Here's Why AI Is Creating The Biggest Monopolies in History
09:00 Share

Deep domain expertise, strong relationships, broad capabilities. It's what makes Stifel one of the industry's leading providers of M&A and capital-raising services in the middle market. But don't just take anyone's word for it. IFR has named Stifel U.S. Mid-Market Equity House of the Year five times in the last 10 years. When it comes to investment banking, Stifel is the name you should know.

To learn more about how Stiefel can help you address your most complex investment banking needs, visit StiefelInstitutional.com. Stiefel, Nicholas & Company, Inc., member SIPC and New York Stock Exchange. This is an ad by BetterHelp. Think about your mentors and idols, the people who inspire you most. While it may look like they have all the answers, they don't. But they do know when to ask questions and seek the support they need. In

Thank you.

It's convenient, too. You can join a session with a click of a button, helping you fit therapy into your busy life. And you deserve that. You can even send your therapist a message any time something comes up. Build your support system with BetterHelp. Visit BetterHelp.com slash podbusiness today to get 10% off your first month. That's BetterHelp, H-E-L-P dot com slash podbusiness.

Switch to Verizon Business and get more from your internet without paying more for your internet. Get LTE Business Internet starting at $39 a month when paired with a Business Unlimited smartphone plan. That's unlimited data, and with it,

unlimited possibilities. Start saving today with Verizon Business. Ranked number one in small business internet customer satisfaction by J.D. Power. Starting price for 25 megabits per second internet plan with savings, plus taxes, fees, and economic adjustment charge. Terms apply. For J.D. Power 2024 award information, visit jdpower.com slash awards. Bloomberg Audio Studios. Podcasts. Radio. News.

I'm Stephen Carroll and this is Here's Why, where we take one news story and explain it in just a few minutes with our experts here at Bloomberg. ChatGPT, fastest app in the world to 100 million users and today well over 300 million users. We're at the very beginning of a 20-year cycle where productivity will really increase using these technologies. It'll change everything. It'll be integrated into every business.

ChatGPT has been a game changer in artificial intelligence. The sudden rush into large language models has propelled a handful of companies into market superstars. So it is NVIDIA Day. Happy NVIDIA Day to all who celebrate. TSMC revenue climbing 39% in the first two months of the year. Korean shipmaker SK Hynix delivering record quarterly profit.

Previous revolutions in technology, like with personal computers or search engines, have created industry giants. This time around, the scale is even bigger. Here's why AI is creating the biggest monopolies in history.

Our technology reporter, Mark Bergen, joins me now for more. Mark, who are, first of all, these megastars of AI, and what do they do? MARK BERGEN The household name now is NVIDIA, which has gone from this sort of, in the past few years, a niche chips company, mostly for the gaming industry, into one of the world's largest multi-trillion dollar companies sort of associated with this AI boom. And they are, of course, the primary company designing what are called AI accelerators. They're basically the chips that go into these data centers that make things.

Things like ChatGPT work.

What we found that's really interesting is that behind those companies are actually three of their suppliers and partners that have a somewhat just as fortified market position. So NVIDIA has about 92% of the market for AI accelerators, which is just phenomenal. There's a company called SK Hynix out of Korea that produces memory chips and some of the highest-end memory chips that go into the sort of GPUs that NVIDIA makes. They have, in estimates, maybe 80% of that market that's

There is TSMC, which produces the chips for NVIDIA and a lot of the rest of the world. They have for the AI chips, maybe 95%. Then there's a Dutch company called ASML that makes these very expensive lithography machines that are necessary to build these chips. And they have for those specific types of lithography machines that are making these chips, 100% market share. No other company in the world can do this.

MARK MANDEL: And is that how we can qualify them as monopolies? I mean, that's a big word in the business world when we talk about monopolies. MARK MIRCHANDANI: It is, yeah. And I think it should be very clear-- and we say this in the story-- that there's a difference here between the sort of definition of monopoly around the market share and the legal definition of a monopoly, which may be associated with anti-competitive behavior. And so far, there have been some antitrust cases opened against Nvidia, but no findings about these four companies that they've

demonstrated anti-competitive behavior in getting these positions. So I think maybe they're not necessarily what we call an abusive monopoly. They're just good at what they do, I suppose would be another way to look at it as well. Yeah, I think that's one of several factors behind why we have these positions. They have huge market share, as you say, in these areas. Are there any competitors that we should be watching that could potentially disrupt that? One of the things that's happening in AI in this moment is everything is moving much faster than sort of

the historical timelines for technology sectors, for something like the personal computing and mobile phones and browsers. NVIDIA is certainly the most prominent and had gained the most from the AI boom. In some ways, it's the most vulnerable here, in part because they have this really fascinating dynamic where their biggest customers are now effectively coming to compete with them. Almost half of their revenue comes from the biggest

giants of Silicon Valley. They have Microsoft, Amazon, Google, and Meta. All of those companies have been exploring to various degrees

to NVIDIA, whether it's through partnerships, whether it's through building their own custom chips. In Google's case, Google's had a custom chip for a long time. Certainly developing that. Amazon, as we reported, has done some very aggressive in-house design. These companies don't want to be reliant on a partner like that for something they see as so mission critical. OpenAI, there's been some reporting, has also been designing its own custom chip for

To be clear, we haven't seen this reflected at all in NVIDIA sales or its projections. There are a number of startups that have come around that are also trying to replicate what NVIDIA has done or sort of produce a next generation of chip.

Intel is probably the most interesting example right now. They just got a new CEO who has signaled that they want to get into this, what we call the Foundry game, basically being a factory like TSMC. Samsung has probably been the closest competitor. SK Hynix has a lot of more competition there in the memory chips. You have Samsung, you have Micron. And at this point, it

ASML is probably, of those four, the most entrenched. Unless some sort of black swan event happens, it's very unlikely that we're going to see another company develop that machinery and those supply chain partnerships to compete. Are competition regulators worried about these companies and the, I suppose, dominance they have in these spaces?

I think the short answer is yes and no. We've seen in NVIDIA's case, the Department of Justice last year had opened a potential case. And as far as our reporting shows, the concerns here is that they are kind of bundling their software with some of their hardware products where it basically makes it really difficult for, say, someone to use a different type of chipset because...

because NVIDIA software is so prominent and so widely used across AI. If you're building a chatbot, if you're building a large language model, one of the reasons that NVIDIA has grown so well is because they were out and ahead of having this support software that works really well with these chips. The company, to be clear, has pushed back on that and said it doesn't bundle its software. It faces a lot of competition. There was some scrutiny around an acquisition they made. This

this Israeli company that makes software called Run AI. And I've talked to some people who have, some critics who point out that this is sort of a traditional move that they've seen of other companies building these kind of walled gardens. In the US, certainly the Trump administration seems to be signaled less interest instead of going after monopolies than the Biden administration. We haven't seen an acquisition of Run AI by Nvidia cleared in both the US and the EU, which has traditionally been much more

aggressive about monopolies. So as far as we know, there's certainly not, say, the antitrust concerns about Google or that Microsoft has faced in the past, or even going back further in the hardware realm that companies like IBM have faced.

Mark, you've been reporting on the tech industry for many years. I wonder how you compare this moment and the emergence of these giants in artificial intelligence chips to previous big bang moments in tech. Is this an iPhone moment? Is this a Google moment in search engines? I'd say it's more of a Wintel moment, which is less familiar maybe, but Wintel is that term for the concept

combination of Microsoft and Intel sort of duopoly around personal computers that formed in the 80s and really solidified that, you know, every single desktop computer that shipped had Microsoft software with it and had an Intel chip inside. And that, to be

be clear is that that's still a fairly dominant market position at this point. A lot of the market's attention, a lot of the legal and political attention has shifted, say, to mobile and maybe now shifting to AI. But we're still, you know, you and I are looking at PCs right now. They're still prominent parts of life, and most of them still ship with Microsoft and Intel. I think just because, you know, the chip industry has traditionally had

companies that have tended to dominate particular components or niches. That's not necessarily new, but what is new is the attention that this is receiving, I'd say globally, politically, and certainly in the market, right? The companies like NVIDIA and its partners, the amount of money tied up with them in the market and the amount of money tied up in this thesis that there's just going to be increased, increased spending on

around AI and AI hardware, that is unprecedented. Okay, Mark Bergen, our technology reporter, thank you very much. For more explanations like this from our team of 3,000 journalists and analysts around the world, go to bloomberg.com slash explainers. I'm Stephen Carroll. This is Here's Why. I'll be back next week with more. Thanks for listening. AI is redefining what's possible for your business. With more unique challenges to solve and higher stakes than ever, Microsoft helps you stay ahead.

Our trustworthy AI tools and guidance can empower leaders like you to drive greater impact. And with Azure's simplified platform management, we're helping businesses go further, faster, unlocking up to 150% improved output. Whatever challenge comes next, let Microsoft help you keep pushing forward. For more details, visit microsoft.com slash challengers.

There are presentations.

And then there are Canva presentations. With Canva, you can use AI to take your presentation to the next level. You can generate dynamic slides and text with a simple prompt. You can drag and drop graphics and charts from Canva's media library and add interactive elements to plus up your deck. And with collaboration tools built in, the whole team can work together better. You'll love the presentations you can easily design with Canva. Your clients and coworkers will too. Love your work with Canva presentations at canva.com.