E45: Sequoia Capital's Roelof Botha and Jacob Helberg on AI and National Security

0:00I think it's fair to assume that what we think are close source or proprietary models will in fact not be because they will have been stolen at some point. And so we're kidding ourselves, I worry, if we think that we can truly keep these things secret. I'd really favor an approach that just leans into our advantage. Because if we operate from a position of fear, where we try to lock down the doors and batten down the hatches, and this is the way we're going to retain our competitiveness, I actually think we're going to lose.

0:28Welcome back to Turpentine VC, a podcast where we discuss the art and science of building successful venture firms, VC to VC. In today's episode, I sit down with Roloff Botha, the managing partner of Sequoia Capital and Jacob Helberg, author, Palantir advisor, and founder of the Hill and Valley Conference, which bridges technologists and DC policymakers. This conversation goes into the role the US government has played in the growth of the tech industry, Sequoia's split from its China and India entities, the future of AI regulation, and the importance of open source versus proprietary tech models.

1:00Roloff, Jacob, welcome. Stoked to be interviewing you both. So the Hill and Valley Conference was a massive success. When we first start with you, Roloff, why did Sequoia decide to join this year? Well, before we get into that, I want to say a big thank you to Jacob, because he deserves the credit for bringing together Silicon Valley and D.C. You know, there's a multi-decade history between Silicon Valley and the U.S. government writ large. I don't think people always appreciate how much of what Silicon Valley became was thanks to a significant amount of government spending on basic research and development for the development of semiconductors and the internet and everything else.

1:38And so that seemed to have softened, I'd say, maybe a decade ago. And it's wonderful to see somebody like Jacob step up to rekindle these conversations where people from D.C. and Silicon Valley can get together, as the name suggests, Hill and Valley, for us to get together in the same room and find out that we actually have much more in agreement than may have been apparent from a distance. So I wanted to say that off the bat. You know, the reason we showed up, when I joined Sequoia shortly after the PayPal exit, I think there were about 200 million people on the planet that had access to the internet.

2:11And today, there are obviously measures in the billions. And so technology has infused our lives way more extensively than I think any of us could have imagined 20 years ago. And so I just think it's kind of quaint and maybe outdated for Silicon Valley to imagine that it doesn't have a big role in society at large. And so it's imperative for people from Silicon Valley to show up and have a voice in the debate in D.C. as technology shapes so much of our lives. And if we're not present, we're not able to have a proper conversation that's fully informed to make sure we have sound policy that enables the U.S.

2:42to continue to thrive and be competitive on a global stage. Yeah, that's well said. Jacob, I want to turn it over to you. You've done a few of these summits now, and this is a massive success. major sort of contributions from both sides in terms of people who presented. We had even former President Trump submit some notes of encouragement. Why don't you talk about how attitudes have changed both on Capitol Hill and Silicon Valley as these last few years as you've been a bridge between the two? Yeah. So first of all, I just wanted to start by thanking Roloff for his incredibly kind and generous words.

3:16And Eric, if I could just build a little bit on that first question you asked, which was about Sequoia's participation in Silicon Valley, and just add a couple of points to what Roloff said and say that a core mission of the Hill and Valley Forum is to promote the partnership between Silicon Valley and Capitol Hill to make sure America has the best technological capabilities in the world. And in my view, we can't fulfill that mission well without having the very best players from both sides at the table. And my boss at Palantir, Alex Carb, has a saying that if you are trying to treat a sick patient, you want the best surgeon on your side.

3:53And I kind of view this through the same lens. And so I think for the Hill in the Valley Forum to be a success, we want the best innovators and the best venture capital investors coupled with the most thought-provoking and committed policymakers. And so I was incredibly excited when Roloff said that he was interested in joining because Sequoia obviously plays an instrumental role in Silicon Valley and has been at the forefront of innovation. And so it was a real treat to have him join us at the Hill and Valley Forum this year. That's awesome. Will articulated, why don't we segue to the second question, which was how have you seen attitudes evolve since you've been doing this work of bridging?

4:33So when I first started going through the intellectual exercise of taking notes on the relationship between Washington and Silicon Valley, My work to some extent started when I worked on the policy team at Google and I saw two things. The first is that the relationship between Silicon Valley and Washington had grown somewhat fraught over the years with an enormous amount of distrust. And there was a pretty noticeable cultural disconnect. Part of that distrust stemmed from an unfortunate series of events that ranged from the fallout of the Snowden leaks, which caused a huge amount of international backlash against large Silicon Valley companies.

5:19And the second was various events from Project Maven to Project Dragonfly, which got a lot of press and attention, which caused a lot of policymakers in Washington to distrust Silicon Valley. The other thing I noticed is that a lot of the people in Washington from Silicon Valley who were getting the most airtime and attention were executives from the biggest technology companies. But at the time, I lived in the Bay Area and I saw that there was a whole vibrant technology ecosystem in the Valley that was getting no attention and exposure to Washington. And so one of the byproducts of this is that a lot of people in DC had kind of pessimistically grown to believe that our best and brightest engineers are focused on optimizing advertising technology while our adversaries in China are focused on building hypersonic technology and what people call hard tech.

6:23But I knew that wasn't true because I saw every day founders undertaking big risks and trying to solve hard engineering problems. And so ultimately, I wrote a book that came out in 2021 that had an entire chapter devoted to the rift between the hill and the valley. And a byproduct of that work ended up materializing in building out a conference and ultimately a community to bring these two communities together and to try to bridge that cultural gap. And the cultural gap is real. I mean, in Washington, people in Congress approach problems from a legal background. Most stakeholders are in their 60s.

7:08In Silicon Valley, people are on average about 20 years younger. They have an engineering background. So people approach the same set of facts with different worldviews and life experiences. But fundamentally, they're both in the business of building the future. And I was really passionate about helping serve as a bridge between these two critically important communities. That's well said. Roloff, I want to turn to you. I'm curious, where are you guys at Sequoia bullish on defense and frontier tech? Where are you bullish and where are you concerned? Well, the job I have is actually a lot of fun because I just get to listen to the smartest entrepreneurs.

7:46And that's what we follow. And so fortunately, founders see the future before the rest of us. That's their special gift. And what we at Sequoia try to do is help them realize that ambition and make that future a reality. And so we've seen more and more of our entrepreneurs, partly for good company building reasons, sometimes for good patriotic reasons, want to address opportunities in and around defense technology. So we're an investor in SpaceX, which obviously has a huge influence on defense for the United States. SpaceX now accounts for 80 % of all mass into orbit globally. It's bigger than all the other countries put together.

8:25And if it hadn't been for what Elon had done at SpaceX, the U.S. would be lagging significantly. And if you think about control of the space race, that's a huge, huge element of our national defense. And obviously, I have a long history with Elon. he recruited me at PayPal many, many moons ago. We also partnered with a company called Muck. Jacob earlier talked about hypersonics. It's a little bit sad, actually. It was DARPA funding that helped the US develop hypersonic technology. And unfortunately, we didn't develop it enough. The US Air Force didn't take up that opportunity. The IP was stolen, appropriated.

9:00And today, our adversaries have far more hypersonic capabilities than we do. And so we're back to a company called Mark. Ethan, the founder, was actually at the Hill and Valley Forum. And they're building hydrogen-powered hypersonic missile systems and defense systems for the US government. So we're thrilled to be in business with them. We also have a long history on cybersecurity, which is obviously a big domain where some of the battleground is being played out. And so whether that dates back to our investment in Palo Alto Networks back in 2005, we've partnered with companies like WIZ and Oasis more recently that are providing cloud infrastructure, cybersecurity solutions, or solutions for non-human identity, machine security, if you will.

9:43And so we continue to see these opportunities. We have two more investments that we've already closed, but they're still in stealth. And so we're super excited. Actually, earlier today, we just approved another investment in the defense space. So it's wonderful that the founders are finding their way here, and we are eager to help them build great businesses. Jacob, I want to pass it over to you. You've been very vocal about why we should ban TikTok here, and you've been very influential, perhaps the most influential in helping make that happen or steps towards that happen. Do you think we're in a new arms race with China?

10:20I do. I do think we're in a new arms race with China. And for a long time, I was very concerned that Silicon Valley, that meant too many elements in Silicon Valley were trying to remain neutral and be Switzerland. The business of geopolitics is not core to the day-to-day work that draws most founders to wanting to build a company. So I think a lot of entrepreneurs understandably prefer staying away from being in the middle of a geopolitical rivalry. With that being said, I think part of what has happened in the mid-2010s was a lot of companies that tried to remain neutral in this deepening geopolitical rivalry in practice ended up creating a policy of equidistance where they were trying to essentially balance competing priorities with Beijing and Washington.

11:18And I wrote an article in 2020 that said Silicon Valley can't be neutral in the US-China Cold War and that that approach would be untenable. I'm very, very happy that I think we've turned a corner. And I think, generally speaking, there is actually broad consensus today in Silicon Valley that the relationship between the US and China is in a place where it requires companies to take a side in a less ambiguous way. And so I'm happy that I do think today the era of Silicon Valley's neutrality is over. And we are seeing, to Roloff's point, we're seeing this incredibly inspiring growing ecosystem of tech founders who are willingly answering Uncle Sam's call and solving some of the hardest engineering problems from hypersonic technology to biosciences and artificial intelligence.

12:17All of these hard technologies and frontier technologies are inherently dual use, and it's incredibly important that our very best founders focus on the hardest problems possible. And ultimately, I think it's incumbent on the U.S. government to do a better job at reaching out to the Bay Area to help convert our best private innovations into hard power. Roloff, I was just giving Jacob a lot of well-deserved credit for his advocacy work. But you also, you and your firm deserve credit for being very open about separating from the Chinese Sequoia entity, focusing your efforts in the US and in Europe.

12:58Can you explain the status of that separation? To what degrees you're still connected to that firm? And why you thought this was in the best interests of the US and Sequoia? Sure. So, you know, we embarked on an expansion period starting around 2005. And this was shortly after China was admitted to the World Trade Organization. And I think the general sentiment in the West was that economic integration would lead to sort of normalization and an embrace of Western democratic values. And that's, you know, the US government explicitly encouraged investment in and development of economic ties with China.

13:35And it obviously led to an amazing rise from poverty for hundreds of millions of people. So we opened up offices in India, Southeast Asia, and China. We did set them up as separate entities that were managed by separate teams, separate decision-making, separate funds that were raised independently. And what we shared was a brand and back office functions like finance and compliance and IT, if you will. And obviously, the world has changed. Some of the data I've been looking at recently, world trade in goods as a percentage of global GDP peaked in 2008. 2008, more than 15 years ago. And so we're in this period right now where you see more economic nationalism.

14:21And it's a very different period from this sort of post-Second World War 70-year golden period that we've enjoyed. And we need to recognize that reality. And so we're not blind to geopolitical pressures, obviously. And honestly, there were challenges in running this very distributed global organization, yet that they were separate entities. And so we all got together last year. We agreed that it was better for us to go separate ways. We announced that in June 2023. And by the end of last year, the full separation was complete. What was Sequoia, China is now called Hongshan. What used to be Sequoia, India is now peak 15.

14:59And we are fully, fully independent, separate brands at this point. Hey, we'll continue our interview in a moment after a word from our sponsors. Jacob, is what's happening with TikTok a bellwether for more technological decoupling between the US and China? I think it is. And to echo some of the points that Roloff was making, I think the arc that Roloff described where in the early to mid 2000s, the prevailing consensus was deeper integration with China was good for the world and good for America was as true in the private sector as it was in the public sector, where to Roloff's point, it was official US policy for a long time.

15:44to promote deeper regional and international integration. And so part of, in a lot of ways, the changes that we've seen a number of different companies make to decide to look to other markets outside of China and to, some companies have adopted a build where you sell model, whatever the recalibration in a given company's approaches, but to adapt to this new landscape, we've actually seen that same process take place in the policymaking community. And the TikTok bill is actually a reflection of that where, you know, TikTok was allowed to operate in the US for a long time. Huawei wasn't an issue until the mid 2010s.

16:33So was ZTE, you know, DJI is a topic of discussion and is still allowed in the US, although is undergoing a lot of scrutiny. And so if you zoom out, I think you're seeing a lot of policymakers go through that same process of recalibration, of how they think public policy should be recalibrated to factor in the fact that the world has changed and therefore updating the policy approach is warranted. And I think that the TikTok legislation is ultimately an important overdue recalibration of our technology trading relationship with China. And it was a hugely important precedent and bellwether for how we approach other technology issues.

17:26And ultimately, I've been a big proponent of the United States government thinking through in a holistic way what should our technology trading relationship as a whole be with China? I think, you know, I would favor a holistic approach rather than a piecemeal approach. And I think you're kind of seeing that take place, you know, bit by bit, whether it's debate that's going on with ZTE, Huawei, TikTok, DJI. I think, you know, the Trump campaign announced that, you know, it could, It would consider the imposition of tariffs on electric vehicles. So you're seeing a 360 recalibration of our technology trading relationship.

18:07And ultimately, I think given where the world is in 2024, I think it's a healthy thing. Roloff, is there anything you wanted to add to that? No, no. Having studied economics as an undergrad, along with statistics and math, I was always a big fan of Ricardo's view of, you know, the benefits of trade. And the problem, obviously, is that it makes simplifying assumptions about the interests of the trading partners, and that, you know, everybody serves a common good has a certain sense of values around the respect for intellectual property, human rights, and things like that. And so I'm reminded of this one joke that, you know, at some point, if you're playing a sport, let's just say playing soccer, you know, if, if the one side is winning, because they keep going offside to score goals, you know, at some point you sort of have to say, listen, can we just recalibrate?

18:55It's not the way we're all playing. So I think this is a very healthy dynamic that we're having these conversations. Yes. Yeah. I didn't learn that. The sophistication that we're talking about now, I didn't learn that in my econ class as well. And Eric, just to build a little bit on what Rolov just said, I think one of the reasons why having open lines of communication between Capitol Hill in Silicon Valley is so important is because I think everyone in DC understands that, you know, being autarkic and reshoring everything to the United States is obviously not realistic. And so actually for policymakers to be able, policymakers need to be able to get feedback from technology executives in the Valley in order to basically say, okay, our objective, China is abrasive to American security interests.

19:47We need to reduce our dependence on them from a supply chain standpoint and from a data security standpoint. What is the best realistic way to achieve that objective? And getting that feedback from technology executives who understand their supply chains, who understand the technology stack and the architecture of complicated platforms is actually incredibly valuable. And policymakers can then, with that feedback, do a better job at actually implementing a trade agenda, whether it's with countries like Vietnam or India or other markets, in order to provide the scale that companies need to develop great products and groundbreaking technologies while addressing the security concerns that policymakers have.

20:33And maybe just to add to that, in addition to the security concerns, there's just basic rule of law around IP, intellectual property. I mean, I'm involved personally with at least three software businesses where our software was blatantly stolen and we're not able to build businesses in China because, you know, we can't enforce our property rights. And so we're in a situation where, you know, these companies have spent, in some cases over the last decade, a billion dollars in R &D spend to build world-class software in certain domains, and yet we're not able to access a certain market where our software gets taken, appropriated, and used to build businesses by rivals who didn't have to make that R &D spend.

21:17It's a little bit like me stealing somebody's factory, not having to pay for all that CapEx and claiming I'm a low-cost producer. It's like, no, it's because you stole something. So there is something fundamentally unfair about the way that trade is conducted, And I don't think one should tolerate that kind of unfair behavior from a trading partner beyond the national security issues, which obviously are even more important. Yeah, absolutely. And Peter Thiel actually had an interesting observation once upon a time that the US and China in some ways are playing a different game where because China's population is roughly four times larger than the US, for parity means America is already losing.

22:00Technology parity. because if they manage to fully catch up to us technologically, their economy could be four times our size just by virtue of the size of their population. And so for us, it's actually even more important for us to stay ahead than it is for them. And building on what Roloff said, it's incredibly important for us to be judicious to close down the pathways that the Chinese have been able to exploit to carry out IP theft, which at the scale of the US economy amounts to hundreds of billions of dollars. And the former head of CyberCom actually referred to it as the biggest transfer of wealth in human history, which is obviously hyperbolic, but he was making a point that the scale was significant.

22:51It's really interesting, Jacob, to that point. It feels like you've been calling these warnings or warning for years. And it seems like people are finally starting to listen and take this threat seriously. But if you knew what you know now, and could go back to say 2010, or something like that, because it feels like we it seems like we had all these expectations, or hopes that not only would China liberalize and play fair, that we had expectations for them to do all these things that they ended up not doing or acting in ways that that we didn't expect and anticipate. And if we knew that they were going to act in this way, what would you have recommended that we do differently or what could have we done differently that would have set us up in a much better position?

23:30I mean, I think moving along the debate about, you know, strategic technological decoupling could have happened sooner. I think there was already evidence of China's nefarious behavior. I don't think I had, you know, necessarily had a crystal ball. I think I was just paying closely attention to the pattern of China's nefarious behavior in a way that wasn't getting a lot of attention at the time. But ultimately, I'm grateful that people have been absorbing information with an open mind and that when presented with evidence, a lot of people kind of evolved their views. And I think that's a healthy thing.

24:09I think you want that debate to happen. I mean, my basic approach over the last few years has been engaging in conversations with almost everyone willing to listen. And I think it's, you know, one of the great things about our democracy is that you can have these Socratic dialogues and, you know, based on the merits of an idea, let the best argument win. And if an argument is compelling, people are open to it and, you know, we'll take that feedback. And I think it's great that the consensus has, you know, evolved in the way that it has. And I think ultimately we're landing in the right place, which is good news.

24:48Well said. I want to zoom out and talk about the AI supply chain. How are export controls affecting tech? Roloff, do you have thoughts on, I'm curious, have people in the Bay Area been talking at all about export controls? And is this a topic of discussion in the startup ecosystem? Yeah, export controls, in and of themselves or not, I think there's, some of the things that are topical in Silicon Valley are CFIUS. So the ability for foreign countries or people, particularly China, to be able to make investments in the United States, those rules have existed for a while. There's a question of whether or not they're sufficient given the changes that have happened.

25:33And also whether existing ownership gets reassessed. If an investment happened three years ago and what today is deemed sensitive, should those be reassessed rather than just forgotten about? So that's a contemporary conversation. And then I think increasingly investors in Silicon Valley are loathe to invest in China. And I will say that in some pockets, one of the phrases I heard was many people are concerned about a return of investment rather than a return on investment because of what has happened in China. And in fact, you're seeing a big tide of Chinese investors actually investing in Europe and the United States because the opportunity set in China has shrunk, given how capricious the government has been and how involved the CCP has been in so many of the leading private and public technology companies.

26:26So that's an ongoing debate. When it comes to export control, I think the way I think about it maybe is in the context of AI in particular, where we need to have something that's the equivalent of energy independence. I think the US would have been in a truly precarious position, if you think about what's unfolded over the last decade, if we hadn't cracked fracking and been as energy independent as we are now. And people here are very apprehensive about a future in which we may not have AI independence. And that goes all the way from the semiconductors that you need, the data centers that are used to build foundational models, and also then to run inference, the software models that are built for those, the applications that are developed.

27:10And right now, I don't think we can claim full independence. Certainly with semiconductor dependence on manufacturing in Taiwan is obviously a huge exposure point. And I think people here are very eager to see the redevelopment of key nodes of manufacturing capability within the United States. And so that is certainly a very, very hot topic here in Silicon Valley. And I think in D.C., I think those observations from Silicon Valley are actually incredibly interesting because a lot of time and attention in Washington has been very focused on chip production. but less attention has been drawn to other parts of the AI supply chain and the technology ecosystem.

27:56I think the trend that you pointed out, Roloff, of an increased level of Chinese capital in the US tech ecosystem is actually underappreciated in the policymaking community, as well as this concept of AI independence I think is incredibly interesting and is in keeping with some of the things that we've seen. For example, KV invest in an Indian-based version of open AI based on the basic thesis that different large countries will want their own quote-unquote sovereign AI capabilities. And so I think this concept of AI independence is actually a really interesting one for people in the policymaking space to study and become more literate about, especially in light of what's going on in the UAE right now, where the UAE, which is obviously a very energy-rich country, has committed to investing billions in large computing clusters.