Nvidia, ASML Drive Global Tech Rout, Lyft Buys Into Europe
46:23 Share

Possibility surrounds us in digital innovation, evolving markets and disruptive ideas. And while promises can inspire dreams, proof is the catalyst for transformation.

Developers like you are building the future, but you need the right tools to move fast and go further, right? That's where Microsoft comes in. With Microsoft, you can do it.

With tools like GitHub Copilot, VS Code, and Azure AI Foundry, you have everything you need to push the limits and bring your ideas to life faster. And with security, compliance, and responsible AI built in, you can focus on what matters most, building the next big thing. Learn more at developer.microsoft.com slash AI.

From the heart of where innovation, money, and power collide. In Silicon Valley and beyond. This is Bloomberg Technology with Caroline Hyde and Ed Ludlow. ♪♪

Live from San Francisco, this is Bloomberg Technology. Coming up, NVIDIA drives a global route in chip stocks after new US government restrictions on tech exports to China. Plus, Lyft buys its way into the European market with an eye on rolling out autonomous driving. And we speak to the president's chief technology advisor and director of the White House Office of Science and Technology Policy, Michael Kratios.

Again, it is the chip sector, semiconductors, where our focus is. New export controls and license requirements on two big names for the export of specific technologies designed for the Chinese market. This is the picture at the index level. We're down 3.5%. Clearly some pain across the industry that's spreading. The key names are this, NVIDIA and AMD.

both of them taking write downs in the first quarter. We'll get to the numbers very soon. Elsewhere, ASML is an earning story out of Europe. It missed its bookings by almost a billion dollars. And it's saying, we don't know how to quantify the impact of tariffs on our business. Their business, they are the biggest and most important maker of the machines that make the chips, chip making equipment. Again, we will go deep on that story in the show. Let's get to our top story. Nvidia and AMD have been hit hard.

by new US restrictions on semiconductor exports to China. Bloomberg technology editor Michael Shepherd joins us from DC. Give us the details, the license requirements, and I guess, Michael, what happens next for these two names?

Well, what happened just now is the write-downs that the companies had to report. First of all, they are not small. Nvidia reported a $5.5 billion charge for the first quarter. AMD, $800 million. And all from these new restrictions that they are now facing on the chips that they have designed specifically for the Chinese market. And these were designed to comply with previous rounds of U.S. restrictions.

on sales of advanced semiconductors to China. The goal for the administration and for the prior administration too has been to try to keep those sophisticated semiconductors out of Chinese hands, where they can help Beijing gain an edge militarily and give Chinese companies perhaps

a way to gain ground on American hyperscalers and American advantage in artificial intelligence. Now, the catch with these chips is that while they have been scaled down and are lower power for the Chinese market, Ed, they are also useful in inference. And this is an emerging area in AI where

It is helping companies develop models faster and more effectively, something that we saw with DeepSeek. And that was one of the wake-up calls for the Trump administration as it took office, as I'm sure you'll recall. And we heard Commerce Secretary Howard Lutnick articulate this and make clear that restrictions like these would be coming. And now we're seeing them today. And the companies, of course, are feeling the pain. And this is compounded, of course, by all the concerns about tariffs and elsewhere, Ed.

Bloomberg's Mike Sheppard in Washington, DC. Thank you very much. The other big downward pressure on the semiconductor space is ASML, the world's biggest chip equipment maker, posting orders that missed estimates in a very big way. It has dimmed the outlook for the semiconductor sector overall, particularly with regards to anxiety about tariffs. Bloomberg's executive editor, Peter Elstrom, is in London, and it's a $1 billion miss, basically, on bookings. But what accounted for the miss, and what did ASML say about

about the impact of tariffs principally and efforts by the United States for ASML to bring some manufacturing to this country, Peter?

Yeah, it's a big miss by ASML. And ASML, of course, makes the machines that produce chips for the likes of TSMC and Intel and Samsung. So they're kind of a bellwether for chip demand in particular. Now, the CEO, Christophe Fouquet, was pressed pretty hard on the earnings call that we were just listening to now, asking questions about what's driving this, what are the implications. One key factor is certainly this tariff uncertainty that we're seeing. It's not exactly...

clear what kind of demand is going to be generated in some of these areas including artificial intelligence and in other areas another key question is whether tariffs are going to be levied on the machines that ASML makes as they're getting shipped into the United States the Biden administration now the Trump administration both want to build up the chip capacity within the United States but that's going to be more difficult if TSMC and Intel and other companies

have to pay more for these machines. At the same time, you're seeing some weakness in some of the key players here. Intel, of course, is struggling quite a bit. Even Samsung Electronics now is having a bit of a difficult time building up its capacity, has slowed things down. So you don't want to add additional costs on top of that. And just to give a sense of the scale, ASML's machines can go for $380 million. If you put a tariff on top of that in the range that the Trump administration is talking about, it could be $95 million in tariff costs.

Now, Fouquet was very clear to say that the customers are going to bear the brunt of those costs. It's not going to be ASML themselves, but even if the customers bear those costs, it may slow them down a bit in terms of buying these machines and then deploying them. They may have to think twice about how quickly they're going to move ahead in building the kind of capacity that the administration actually does want in the United States.

Bloomberg's Peter Elstrom in London. Thank you very much. Denny Fish is a tech investor with skin in the game. His funds at Janus Henderson count NVIDIA and ASML among its top holdings. There's also TSMC, Broadcom, some of the hyperscalers, and Apple. Let's get the investor's take on technology export controls and tariffs. Good morning, Denny Fish. Were you ready for those headlines on NVIDIA and AMD?

No, I wasn't, particularly because there was reporting just a couple days ago that they had come to an agreement, the administration with NVIDIA, that they weren't going to have these restrictions. Obviously, that was just a report, but it definitely took the market by surprise.

I want to show a chart which is the proportion of Nvidia's revenue that comes from China. And clearly the chart tells a clear story. In 2022, China revenues around 26%.

Currently 2025, we're down to 13%. I think many accept that this is headed towards single digit because of the impact of technology export controls. Does that matter to you as an NVIDIA investor, Denny? Do you sort of worry about that and lay awake at night thinking about NVIDIA's business in China?

Well, two parts to that question. One is whether it matters to me as an investor, and two, do I lay awake thinking about China? Indeed. I think the important thing is it's actually good to just get it out of the model, okay? So, I mean, we're kind of wiping it clean. We're taking a clean cut to estimates. And to your point, the revenue contribution from China has come down materially over the last few years. I think something else that's important is that, you

You know, there's something that's called the diffusion rule that is supposed to come into effect in the month of May. And effectively what that is, that just puts a quota on the number of GPUs that certain countries can actually buy. Now, one might think that potentially the diffusion rule could be relaxed somewhat. That could benefit NVIDIA. The other thing that's just kind of puzzling about all of this is that

In some ways this makes no sense at all for the administration to do and why is that? Huawei already produces a chip that's somewhat comparable to the h20 that is actually the customized chip that Nvidia has been selling in to China and as a result They're just giving the market to Huawei and so they're penalizing a US company in

by restricting a chip that's not even that competitive relative to a locally supplied chip. So, we've already thought about this pretty extensively. We didn't know if we would get a full cut

But I think, you know, the direction of travel was clear. Contribution from China was going to go way down. And, you know, because we're still supply constrained, NVIDIA has the benefit of being able to redirect wafers to other areas where they can actually sell more powerful chips and chips with higher gross margins. The goal of this administration appears to be...

to shift more to the United States, be it localised fabrication of the core chip, assembly, whatever it may be. And so the companies will face the higher cost of doing business in this country, or they will face 25%, 35% net tariff somewhere else. What are the pros and cons of just going with what this administration was and saying, "Okay, we're going to do this in America"?

Well, the pros are that to the extent that we actually have the capacity to manufacture leading edge chips, that's a good thing. And it's a good thing because those are the types of chips that are used in our most important industries for national security and for maintaining our competitive advantage globally as we race towards artificial general intelligence. The cons are it is more expensive.

But nonetheless, you would think that the productivity gains that we get in aggregate from deploying AI probably more than offsets what the aggregate costs would be for manufacturing the chips in the United States. And I just say, you know, thank goodness TSMC put a shovel in the ground

you know, several years ago in Arizona that they're actually in a position now for companies like NVIDIA and AMD to actually announce that they're going to be producing leading edge two nanometer chips with TSMC in Arizona as well as configuring full systems in the United States from a manufacturing standpoint.

a story we put heavy emphasis on this week. So, Nvidia is your top holding across a couple of funds, right? TSMC is always in there. You said, "Thank goodness that's the case," basically. But how does your thinking now change, right? Do you stick with this idea that capital expenditures will continue to grow in '25, '26, '27, the hyperscalers will continue to take from Nvidia, or are you now going to adjust where you put your emphasis, Denny?

Well, we take the data as it comes, but most importantly, I think if you just look at the most recent comments out of Nvidia, where they just had their GTC conference, if you actually parse through the comments from ASML today and you parse through the comments that we've heard from hyperscaler CEOs, we don't see signs of

a slowdown in AI CapEx. And so now the growth may slow, but nonetheless, the signs still support healthy growth in AI CapEx. But like I said, we'll take the data as it comes, but ASML was still optimistic about AI and it's really the lagging edge stuff, analog chips and so forth that go into industrial and auto and things like that, that still continue to be soft.

And I will remind you, you know, I know this is different with the tariffs and the geopolitical situation, but just rewind the clock one year and ASML, same thing happened. They missed bookings. The stock was down and then all they did because the bookings are really lumpy and then bookings just reaccelerated throughout the year. And they're doing some really cool stuff. And, you know, one of the things they talked about is, you know, one of the reasons that, you know, they can continue to drive higher pricing for their EUV and high in a.

EUV is because they're starting to do what's called single layer and what that effectively does is it increases throughput, it's more valuable and therefore they can charge a higher price. Better technology, charge a higher price. Denny Fisher, Janice Henderson, it's great to have you back on the program. Thank you very much. Now coming up here on Bloomberg Technology, Lyft is expanding outside of the US and Canada. We have CEO David Risher on the company's European acquisition. That's next. This is Bloomberg Technology.

Some deal news. Lyft has agreed to buy the European multi-mobility and taxi hailing app, FreeNow, for about $197 million. It marks its first global expansion beyond the US and Canada. Lyft CEO David Risher joins us now. I've been using Uber, your main competitor, in Europe for a really long time. I've been going in taxis in Europe for a really long time. Why acquire FreeNow and not just launch the Lyft app in that jurisdiction?

because free now is a great company that's got great relationships uh all across europe in the really important taxi industry i mean look so lyft as you know ed you've been following us for a while we're in a stronger position than we've ever been we're picking riders up faster than we've ever picked them up before drivers are making billions on the platform in the us and canada and so now's the right time for us to go global and what's great about it is it expands our market doubles our market and it allows us to plug into as you say the taxi world

which is a huge, huge part of the European mobility scene. Tomorrow, I think I'm right in saying, is your two-year anniversary at Lyft. Good memory, you're right. And you have fixed the finances of that company to an extent, but the data is so key, right? Uber has what? One million drivers in Europe, and I think FreeNow has about 150,000 registered taxi drivers, nine cities.

How do you bridge the gap? Are you confident you can bridge the gap? Yeah, I am. So, FreeNow is a great company. And one of the things that makes them so strong is they really understand the local dynamics market by market by market. Here's the thing. The mobility market in Europe for this kind of ride-hailing is about 40 billion euros.

Half of that are offline taxis. These are literally still people calling up a taxi company and saying, "Come take me to the airport. Come take me. I'm going on a business trip," whatever it is. So while it's true there are other companies there, FreeNow is the leader in the taxi space. And remember, as you know, taxi is kind of an elevated experience to a lot of Europe. Think of the London Black Cab. So we think the company is really well positioned to really sort of continue to grow and take over the much bigger space.

But it is a company that's only recently broken even. It's had historic losses. What are you modeling for in terms of what it adds to your top line and bottom line and how immediately? Yeah, so it's a profitable company. It has about a billion dollars in bookings, which is significant. Now, Lyft, of course, has 16 billion in bookings. So, you know, there's a lot of growth ahead of both of us.

But it'll be financially creative for us, which is wonderful. And I think as we, you know, they've been owned by Mercedes and BMW, two great sort of storied auto manufacturers. But I think combining with us is going to give them some new energy, some new expertise to really allow both of us to grow. There is zero mention in the release about autonomous driving. And every sell side note about this deal says it's your plan to roll out autonomous driving in Europe with those two OEMs.

that you just mentioned? Yeah, so here's how we think about autonomous. As we've said, I mean, first of all, of course, it's happening in the US. We're actually a little bit further ahead. Europe, a little bit not so much. One of the great things FreeNow has is, well, two, they've got great relationships with European regulators and they've got great fleet management.

And as you and I have discussed before, fleet management is so important to autonomous. So yeah, we think it sets this up really well. Let's be clear, this is going to be years down the road, but it's a really important part of the overall strategy for being a global player. Two years in, you've done some M&A. Will you do more M&A and where?

I mean, never say never, but this really is our focus right now, right? I mean, the good news is, you know, again, Freenow is a great, strong company and we've got so much to build on. So we're going to be really focused on our European expansion, obviously making sure it's a big success. And then who knows? We'll talk later about something else. Engine Capital, activist investor, they want to change your board. Your reaction, please. I mean, we interact with shareholders all the time and, you know, listen to their feedback, take it seriously.

There's a whole process for dealing with activist investors that we're involved in, but broadly speaking, if you look at what shareholders want, you know, capital return, we're giving out $500 million in a share buyback. Growth, we're growing, we're now expanding, we're strengthening our business by going international. I think we're really well positioned to address all of the issues that shareholders bring. Sorry to interrupt, David. Have you spoken to them about the FreeNow deal since it happened? No.

No. Okay, that's kind of interesting. Final questions on technology. Atlanta, technology partnerships, May Mobility. Operationally, give me an update, please. Yeah, so you're referring to the partnership we have with May Mobility, an autonomous vehicle group or company. That's on track for this summer. So super excited to launch right in the middle of downtown Atlanta. On track for this summer being commercially on track or this is a kind of toe in the water? Well, it's on track for us to start launching

operationally this summer in Atlanta and then you know it'll grow from there but first we got to get you know riders in the in the cars and that'll be sometime this summer because I have the time to do so two years in what would you score yourself

or grade yourself on the job you've done as the Lyft CEO? Look, I am a self-critical person. I am. I guess maybe one of the things that I do is always look for the next thing. But I do have to say I'm proud of what we've accomplished as a team over two years. You know, getting to profitability, $700 million plus in free cash, and maybe most importantly, picking up riders faster than ever and drivers making more on the platform than ever. So if I look at it from the customer's point of view, we're doing super well. It's up to other people to tell me how I'm doing as a CEO.

And the street gives you a lot of credit for that. M&A, maybe they didn't see coming so much. Maybe we'll see some more. Lyft CEO David Risher, thank you for joining us here on Bloomberg Technology. Now, coming up, Americans are still using Chinese e-commerce apps to shop cheap alternatives to some high-end brands. We're going to drive into the surge in sales for Timu and Xiyin. That's next. This is Bloomberg Technology. Bloomberg Technology

♪

A trade war won't stop Americans from shopping on Chinese apps. Revenues surge for both XIIN and TIMU this month as US shoppers anticipate price increases due to tariffs. Bloomberg's Liddy Meyer joins us and covers the retail beat. This is really interesting data. A lot of people in my world use both of those apps. I don't. What are we learning? Yeah, it seems like these apps are seeing a lot of success and people are going to them as tariffs are starting to loom.

The eco data kind of comes in here, right? When you think about tariffs, we think about passing on the costs to the consumer. Through your beat, through the tech companies that are in the e-commerce space, what are we learning about the consumer and that March data? Yeah, absolutely. So it seems like the consumer is looking to spend right now. The eco data showed a lot of spending and it seems like people might be stocking up or buying now before prices might get more expensive.

Lily, you cover retail, specialty retail in particular. What are you learning about tariffs? How do companies really think about what happens next? Do they even know?

Yeah, so a lot of companies are doing different things. A lot of the companies I cover report in the next month. So that will be really interesting to hear, you know, what they're looking to say to investors and shareholders. But it seems like a lot of people are in kind of a wait and see mode. They're not exactly sure how tariffs are going to impact their business. And they're trying to figure that out. Bloomberg's Lily Mai. Great to have you on here on Bloomberg Technology. Thank you very much. Now coming up,

Mark Zuckerberg's email reveals his Instagram and WhatsApp strategies from the past. We're going to go back to that antitrust case next. Let's get to markets as well. It is the semiconductor space that's dragging down technology shares broadly, but the Nasdaq 100 is under pressure, generally speaking, and by association, some of the other key names in our world, the hyperscalers,

consumer tech under pressure. Bitcoin is going in the opposite direction. $85,000 per token, up a percent or so in the session, but it's not behaved uniformly like as risk on, risk off. It doesn't sort of take into account the same factors tech does. We'll dig into that later in the week. Be right back. This is Bloomberg Technology. Bloomberg Technology

♪

Welcome back to Bloomberg Technology. I'm Ed Ludlow in San Francisco. The market's picture is very clear. We are under pressure, and that largely comes from the chip sector. New U.S. restrictions on technology imports to China, specifically for NVIDIA and AMD, that is causing pain. But by association, you see a lot of the MAG7 mega cap names down. Earnings season is right around the corner. The underperformance of the Philadelphia Semiconductor Index distills that chip pain.

The single names we know about, NVIDIA, AMD, very interesting discussion throughout the show on how long that pain will last. But ASML is an earning story. It missed bookings by $1 billion. Later in the program, Pierre Ferraghiu, New Street Research, is going to come on and give us deeper and more granular context about this name. There is other news. Elsewhere, in his second day on the FTC's antitrust trial, Mark Zuckerberg revealed that he once considered spinning off Instagram back then.

back in 2018. In a letter to senior leaders, he writes, quote, "I'm beginning to wonder whether spinning out Instagram is the only structure that will accomplish a number of important goals. We should keep in mind that there's a real chance that all our work to build a family of apps may be something we don't get to keep." For more, Bloomberg's Kurt Wagner, who has been in the courtroom at that trial, joins us.

Let's go back over it and explain it. At one time in 2018, of his own volition, Mark Zuckerberg considered spinning off Instagram. What does that mean in the context of this trial?

Well, it's very prophetic, right? Obviously, he was sort of foreseeing the issue that he's facing right now coming down the road for him. That's, you know, six, seven years ago. I don't think the fact that he thought about doing this really, you know, hurts him so much today. I think it was shared in the context of the FTC trying to show that that Instagram has perhaps.

armed users, right? That's part of their play here is that they want to show that buying Instagram, buying WhatsApp was actually bad for consumers. And if they can show that, you know, Mark Zuckerberg even thought, hey, maybe we shouldn't be doing this seven years ago. Maybe that's a sign that they weren't, you know, employing the best strategy for consumers. But it is a very interesting thing in part because it was so foreshadowing of where he is today.

This is a multi-week, potentially multi-month process. What else does the audience need to know about the trial and proceedings? Where has Meta focused and where have the FTC focused?

It really hinges on this definition of what the FTC is considering the friends and family sharing social networking market, right? So they're saying, hey, look, there's a lot of places you can spend time online. There's very few places where you actually go specifically to share with your friends and family. And that is where Meta has their monopoly. This is the FTC's case. Meta, on the other hand, is arguing, hey, wait a minute, we have tons of competitors, right? TikTok is a competitor. Well,

Apple's iMessage is a competitor. Obviously, Elon Musk's X. And none of those, by the very narrow definition that the FTC is going with, are considered competitors. And so this is basically a fight to determine what the market should be. And if the FTC succeeds in keeping it very, very narrow, they may have a good case. But as Facebook has pointed out repeatedly, they consider their competition very wide.

Bloomberg's Kurt Wagner, who I am sure we will come back to multiple times over the course of proceedings. Thank you very much. Let's break down more on what the FTC's trial against Meta could mean for innovation, regulation, the future of digital markets. Joining us now is Jennifer Huddleston, Technology Policy Senior Fellow at the Cato Institute. And as a piece of documentary evidence in a trial like this, the CEO at the center of it

in a letter six years ago, seven years ago, saying maybe we should spin off Instagram. How does that play in this environment? What is this environment of antitrust that this trial represents?

We've seen a lot of animosity towards some of America's leading tech companies and this presumption that big is automatically bad. One of the key elements of the FTC's case is it's basically imagining a world that never existed. It's trying to ask for a do-over on an acquisition it approved more than a decade ago.

the same time the social media dynamics have been changing. They were changing in 2018 when that letter came about. They were also changing more recently with the emergence of things like TikTok. We have to remember that at the time Meta acquired Instagram and WhatsApp, these were not surefire successes. In fact, there were jokes particularly about the

of Instagram at the time, kind of why are they doing this? Why are they paying all this money? We don't want to see a world where government enforcers penalize businesses for taking risky chances that turn out to be good bets and turn out to allow more innovation and more benefits for consumers in the market. This is a trial and it is contested between the FTC and Meta. But we are, you know, a few months into this new administration in the White House.

Prior to the election and after the inauguration of President Trump, many put emphasis on Vice President J.D. Vance's views on big tech and antitrust. How does that context play here in the trial? It seems as if this administration will follow through with the general attitude towards big tech that the prior administration did.

If you look at the history of many of these cases, many of them began in the first Trump administration. Then we saw further gasoline on the fire during the Biden administration, particularly under Chair Lena Kahn at the FTC, and this real push towards this big is bad mentality.

It seems like so far, both with the remedies proposed in the Google cases as well as with what we're seeing in the Meta case, that this administration seems determined to continue that shift away from the kind of law and economics, consumer welfare-based approach.

approach to antitrust to something that makes more presumptions about competitors than consumers and something that perhaps is using antitrust enforcement for other policy goals. And that should be concerning because if we lose focus on the consumers whom antitrust was designed to protect, we risk losing the purpose behind antitrust and making this powerful tool more subjective. Jennifer, why would a breakup of Meta and its family of apps be bad for the consumer?

I think that's what's really at the heart that we need to remember when it comes to this case. You'd lose certain cross-functionality that consumers like, but you also would have companies with fewer resources to engage in research and development technology, to engage in rolling out new features that consumers want in the time that AI is disrupting things, but also fewer resources to do some of the things that consumers have expected.

it around things like young people's online safety. And finally, you might have companies that have to rely more on data and advertisers because they no longer can share those resources. And so they would have to actually shift more towards using their consumer data and more on doing what the advertisers want rather than what their consumers want.

Jennifer, the core of Meta's argument or defense is that the FTC has this narrow definition, but they're misunderstanding the marketplace, that Meta's competition is TikTok and that it is YouTube. How successful do you think that defense will be?

The FTC is seeking a really narrow definition of personal friends and family social media networks, which is a definition I don't think the average consumer thinks about. And it's something that if it's successful could potentially lead to larger disruption in the social media marketplace.

Not only does this definition exclude companies like TikTok that are incredibly popular with Gen Z, it also excludes other messaging apps like Signal. It excludes platforms like Snap that have grown in popularity. So it doesn't seem to reflect the experience of the average consumer when it comes to social media and messaging, let alone when it comes to broader conversations around things like advertising or entertainment more generally. In your research and your review of policy,

pending the outcome of this trial, do you expect actions against other big tech names? And if so, who? We have already seen cases against four of America's leading tech companies. Google, which is now in a remedies phase in one of its case. There are also pending cases against Amazon and Apple. And of course, this case against Facebook and Meta. All of these cases really raise concerns about the potential government intervention into what has been a very dynamic ecosystem.

Not only could these cases have consequences for consumers, but there's going to be a question as well of what are we missing out on while companies spend their time dealing with enforcement actions instead of innovating and trying to provide better things for consumers. What sort of message does this send to the startup ecosystem? Jennifer Huddleston of the Cato Institute, thank you very much. We'll be right back. This is Bloomberg Technology. ♪

this is bloomberg technology and you're looking at a live shot of the principal room check out the bloomberg technology podcast you can find it on the terminal as well as online on apple spotify and iheart this is bloomberg

the united states will have to make a creative r d push while making quote smart choices with budgets to stay ahead in the ai race those were the first public remarks by michael cracios advisor to the president and white house office of science and tech policy director since his confirmation in march the main message on strategy for tech

promote and protect. Mr. Kratios joins us now. Specifically, what do you mean by smart choices? This is a time where budgets across federal government are being cut. Some tools and examples please, Mr. Kratios.

Thank you so much for having me, Ed. I think if you zoom out a little bit, the message that we were trying to share in Austin a few days ago was this question of how does the US ensure its position as the global leader in critical and emerging technologies? And in order to do that, you have to do two major things. You have to do promotion and protection. And the key pillar of the promote agenda is around being more creative in the way that we spend fairly funded research and development dollars.

For far too long, these R&D dollars have been spent in the same ways by the same agencies, going to the same types of research. And there's a lot more that we can do in all sorts of things, like prizes and challenges, advanced market commitments, or even fast-track action grants, where we can actually worry a little bit less about how much and more about how we actually deploy these dollars in ways that we can ensure we're making the best discoveries here in the United States for years to come.

We've reported a lot on DOGE, that's the context right, the work to cut waste in government, but very specifically Mr. Kratios, do you call for support and increase in R&D budgets, growth in R&D budgets?

Again, the most important thing we have to do is make sure that we are prioritizing our R&D budgets. Spending a lot of money on the wrong things is far worse than spending less money on the right things. So what we do at the White House in coordination with the Office of Management and Budget is help set research and development priorities for the country.

The President spelled these out in a letter that he sent to me two weeks ago, and the ones that he did call out were artificial intelligence, quantum computing, and nuclear, to name three. And you can imagine that in future budget cycles, these are the areas that we're going to be prioritizing because those are key to our national economic security.

Are those priorities achievable, AI leadership from America with the tariffs that we have in place and the technology export controls, some of which were amplified overnight on the likes of Nvidia and AMD?

Absolutely. If you think back to the strategy that we discussed a couple days ago, in order to achieve this leadership position, we have to do both promotion and protection. On the promote side, we talked a little about the R&D pillar. The second piece of that is around deregulation. For far too long, especially in the last administration, a lot of our regulatory policy has been driven by the spirit of fear, this idea that we should be worried about these employees.

these emerging technologies. What the President and the White House has said many times is we have to make sure that we lead in this technology and actually deploy it.

And then on the protect side of the equation, we have to make sure that we're not giving our adversaries the critical tools that could help them try to catch up to us in this race. We know that the PRC is attempting to catch up with us in things like artificial intelligence, and there's no reason why we should be helping them do so. So that's why in order to kind of achieve this U.S. position of

leadership in artificial intelligence and other emerging technologies, you have to have a balanced promotion and protection agenda.

NVIDIA and AMD now face the situation where they have to apply for licenses to be able to export H20 and MI308 to China. And the administration's argument was they could end up in supercomputers. The company's argument is these are much lower performance. And one idea is that if they haven't got access to American technology, China will simply champion a domestic name. Huawei has a similar accelerator. How will you advise the president on whether or not

AMD and Nvidia should be granted those licenses to export lower performance accelerators to China? Yeah, I'm not going to opine on the specifics there, but I will say that there has been a track record of a wide variety of chips being used to drive the development and training of what are almost frontier level open source models by companies within the PRC. And as we think about our

protect agenda, it's critical that we develop an export control regime that is simple and clear and is one that we enforce very, very strongly.

One of the elements of PROMOTE is to bring manufacturing of lead-edge chips and assembling of compute systems to America, right Mr. Krasios? And ASML is in the spotlight as an example because of its earnings report overnight. Would this administration consider an exemption of tariffs on the leading chip equipment maker in order that they might be able to do that, bring more of their activity to the United States?

Again, I can't opine on specifics, but I think what we saw from the news last week on NVIDIA, there is a continual effort, and we see that from all around the world, of people wanting to bring manufacturing and these core sort of bleeding-edge technologies back here to the United States. And that's something that the White House is very proud of and something that we have seen across industry. The policies that the president and our trade team have implemented –

is one that is actually proving true. In just a few short days, we've had so many great announcements about companies coming here to build and manufacture these critical technologies in the US. And I suspect that we'll be seeing a lot more of those in the months and weeks ahead.

There is some confrontation between the administration and leading academic and research institutions. Harvard is an example. How does the potential of cutting budgets at institutions like that impact America's competitiveness in the field of research? I reflect on Jensen Wang's comments

I believe in March at GTC that, well, more than 50% of the AI research is either being done in China or by Chinese nationals. And yet here in America, we have some debate about funding for these institutions.

Yeah, I want to make sure we don't necessarily conflate two separate issues. To me, I don't think this is particularly hard. There is no reason why we should be supporting anti-Semitism on college campuses or tolerate that. I think we can be intolerant of anti-Semitism on university campuses and also support the academic research enterprise. Those two things can happen simultaneously.

And I think first and foremost, these universities think very carefully about Title VI, and I leave that to the DOJ and to DOE, Department of Education, to think about. On my end, and what I've advocated very strongly, is we continue to need great universities to work on basic research that has funded and fueled our innovation ecosystem for decades. And that's something the White House will continue to advocate for, while at the same time, we can never stand for anti-Semitism on our university campuses.

Mr. Kratios, as part of your office's request for information, several AI companies submitted responses to you asking basically to preserve their ability to learn from copyrighted material. We're talking about those training frontier models, foundation models. Very simple yes or no, if you will. Is copyright reform on the table as part of this president's AI action plan? And how would you advise him on that, please?

Yeah, we were so delighted with how many responses we got to the AI Action Plan RFI. I think in the coming days we'll hopefully be able to release all those responses. It was many, many thousands and it just showed how far and wide the community is that wants to think about and help shape the national agenda for artificial intelligence. And obviously copyright is going to be one of the issues that many people commented on over the last few months. Mr. Kratios, copyright reform, yes or no?

Again, I think we could not be more thrilled with how many people commented on all sorts and all facets of the AI agenda. And this is one that we had a lot of commenters send us some very interesting and detailed thoughts on.

The president has spoken with admiration about Jensen Huang and NVIDIA's leadership. Just as our audience, Bloomberg Technology, gets to know you, Mr. Kratios, and your work in the White House, have you had the opportunity to meet with Jensen Huang and his peers about what they plan to do in America and also critical markets for them, which once included China? I have, yeah. In the first two weeks after my confirmation, I had the opportunity to meet with most of the leading technology companies

industry CEOs and chat with them about how we can together usher in this golden age of American innovation that the president has called for. With Jensen specifically, he's a tremendous thinker. And one of the great things that we connected on is the question of AI for science. There's been this almost this extreme fixation on thinking about the protect side of the agenda when it comes to specific semiconductors.

But I think another equally more and maybe even more important part of the problems we face ahead of us, or the challenges, is how can we actually implement this technology to drive transformational change for the American people? And Jensen and NVIDIA,

He was extraordinarily excited to help us think through and see what we can do more on driving scientific discoveries through AI. And it's something that I will be encouraging Secretary Wright with our labs and Ponch at National Science Foundation, so many of our leaders to take more seriously because we have to do better.

Michael Kratios, White House Office of Science and Tech Policy Director. Thank you. Look forward to speaking more over the coming years. Okay, I mentioned ASML shares trading significantly lower. The company posting a sales miss that was almost a billion euros less than expected. Pierre Ferreger of Newstreet Research, head of global tech infrastructure, is with us. There are people out there, Pierre, and it's good to see you again, that are a bit more sanguine, right? Orders can be lumpy. But the thing that's spooky to many is...

the inability to quantify tariff impact. How are you modeling for that in ASML's case? Yeah, I think you're making a good point. The way I model it is by actually modeling orders coming down in the near term because while you're not too certain about tariffs,

In some industries, you could see people like front loading orders before time gets into place. It doesn't work very well with these very large pieces of equipment. And then the second approach is to actually wait and see and be more careful. In the long run, modeling like tariffs requires you to take a view on what the tariffs are here for, what the US administration want to do with these tariffs. And so if you listen to them,

You know, they want to encourage changes in behavior with trade partners. They want to bring back to the US or to reliable partners manufacturing, and they want to create additional

And if you put that into motion around semi-cap equipments, what you see is that none of these objectives are going to go against the ability for the U.S. to build actually the AI infrastructure and to continue to let, you know, the manufacturing of leading-edge semiconductors like memory chips and so on.

Logic chips, as they built up in the US, you still need to continue to have technology advancing everywhere in the world. So I'm not too worried about the impact of tariffs on semi-cap equipments. You heard Mr. Kratios decline to answer my question on whether there would be an exemption for chip equipment makers. But Mr. Fouquet was very clear, we will assemble in the Netherlands. Does the administration have a say on what ASML does in this case, Pierre?

They definitely have a say. I think if they give a phone call, the phone call will be taken. The specifics of ASML tools is that it's very unlikely to see the final assembly of these tools moving away from the Netherlands the same way it's very unlikely to see the most leading-edge manufacturing node of TSMC moving away from Taiwan. It's probably a hard limit.

that cannot be surpassed. Now, what Monsieur Fouquet said as well is that increasing the amount of manufacturing done in the US, which is already very significant, very near to where I live, you already have very critical stages that are being manufactured, like the light source, which is the most critical piece of the EUV tool is manufactured in the US. And seeing that going up is a possibility, like changing the final assembly of these tools

is, yes, probably a stretch and practically, in practical terms, not doable. Right. Pierre Ferrigu of New Street Research. Good to catch up, have you back here on Bloomberg Technology. Well, that does it for this edition of Bloomberg Technology. A huge thanks to those of you that submitted questions for some of our key interviews throughout the hour. There is a lot to recap. Don't forget to check out the podcast. You know where to find it, on the Bloomberg Terminal as well as online on Apple, Spotify and iHeart. What a week. From San Francisco.

This is Bloomberg Technology.

Developers like you are building the future, but you need the right tools to move fast and go further, right? That's where Microsoft comes in. With tools like GitHub Copilot, VS Code, and Azure AI Foundry, you have everything you need to push the limits and bring your ideas to life faster. And with security, compliance, and responsible AI built in, you can focus on what matters most, building the next big thing. Learn more at developer.microsoft.com.

You're the owner of a small business, which means you're also the tech guy and HR and personal assistant and head honcho and intern. You could use another pair of hands like the experts you'll find at Verizon Small Business Days, April 21st through 27th. Get a free tech check, special deals and more. Call 1-800-483-4428 or visit verizon.com slash small business to book your appointment. Verizon Business.