

4:14Right. Because each one of these boxes is like six figures. They're pretty expensive. And we have learned a lot about how to quickly deploy these because we have this demand coming from customers. Let's talk about lessons from your big like NeoCloud, like hyperscaler customers, the folks that are standing up these big data centers that we read often about in the news. You're transferring these lessons now because companies are coming to you and saying we want our own AI, which means we have to have our own AI infrastructure. Right? Yeah. Companies want control of their data, their economics, their privacy, their security.

4:57So let's say you have 100 factories, you're a manufacturer. You actually want to have the AI inside each factory, doing inference to help you be very effective. And so, yes, we're building these out. It took us eight quarters to have 3 ,200 customers building these Dell AI factories. And in the last three quarters, we've added 3 ,300 customers. So it's doubled. It's more than doubled, 6 ,500. And what's also interesting is when you look inside these companies and actually what they're doing, they're just kind of at the beginning of figuring out how to do this. So I believe we're in the early stages of how this technology is going to be diffused and used around the world.

5:44And Dell makes all the things that are needed to make it happen. What would you like to know about the business of building AI supercomputers? If you're a listener on Spotify, let us know in the comments. Coming up, Michael Dell was also running his company during the dot-com bubble. How is that shaping his approach to a frothy time for AI? That's after the break. Thank you. CitiWealth investment products and services are provided by Citigroup Global Markets, Inc., member FINRA and SIPC.

6:47You're one of the few CEOs who was also a CEO of the same company, obviously, during the dot-com bubble. We're in the middle of an infrastructure build-out that, according to economic historians, is bigger than any in history. Like, as a percentage of GDP, it's bigger than the railroad boom. Do you have any concerns about overcapacity? Having seen a lot of these waves over the last 42 years, I think every one of them is met with excesses and optimism and skepticism at the same time, right? And that's actually pretty healthy, right? You don't want everybody sort of believing that something is fantastic, right?

7:32And so I think we take a pretty measured view of this and we think about the risks and the challenges and how we balance all that. But certainly, having seen the dot-com period. Which Dell survived, not just intact, but pretty well. Sure. And our plan is to survive this as well and to come out stronger. Are there similar strategies? I mean, Dell famously, you've always kept inventories low. You were building on demand. Are you still doing that? Sure. I mean, we're doing it to the best we can, right? There are some new challenges that this presents, but if you look at our cash conversion cycle, it's still negative.

8:15And we have an incredible focus on a lot of the fundamentals that have built the business to be a durable business over time. And so those are all the things that we think about. Well, let's talk about those challenges. I think on earnings calls, some of the other executives at Dell have said, you know, our number one challenge is DRAM, DRAM, DRAM. In other words, memory, memory chips. Right. And of course, this is a problem that everybody has. Like if you're trying to buy a laptop, this is a problem. If you're trying to buy a PlayStation, this is a problem. If you're trying to buy phones, the price of those cars, cars, right?

8:52Because they need memory. Do you see that bottleneck easing at all? Well, I think it's going to get worse in 2027 based on what I see right now. The memory crunch. Yes. And the reason is that it takes about four years to build a new memory factory. There are a number of them under construction. And those factories will start producing in 2029. And who knows where things will be then. And so maybe 2029, everything will be fine. Right. But what I see is the way people are using this technology now is very different than it was even a year or two ago. With AI, you mean? With agents. Yes, exactly.

9:32And so the demand is growing faster than supply. It takes four years to build these factories. So I don't think that the shortage is going to quickly resolve itself. Now, maybe conditions change somehow and that's not the case, but that's kind of what we see. There's been a lot of talk about, you know, in a future in which AI takes a lot of jobs, we should have so-called universal basic income. This is something Sam Altman at OpenAI has talked about. Other folks have said, no, we should have universal basic capital where people have ownership of the companies and therefore ultimately the infrastructure that is creating all of this value.

10:13You are making record revenue, record profits, building this AI future. Do you ever think of it in those terms where it's like, we're going to give people ownership of this thing that could be disruptive and that is kind of a bridge to a future in which we maybe have fewer jobs or our economy is structured differently? This technology is very powerful. And with that, there will be some incredible things that happen, right? And I think it will accelerate sort of human progress and human flourishing. At the same time, there's probably a period of time that's pretty challenging to get through as people, you know, reskill and retool and reimagine, you know, how things work and all of that.

11:02And we have some things that we're doing to address that, both on the philanthropic side and inside our company. You know, we think about that a lot. It's a big cycle, but we've seen cycles in the past. Yeah. And we've got through them. And, you know, if we didn't improve our processes and our systems and invent new things, we would go out of business, right? And that's it for Tech News Briefing. If you're a listener on Spotify, be sure to leave us a comment. Today's show was produced by me with supervising producer Katie Ferguson. I'm Danny Lewis for The Wall Street Journal. We'll be back tomorrow with a new episode.

11:38Thanks for listening.

11:45Weren't agents supposed to take things off your plate? Building another to handle the others is a bit much. Don't make more work out of work. Gemini is one agent that knows all parts of your business and actually gets stuff done. You give it a project, close your laptop, and when you come back, the job's finished. No matter your industry or role, your agent has the skills to help every team stay in their flow. Seriously. Make work less work with Gemini on Google Cloud. Check response. Subscription required.