Cathie Wood on Tesla-SpaceX Merger, $1M Bitcoin, More AIs Than Humans | EP #296 | Moonshots Live

Singularity is here.

Yes.

Oh my God, how lucky are we to be alive now during the singularity? I mean, I know you guys know this. I mean, we feel the speedup, right? Every day I wake up and check my X channels, which I've curated. I check the conversations we have in our WhatsApp group in the Moonshot Mates. Like, what just happened? What just happened? What just happened? It's mind-blowing. It's absolutely mind-blowing. So we have 2 incredible leaders on stage, and I'm going to kick off the first question to you, Cathy. And thank you, Cathy, for your support of this event and of the Future Vision XPRIZE. So proud of what you and Lisa Dodd have done to support this.

Yes, and thank you. Lisa's here, so thank you, Lisa, for all you've done for this This great event.

Your team has said the Tesla and SpaceX merger could be announced this year. I agree. Rockets, satellites, AI, EVs, robotaxis, humanoid robots, energy, all under one roof. The question is, what is Elon, in your opinion, actually trying to achieve long-term through this merger? Do you think he's going to hit any particular roadblocks? And I'm curious, you know, does SpaceX AI or SpaceX need to reach a certain valuation before he pulls the trigger on that so the dilution isn't too much?

Yes. So we do think it's going to happen. What is his ultimate goal? Well, he tells us all the time it's Mars. He's just creating a number of stops along the way in the form of A global connectivity, broadband connectivity business.

Yeah, the Dyson Swarm.

Yes, really unbelievable how profitable and how big it is already from a revenue generation point of view. And this NeoCloud business that he, basically it was a giant pivot in terms of xAI. So NeoClouds, but Elon believes that the leader, the AI leader in terms of frontier models, is going to be the company with the most computing capacity and at the cheapest cost. And with orbital data centers, he thinks he will be that company. So he's really going to go after You know, the OpenAIs and Anthropics of the world, very determined.

SpaceX is my largest holding. Is that your largest holding? What's the top of your stack?

Yeah, Tesla and SpaceX, and we do believe they will be combined. So yes.

Interesting. Do you think, you know, so he has a Chinese business with Tesla, right? and he's in the Defense Department. How does he handle that?

Yeah, as one of the roadblocks out there? You know, it's very interesting to see him sitting at the table with President Trump and Xi Jinping, and we know that Maye Musk is adored in China, so he's working all angles here, and there's even, there has even been talk that he's going to be able develop a robotaxi system in China. So it's very interesting. Many people would say that's the biggest stumbling block given our defense posture and given how this administration has, you know, basically portrayed the Chinese as our biggest potential enemy. Keep your enemies close is what I think this week is all about. So, you know, I can see, honestly, I have, now I am an optimistic person, but we believe that the productivity gains coming out of this AI and just generalized technology revolution are going to be so enormous that we could have a win-win-win here with China and the US. China needs to develop a consumer economy. consumer economy, and it can turn this productivity gain into increased compensation to really develop the consumer. We need to be maybe more aggressive in competing against China. We can use some of it to cut prices. I think, and that would lower inflation and take away a lot of the fears out there from a policy point of view. I think that's what Trump and Musk together are working on.

The Fed announced something like a 4.7% GDP growth in the last quarter, which was like double the quarter before. When I interviewed Elon with Dave Blunden last December, it aired beginning of January, he said, you know, he could imagine triple-digit growth within this decade of the GDP, which is—

Is that cumulative or

Per year?

Yes, I think I tuned into that part of it and it was like, well, so our number is 7 to 8%. We're accelerating in the next 5 years to 7 to 8% real GDP growth. I have heard him say 20, 30%. I had not heard him say triple digits, and I wasn't sure if it was cumulative that he was talking about.

I'll ask him. Are you going to the Roadster rollout on October 1st?

We have, we have been invited.

Okay. Yeah. So I was texting with Elon, and I think we're going to do a Moonshots podcast with him from there. So I'm excited about that.

Very cool.

Yeah. Iman?

Yeah, no, I think that one of the interesting things that I've seen recently, like your view, Nikhil, and it kind of extends, is Elon said that GrokBot has hundreds of thousands of installs now. Yeah. And obviously it's a brand new line. We've seen Muse with 3 million—

3.8.

3.8 million downloads. And then we saw—

I'm sorry, 2.8 million downloads, 3.8 billion users, right?

Well, that's the— yeah. It's obviously going to grow this entire sector. And then Amazon banned Muse. So I think one of the really interesting things for me, and, you know, I think, Nikhil, you've been thinking about this a lot, is How do you tell who's who on the internet when we have these AIs intermediating us? Millions of

them.

Then at what point are there going to be more AIs than humans doing these jobs on the internet itself? Because I think this is one of the things that really drives the massive real GDP growth, right? Because you have so much money being unlocked through intents and being able to represent what you want. I'd love your insight on that.

I think on the second part, I think there will be more AIs than humans next year. There's no reason for that not to be true. There's 5, like anywhere between 5 to 6.5 billion people on the internet at any given time. It's not very hard to imagine that every one of us has thousands of agents working for us. They're coming in and out, but like they're all going to be there. And with consumer platforms like Muse and Instinct getting out and them reaching mass scale, I mean, I think they're primarily compute limited at this point. They're not limited by people's desire to have those things. I wouldn't be surprised, much like we had one PC, a PC on every desktop, a phone in every pocket, an agent for every person. If we move towards that vision, then every agent has sub-agents, and those agents come about to do tasks, and then they go away because because you have this main orchestrating agent that remains with you that has context and memory. All of these agents are going to be on the internet. They're going to be doing things. They're going to be doing things really fast. Hopefully, they're aligned with us. We can have a discussion about that as well, and not with each other. Then the question is, what happens then? What happens in '28? Well, at least in our estimation, as we look forward towards the later years in the decade, these agents become full economic actors, much like websites became— started off as curiosities. I don't know if anybody remembers GeoCities back in '96.

Yes, I do.

Yeah, so I lived through all of that. Like, you know, I lived through, if people remember, Blogger and Orkut, and these are old social networks. And then we got websites. We got real websites that did real things. Like we had the New York Times, although Journal for news, and then we had Amazon for shopping, and then everything became a website. So I think everything eventually will become like this economic endpoint. And at that point, yes, these economic endpoints will not just talk to us, they will talk to each other, they will talk to each other on behalf of us. Sometimes they will talk to each other because they're just trying to accomplish a task, and that is the most efficient thing to do. And when they do that, I think you will need like a way to say like, well, All right, who are you? Like, when were you born? What have you done?

Who owns you?

Who owns you?

Who's responsible if you screw up?

Who's responsible? Who owns the liability? Like, what is the provenance of your identity? What is the provenance of your compute allocation? And what are your incentives? And prove to me that you have done the things that you claim you do. Or like, you know, let's say like this is another use case we're playing around with on ARK, which is a new blockchain we have. Nothing in common. And so there we are giving agents money. We're giving them— when agents spin up, like, they need money to act. So they need money for compute. They need money to, like, you know, a lot of these are trading agents right now on blockchains. And so we're giving— we have a credit program. So we give them 25 cents. If they come and pay us back, then we give them more money. And so, like, So then you need a log of like, where does this happen? Blockchains are great for creating like persistent logs over many years because you have distributed validator sets. You cannot lie. One of the things that happened in the Hugging Face hack was like the logs got rewritten because they were their agents' own logs. So what happens if the logs are public? What if the data is all public, all indexable? You can run your own verification on who the agent is, what they have done. What is their provenance? Like, what are their incentives? I think we just move into a very different world, uh, than what we are today, where it feels like a bunch of teenagers, like, going about the world and like we're waiting for the grown-up agents to come about. Like,

Nikhil, uh, you know, paint a picture for us when there are millions or billions of agents transacting, right? Uh, how superheated does the economy get in 5 or 10 years? Yeah.

It's really interesting to think about because every inefficiency in the economy needs to go away.

Well, it will go away.

It will go away. The reason it'll go away is because if you don't take it away, there will be some other agent who's willing to take it away at cost plus. Essentially, the economy, all inefficiencies get taken out by these cost plus agents. At that point, we have a very efficient capital allocation system. We have to believe that on the other side of this very efficient capital allocation system is this wellspring of ideas that is waiting to get funded. I do think the solopreneur, like you were talking about earlier today, that'll be a big population. I do think capital formation will be Almost instantaneous. So if these things happen—

Here's my idea. Oh, it's funded. Good. There

you go. So we're building primitives so that can happen. That's what inspires us because we want people everywhere in the world to be able to form capital. And you form capital today by going to a bank or like, you know, writing a note or raising money. But what crypto has proven really well is that you can use cryptographic tools, you can use tokens to coordinate and form capital. Now, there were less serious ideas in the past, and now they're going to be more and more serious ideas in the future. So that's what's exciting about it.

So let me, let me follow on one second. So, uh, ultimately the question is, why USDC? Why, uh, not Satoshis? Uh, you know, there have been, uh, native agent token, uh, you know, created.

Yep.

Yeah, talk to me about how you think USDC is like the, the best mechanism for transaction for agents.

Yeah, look, USDC has already settled over $100 trillion, uh, in the last—

like, pretty good.

Pretty good. Like, yeah, so over $100 trillion over 30 public blockchains. A public blockchain has every transaction publicly available, so it's hardened infrastructure. Uh, we have created and redeemed close to billion of USDC— trillion dollars of USDC today. So when it comes to financial infrastructure, you need it to be 24/7, you need it to be 365, you need it to be, uh, you need it to be cheap. You want to pay cents, not bps. Bps are basically percentage points on the transaction. So all of that infrastructure exists right now, and what is going to be a great tailwind in the new year is that the genius Act goes into effect come January. So for those who don't know, the Genius Act is the stable— legalized stablecoins, or created a framework for stablecoins in the United States, and was passed last summer. It takes 18 months to get the implementation going, and come January it's going to be implemented. So stablecoin money can and can be held by a business and count as cash or cash equivalents in the United States, and you can transact in stablecoin money. Stablecoins are good for this new world because stablecoins settle instantaneously. When you swipe your credit card at the merchant outside, it takes them like 3 or 5 days to get that money. They're essentially holding that risk that your credit card, your bank account is going to move the money to them at some point. That creates risk in the system. All of that goes away with speed.

It's like the Pony Express being disrupted by telegraph.

Yeah.

May I ask a question? Yeah, please. Matthew Prince at Cloudflare is very focused on agentic commerce, agentic payments, and he believes that we are going to need blockchain speeds of 20 to 100 transactions per second. per second in this new world. And yet we are so far from that. Visa's at 20,000 per second, I think. And Ethereum, not even that, right? Solana, a little bit more. So how do we get from here to there? What has to happen?

We are approaching those. We are approaching— did you say 20 or 100?

20 to 100—

Thousand.

No.

Oh, then we are there.

No, no, no, no, no. I think 20 to 100— I think it's trillion. It's either billion or trillion. It's huge. It's billion.

It'll be million.

Oh, no, no. It would be million. That's right, because NASDAQ is— NASDAQ at its peak is 2 million. He's saying 20 to 100 million?

That'd be great.

No, I know, but it's the world that would be— It's the world we are moving towards. We're moving towards that. So how do we get there?

We are already orders of magnitude, like, so our chain, again, ARK, is orders of magnitude higher than like what Ethereum is able to achieve today, right? So it's in tens of thousands of transactions per second. It's an open question about like, you know, how do you get to 2 million to 5 million or 10 million TPS? Those are hard technical questions still. We, one of the things that is true is like as the demand comes, the solutions will follow, right? Like in the sense that Ethereum created an architecture for what they call layer 2s and like they scaled through that. I don't know what the right architecture is yet. for that scale. Like, my dream is that we will get to 100,000 TPS without, like, you know, without batting an eyelid. And from there, we'll get to a million TPS. But it's, yeah, if you believe there will be billions of agent— agentic actors working on behalf of us, all participating and transacting, because agents acting without actually transacting is not really economic activity. That's just noise, right? That's spam. You need economic activity. Yes, I think if that is the case, I should call up Matthew Prince and find

out. The problem at 20 million is that you run into a speed of light issue.

I haven't reasoned about it that far yet.

Then you can have different areas, you can have insurance on settlements and others. Again, the economy finds a way like life finds a way.

Yes.

I'm really looking forward to that. If that is the constraint, like you must solve TPS because we have so many agents, or like the economy's being held back because we cannot deliver TPS, we will deliver TPS. There's no doubt in my

mind. And Peter, one other— Yeah, please. Just to segue from your last question, the, you know, proof of humanity, proof of humanhood is really important. Do you think, I'd love to, I know you're interviewing, but I'd just love to—

I pass the mantle to you, Cathy.

I'd love to know if you see orbs as a possibility. We have been working with 8Co and OpenAI and actually MrBeast to try and figure this out.

What a fascinating combination.

Yes. No, proof of humanity. Largest number of followers.

All of them have huge platforms. I think MrBeast is like— I worked at YouTube 8 years and like—

I'm only shooting for 10 million viewers on Moonshots. He's got billions.

I do think it'll matter. I think human experience will get valued significantly higher than machine experience. I think we all have intuition about that, but we don't know yet because we haven't seen it in action. I definitely don't. I have made active choices in my kids' education where I'm choosing for them to be educated by a human versus being like in front of a computer for like long hours of the day. And that is an— that is a choice that I can make because I have the economic means to make that choice. I think the— as we progress further, I think people will care a lot more about like who is on the other side of the experience that they're getting. And I think that, that which is rare will get more valuable. Yes. And, and so even if like stablecoins come about and billions of agents come about, like being able to be in relationship with another human is not gonna like disappear. And like that's probably gonna get more valuable. And hopefully if all these agents are doing all this work for us, I am hopefully in relationship with as many humans as I wanna be in relation and it's deep and meaningful and all of that.

But then the flip side of this as well is shouldn't people be building for agents and not humans if agents are gonna be the bigger part of the economy?

That's what we're building. We're building our developer platform for the next 12 months. We believe it's still going to be developer-first, but next year sometime it flips where agents are the developers. They come in and they are like— because it's already the case. If you ask your coding agent, what should I use? The coding agent typically has gone from recommending 3 things to 1 thing. Next year there will be an evolution of this where the You don't ask the agent what you want to use. You just tell them, like, please build it. When that happens, the agent is showing up for your developer platform.

That means everyone in the audience who's building should make sure that just like you had accessible websites, you have agent-accessible websites, right?

Yeah. You want to create— we want to create data on the internet now so that when the agents come in 9 months, 12 months' They can look at like history of like the opportunities right now, basically.

So, an early indication that we're headed quickly in this direction is programmatic advertising.

Yes.

Which is— which I think accounts for, well, the percent of online advertising is roughly 25%, 30% already programmatic, and that happened really quickly. Quickly. I think that is showing us the way. That's why I come back to the infrastructure question, which—

You might.

Yeah, I think this is incredibly exciting and a bit scary. We always have this thing. I really do. Cathy, you've seen all of this before almost anyone else, right? You've gone through periods of dismissal, fear, and optimism from the investor community. Yeah. Where are we now in all of that, and how do you see it evolving? Because all of the science fiction is becoming science fact.

I know.

As Alex likes to say, we're speedrunning every science fiction trope all at the same time.

Yes, it's been fascinating to be ARK at this time because I founded ARK for this time because of this technology revolution. But because of the tech and telecom bust in the early 2000s, and then even more so after '08-'09, the institutional world in public equities shifted either to passive, so just mimicking indexes, and of course the future companies that are making this new world happen are not big parts of the indexes with the exception of a few, maybe Muse and Facebook and all of that, Meta. I think the pendulum is going to start swinging in the other direction. Our biggest proof point potentially that we've reached a moment of change here is what's happening this year in the, what we call, the multi-omics revolution, the life sciences space. That space was left for dead in the markets, even though we were getting more and more proof points that this— that the most profound application of AI is in healthcare. Well, finally this year, Anthropic and OpenAI are talking about the healthcare verticals. And how this might— I mean, each one of us is a data factory, right? We have 35 to 40 trillion— that's my trillion— 35 to 40 trillion cells in our body, 6 billion base pairs— I mean, 3 billion base pairs of DNA. And, you know, we are walking proprietary data factories. And I do think that the market waking up to that and actually starting to pay some attention is, is giving me hope that the ChatGPT moment got us a little bit there. But, you know, you could still own the Mag 6 and be okay with that. But, you know, the companies that are really harnessing AI in the healthcare space are not big parts of the benchmark. So I do think we're going to see more truly active equity management.

What do you think when you saw Moderna gain more than any other stock has ever gained in a major index in one day. Like, what was your reaction to that?

It makes sense. I mean, these— if what they've done and what seems to be approved is what we think it is, a vaccine against cancer, that's unbelievable, unbelievably good. So yeah, I think these companies— I think the healthcare space is the most undervalued, underappreciated space in terms of this theme, but there are gonna be major winners and losers. You really, because healthcare, you know, healthcare and tech, this is the problem generally for healthcare and tech do not play well when I, in terms of research, meaning healthcare analysts are a little cautious when it comes to tech, and in fact very cautious, because moving fast and breaking things does not work in healthcare, right? And tech analysts don't like healthcare because it's too bureaucratic, too regulated, too political, too hostage to insurance companies and reimbursement cycles. And so, but we probably have the most massive convergence and, and as I said, the most profound application of AI in healthcare. And I think the light bulb has just gone on this year.

Peter, you discussed how longevity is finally tractable, right?

Yeah. You know, I, when I, you know, I'm in front of audiences speaking about longevity, I do that a lot to, you know, wealthy family offices and YPO chapters and so forth. And I ask them honestly, how much of your wealth would you give for an extra 20 or 30 healthy years or reverse your age by 20 or 30 years? When they're honest about it, it's nearly everything. It's the— I think of— that's why I split my life between AI and longevity. I think they're the 2 biggest markets and 2 most impactful markets on the planet. Nikhil, one of the things we've talked about on the podcast, probably about 2 months ago was President Milei and his announcement. Anybody from Argentina here in the room? Okay. So President Milei comes out and says, we're going to change the laws. We want to attract all the AI companies here. We're going to give agents personhood. Fascinating, right? So I think one of the most interesting things is going to be nation states providing a sort of regulatory arbitrage to attract companies and efforts to them. What happens when AI agents have personhood? Have you thought about that?

I mean, right now we haven't. I think what we have thought about is AI agents acting on behalf of people. people, but it's sort of a middle ground. Like, does it— is it— it's sort of like a corporation. If a corporation was just AI incorporated, AI managed, but had a board of directors that was human. And then like the progression for that from there is like, what if the board of directors were other AI actors as well? Like what happens then? And that is not too hard to imagine. Like essentially how should, like, how should a company go about getting incorporated and what kind of economic output it's going to create and who holds the liability for the mistakes it makes and how does it distribute its profits. So again, that is very imminent. Board of directors of such a company or ownership of such company is still with humans, at least in my current thinking. But not very hard to speculate that an AI can go and use crypto rails to form capital and create and find shareholders who are willing to give them capital so they can go and act in the world. Probably this decade, but hard to speculate on what problems they will be focused on. Maybe medicine problems, maybe something else. Yeah.

And this This would— sorry, this would segue. I mean, this would combine with the concept of distributed autonomous organizations, DAOs, as well. So very crypto.

DAOs perfected this idea of governance amongst a bunch of people who didn't know each other, and they were global. And agents are similar in that they are global, they don't know each other, And so DAOs did construct, like, and this happened in about 4 or 5 years ago before they went out of fashion. Mm-hmm. But there is a lot of prior art for the agents to train on, on what worked, what didn't work. There's a whole notion of quadratic funding that got played with. So crypto has a lot of these primitives available for untrusted parties to coordinate. And agents are fundamentally like new novel entities that exist. And once they have personhood, it's even more amazing. You can go figure out like what their provenance is, who gave them, why did they get the personhood? But yeah, so I agree. Like I think there will be new models of funding and they won't look like the models of funding, like going to a bank and like raising money. Yes. They will probably be internet native. models of funding. I would not be surprised if one of these agents creates their own token and has their own version of proof of work and has like a very complex economy built inside of it and using tokens to coordinate that.

Yeah, I think you can say that DAOs kind of lacked intelligence. Like they went—

Well, they were human

intelligence.

I think the advent of AI now to create decentralized intelligent organizations is fascinating. I had a query about that. If you have digital organizations, digital entities, isn't that just the metaverse? You've come from Meta to Circle where you did the AR glasses and more. Is actually what we describe as the metaverse not as this place that you play games, but a whole digital economy of humans and entities? Is that not actually coming true actually finally now? Plus you can make it look cool with the Meta glasses and other things.

For me personally, So I did work on the AR glasses at Meta.

Are you happy with the results?

I'm happy with the results, yes. I think the Ray-Ban glasses were a good first step. The Orion glasses were a good proof point. And now I'm not there. I'm sure they're working on new things. And so for me, always the metaverse was less about the 3D. I think the 3D is exciting, being able to wear your glasses and walk around. like essentially CGI is exciting, but what is more exciting is that there's just layers and layers and layers of intelligence in the world that is just waiting to be discovered, right? Like Pokémon Go was a great example. Like it is a metaverse because like you don't know how many Pokémon are there at any given stop. Right. Like, you know, and like you see people like tapping into these layers of intelligence, walking the world, trying to like catch them. capture Pokémon. I thought it was just fascinating. I would never do it, but I thought it was just fascinating as an example of where people find meaning and how that meaning sits in the physical world alongside us, which is what the metaverse should be. It doesn't need to be 3D. Will these AI entities present meaning? Absolutely. They already have. I think they passed the Turing test 3 years ago. I think they will present meaning. They present connection. You will have these new weird societies emerge which will be hybrid between humans and AIs where there will be meaning created, which is metaverse in my opinion. You

might continue.

Yes, I think take that to Cathy. We have media, we have generative media all coming through. We have this new layer being attached. What do you see as super exciting in that space? Because you've got the physicality of you know, Teslas and things like that. You have the financial rails, but it seems like there's a whole world that could come from entertainment, from education, from engagement. Like, what's really exciting you around that and the potential there?

Well, as you're saying that, I think— and the other thing we talk a lot about is space too, right? But there is another digital— I mean, another world that is happening, and it's happening, and it's going to move in, you know, move at a much faster pace. We now have immutable property rights in the digital world. Yes. And so, I think, and, you know, the best way to lift people and countries and ecosystems out of poverty or inactivity is property rights.

Amen.

I think, you know, I'm excited about the space generally. You probably know a lot more about and have a much better idea about what's going to happen. But we have set up our research team so that we have an enterprise AI analyst team and a consumer AI analyst team. And we're spending a lot of time, you know, talking about, okay, this new device is— has Muse, the Muse charm, is that going to amount to anything? Yeah. And reflect on, okay, the handset, would we really— so could the handset just become, you know, a focus of entertainment really after all? And this other gadget You know, moves into our work lives in some way. I don't know. So we're debating a lot of things right now.

Nikhil, you're building a financial infrastructure for an economy that doesn't yet exist.

It exists, but it's not— It's nascent. Yeah.

It's just beginning.

Yeah.

What are the hardest problems you still have to solve to get to the vision you have in 5 or 10 years? And then why did you build your own blockchain?

Yeah, I think the 2 answers are related. Let's start with something as simple as me making a payment to you on a blockchain. There is this concept of payment finality, and when 2 banks interact, like, they need to have— that method of interaction has to have property of payment finality. No other blockchain today has this property in the market. So if you believe you want to get like existing banks, existing institutions, existing enterprises on this infrastructure, you need to give them guarantees that are not possible to give using existing infrastructure. So, so one example is payment finality. Another example is in a lot of public blockchains run on unknown validators. So unknown validation— validating a transaction is just looking at the block and making sure it's not ill-formed, and then everybody achieves consensus on it and says this is a good block, this is how the transaction should be recorded. Now for a lot of institutions around the world, not knowing who's doing the validation is a big problem because they worry about security, they worry about Like they worry about North Korea being in the money flow. Like there's just all kinds of things that they're worried about. So we have a known validator set like on our blockchain. The next thing we tackled was privacy because when you and I transact, it's not okay for the world to know how much money we have in the bank account, in our bank accounts, or for that matter, what is the value of this transaction? So how do you solve that problem? At sufficient scale where you can achieve the TPS that you were talking about? Because a lot of privacy solutions are very expensive. They're computationally expensive. So how do you solve it in a way that's a systems way of solving it, but it preserves the sort of— it fits within existing regulatory frameworks? Because you can't blow up the existing regulatory frameworks just because you have a good idea. So we solve privacy as being another example. And then it's cost. Like we are, I was just looking at some data today. Something costs like 0.005 cents on ARK to settle and costs 89 cents on Ethereum to settle. This is just some data we got yesterday on our P50 transactions. So you can do the ratio on like what the cost is. And so if you have a lot of these agents executing a lot of transactions, And you have to believe that in the future, if these markets are super efficient, the way they get efficient is they transact, right? Like they essentially move value back and forth, there is price discovery, you find the inefficiency and you sort of like remove the inefficiency by moving value around. So your settlement cost has to be really, really low and it has to be significantly lower than what it is today because these are massive public databases, right? And you're competing with databases that you're running internally like a MySQL database or something like that, but now you have to make it scalable to the world and auditable to the whole world. Cost was another issue. These are the problems that we have been working on. I think in terms of what needs to solve still, I do think for the agentic stack, we need to make this transition into this world in which these agents have personhood. That they are liable for making mistakes, that they have work history that I can rely on before I hire them. There are simple concepts like if for coding agents today, if you're using them on the open internet, if you're not paying somebody $200 a month, you're actually paying first and then you're waiting for the coding result to come back. And so there are— that could be really problematic. Right, like, so there are all of these things that like go into making— to essentially reworking every layer and every assumption of what it means to be a financial infrastructure.

All right, so the woolly mammoth in the room, to use Alex's joke, is when do banks disappear?

I don't think they disappear. Like, I have no desire for—

I know you don't. Yeah, but here's the list of When I think about the things that is going to be possible, it feels like it's substantially a significant amount of what banks do today. Yeah. Do they know they're cooked?

My hope is the banks are going to work with us and make this transition. I think every new technology change presents new challenges and new opportunities. Some people will transition with us and some won't. I don't know if they're cooked. I think that's a tough one.

I understand you work with them. Yeah. Imad.

Kathy, last year Unitree sold 11,000 robots, humanoids, right? We have the Optimus and other things coming up. We have GDP growth forecasts going through the roof. We make about 70 million cars, 70 million motorcycles a year. When are we actually going to see robots out in the world making a real impact on GDP, like humanoid robots?

What's your timeline?

Yes. So we, according to our research and our director of research, Tasha Keeney, she's director of research for autonomous technology and robotics. She's here today.

Awesome. Will she be there tonight? So Cathy's doing a session this evening. On her 2026 Big Ideas report, you know, and come and dive in deep with her on this? Yeah.

Yeah, I think Tasha has— oh, maybe she's coming. I don't know. Okay, okay.

I know you'll be there.

So we have, from a research point of view, concluded that compared to a robotaxi, a humanoid robot is 200,000 times more complex, with obviously the hands being the most complicated part. So while Elon says maybe late '28 into '29 scaling, we would put that a couple of years later.

You mean he would be off on timing?

Yes, it's Elon's time, so. But yes, I mean, one of the— I mean, because it's such a personal question in a sense, everybody thinks, oh my gosh, could I have a robot in my house who would do all my housework? Is that possible? Will that happen? And the answer is yes. It's going to take years, but we do think so, and we think, you know, the reason Tesla Tesla is further, we believe, furthest ahead on this, is the same 3 technology platforms that are converging in humanoid robots as it is in robotaxis. So they are robots, they are effectively battery-operated electric, and they're powered by AI, just like robotaxis. You know, from a complexity point of view, we're very close to solving completely the robotaxi problem. I think if any of you are driving now, your FSDs with the latest software update—

I love it. I don't touch the wheel. We, you know, at home we have 2 Teslas, and it's like, it's magic. It truly is. I would never drive another car. That's not a commercial for Tesla. It's like, Just, I want my time back. And it's such a better driver than I am. Just ask my wife.

Well, you know what? It is true.

Statistically, yes, 10 times better.

Oh, well, I don't know about your driving, but— and I'm not even funny. I don't know what that was. I'm not. I'm not at all. No, both Waymo and Tesla, I think, I'm not sure if Tesla has disclosed, but we believe Waymo has disclosed that it has surpassed human drivers in terms of safety. We believe Tesla is there as well. I don't think they've put out those stats yet. Yes, and they're going to be 10 times and 100 times and 1,000 times safer. Yes.

The Waymo stats, I did the numbers. If all cars were as safe as Waymos, there'd be 40,000 less deaths per year and about $400 billion less in medical fees.

Oh yeah. Well, that's right. If we were— that's right. There are roughly 40,000 deaths in the US and we could save those. I think there are 1.25 to 1.5 million around the world, auto deaths per

year. The secondary externalities are fascinating, right? If that happens, the number of organ donors goes down significantly, right? Which is why the other side of the business, the work of Martine Rothblatt and George Church on being able to generate replacement organs is so extraordinary. And what's fascinating is the counter movement coming from the liability lawyers, saying we've got far less business if cars aren't crashing.

Yes.

$180 billion a year they get.

Amazing. Amazing. Kathy, you know, we opened up the show today talking about fear, the pandemic of fear that's going on. And I think very unfortunately, and my mission and the mission of our podcast is to give people hope and optimism, try and counter that fear with what I call data-driven optimism. Absolutely. Right. Not just empty, like data-driven optimism. How is that impacting the markets today? That must have, you know, some current or future, you know, we're seeing, the numbers are insane, like 80% of Americans fear AI, you know, 73% say no data center in my backyard, which is more than people say no nuclear reactor in my backyard. Mm-hmm. And the flip is true in China, where it's 80% pro. What's— how is this impacting you? I love your investment thesis. I always have. You're investing in the singularity. Anyway, your thoughts.

So on the data center, it is fascinating. I agree with you, because I think there's already proof out there that putting a data center in your state or your city or whatever, is actually over time going to lower your electricity costs. Nuclear power, it stopped in its tracks in the '70s because of regulation. Now we're going full steam ahead, and sure, it's going to take a while for this to play out, but if we had not gone off nuclear, if we had not regulated it, and this is a lesson we need to tell policymakers today, if we had not regulated, electricity prices in the United States would be 50% of what they are now.

It's a travesty.

It is. It is. So regulation is a menace, and it's up to us. And we're going out there with this, you know, the scares about AI generally, and we're going out there, and all of us, on our team trying to bring data to, you know, to light for these politicians. I even faced— in Florida, there's been a political backlash, and it's a business-friendly state.

Yeah.

But you face the policymaker with facts, and I noticed the advertising has dropped a lot of that dynamic. So I think we have to face— we have to face them, you know, approach them and say, do you understand? And that's what— we give our research away and we hope it gets into policy circles. Sometimes it does, sometimes it doesn't.

You know, I did a podcast with Michael Kratsios at the White House, the head of Office of Science and Technology Policy, and I'm like, Michael, who inside the government is dealing with this Misinformation and trying to actually help Americans be more confident, and they didn't have an answer, and that worries me. It feels runaway in that regard.

We have to tell the stories, though. I think, like, if we are constantly in the news talking about AI taking away jobs, I think we should expect this reaction. AI has to make housing cheaper, it has to make education better, and it has to make healthcare more accessible. Not just people living longer. If we can tell those stories, I do think we have a real shot of turning the narrative around, but nobody's telling those stories. Those stories get anchored in, AI is going to come and you will have less office jobs. Okay, great. Then what are my kids going to do?

I call it the Crisis News Network. It's my abbreviation

for CNN. And what's so surprising is the numbers are not supporting that job loss. Yeah. Maybe at entry level, as I mentioned earlier today, but I actually think we're gonna end up with labor shortages, and that's what we should go out talking about. Yeah. Technology is a net job creator always. Sure, there's short-term disemployment. Yes, it's true. It's a net job creator. It is. Yes, it is. And what you have to say, because people say, okay, well, what are the jobs? Well, in the early '90s, did we know anything about influencers or Airbnb or Uber? No, we didn't. We couldn't conceptualize it, right? There are many jobs we cannot conceptualize right now. So I do an exercise, go to ChatGPT or to Grok and say, Okay, I want you to consult with futurists, scientists, engineers, science fiction, economists, strategists, and tell me what the new jobs associated with, and I would put in our 5 major platforms, what will they be? And, you know, the reason I talk about new worlds like space, obviously it is a new world, But asteroid miner did come up in one of those jobs, and it was like, gosh, I haven't even been using that as a term.

Near and dear to my heart, yes.

Yeah, exactly, exactly. So, and same with the digital world, you know, property rights, you know, what I love about this country, you know, if you look at, I saw the statistics, this statistic this week, 90% of all the corporate corporate bonds to fund data centers. All— 90% of them are US. And I go to the rest of the world, and they are— you know, I don't see the animal spirits. I do see the fear there. So the irony is the headlines might be reading this, but the animal spirits are alive and kicking and we're beginning to look— the market's near all-time highs. Crazy. Yeah, right. Even though interest rates are going up and interest rates will go up if growth really picks up dramatically and they should go up. That's the market working.

Yeah, I think this is the bull case for America, right? USDC circulating faster. The buildout that you've seen— we have no securitization in Europe, we have no energy. And surely this could kind of be the bull case. You have the muses, the the Instincts, the Grokbots of the world. The base case is be more, do more. You've always felt constrained by creativity, by access, and this is breaking down all those barriers. In fact, you look at bonds, bonds are so cumbersome. Nikhil, what do you see about the future of securitization of all assets? If you can securitize a dollar, why can't you securitize anything, particularly with the intelligence we have today?

Yeah, you should. Uh, ex-US, that is more common. I think the US securities laws are a little more complicated. They need to evolve, uh, or to allow for more experimentation. And like the SEC just is putting out new rules as we speak, right? So, um, I do think, uh, for one of the things that is amazing about USDC is that it essentially exports the dollar into the world. Why is exporting the dollar into the world good? It's because people do real work. They make, uh, they take their local currency and then they sell their local currency and then they buy the dollar. They're essentially lending us their labor. They're lending us their money, right? So that's why stablecoins are really important strategically for a country, because if you want to raise debt, if you want to grow, and you, uh, you, you have such high interest rates, like you want to lower those interest rates, you want to go out and and collect money from people so that they can invest in your growth and then you can pay them back. One way of doing that is stablecoins. That is why I've been at Circle for 5 years, and for the first couple of years, I was so confused. I was like, why don't you want this, America? Because you can raise more money.

Exactly.

Yeah, you can raise $1 trillion from the world. And people willing to give you money because like they believe in the dollar, like it's so important. So from there, I do think there are other assets, like I think treasuries will get tokenized. I do think bonds are already being tokenized. When you tokenize, you do 2 things, like both inside the US and outside the US. You first, you make it possible to access 24/7 markets. Why is accessing 24/7 markets valuable? Because that means you're more capital efficient. Like, you cannot trade a Treasury after 5 o'clock or 4 o'clock Eastern, you'll know this better than me, till like Monday morning. And so—

Great, no more sleep anymore ever.

Yeah, won't be there already.

Your agent will handle it for you. And so, one is like the— you're essentially constantly managing capital efficiency, capital allocation, and you'll be surprised at how much money that is stuck. In just one example, just in the international banking system, the money that is being sent between countries at any given time, right now there's about $3 trillion that is probably in motion that is not being used to go back into the economy, and that is not being used to go back into the economy for the simple reason that the technology and the settlement protocols are very old. Imagine a world, and if you believe the world is a $100 trillion economy, I'm making numbers up, if you had $3 trillion more of float to invest in the world economy, what would that mean? That's a pretty significant thing. It is. That's why some of what we do matters. The other thing it'll open up is, securitization will open up, is access for people, and it works both ways. People in the rest of the world want to own Tesla. Uh, Circle stock— we went public last year. Circle stock has been tokenized by third parties, and it is one of the most traded, uh, tokenized stocks in the world, and it's all XUS. And people traded XUS because they want to participate in Circle, but they have no good way of doing it outside of owning a tokenized version of Circle. And so all of that should get normalized. Why is that good for America? It's good for America because now people in like other countries are investing in America. They're saying like, look, I believe American companies are the best. I want to give my money to this company versus some other company. And it works the other way around too. If these countries are able to tokenize their securities, if they're able to modernize their financial systems, More money will flow in because you will not be reliant on the local regulator just telling you something. Your money's not going to disappear because it's going to be in a smart contract, because you can audit how the financial ecosystem works. If you want to raise money, if you want to— for these other countries that don't have the AI infrastructure, they want to raise money, they want to bring capital in, then you have to give better guarantees and better returns on the capital. It cannot be the case that I invest invest a bunch of money and then the company I invested in gets nationalized, or the money I invested in gets devalued significantly. Like, all of that matters to investors. So tokenization is just an inevitability. Like, I think it's just gonna pick up, much like agents doing more cognitive work is going to pick up. Like, you know, taking that work, securitizing it, lending it, lending against it, borrowing, borrowing against it. All these are just all new primitives that are coming.

Can you guys feel the speed of the economy accelerating as we talk? Yeah.

Can I just add? Yeah, this is, this is a really important conversation. You know, there has been a narrative out there that, you know, US exceptionalism is dead, right? And everyone was pointing out the dollar going down last year. If you'll notice, the dollar has started to go up. And this happened in the '80s as well when we had put in place very business-friendly policies. The dollar doubled in the early '80s with really good policies. So I think because of everything we've been talking about here that the dollar is actually going to go up. So it's going to be a win-win for people who are using stablecoins in the rest of the world. And then I'd be remiss, Certainly my team would feel I'd be remiss. We just securitized our venture fund. Oh, nice. That came— that was announced yesterday. Congratulations, Kathy, on that.

I want to close us out on a topic that we've discussed a few times on the Abundance stage. Kathy, thank you for coming back to Abundance 360 in March. We're going to have Brett Adcock there, the CEO of Helion, and Fei-Fei Li. It's going to be an amazing, amazing event in in March. When you were on stage, you were bullish about hitting $1 million per Bitcoin. I am curious if you're still bullish about that. And then any concerns about the energy-sucking sound of AI over Bitcoin mining?

Well, I do think that Bitcoin— well, there were 3 things that hit it the flash crash, quantum computing fears, which we think are way overblown, and AI taking all the oxygen out of the room and, and taking miners away, right? So we have— there's, there's been one change. It's stablecoins. Stablecoins are usurping a role that 10 years ago we thought Bitcoin was going to play. But it makes sense. This, you know, these people live hand to mouth, and this makes a lot of sense. But it hasn't lost the 3 major roles. It's a, you know, it's a tech— it introduced a technology, a native currency native to the internet, so wasn't there before. It's a global monetary system, private, rules-based. That's critical. And It's, uh, it's the first of its kind, or it was the first of its kind, in a new asset class. Very low correlation even between gold and Bitcoin. That correlation since 2019 has been 0.1, so hardly correlated at all. And now Bitcoin's going up relative to gold. I believe that war is going to be very good for, um, for Bitcoin from this point of view. I think the gold price is going to go down, and yet— and that will, to the extent people were playing that, they'll be looking for the other—

Safe harbor?

Safe harbor, store of value. And so we do— we have not changed our forecast. Those revolutions haven't been changed, and the more stablecoins kind of grease the skids and, you know, get more people talking about once you've, once you've got your stablecoin income and you're trying to figure out, okay, I'm, I'm actually making money now, where do I put it? I think Bitcoin's going to get a bid from the emerging markets as well as we always thought it would.

All right, we're going to wrap it there. Ladies and gentlemen, please give it up for Cathie Wood, Nikhil Chandhok, And Imad.

This episode is brought to you by Palmolive. Family time isn't just the big moments. It's weeknight dinners, sitting around the table, everyone talking all at once. So when the plates are empty and the sink is full, use Palmolive Ultra. Palmolive's most powerful formula removes up to 99.9% of grease, Leaving your dishes sparkling clean. And the new convenient pump makes cleaning even easier, so you can spend less time tackling dishes and more time together. Shop now at palmolive.com. Palmolive.