AI、中东、中国、关税、和解法案、Invest America｜BG2 与 Bill Gurley、Brad Gerstner

Can you imagine if, in the year 2000, Washington, D.C., had said, “Oh my God, Google is so powerful that we’re not going to allow any other country in the world to have access to this Google machine that might give them answers to questions unless Google comes to Washington and gets federal approval before it launches in any of these countries.” That’s what we were doing in AI. I think we’ve ripped the chains off of that, and I think now we’re going to allow our companies to go compete and win.

Hey, Bill. It’s great to see you.

It’s been a while.

It has been a while. It’s been, I think, 3 weeks since we did a podcast, honestly. And thanks, as an aside, to hundreds, if not thousands, of listeners who remind us that they love the pod. It is hard to get it scheduled.

Part of the reason it’s been hard is I’ve been on the road. I’ve been in L.A., I’ve been in Washington, D.C., and then I spent 10 days in the Middle East. I think about what’s happened, Bill, over that period of time. We basically landed the plane on tariffs, we’ve had these huge deals announced in the Middle East, the reconciliation bill is on track to passing, and we have some talks of a ceasefire in Ukraine.

The market, which was down 20% for the year, has now bounced 20% in the last 20 trading days. Maybe today we could just unpack this flurry of activity that’s occurred over the last 3 weeks.

That sounds great.

You were at the big event, the one that everyone’s talking about. Why don’t you give the listeners, at first, just a sense of what it felt like? What was different this time? You’ve been there before, but frame what happened. What did this look like? Then I’d be interested in your takeaways.

1. AI Diffusion Rule and Middle East Deals

I have to say, the whole orientation toward the Middle East over the last few years has been about control. We had CFIUS blocking all these deals in the Middle East. We had the Biden diffusion rule. Remember that, in the final weeks of the Biden administration, they passed a rule known as the diffusion rule that really created this complicated regulatory framework for selling advanced chips and models. It broadly restricted the ability of the U.S. to sell chips to about 100 countries around the world.

I think it felt like a massive 180-degree shift from what I would call the Washington approach of control and preventing the diffusion of American technologies to much more of a Silicon Valley approach to partnership and openness. That was the language that I heard.

I was happy to be part of an AI delegation with a lot of CEOs who were over there helping to craft these deals and getting to ride shotgun on some meetings with David Sacks, who is obviously leading the AI initiatives for the administration. I have to say, it was incredibly well received in Riyadh, Doha, and Abu Dhabi. There was a spirit of partnership in every one of those places.

You saw the fruits of that: major announcements of investments in the United States and major deals getting struck about AI data centers and so on that are opening in that part of the world.

That’s amazing. Of all those things, which one surprised you the most?

It’s a really big change. You can’t underscore enough what it means. The Trump administration just repealed the Biden diffusion rule in May of 2025, so it literally is fresh off of that.

If you had told me that, in the first 5 months of this year, we would have gotten a trillion-dollar deal signed with Saudi in terms of their investment in the United States, another trillion dollars with Qatar, and another trillion dollars with the UAE, I wouldn’t have believed it. Those are all investments into the United States.

Let me give you a sense of the scale and magnitude of the deal that was announced in Abu Dhabi. They announced a 5-gigawatt U.S.-UAE AI campus. It’s an incredible architectural campus in collaboration with NVIDIA, OpenAI, and Oracle. Think of it as almost like a global Stargate.

To put that in perspective, every gigawatt is about 500,000 GPUs. So that’s about 2.5 million GPUs’ worth of compute power to power the AI initiatives out of the UAE around the world. If you had told me that would have come together in this short a period of time, I wouldn’t have believed it.

Howard Lutnick played an incredibly important role. Scott Bessent was there with the president. Of course, David Sacks was helping to put those deals together. You had pretty much every major CEO in U.S. technology there. Jensen Huang was there, et cetera. They were putting these deals together.

There was just a feeling of the art of the deal, of what is possible, partnership, and acceleration, as opposed to the deceleration and control that I think you and I have felt over the last few years and, frankly, have been worried about.

Yeah. Right. This idea that the U.S. was going to somehow shut down our AI and not allow the rest of the world to have access to our AI—I think we not only thought that was bad policy from an economic perspective, but also just dangerous for the world.

It was basically going to allow for this Huawei Belt and Road, where Chinese full-stack technologies would move into the Middle East. I’ve read that was basically underway prior to this big event—literally underway in Saudi alone. Huawei’s market share has gone up a lot over the last few years, right?

It’s not as though these countries have the ability to do nothing, and they’ve felt a little, frankly, betrayed by the United States, I think, over the course of the last few years. Here they are. Take Saudi Arabia as an example: They’ve been our ally for 92 years. These are not new relationships.

All of a sudden, we have this technology that we’re telling everybody in the world is existential to your national development, and yet at the same time we’re saying, “But we’re not going to give you our best technologies.”

While I was in Saudi, I had the opportunity to travel with His Excellency, the minister of technology, as well. I got to go to incubation labs and see what was happening there, see the data centers on the ground. They have a huge AI project there called HUMAIN that they launched in partnership with folks like NVIDIA and Groq.

There’s an incredible level of enthusiasm and investment going on because they know how important AI is. I think, frankly, the U.S. was on the verge. If we had continued down the path that we were on, one year from now they would have made commitments to build on Chinese AI stacks because they couldn’t afford to do nothing.

I think the timing of this change was critical. You could tell just the appreciation for the spirit of partnership. I saw that in every capital that we were in.

Let me ask you a question. What do you think the competitive advantages of these national data centers in the Middle East are? I have an answer, but I’m going to ask it naively first. What do you think the competitive advantages are?

First, you and I talk about it all the time: A primitive to AI is power, and they have cheap power. Eventually, the price of producing tokens—we’re effectively converting electrons, right?—is tied to cheap energy. They have cheap energy.

By the way, it’s both solar and nuclear and natural gas. Some of the largest natural-gas fields in the world—those things are being converted into intelligence.

More broadly, for the last 50 years, this part of the world has been powering the industrial age by exporting oil. I think their view of the future is that they want to power the age of AI through the production of tokens and the export of tokens. That’s just converting this power into tokens, and I believe that is a critical element.

Here’s the other thing, by the way. Let me poke at that a little bit, because I’ve been, out of curiosity, thinking about maybe trying to break down in my mind the percentage of cost—if you’re out buying a token on the open market—that’s tied to power.

I’ll tell you what I discovered, but if anyone in the audience has better numbers, let me know. What I’ve seen is cooling and power might be 20% of COGS, which, depending on your gross margin on your hosting, let’s say maybe power’s 15% of COGS. I don’t know.

That would equate to maybe 10% of revenues. So if you had, let’s say, a 2-to-1 power advantage, you might pick up 5 percentage points of operating margin, something like that, or be able to price 5% lower. I’m just trying to frame that advantage.

I would throw a few other things in there. Number one, you just have the issue of latency. This part of the world needs local data centers, whether it’s for Europe, whether it’s for the Middle East itself, or whether it’s for India. There’s just proximity, which is an important consideration here.

Then, of course, there’s the cost of labor. They’re building and investing in these robotic futures, where they’re building very futuristic data centers.

I would tell you, listening to Sheikh Tahnoon in the UAE and hearing about their commitment, they are as technology-forward as anything here in Silicon Valley, right? Partnering with G42, et cetera. All 3 of these countries are going to invest aggressively at a level at which really no other sovereign on a global basis is investing, to ensure that this part of the world becomes a center for technology and AI. I think it's critical that the US was their partner in that regard.

Sacks tweeted something when we got back that I thought was important. He said, “This AI acceleration partnership is not just a single deal. It's a new framework for advancing American AI both at home and abroad, and positioning the US tech stack to be the partner of choice in this region for decades to come.” I think that's right. This wasn't just a one-off deal.

A lot of people who are hearing about these deals thought, “Oh, maybe Trump went over there, worked his art of the deal, and got a deal done with the Saudis or the Qataris.” But no, I think this is a framework that the world can depend upon—that, in fact, they are going to have access to these technologies.

Let me give you a comparison, Bill. Can you imagine if, in the year 2000, Washington, DC, had said, “Oh my God, Google is so powerful that we're not going to allow any other country in the world to have access to this Google machine that might give them answers to questions unless Google comes to Washington and gets federal approval before it launches in any of these countries”? It would have prevented Google from ever becoming the global powerhouse that it has become, which has been so advantageous to the United States: hard power, soft power, economic power.

That's what we were doing in AI, and I think we've ripped the chains off of that. Now we're going to allow our companies to go compete and win.

To frame that a little more, I just put on X earlier today an article in The Wall Street Journal that just came out, talking about how far behind Europe is, broadly speaking, in terms of business culture and tech. The article goes into a lot of detail about how complex regulations have limited or stifled innovation, both on the labor side and on the technology side. The policies that were being considered, let's say, 3 months ago, to me, were all mirror images of the ones that are listed in this article.

I think we were 100% on a path toward building a wall around America, not building a wall around China. If we had continued on that path, I think I would have predicted that, if you look at the internet era, there really was the rest of the world using the American internet, and then China built its own internet.

Correct. Correct. I think we were headed toward a path where we were going to have a wall around America, and there was going to be American AI, and then China AI was going to have the rest of the world—the opposite of the internet. Bingo. Had we continued on that path, or if we go back toward that path, because I don't think this is over—I think some of the rhetoric coming out of the Hill & Valley Forum was more consistent with a wall around America—but it's a great sign that we're not holistically committed to that strategy.

I think you said it incredibly well. It would be a disaster for this country if we embarked on a path that was the opposite of what we did in the age of the internet. The United States was the greatest beneficiary from the internet boom of the last 20 years. The AI boom is going to be even bigger than the internet boom, and we need to follow the pattern, protocol, and approach that has made the United States a global leader in technology.

We do not want to be copying Europe, and we do not want to be in the situation that China was in during the age of the internet. I think this was a major and important first step in that direction.

But you're right, Bill. I saw a headline this morning that said, “Trump's rush for AI deals in the Gulf opened up a rift with China hawks in the administration.”

There are still these effective-altruist types—these decelerationists in Silicon Valley—who don't want to see the diffusion of any of this AI. It's creating this weird coalition of folks who have resisted the diffusion of American technology. I am firmly in the camp that we are not only safer, we are more prosperous, when the world runs on the American AI stack.

I was celebratory last week because it was planting a major flag in this incredibly important part of the world, with world-class allies who have major resources, desire, commitment, and passion to invest with America in this.

I also think one of the side benefits, Bill, is that this has been a part of the world that has been unstable for a long time. We still have issues in Iran. In fact, I think we're perhaps on the precipice of a major fork in the road with Iran, where they either sign the nuclear nonproliferation deal on the table, or we could have a hot-war situation with Iran.

I would say I expect that the president's lifting of the sanctions on Syria probably gives way to the Saudis signing the Abraham Accords and other positive developments in this part of the world. I don't think it was just a business deal. Had we alienated this part of the world in terms of business, I think it would have been a major setback for American diplomacy in this part of the world.

I think it was a good day for American businesses. This is going to bring in a ton of business for companies like NVIDIA, companies like AMD, and folks who are suppliers to those companies. I think it's important for companies like OpenAI that want to build out capability there. A major development.

Two things I would mention at the end of this that relate to global AI—the global AI race, let's just call it that. First, this Hill & Valley Forum, which I watched a little bit online. I didn't go. You weren't there, were you?

I wasn't. I wasn't. One of my partners was there, and I went last year. There's a lot of good that comes out of that conference, there may be, but the tone—the AI tone, I would say—was the opposite of everything we just talked about.

In fact, there's an interesting write-up where one blogger titled it, “Welcome to the China Hawk Industrial Complex.” I'm not trying to pour cold water on everything we just talked about. I'm just saying there's still a large group of people in our community who have invested in either software that they sell to the military or now products and services they sell to the military, and I think they need to be China hawks in order to justify the business they're in and in order to root for their companies.

After spending 25 years in venture capital, I can tell you there's just this instinctive thing—it's like defending your children. Once you invest in something, you just start to adopt the mouthpiece for it and make a lot of arguments that are central to that company's success. I think that's happening now that we have a large number of VCs in the military space.

I don't know. We're big investors in likely Anduril. We're big investors in ByteDance. I have an intellectual capacity to understand these investments for what they are and these partnerships for what they are. But most importantly, I think everybody has to figure out what's best for Team America, right?

We need to put America first when it comes to thinking about our global diplomacy, rather than what's best for my investment, maybe in ByteDance, or what's best for my investment in Anduril.

I do think there is a valid disagreement here. I've had this debate with Josh Wolfe, who's a good friend, and Josh is much more skeptical when it comes to China and open and free trade on issues of American AI, I would say, than I am. I think he has some thoughtful arguments on the issue, but on some things we can just agree to disagree.

As it pertains to the Middle East, I'd welcome Josh's view on this. I'm not sure what they are, but I think it's unquestionable that the US is better off having American AI in partnership with all of our allies in the Middle East. If we can't be in partnership with our allies, like India, like the Middle East, et cetera, then what are we doing?

Yeah. Right. What are we doing? I totally agree with that. I will agree to disagree on the other thing. I do think that once people start backing this stuff in a big way, it's natural that they would take on that point of view.

Let's talk a little bit about the Middle East. Maybe you know the other big news. By the way, I did want to finish with one thing. It appears the markets reacted positively, specifically to the Middle East talks and events. Is that correct?

2. China Tariff Deal

Well, the Sunday before the Middle East visit began, you had Scott Bessent, who had concluded the Chinese negotiations in Switzerland, and I think that was the bigger catalyst, right? We had this huge bounce-back in the markets from these 2 converging events.

You had the walking back of tariffs on China, and Bessent said, “Our goal is not to decouple from China, especially in non-strategic goods.” Remember, with China, almost everything is non-strategic goods. “But to open markets and restore balance. We will continue trading with China, especially in non-strategic goods and at lower tariff levels.”

At the same time, we're focused on reshoring critical industries like medicine, chips, and steel to protect national security interests. So, remember when we talked, Bill, 3 weeks ago, I was very vocal in saying there were 2 paths being presented to this administration. There's what I described as the nuclear Navarro path, which was high structural tariffs on everybody in the world, generating $2 trillion of tariff revenue and getting rid of the Internal Revenue Service. That was door 1, which the markets abhorred, right? That's what sent the markets down 20%.

Because we knew $2 trillion in tariffs is probably a 600- to 700-basis-point headwind to GDP. Door 2 was what I described as the Bessent or the Hassett approach, which was more consistent with the fair-trade argument the president had previously outlined: to reshore critical industries, to have some tariffs on the rest of the world, but to be modest in the scope of the overall economy.

And so it looks like—and I think what the market was reacting to was—that the president has leaned in the direction of the fair-trade, Bessent approach to China and to the rest of the world on tariffs, and away from the Navarro approach. In fact, we haven't seen much of Navarro in the last 3 weeks. And who's been in the lead on everything? Who was on the talk shows again this weekend? It was Scott Bessent.

And so I think when the tariffs started, you and I talked about the fact that chaos is very difficult, both for investors and for executives—committing to hiring programs, committing to capex, all these things. There's no doubt that Bessent brings a notion of calmness to the table. Almost every time he talks, he just talks in a calm way. In fact, the president said, “Every time you go on television, the market goes up.”

I think the president agrees with you.

Yeah. So that's all good. And just because we didn't mention it, I know there was a lot of concern among some of the companies that use strategic inputs—the rare-earth issues, the magnet issues. Those restrictions were removed as part of this first 90-day pause.

So, as I understand it, we're in a 90-day suspension on drastic tariffs. We have a 30% import tariff on Chinese goods, though. So, if you're importing 80% of your COGS from China, that's still a big deal. It's split: I've been told 10% tariff and 20% this odd fentanyl tariff. I guess the second part leaves open the door that if China agrees to something on fentanyl precursors, it immediately drops to 10%. That's my interpretation. I don't know if you know more than I do.

Well, what I would say is that going into this year, our tariffs on China were 15%. But look at the tweet that Bessent sent. He said, “Especially in nonstrategic goods and at lower tariff levels.” So there is the possibility that on certain nonstrategic goods, you could actually be back to 15% or maybe even lower, Bill, as you suggested.

But I think on the nonstrategic, I believe—textiles, home goods, correct? Toys would all fit in that, right? Backyard umbrellas.

Yes. All the stuff that people want to buy on Amazon—that is nonstrategic. But when it comes to steel and aluminum, chips, and items that we consider very strategic, certain medicines—solar panels might be a borderline issue, right? They've said those are going to be higher tariffs, so maybe those stay at 30%, Bill.

When you blend it all out, I had my team rerun the math. Remember, I came on here on Liberation Day and held up the list of tariffs that were on the poster board that Trump showed, and I said if you just add all those up, it comes up to $800 billion in tariffs, right? We have $3.3 trillion of goods that are imported every year into the country from abroad, and so $800 billion of tariffs is pretty substantial.

I had my team rerun the math today, and if you put China at 30% on strategic goods, at 15% on nonstrategic goods, and then the rest of the world roughly at 10%, you come up with about $300 billion in tariffs. I just told you that last year tariffs were $77 billion. So that's still a 4× increase in tariff revenue to the United States, right? But it's $300 billion on a $28 trillion economy.

I think what the market was saying is, okay, $300 billion: half of that will show up as increased prices and taxes on consumers and businesses, and half of it will get eaten by the producer of the product in China, in India, wherever. That's not that big a headwind to the U.S. economy, and that's why I think you've seen the markets bounce.

Now remember, we're just back to where we started the year. The markets were down 20%; now they're up 20%. I think this is about giving a path and giving some definition around the path forward—the flight path forward.

If you take these 2 things together, we went to the Middle East and landed the plane, did a bunch of deals, got trillions of dollars of investment into the United States, and unlocked hundreds of billions of exports to the Middle East. Then you landed the plane on tariffs. Those 2 things together, Bill, are a very positive orientation toward the world.

Again, you were saying this is still not a settled issue within the administration on the diffusion of technology, but these 2 things together are not an isolationist approach to the world, right? These 2 things together are still somebody who—I think somebody asked me on CNBC in March—I was saying there are these 2 doors, these 2 paths we could follow, and they said, “Well, what do you think Trump is?” I said, at the end of the day, he wrote a book called The Art of the Deal. This guy wants fair trade, but he's a dealmaker. On tariffs and on AI diffusion, he's been a dealmaker, and I think both of those things are positives for the economy. That's why you've seen the market bounce back the way it has.

Oh, by the way, one other thing I wanted to mention, just because we didn't mention it: I know there was a lot of concern among some of the companies that use strategic inputs—the rare-earth issues, the magnet issues. Those restrictions were removed as part of this first 90-day pause.

Yeah, I think that's going to be a critical question. Remember, we still have a 90-day pause, so it could come back if things were to go off the rails. And remember, we still have a ban on the export of H20. So there are no chips currently being sold by NVIDIA into China.

Yeah, right. We have a complete ban on AI chips into China. I think Jensen Huang was on social media again in the past 2 days expressing dismay about that. He even came up with some math: he would have sold $15 billion next year, and he said it would have created $3 billion in U.S. tax revenue that the U.S. won't have.

I think if you forecast out 2 or 3 years, they were on track to be a $50 billion business, I think, in China. If you just apply a normal margin and tax rate to that, that's billions and billions of dollars lost to the U.S. Treasury that Chinese consumers would have effectively been paying. It's probably another $10 billion of profits to NVIDIA that were lost—profits that they would have otherwise plowed into research and development to keep America at the forefront of chip technology.

So I think we could still see, as part of the negotiation over the next 90 days, NVIDIA allowed to sell some form of deprecated chip into China. I know, again, different people have different opinions on that. My opinion—and I think you share this—is that China already has frontier AI. Huawei is already there. We have demonstrable evidence of that, and their models are already there. Arguably, their open-source models might even be in front of the United States at this point in time.

There's no keeping China from frontier AI. So the better question is: are we better off competing against them? Are we better off selling to those companies, keeping companies like ByteDance and Tencent, et cetera, in the CUDA ecosystem, rather than allowing all of that data and all of those profits to flow right into the Huawei ecosystem and benefit the Chinese AI stack?

I think that's a closer call than selling chips to the Middle East, but when push comes to shove, I would sell chips into China because I think it's a net benefit to the United States. And we've talked about that in the past.

Yes. And there are many people who now believe that restricting our technology into China just gives them more and more incentive to implement their own technology faster and to invest behind it, which has been happening.

Well, if you look at autos, robotics, et cetera, there's plenty of evidence about that. Another—I would say the third big thing that's happened since you and I were last on—I can't believe, by the way, you step away for 15 days and all of this stuff happens. It's really crazy. AI time has new data every day.

3. Reconciliation Bill

But the other big one underway is the reconciliation bill. As a reminder to folks, we have to pass a budget for the United States. The reconciliation bill is a special type of legislation in Congress designed to expedite budget-related laws.

The way it does this, Bill, is it basically suspends the filibuster. So long as you're complying with these budget rules in the Senate, known as the Byrd rules, the only things that can go into this package have to be related to the budget. But if that is the case, then you only have to get to 50 votes in the Senate rather than the 60 votes otherwise required to be filibuster-proof.

You basically take all of these smaller pieces of legislation that might have otherwise touched the tax code or the budget, and you roll them up into this huge package called a reconciliation bill.

Now, the reconciliation bill started in the House with a draft. The Senate will then review that draft, and the White House is weighing in on it. The expected timeline—I think the president spoke to the House caucus today—is that it’s expected to be passed out of the House this week. Then the Senate will weigh in on it for 2 to 3 weeks. It will amend some of the language in it, add things, subtract things, and so on. It’ll send it back to the House at the end of June, and then the president is expected to sign it into law somewhere at the end of the month, or around the 1st or 2nd of July.

So maybe we can break down a little bit of what’s in the package, the bill. I think the biggest thing is the extension of the Trump tax cuts, right? Remember, when those tax cuts were passed in 2017, they only had a 10-year life. So they’re set to expire in about a year unless they’re extended, and that would be a major $3 trillion to $4 trillion increase in taxes at the end of this year unless they’re extended. The first thing they do is extend those tax cuts for another 5 to 10 years, whatever they agree on in the reconciliation package.

But then you have all these additional tax cuts that he talked about on the campaign trail: no tax on tips, no tax on overtime, no taxes on Social Security, and the immediate expensing of capital investment. This is huge for business. If you’re buying capital equipment, plants or equipment, or even software, you’ll be able to immediately expense those items from a tax perspective, not an accounting one. It’s important that those are 2 separate things, but yes, from a tax perspective.

By the way, I know a lot of people talked about this, but they’ve got to put a cap on this tax-on-tips thing because it’s going to leave a loophole that everybody and their brother will walk through. I think it needs to be de minimis tips, because otherwise you just redefine your consulting company as all tips.

Correct. I think they already have that in there. They’ve done a lot of work on this. Remember the 2 people in the House who are relevant here: You have the speaker, Mike Johnson, and Jason Smith, the chairman of the Ways and Means Committee. Obviously, everybody’s relevant, but those are the 2 leaders. In the Senate, it’s Leader Thune and Mike Crapo, who’s the chairman of the Senate Finance Committee. In the White House, you have Kevin Hassett and Scott Bessent.

The 6 of them—think of them as the Big 6—are putting together this package. They each have a list of their priorities, and then they get together and negotiate those priorities. We don’t even know exactly what’s going to be in the House bill yet, let alone what’s going to survive in the Senate, but we certainly have the contours of what’s going to be in it.

I saw this morning that the Council of Economic Advisers, which of course is in the White House, is saying, “Yes, although this will increase our spending—our CBO scoring reflects that the tax cuts cost money—it also will lead to much higher economic growth than we would otherwise have had.” That’s a normal argument that you see from those who support tax cuts.

And where do you stand?

Of course, higher taxes mean lower economic growth, and lower taxes mean higher economic growth. I think I’m firmly in that camp, but I do worry that we just had a $2.2 trillion deficit. We have a $38 trillion debt. We all got excited about Elon and DOGE being able to cut a lot of costs out of government, but I suspect when you look at this package—and again, I’m just saying this based on how the Congressional Budget Office will score it, and this is kind of arcane scoring—it will probably increase the deficit. At best, it would reduce it by $100 billion or $200 billion, depending on where our economic growth comes in, Bill.

Obviously, if you have higher economic growth, you’ll have higher tax revenues. There’s a huge benefit to driving higher economic growth for the country. But there are no big areas where we’re taking out $500 billion, $700 billion, or $1 trillion.

When you look at the tariff revenue, there are a couple of nuances here. We talked about the tariff revenue being $300 billion. They don’t count that in the budget. Think of that as part of your cash flow, but it’s not really in your P&L for the year. It is money to the federal government, to the U.S. Treasury, but it’s not included as part of our budget scoring.

I think another thing is the DOGE cuts. Those are real. There are a bunch of real cuts coming to fruition. I don’t know what those will total—maybe $100 billion—but they’re not going to be counted in the budget bill unless there’s an article of rescission. The White House basically has to get Congress to agree to rescind the money, and that hasn’t happened yet. So those cuts are also not added into the budget math that we’re seeing.

People should not get their hopes up that they’re going to wake up next year and our budget deficit is going to magically be cut by 50%.

One thing I would reiterate, just because it would be such an amazing scenario: There was a moment in time in the past 6 months when Trump mentioned potentially discussing with China and Russia cutting all of our military budgets in half. If that were to happen, I think it’s the biggest thing he could possibly deliver as president. If the Ukraine-Russia situation can be taken care of and things calm down in the Middle East, maybe that gets on the table, and that could be a big cut that would meet this agenda. Our neocon China hawk friends wouldn’t like it, but I would like it quite a bit.

Well, I think the Department of Defense budget goes up by about $150 billion as part of this package, and it’s about $1 trillion for the year. Part of the reason we have this challenge is that you have a lot of members of the House and a lot of members of the Senate, and it’s just very difficult to cut our way out of this problem.

I do think David Friedberg had a good rant on this on the All-In podcast. I give him a lot of credit for continuing to beat this drum. I’ve long supported a balanced-budget amendment. I think we have to come up with some structural changes that will allow us to get this under control.

I would love to believe that we’re going to grow our economy at 4% or 5% per year instead of 2% per year, because that would grow our way out of the problem. But I think it would be aggressive to forecast that kind of growth. We haven’t seen that kind of growth in a while.

One data point that’s kind of interesting: I listened to that same episode, and the team at All-In was talking about potentially monetizing the balance sheet. The question is, are there assets that exist on the United States’ balance sheet that could be turned into offsets for the deficit?

One thing that obviously comes up is land. Living here in Texas, many people probably don’t know this, but the university system here was granted, as a gift years ago, 2.1 million acres in the Permian Basin. That spits off, I believe—I could have this number wrong—$894 million a year.

Now, the U.S. has way more than 2.1 million acres, but that is an example of an endowment-like situation where land is monetized and creates cash flow. So, yeah, maybe there is something to that.

I would say one thing I was happy to hear the president say over the course of the last couple of weeks: He said that if we do anything like that, rather than putting that money in a sovereign wealth fund, we’re just going to use it to pay down the debt. Now it sounds like the sovereign wealth fund is on the back burner, which I think is a good idea.

I do think that we ought to have a task force in this country specifically coming up with ideas for how we achieve deficit reduction and debt reduction. Monetizing the balance sheet should be one of the things on the list to be discussed, along with other mechanisms to allow us to have permanent cuts—things like the defense cuts that you’re talking about, Bill. I think everything has to be on the table.

We have to come up with a plan that, over a reasonable period of time—let’s call it 10 years—even like a family, puts us on a path where you start small and work your way over a longer period of time toward deficit reduction.

I think it is morally unacceptable to saddle our kids with what will be over $50 trillion in debt. We effectively increased our quality of living during our lifetimes, and it will only hurt theirs because we’re saddling them with that big pile of debt. We’ve got to get around to whittling away at that debt, and hopefully we can get it done in the next couple of years.

A hot topic always in Silicon Valley is carried interest. Is carried interest on the table here?

I think the president has come out in support of eliminating carried interest, as well as—you heard him come out in support of raising taxes on the highest earners.

I don’t know. The last I heard, neither of those provisions are in the House version of the reconciliation bill. They may get added in the Senate, but I think those are fairly controversial. Remember, Republicans generally in the House and Senate are against tariffs, higher taxes, and all forms of higher taxes, whether on all these different brackets.

I heard somebody come out and say—maybe it was Phipe[?] on the All-In pod—that in New York, he pays 57% in taxes, which is, I think, what I pay in the state of California. There is a tipping point, right? People are familiar with the Laffer curve. At some point, taxes go up to a point where your tax revenues actually go down, right?

I would argue that at 57% all-in for the highest earners, you’re getting to that point where you disincentivize additional work, right? We may not be there yet, and you may be able to increase that tax bracket even more, but I would love to think that our country could come up with more creative solutions to solving our debt and fiscal crisis than just raising taxes. It seems like the easy way out, and I think there’s a better way.

4. Invest America inclusion in the bill

In fact, I’m happy to report that one of the things that got included in the House version of the reconciliation bill was Invest America. One of the purposes of taxes on high earners is redistribution, and I think there’s a much better way to achieve redistribution. That’s by getting everybody in the game from birth. So, it was a huge breakthrough that we got Invest America included in the House version of the bill.

Well, congratulations, Brad. I will tell you, I’ve talked to people who have pivoted later in life and dedicated themselves to philanthropy. The number one thing I hear from all of them is how hard things are and how slow things move.

So, for you to tilt at this and get it included so quickly—in D.C. time, this is ultra-quick—is a huge accomplishment. So, thanks for that.

We’re not done yet, but I would say this—maybe just a little behind the scenes on it, Bill. Listen, this is all new to me, and I learned a lot, as most people know. I started Invest America, the not-for-profit, about 2.5 years ago.

The idea was very simple: We needed to attack the wealth gap. We needed everybody to feel like they were part of the system. The way to do that was through the magic of compounding. Every child born in America—there are 3.7 million kids born a year—would get $1,000 in a seed investment account in the S&P 500.

It would act like a 401(k) from birth. Then companies and parents and others—church groups, whatever—could add to those accounts, such that by the age of 18, you could have about $50,000 in that account. At age 30, you could have about $150,000 in the account. You couldn’t take the money out. You couldn’t trade it. It would really just be that 401(k) from birth.

I will tell you—and you’ve heard me talk about this on the show—it had such high product-market fit, bipartisan from the left to the right. I thought to myself, maybe this could actually happen.

I was in Washington 2 weeks ago. I was at the White House and on Capitol Hill, meeting with the speaker, meeting with Jason Smith, the head of Ways and Means, and my friends Ted Cruz and Mike Crapo and others over in the Senate. I have to say, I was actually surprised that by the time I landed in the Middle East, I heard it had been included.

Now, of course, they tinker with it, right? One of the things I learned is you lose ball control. The name of this was Invest America. In the House version of the bill, they renamed it the MAGA account, which stood for Money Account for Growth and Advancement.

But, as you and I both know, a MAGA account in the minds of many is deeply polarizing. We’ll see what happens. I think when it goes to the Senate, it may very well get amended and renamed back to Invest America accounts. I imagine some changes will occur there. They’ll ultimately agree on what the rules are around distribution of this.

But the great news is this: Whatever you call it, whatever you call this, at the end of the day, if we do this and launch this in 2026—$1,000 for every kid born in America—we get every kid from rural Texas to rural Indiana to inner-city Trenton to the east side of L.A. into the game.

They can open up their phone and see that they own a little bit of Berkshire Hathaway. They own a little bit of Microsoft, a little bit of Apple. They feel like they have a shot at the American dream. I think it’s a game changer for the country.

The president weighed in here with the speaker of the House and said he would like to see it in the bill. That’s what eventually broke the logjam and got it in there. We’re working really closely, and I deeply would love to see this be something that unites the country, right?

This disproportionately benefits the 60% to 70% of people who are left out and left behind. Those are folks—it doesn’t matter the color of their skin, it doesn’t matter where they live. There are poor folks everywhere, people who will never have a shot at an account that compounds. This is an absolute game changer for those people, and it unlocks their human potential.

I know I have a ton of Democratic friends in the Senate and the House who support this. If everybody reaches across the aisle in the spirit of bipartisanship, we can get it done in an even bigger, unified fashion. But either way, we’re thrilled it’s in the House version. We hope the Senate will take it up, which we expect it will, and we hope it’s in the final version of the reconciliation bill. It’ll be a game changer for the country.

5. Delaware Incorporation Rights

Thanks, brother. A couple of speed-round topics, then. One is our good friend Kathaleen McCormick, the chancellor in the state of Delaware, Bill, who overturned the Elon pay package.

She caused this firestorm in Delaware that is literally going to potentially bankrupt the state, with every company in the state of Delaware now heading for the exits because they don’t want to be a part of this capricious decision-making by the Delaware Court of Chancery. I know that you have some updated thoughts based on some stuff you’ve been reading.

Yeah. I was made aware of something that I want to share with all the readers, and we’re going to put a pointer in the show notes. I’d encourage them to go read it.

There’s a professor at Stanford named Joseph Grundfest, whom I’ve known for many decades now. Joe is a former commissioner at the SEC. He was appointed by Reagan. He is the creator of the Stanford Directors’ College, which I think most people consider to be the number one directors’ college in the world. This is where, if you’re an independent director, you go and learn how to be a better director.

He’s considered to be one of the most knowledgeable people on the planet when it comes to corporate governance. Anytime I’ve ever had an issue—and there have been some where I’ve really needed help—I’ve always called Joe. He’s had me speak at some of his classes. He’s just a wonderful individual.

He recently—and I mean May 14 recently—published a piece of research that I think is super interesting and that everyone needs to understand. What he highlights is that the award multiples—and apparently there’s a thing in business litigation where the judge can decide what multiple of the standard hourly rate a lawyer should, or is able to, get—which is an interesting concept. I don’t know that anyone else gets awards as a multiple, but I’m sure a lawyer wrote it, and the base rate’s pretty high, right?

Anyway, what he found was that he compared a bunch of Delaware judgments with a bunch of judgments in the federal courts. He looked at a bunch of different breakdowns, but one of the breakdowns was: What’s the frequency of awards 7 times or higher, and what’s the frequency of awards 10 times or higher?

What he found was that in Delaware, 7 times or higher is 23 times more likely than in a federal court, and 10 times or higher is 57 times more likely than in a federal court.

You might say there could be a tail of scenarios where a judgment deserves some kind of outlandish result—some level of fraud or whatever. But what he’s showing is that there’s an activist mentality in the Delaware courts, that they’re giving out these super-high payouts at a much more frequent rate—23 times and 57 times more than at the federal level.

The big eye-opener to me is, first of all, I would tell everyone to go read this. But if you’re sitting on the board of a Delaware company and you’re not aware that this is happening—and this is new; these are all cases between 2009 and 2024—if you’re not aware that the awards being paid out to lawyers for cases in Delaware are increasing at this level and are being paid out much more frequently at very high payouts, you need to be aware of this.

To me, this is more damning than just the 1 Tesla thing. Delaware was known as a place where—you know, we talked earlier about tariffs, chaos, and calmness—it was supposed to be business calm, where you didn’t expect chaos, right? This shows that chaos is being built into the system, and it’s a recent development.

One more thing that I want to highlight. It turns out that in the 20 cases where you have this super-high multiplier, 55% of the cases are just 2 of the judges. There’s something else that I learned in reading this that I didn’t know: the chancellor, who is one of the 2, gets to pick who the cases are assigned to. And so the chancellor’s creating these super-high multiples, and then the chancellor’s in charge of handing cases out. She could hand them to herself or to this other judge who’s also doing the same thing.

I just think: eyes wide open. I read this and I think any company that I’m involved with, I’m going to encourage to leave, because this is radically different from why I was told we were supposed to go to Delaware. And, Bill, these awards go to who?

They go to the lawyers, and they go to whoever the plaintiffs are. Well, get this. We’ve talked about this, so I don’t need to pound the table on it, but in the Tesla case, the plaintiff had 9 shares. The plaintiff made $0 on a relative basis, right?

Right.

And the lawyers made $380 million or $345 million or something like that. And yet you had a plaintiff that didn’t get any recovery. Why should we have any law in this country? PAGA works this way in California. Shouldn’t the lawyer fee be a small percentage of the plaintiff fee? Why should it be 99% of the fee? It makes no sense.

I think the significance here is that what you’re arguing is what tends to happen is people heard about the Elon case, they treat it as a one-off, and they may not have made any changes. What you’re saying is that this was not a one-off in the state of Delaware. This is a structural problem that has emerged in the state of Delaware, which was the state that everybody went to incorporate in because they thought it had the most predictability.

And now you have a study that shows not only does it not have the most predictability, it actually has the greatest risk of long-tail adverse outcomes to the company that are highly unpredictable.

Correct. And I would encourage—we’ll put the link in there, and I’ll tweet it out after we post the podcast—but I would encourage everyone to go read that.

And following on that, and trying to keep with what you just said about the speed round, last week the 2 states people said you should consider other than Delaware were Texas and Nevada. Last week, the Texas state legislature passed a bill, Senate Bill 29, trying to improve Texas’s position, if you will, in this competition with Nevada over where you should incorporate. I just wanted to mention a couple of things that they put in there to hopefully make Texas better, which would make everyone better if they choose to incorporate there.

The first thing was that they codified the business judgment rule there. In some of these states, you can hop around the business judgment rule. This put a hard line in place and said, “No, you’re not supposed to. You can’t hop around the business judgment rule.” For people who know what that means, that’s a positive for businesses.

The second thing they did was put a limit on opportunistic legal claims. Some of that has to do with the multiple point that we talked about. But the big thing they did was that, in your corporate documents, you can put a minimum threshold of up to 3% on what’s needed to bring a derivative action. So you can basically say, in order to bring a derivative action, you need 3% of the shares outstanding.

This Tesla situation, where you basically just called around and found someone who held 9 shares, is ridiculous at face value, right? The fact that someone brought a whole case with a plaintiff who had 9 shares makes no effing sense whatsoever. And this gets rid of that. So I think that’s huge.

Basically, you’re flying the Texas flag and saying not only is Delaware more capricious than we otherwise thought, more activist than we otherwise thought, but that, in fact, it’s leading to new laws in states like Texas and Nevada that are going the further distance to try to encourage companies to come there. And they’re giving them predictability—not in the courtroom, but they’re giving them predictability in the code.

Yes. And there are a few more quick things. You can, in your documents, make it very clear that the new Texas business courts are the only place that you’ll have to appear. You can waive jury trials. There are some limitations on books-and-records requests. If you’ve ever been deposed, sometimes they grab everything, and this will limit that too.

So there are other little things, but those first 2 things are the big things. And, you know, I’m really trying to look after everybody—all the startups out there, all the venture-backed companies. If Nevada wants to raise the bar and wants to send us some information, I’d be glad to share that as well. I think the important thing is to get out of Delaware.

Well, the truth is, Bill, after this case, I will bet you that inertia kicked in and 90%-plus of companies and lawyers went back to incorporating in the state of Delaware, because that’s just the way momentum and inertia work. I think it’s important that you continue to beat the drum so that people understand—and we’re going to put a link to this study—that this was not a one-off.

I think the onus, the burden, is on the state of Delaware. We’re not trying to unfairly attack the state of Delaware, but the onus and burden is on the state of Delaware to explain and make the case as to how they’re going to change their system to provide the predictability and the protection that people thought they were getting, that clearly they’re not getting there.

6. Major Momentum in Crypto

Maybe to continue on, Bill, in the speed round here: There’s been a ton of momentum recently in crypto, and you sent me a paper that was recently written. In this regard, why don’t you tell us about it? Then I want to talk for a second about the GENIUS Act and the stablecoin legislation that took a major step forward last night.

I was told to read something, and I’m holding it right here. There’s a member of the SEC, Hester Pierce, who all the crypto people love dearly because she’s been very outspoken in her support of crypto. She’s actually the longest-serving member of the SEC at this point in time.

She put out a paper on May 8 called “A Creative and Cooperative Balancing Act.” I was surprised, and it’s not that long. Everyone should go read it: 5 pages. But she makes a very strong argument that crypto and blockchain specifically may be a better mechanism for tracking securities, including the tokenization of companies.

She goes on to say that you can have regulatory capture that can unfairly protect people. That regulatory capture may lead to market fragility because you end up with a single player or just 2 players. And she even suggests maybe there should be a regulatory sandbox, which would mean letting some people play around with this before you jump on top of them and kill them.

I’ve been this huge Delaware proponent, and I do believe that the number of public companies—she also brings that up—is way down from its peak. That could relate to regulation, but I also think it relates to inefficiency in our markets, in our IPO process.

I haven’t been a big crypto bull, but I’m going to pay attention to this. This is interesting to me.

In that regard, I think it may be time for you to revisit your crypto bullishness-bearishness thesis.

Yes, because the GENIUS Act, as it’s called, is pretty historic legislation. It effectively crossed a hurdle in the Senate last night that I think now puts it on a fast track to being signed into law. It was led by my friend Senator Bill Hagerty in the Senate. He’s just done an unbelievable job waging this battle to basically say: We’ve been persecuting everybody involved in crypto for the last 4 years, and rather than doing that, why don’t we just bring it under the federal regulatory auspices, put whatever protections we need in place, and then allow this innovation to occur in the United States?

In the case of stablecoins, we know that this could be the next generation of money management, of financial transfers, of updating the rails on which our financial system has been built over the course of the last 40 years. It was beginning to develop in other parts of the world because we had not developed a system that people felt comfortable innovating in.

What does the legislation require? Number 1, it requires specific reserve backing. It requires stablecoin issuers to maintain reserves equal to the value of their issued tokens. Number 2, it provides regulatory oversight. Now you can’t have this patchwork of people attacking folks who are trying to innovate in this area.

It gives them real certainty, puts in place a bunch of consumer protections, but fundamentally it allows this innovation to occur because it provides certainty and predictability as to the form of regulation. I give a lot of credit to Marc, Ben, Chris Dixon, and Andreessen Horowitz. I’ve spent a lot of time on Capitol Hill working on Invest America and other things, and every time I’m there, I see those guys on Capitol Hill working hard on the GENIUS Act.

I think this is a significant step forward for our financial system. It's a significant step forward for crypto. But really, we ought not think about this as crypto in the form of speculation. This is really about something that's fundamental to innovating in how money transfers in the world.

In fact, I heard somebody say that in 5 to 10 years, stablecoin issuers will become the biggest holders of U.S. debt on the planet because they have to own that debt to back the stablecoins—the tokens that they're issuing. And so think about this, Bill: people today pay ridiculous fees to move money around.

Yes. Right.

Between businesses. I was having this conversation over dinner the other night with John Collison. He's super bullish on this. They're getting into this fast-growing part of their business at Stripe.

Because the reality is, if you have 2 businesses—maybe 1 in Mexico, a small business that's a scuba business providing some tourism attraction, and you have somebody in the United States who wants to purchase that—rather than having to jump through a bunch of hoops and pay a bunch of fees to a bunch of intermediaries, you can just issue stablecoins back and forth. You can provide seamless transfers in the middle of that, so it reduces the overall tax burden on the economy, probably leads to higher growth because we're reducing that friction, and moves us into the 21st century.

I think that this is just the start. You mentioned something else, the tokenization of IPOs. I think you're going to see the tokenization of stocks, maybe of real assets.

So a lot of the promises, honestly, that we heard about, Bill, in 2019 and 2020 in crypto that got me so excited, then they all stalled out. And frankly, they stalled out because, again, of the capriciousness of regulators. They just wanted to attack and shut this down rather than coming up with a regulatory framework that was safe for innovation to occur in.

We're all in the middle of this AI supercycle. But if we look at the way that money is transferred on the internet, not much has changed in 25, 30 years. I think this is a massive breakthrough that's super important, and I think it's going to lead to a lot of unlocks, a lot of business opportunities, and it's going to be good for consumers.

7. Open Source AI in China

So credit goes to all the people who've been working on this for so long. Okay, we have a couple more on the list, but one I wanted to ask you about was all this open-source development in China. Tell me a little bit about what you see in terms of open source there versus some of the activities you see going on here.

Yeah. Well, the one that really caught my eye was an article about likely Baidu. Likely Baidu is scheduled to release the latest version of its model on June 30. They preannounced it for one reason or another.

And when they talked to likely Robin Li, who's been the longtime Baidu founder—I met him back in 2005, 21 years ago—he had originally been a proponent of closed models. He had spoken out publicly against that. And people are like, “Well, why are you going open now?”

And it turns out DeepSeek led to Qwen, led to Xiaomi—they have their own model as well. And so this will be the fourth deep-pocket-funded model in China that's open source. When you consider that this is a competitive dynamic that leads to that, and when you consider that it's already been proven here in the U.S. that different models can help improve other models, having 4 simultaneous models out there, all open source, I think is pretty damn interesting and will be tough to keep up with.

And I don't know what America should do about it. I mean, I guess I would love it if everyone—OpenAI, Anthropic, and Google—followed suit and chased Meta. But I don't see that happening yet. And so, just from a country-versus-country standpoint, boy, I think that's powerful for innovation, for speed, for speed of innovation.

Yeah, I think it's super interesting. I do think you're going to see the OpenAI open-source model. Sam's talked about it many times. I think you're going to see that in the next 30 days or so, certainly by the end of summer, but I would guess sooner rather than later.

He's indicated he wants it to be even more open, that he wants it to be the leading open-source model in the U.S. You and I both read about some of the challenges maybe going on at Meta. I have no idea if any of those things are true, but there's definitely been some blowback about Llama 4 in terms of its capability.

So it does appear that everybody in China is dedicated to open source, and in the U.S. I certainly hope that we have an open-source model that leads the world. I suspect that we're going to see more of these out of our leading labs. Elon has also said that they're going to open-source a model.

So I think this is one where we have to stay tuned, but the early leadership here probably does go to China.

Yeah. And I wouldn't be surprised if the people with proprietary models will use these to train them because they're available. And for everyone that wants to outlaw DeepSeek, guess what? You're playing whack-a-mole because there's 4, not 1.

Bill, one thing I wanted to ask you about—feel free to comment on it or not. I saw some attacks that I thought were somewhat unfair on you around Manus. I know I've talked with you a bit about Manus and the background associated with that. So would you share with us a little bit of your perspectives on Manus and maybe Benchmark's decisions, too?

Yes. So, yeah, I was surprised a bunch of people came after me on the socials, as they say, and they implied a bunch of deep-seated planning and whatnot. And I think part of it just comes from—I think, in not being a China hawk, people accuse you of being a Sinophile. And there's a lot of room in between those things.

But one thing I think these people don't know is that I'm no longer a GP on the new funds at Benchmark, which means I'm not involved in new investment decisions. So I wasn't part of the decision to invest in Manus, and I found out about it after the fact.

I'm obviously a big LP at Benchmark, and I'm a GP in the older funds, and I believe in Benchmark, so I do support the firm. But some of the theories involved scenarios that just didn't happen.

After the fact, I went deep on it, and I think there are a few things that it would just be good for people to understand about Manus.

First of all, they've only operated on U.S. models. They're a wrapper company. They don't have a foundational model, and they've only operated on top of U.S. models. So they've actually never operated on top of DeepSeek or any of the other models we were just talking about.

The second thing is they have offices around the world. They're in Singapore, Japan, and the U.S. Third, they host all of these on U.S. hosting services. So they're not actually operating any of their customers' workloads in China at all. And the data is all resident on those U.S. hosting companies' servers. And so there's no customer data in China either.

And I think people rush to judgment. I've already spoken about why I think they rush to judgment. I think there's a China hawk tilt among a lot of these people. But things aren't always as they seem.

And so the company has some leading agentic technology. I think they have some leading browserless, headless-browser technology. And they've got a long list of customers that are excited about what they're doing. But I just think it's important for people to understand the facts.

Well, I think another thing is, I saw a stat the other day. I think 50% of AI researchers in the United States of America are Chinese, right? There is a danger here and a drift that is very xenophobic, very anti-China, that is not healthy for our own relationship in terms of desirability to work in America if you have a Chinese background.

I think American investors should be free to invest in companies like Manus. They can weigh it into their decision-making. Everybody's free to choose not to invest in a company like Manus, but I think attacking firms simply because they're investing in this without full information about what it's about—I don't know. It struck me as jumping the gun a bit.

And I just think, like you said, there's a lot of daylight between being a Sinophile and being a China hawk. And I think people are entitled to having differences of opinion about the best way forward for Team America.

Jensen Huang thinks the United States should be engaged in competing in China. And I've said, for a long time, the number 1 thing we could do would be to open up our border for skilled immigration from China as wide as we possibly can. That would be recruiting the best and the brightest. I've talked about an AI visa for our Chinese researchers and their families so they feel comfortable and safe and desired to stay in the United States.

I don't see people attacking Elon for having Tesla in China or Tim Cook for having Apple in China. Right? The U.S. has benefited greatly from having our companies compete in China.

In fact, if you ask me what's more desirable, it's to have China open back up to our markets. I would love to see our internet companies allowed in China. I would love to see our AI companies allowed in China. I would love to see that form of reciprocity, right?

In a perfect world, I'd love to see tariffs come down on both sides and competition allowed more vigorously on both sides.

What you basically just outlined was, hey, we should be doing with China what just happened in the Middle East, right? Try and get our differences on the table, try and get the things that matter to us on the table. But on the other side of that, get everything else to a point of collaboration.

Listen, I don't think we should be naive. China—we are in a great power struggle with China. There's no doubt about that. There are things that we're going to compete like hell on, things that we're going to disagree on, things where we're going to be strategic adversaries, and things where we say we're not going to give it to you and you're not going to give it to us. That's fine.

But listen, I'm on the side of Scott Bessent. He said we're not decoupling from China, right? So we need to figure out where we want to engage, and on that playing field, we ought to be all-in on non-strategic trading with China. We ought to be all-in, and we ought to find ways to reduce tariffs and to find more ways, certainly, to recruit their best and brightest—not only here to be educated, but here to stay and help us build great technology.

Thank you. Anyway, I've been wearing this Altimeter hat for over an hour and you haven't commented on it yet. I obviously did this just for you. So, little Team Altimeter today.

I love seeing it. Love seeing it.

Well, thanks, and good seeing you. I look forward to seeing you in person next time.

All right, man. Take care.

All right. Take care. Bye.

Bye-bye.