
All-In with Chamath, Jason, Sacks & Friedberg
Dueling Presidential interviews, SpaceX’s big catch, Robotaxis, Uber buying Expedia?, Nuclear NIMBY
Fri, 18 Oct 2024
(0:00) Bestie intros (2:01) Polls vs Prediction markets, dueling interviews, election update (16:06) Tesla's Robotaxi event and SpaceX's Starship catch (27:36) Uber reportedly looking into acquiring Expedia (45:19) Nuclear Vibe Shift? Big tech is looking toward nuclear solutions to power AI (1:11:10) Lawfare from the California Coastal Commission Follow the besties: https://x.com/chamath https://x.com/Jason https://x.com/DavidSacks https://x.com/friedberg Follow on X: https://x.com/theallinpod Follow on Instagram: https://www.instagram.com/theallinpod Follow on TikTok: https://www.tiktok.com/@theallinpod Follow on LinkedIn: https://www.linkedin.com/company/allinpod Intro Music Credit: https://rb.gy/tppkzl https://x.com/yung_spielburg Intro Video Credit: https://x.com/TheZachEffect Referenced in the show: https://polymarket.com/event/presidential-election-winner-2024?tid=1729285428575 https://x.com/elonmusk/status/1846826782797799580 https://x.com/collinrugg/status/1845472475322462468 https://x.com/SawyerMerritt/status/1839424008900477354 https://www.ft.com/content/94a25bf7-e62b-462a-a4f0-e4feb6e244f7 https://www.google.com/finance/quote/EXPE:NASDAQ https://companiesmarketcap.com/expedia/revenue https://x.com/Jason/status/1847016512583786921 https://www.cnbc.com/2024/10/16/amazon-goes-nuclear-investing-more-than-500-million-to-develop-small-module-reactors.html https://www.cnbc.com/2024/10/14/google-inks-deal-with-nuclear-company-as-data-center-power-demand-surges.html https://www.cnbc.com/2024/09/20/constellation-energy-to-restart-three-mile-island-and-sell-the-power-to-microsoft.html https://www.politico.com/news/2024/10/16/california-coastal-commission-elon-musk-00184017
Freeberg's channeling Tim Walls over there. I know.
Wow. He's as exciting as Tim Walls.
Got your flannel on. Do you know what a venture capitalist is, Freeberg? Oh, he's shilling Supergut.
Well, as of last week, when J. Cal decided to turn All In into a commercial, I was actually going to do a Supergut background. We're launching Supergut nationwide in Target this week. Any Target in the United States, you can go into and pick up Supergut. You can buy... The GLP-1 booster. You can buy the prebiotic shake. I have that, actually. Is that the chocolate? Or do you have the mocha?
It is the chocolate. I mean, I like the mocha. This one's chocolate. Okay. Mocha's good, too. All right.
Let's get started. Well, thanks for the support, J. Kel. I appreciate it.
Of course, of course, of course.
We're cutting all this out.
No way.
This is why I do this. Next time, plug a company I have a stake in.
Let your winners ride.
Rain Man, David Sasson.
Also, All In election night live stream is coming November 5th. You can watch live. Saks will be hosting. We're doing it. You're hosting it. Your team said you're doing it. And so you'll either get to see Saks.
Are you not going to Mar-a-Lago, Saks?
Well, if things continue to look good for Trump, I might go to Mar-a-Lago. I might.
Yeah. Okay, so you're maybe. Let's not commit sex. You're maybe. If you go to Mar-a-Lago, you're excused.
I could live stream from Mar-a-Lago.
Oh, that would be amazing. Absolutely.
That would be amazing. I'll go to Mar-a-Lago. That would be fun. Yeah, if it looks good, I'll go.
Of course I'm invited. I talked to Jared. If things look as good as they do right now, then I think I'm going to have to go to Mar-a-Lago.
I think we should all be in Mar-a-Lago.
It's going to be a unique experience.
Oh my God. Can you imagine being in Mar-a-Lago and he loses?
Oh my God. That would be dark. Unless this thing is in the bag. It's got to be too big to rig. If it's too big to rig, I'm going to Mar-a-Lago. Too big to rig.
Do you guys think polymarket is like, why do you think it's different from the polls? Are we talking about this today? Polymarket's showing like 60, 40 or 65, 35 now, right? Yeah, because they're measuring different things.
I've explained this before. Polymarket is people betting on the outcome. So 58% think Trump's going to win. Whereas the polls in a particular state show the percentage of how each person's going to vote. Hmm. So if for sure you knew the election was 51-49, the betting markets would swing to 100-0.
But let me ask you this. So Nate Silver's model, which takes the poll from each state and builds in a kind of a Monte Carlo of super poll, like a super model for the whole country. Why is his estimate 50-50 right now while the poly market is betting at 60-40?
It's possible he's laggy in his estimates. The betting markets seem to go based on momentum. It indicates the swing in momentum.
How do you think they're going to change after the interviews the last couple of days? Trump on Bloomberg and Kamala on Fox. Do you think those are going to change anything? I don't think so. I think it's all baked in now.
Well, Trump over the past few weeks seems to have had a surge owing to the fact that Kamala's interviews generally don't go well. So I think she started off a little behind, started doing interviews to catch up, and now she's a lot behind. I don't think the Bear interview is going to help her.
Well, let me ask you this. So my observation as, I don't know, I'm not like a super political person or whatever, a party-oriented person. I looked at a lot of the media on both sides, and it seems like everyone on the left says Kamala did an amazing job on Fox. She defended herself. She showed her skills and her competency. And then everyone on the right is like she embarrassed herself.
She fell apart. And then the same thing happened with the Trump interview on Bloomberg. People are like, On the left, they say, look at how he couldn't handle the interviewer and he fell apart and all his lies were exposed. And everyone on the right's like, look at him, he got a standing ovation.
It's almost like everyone's just kind of like self-asserting their beliefs that they already hold when they judge these people on these interview shows at this point. Is it already baked at this point? Like, is anyone actually going to change their view? based on these interviews happening?
Well, the question is what appeals to that small sliver of independence.
Yeah.
The question I would ask back to you is, if the Brett Baier interview was going so well for Kamala, why was her staff on the sidelines waving to try and end the interview? Apparently, they had like four people waving and trying to cut the interview off. Who said that was the case? He did. Brett Baier said it. So it was like in Rocky IV,
when Apollo Creed's corner is like yelling, throw in the damn towel, throw in the damn towel. They couldn't wait to get off the stage after 26 minutes. I just think that if it was going that great. Allegedly, allegedly.
Yeah.
I don't think Brett Barrett's going to lie about that. I don't know why he would lie about that. Why would they get her off the stage after 26 minutes if it was going so great? I'm not saying it went as horrible as some of the partisans on the other side are saying, but I don't think it went that great.
Do you give her any credit for going into the lion's den like she did?
Well, I think that she went, she did the interview precisely to get the talking point that she does adversarial interviews because that talking point was earning them. And so you saw like all of her fans in the media were saying, well, see, she can walk into the lion's den. But again, she did the shortest interview possible. I don't think she answered the questions directly.
I think she filibustered a lot. She deflected a lot. I don't think she was particularly persuasive. I don't think she convinced anybody. So I think that what you saw there was somebody who just wanted to get it over with as quickly as possible to check the box on, okay, does adversarial interviews. Trump, on the other hand, he actually likes doing these things. The Bloomberg interview.
There's no filibustering there, right? Well, no, he's crisp in his answers.
It's all filibustered. Come on. It's all anecdotes. He can do the weave. He can do the anecdotes. But he's also very good at coming back on the interviewer when they get adversarial. And the audience was with him. They gave him a standing ovation. He went for 64 minutes compared to her 26. I just think there's no comparison. I think Trump is someone who relishes.
walking into lines and doing those interviews. I think Harris did it because she felt like she had to.
What do you think, Chamath? Do you see it or no?
Do you have any opinion? I watched the whole interview. It was clear in the interview, he mentioned the fact that he was being waved off. And then he said it after the fact as well. That's not alleged. I think that that did happen. I would say two things. I thought that she was composed. And she maintained her cool. So I think from a stylistic perspective, I thought that she did well.
From a substance perspective, it was pretty lacking. Because if you actually listened to the answers, there was just a ton of non-answers. And they were two very basic questions that I think a lot of people, even if you're not a swing voter, I think would probably want to know the answer to. Meaning, Did she have any regrets about what's happened in the last three and a half years?
Did she have any regrets about what she's done on the border? Has she not noticed that Biden was wavering before he was hot swapped? I think that you could have predicted that these questions were going to come. So I think I was surprised that there wasn't a crisp answer that they had practiced for that. The second thing I'll say is then David is right.
Everybody then gets very tribal in how they interpret it. I think I saw one tweet from Elon about how all of the newspapers characterized her interview with Brett Baier as quote unquote testy. And it was sort of like that was the way that the mainstream media framed it.
I suspect if somebody looked at how Trump's interview with Bloomberg was analyzed, it probably had some similar verbiage that was repeated there as well. So I think you are right, Jason, that the mainstream media can't be trusted to tell the truth. I would just encourage people to watch it. I think, like I said, stylistically, I think she did well and remained composed.
Substantively, I think it was non-existent.
Yeah, it would have been nice to have another debate between these two.
She still can't really explain how she's different than Joe Biden, other than the fact that he's a white male and she's a woman of color. So beyond just sort of the superficial differences, she can't explain on a policy level what she would do differently. She's had so many opportunities to say that. They asked her on The View. They asked her on Stephen Colbert. Brett Baer asked her in his way.
And she still can't explain what she would do differently. And I think that is the fundamental problem she has in her campaign is voters still don't know who she is or what she would do.
Yeah. What did you think of J.D. Vance saying he wouldn't have certified the election? They seem to be going after him on that over and over again.
You're the only person talking about that.
No, no. Literally every interview, they've been chasing him down the hall asking him. He's like, I'm not the only person. I may have started it, but what did you think of him saying he wouldn't have started it?
That is kind of like when he's in a combative reporting moment, that is the question he gets a lot.
That's not the interview I saw. I saw the interview he just did with Martha Raddatz was she was saying that Trump was exaggerating.
No, no, but the question I asked you, Sachs, was... ISU sacks about him saying he wouldn't certify January 6th.
You're the only one who's fixated on it. No one who is persuadable who doesn't have TDS cares about that topic anymore.
What do you think, Friedberg, about him saying he wouldn't certify on January 6th?
It's not what they're asking, J.D. If you want to talk about interviews that J.D. Vance has done, talk about the one that's actually going viral right now. And that was the interview he did with Martha Raddatz, where she starts saying that, you know, we've only had a few of these apartment buildings taken over by foreign gangs. And he's like, do you realize what you're saying?
You know, there's no comeback from that. He destroyed her. It was very compelling what he did. And every interview he does is like that.
Yeah. I mean, she was basically saying that she spoke to, what was it, the city manager, and she's like, he said only a handful of buildings have been taken over. And J.D. Vance was like, what do you mean? Yeah. Only a handful of buildings? Like, isn't that anything more than zero, like, too much? Or anything more than one is obviously a problem? Like, it was just such an obvious...
rebuttal to the narrative that they're kind of over-exaggerating a particular issue. I have no data on this, but he was very compelling in that response. I thought it was pretty strong. But I will say, generally, neither candidate seems to be introducing a new message or seems to be introducing new content.
They're just kind of standing up, you know, kind of repeating things that they've said, showing that they can handle and manage different kind of combative reporting tactics. And that's kind of what's going on. And everyone seems to have made up their mind. I see a lot of people on both sides say, Again, this side, this person did great.
My person did great against this combative reporter and the other person did poorly against their combative reporter. And everyone's kind of biased in their view. It just feels like this election is baked and we should just go to the polls and be done.
Yeah.
What did you think, Freberg, though? But there's no October surprise coming out, right, Sachs, Chamath, Jake? Oh, still three weeks. Yeah, anything can happen. But there hasn't been anything, right? Like that's kind of... a shocking moment yet this month.
But Freeberg, the question I was going to ask you is, since you're not like hosting Trump, you know, fundraisers, do you think, what did you think when J.D. Vance said he didn't think that Trump lost the 2020 election? Does that concern you at all?
There's no way to answer this with the kind of clean framing I think you're looking for.
What I saw from JD is that he wants the reporter and the people that he's talking to, and I hear this from him, to zoom out a little bit and recognize that there are significant control systems and biases that he believes and others believe are strongly affecting the election process and as a result, the election outcome.
And I think that that message is lost because people want him to say, Trump lost the election, you're not admitting it, you're bad. But those people also aren't hearing the point that he's making, which is that there are biases.
And we heard these biases, by the way, with Democrats in prior elections as well, where they highlighted that they believe that there were biases with respect to misinformation being amplified on social media. And then the next election cycle, they were able to step in and influence what was being changed on those social media platforms.
And so there's this big kind of war, media war going on through social media platforms. And I think that that's what both sides are highlighting is their big concern. And now there's this other big concern about is there appropriate voter verification that the people who are voting? And it's a question to ask that shouldn't be dismissed. It is a good question to ask. Yeah, that's a great question.
As a person who doesn't have a strong bias for a political party here, I feel like I want to hear answers to those questions. What is the structure of how the way that most people are getting their media today, which is through social media platforms? What is the mechanism for censorship? What is the mechanism for filtering for moderation and be public and transparent about it?
And then separately, what are the mechanisms for for deciding who gets to vote and how they get to vote. And I think those are both really good things to ask.
I would just like to take a step back and say that that was one of the most incredible answers I've ever heard, Freeberg. Unfortunately, it may not land for the reductive masses, but it was exceptionally powerful and thoughtful. Thank you.
That's what I'm here for, Chamath. I'm here for you. Well, I mean, independent of who wins, we need to get this rules of elections really tight starting next year, I think. Make it a federal holiday. Require people to have ID. That doesn't seem like such a big deal. I don't know. Sacks, what else should happen?
federal holiday, make it- Well, right now, you've got Biden's DOJ is literally suing the state of Virginia, which is required by Virginia law to clean the voter rolls of illegal immigrants. And they've been doing that. And Biden's DOJ has sued to stop that. In California, like you said, we now have a new law signed by Gavin Newsom to make it illegal to ask for voter ID.
So Democrats seem to be undermining the integrity of elections, not fortifying it. So when you ask, you know, why do Republicans distrust elections? Maybe it has something to do with the way that Democrats are acting. But I agree with you. I think that cleaning up the voter rolls, having a minimum standard for voter verification, is something that I think should be done.
According to the Constitution, the states basically run their own elections, but it doesn't make sense to me that in a one-party machine politics state, where basically one party controls the state, that they could set up a system that effectively entrenches their power forever in federal elections.
It just seems to me that the federal government has a compelling interest that must be constitutional in ensuring a minimum standard of honesty in federal elections. So I think it would be great to do something about this next year. I think that if you want people to stop questioning elections or engaging in election denial, you need to make the elections above reproach. So let's do that.
So anyway, Heritage Foundation, which is obviously right-leaning, has a bunch of election fraud cases they've been documenting, and they basically cannot come up with actual evidence that this is changing any election results, but we should make it above reproach. I agree. All right. Our boy Elon had a big week. Tesla unveiled two new concepts at its WeRobot event.
And Elon caught a 23-story rocket, the Starship. Here's the robo-taxi and the robo-bus. Both of them look really awesome. And he caught one of the, I think this is the fifth Starship or the fourth launch? Fifth.
The fifth, right?
So f***ing incredible. Look at this. It's so incredible. It's like chopsticks catching a 23-story building.
Forget about whatever your f***ing issues are with Elon and his politics. Just to appreciate, and we can talk about why this is so important in this segment. But technically, the achievement of this skyscraper falling out of the sky and perfectly aligning itself to go into that chopstick-catching device, it is an absolute marvel of human ingenuity.
And the work and the effort that people put into this over several decades, it's just such an incredible feat. Look at this thing. I don't know if you guys were as emotionally moved by this as I was. I thought it was incredible.
It was incredible. I think I probably watched this 100 times. Totally. From every angle. Every angle.
And so the reason this is so important is because these things cost a lot of money. And when they land here, you can clean them up. And I guess his goal is to have them take off again after he fills them with propellant an hour later, Friedberg. So on a science basis, this is extraordinary. If this works and you can start lifting these rockets.
To be more specific, you don't want it to have feet. A, it's heavy. And then B, you have to lift them up in a way that just complicates the entire refueling and cycle time process. So by catching it, you put it right back into place and just go again. Unbelievable. Unbelievable.
Right, you just catch it and go again. I can kind of walk through these numbers. So... Obviously, the big objective over time is how cheap can you get it to put material into space, we need a lot of material to go into space, if we're going to do things in space, particularly if we're going to go build a colony on Mars.
And so this shows you over time, the cost per kilogram, which is the key metric in this industry to launch material into low earth orbit. And you can see here how SpaceX has dramatically reduced the cost. I remember when the small sat era began in the 2010s. Do you guys remember all these startups that were starting to build like little small sats and put them up to do imaging and comms and stuff?
When this took off, It was about 10,000 bucks a kilogram to put a small sat into space or to put material into space. And then SpaceX has dropped the cost to the point that it's now close to $1,000 a kilogram. So a 10X reduction in cost in just the last decade or so. And that's why SpaceX just dominates the launch market. But Elon's always said that $1,000 a kilogram is too high
But his objective has been to get the cost down to 10 bucks a kilogram. Because at 10 bucks a kilogram, you could launch what some people estimate is needed to get to Mars, which is about half a million tons of material and people to set up a colony on Mars. And it actually becomes feasible to get, you know, half a million tons of material at 10 bucks a kilogram. So if you look at
this new Starship and Starship heavy booster, it's about 150, 200 ton payload. The booster holds, you know, 3,400 tons of propellant. And, uh, the cost of that propellant is pretty low. You know, it's, uh, It's only about a million dollars in fuel.
So then if you can get the cost of the booster and the Starship down enough, and you can reuse it enough, and you amortize the cost of making that device over the lifetime of the device, the cost per launch comes down. And that's what brings the cost per kilogram down. So the booster, there's a group called Payload, and they do estimates on this.
So I won't speak out of turn in terms of like having inside knowledge. But the Payload has estimated that Starship And the booster cost about 90 million bucks today. And they think that they have a path to getting it down to 35 million. So if you can reuse that thing 10 times, that's a $3.5 million cost per launch, plus a million for fuel. You could easily see, and this thing can launch 200 tons.
That's how you start to get to 10 bucks a kilogram. over the next couple of years, but it was critical to be able to reuse that heavy booster. And that's what Elon just demonstrated. It's we can actually catch that heavy booster, refuel it and launch it an hour later. And if you can do that over and over again, you're spending 10 bucks a kilogram to put material into space.
You can get fuel into space and then get those starships to fly off to Mars and deliver. all this material, including setting up a base that would allow you to actually make more fuel on Mars, because everything we need to make fuel is on Mars. So it's the beginning of the next series of really important milestones that'll hopefully get humanity onto Mars.
It was just so amazing to see it come together. The economics are legit. I mean, this is like a thousand X reduction in cost. It's incredible.
Yeah, it's going to be amazing. And they're going to do some, I guess, new stuff with Starlink, some even lower Earth orbit satellites that go even faster and have less latency. So that's going to be super exciting.
Starlink's apparently, I mean, I know everyone here is a shareholder in SpaceX, but Starlink's running at 4 million subs right now. That's like 100 bucks a month, 4 million subs. And if you do the math, I mean, how many people... have ISPs that are slower than Starlink, right? How many people have cell phone providers that they're paying roughly the same amount that aren't as good as Starlink?
If we can get satellite to phone and you can get Starlink more broadly available, this could be 100 million subscriber business. I mean, this could be one of the biggest businesses on the earth.
It could be the largest subscription business in the history of humanity. I think the largest ones right now are like Netflix, you know, 250,000. Disney Plus, 150, Verizon, 100 million. So yeah, it could be hundreds of millions of subscribers. It could even be the first 500 million subscriber product in the world.
We could look back one day and be like, why did we run all this copper wire everywhere? We don't need it.
Yeah, obviously. It's crazy.
Especially if it can be like crazy that we were like ever. I mean, the whole nutty thing about this past week, it's like we could look back one day and be like, why did we ever drive cars? And why do we ever have copper wire laid all over the earth to like move internet signals around? You know, this, this efficiency game that's going to be realized over the next decade is just incredible.
Just incredible.
Chamath, any thoughts on the robo van or the A cyber cab, the Model 2, I guess some people are calling it, but it's, you know, the cyber cab specifically not calling it number two and doesn't have a steering wheel or pedals. I would have bought two of those immediately if it had a steering wheel and pedals. I want to drive it. Yeah, it looks like the hybrid of like a Model Y and the Cybertruck.
So I kind of really love the aesthetics of it.
So beautiful.
Yeah, you like it?
My reaction was actually... I don't know, just seeing these releases now over 10 or 15 years, plus of knowing him, nothing... I guess it's not that surprising. It's weird to say. I just expect him and his teams to figure it out. They're just all so good. The thing to remember, it's not just him that's incredible, but he attracts a kind of...
technical and operational wonder kind people for sure and that's just that's just a really special thing so i had that reaction which was i was really proud and happy for them for the team yeah for sure for the team and for him these guys are like incredibly fearless fail bigly right yeah if you're gonna fail fail bigly yeah and then the other thing that i thought was crazy was how many people were trying to dunk on him this weekend
And that surprised me, caught me off guard, because I think if they were personalizing A lot of anxiety that they are feeling through these companies' successes, which didn't make much sense to me.
Well, in fairness, he did hurt some people's feelings with posting of memes. So yeah, I mean, it's, it makes no sense. Like the guy's like gonna save 30,000 road debts a year in the United States with self-driving and people are losing their minds over a couple of memes or who he's voting for for president. I don't think you have to worry about that.
You can just look at the products, they speak for themselves. Anything, Sachs, any response on the Tesla front? Any thoughts on the bus or Optimus?
I mean, they're both very exciting products. I don't think I've got a lot to add.
Yeah. I love the bus, Freebird. I think that thing could become like mobile homes are, you know, ADUs, and you could just send them to... Can we buy them or no? Well, no, not right now, but I think that might be, you know, that's going to be a big question.
I need it for all my kids.
Yeah. Well, see, if this was a platform like the Mercedes Sprinter vans have become that you see a lot in Europe, then you could buy an empty one of these. It's got enough battery life to last a month. And then let's say you had your in-laws over and there was one that was set up as like a one bedroom.
You could click on Airbnb or, you know, Tesla BNB, press a button and the thing could drive to your driveway. You could rent it for a week and then it could leave. Or let's say 1,000 people or 10,000 people were displaced because of a hurricane free bird.
You could send 100,000 of these to the parking lots at Walmart, which typically does a good job in feeding people and getting them supplies after hurricanes, since those are so ubiquitous. You could put 100 of these in every parking lot and have a place for people who are fleeing natural disasters to stay.
So I thought that was like the most compelling product of the whole thing for me was the possibility of a sled, like a skiff that you could do anything you want with. would be really exciting for society. So congratulations to the team. And it's going to take a while, but I could see them having that robo-taxi.
Also, I think congrats to Amid. He just got promoted.
I saw that. He's in charge of all AI?
I think he's in charge of all manufacturing and sales in North America.
Oh, okay. Well, there it is.
Shout out to Amid Afshar.
Guys, I have big news. I just bought my first Tesla.
Oh, you did?
Did you go with a Plaid Model S? Model S Plaid, yeah. I test drove it for two weeks and sold itself.
And are you using the FSD? I use FSD every day.
I use FSD, and it was, like, really impressive. Super impressive. I've tried Tesla a couple times over the years, and it never really worked for me. The quality just didn't feel like, given what I had before, the car-wise. You were an Audi guy, right? Audi guy, yeah. Always loved Audis. Audi guys are neat guys.
Yeah, that's...
Anyway, it's a big milestone. I thought the FSD was the selling. And then the speed on the Plaid is just insane. It's better than my RS7. It's incredible.
With all of my Teslas, I put it in chill mode because when it's in that Plaid mode or whatever, like coffee goes fine.
It's my favorite. I love it.
But if you have passengers, the kids in the backseat would get literally nauseous because it's too fast. You got to be careful with the passengers there. It's so fast.
It was awesome. Awesome.
All right. Well, there you have it. Robo Taxi Star. We didn't get to it last week. We almost put the show back a day or two just to do it. In other news, Uber is exploring a bid to purchase Expedia. Breaking news. This was dispelled as we got here on the show. They said this was like very preliminary third party talks and that there's no serious talks going on about this.
Financial Times reported that that advisors were trying to look at if a deal structure would be possible between Uber and Expedia. Expedia's got a $20 billion market cap. They popped 8% on the news, obviously. Uber, on the other hand, trading at $170 billion market cap or so. That dropped 3%. If you didn't know, Dara was the CEO of Expedia from 2005 to 2017. He's still on the board.
And it looks like this was a trial balloon. You know Uber's two biggest businesses, Rides and Uber Eats, but they also do freight and train bookings. Dara's been pretty clear he wants to create a super app like you have in China or some other markets.
Expedia's got a lot of cool products, Hotels.com, Orbitz, Travelocity, and I think the most interesting one, Freiburg, you and I were talking about, this is VRBO, Vacation Rental by Owner. It was like Airbnb before Airbnb existed. And if you look at this chart, since Dara left, the Dara effect, Expedia has gone exactly sideways. The revenue has grown modestly. What do you think of this deal?
Chamath, I'll just go right to you with this one, since you like to... Stupid. Stupid. Okay. There you have it, folks. Reason number one, it's stupid. And reason number two, it's stupid.
I mean, this is a $20 billion market cap business. You probably have to pay a control premium of 50%. So the question is, if you were going to spend $30 billion today in the public markets, what would you spend it on? And I think the most important lens that you have to use to answer that question is, what reinforces a moat that I have while also being inoculated from the risks of AI?
And I cannot think of... a more fragile business model than the UI layer on top of widely available data. So the problem that Expedia has is the same that Booking and a bunch of these other folks have, which is that the principal heartbeat of the company, flight information and other things, are licensed to them by third parties. And so what they are is a UI and a front door.
I think it's way too early in the evolution of AI to know that that's safe. And in fact, I think a more reasonable assumption is that those things are pretty fragile. And part of what may explain the doldrums of the stock is that I think people are anticipating a world where, for example, I don't know if you saw, but Perplexity launched something this week. It's just in test mode.
They whitelisted me into it, but it's basically a checkout concept. So you tell Perplexity what you would like it to buy, and then it will go and complete the transaction for you. So in the example of flight bookings, you could go directly to United and Because A, perplexity will just show you all of the flights. They'll show you the exact prices.
And then it'll go and execute that for you with your payment method. In a world that looks like that, where these companies have the money to pay for the data feeds, The existing v1.0 generation UIs, I think, are in trouble. So it would just be a very bad capital allocation decision. Now, that's okay to get things wrong, but not for $30 billion wrong.
You can probably do it for a couple hundred million dollars wrong, or maybe even a billion dollars wrong, because you can absorb that as a $150 or $60 billion company. But $30 billion is too big of a price to pay for that kind of risk.
I would agree with you. And there's other things they could buy, like WeRide or Pony AI and a bunch of these AI companies that are doing self-driving. So why not double down on that? Freeburg, the one thing you and I talked about was kind of VRBO, which is a very cool marketplace. And that feels directly in the Uber kill zone.
What do you think about them just maybe carving out and buying VRBO and having an Airbnb?
Why don't they just buy Waymo? Why don't they just go to Google and give them $30 billion of Uber stock and just carve in Waymo? Isn't that a better idea?
I think that's what's going to happen. I've been hearing rumblings of that.
So I think that Dara knows Expedia better than anyone. He ran the business for, what, a decade or so?
Yep.
And so he knows how that business operates. And so if he's looking at this thing and the stock price has been flat roughly since he left in 2017, if you look at the underlying financial performance, you could kind of start to construct a rationale for buying Expedia this cheap. And it would be very accretive to Uber, even if there are these big strategic risks on the horizon.
So just to give you some numbers on it all, Uber has got about 150 million monthly active users. Expedia has about 45, 50 million customers a year that use the service and pay for stuff. So there's a real opportunity to think about the Uber customer base that's installed and as being almost an opportunity to market to them Expedia services and cross sell.
So Expedia on an annualized basis is spending about $8 billion a year in sales and marketing and about $720 million a year in G&A costs. And they're running about $3 billion EBITDA right now. run rate.
So if you cut about half the GNA in an acquisition, because you don't need all the people that overlap with Uber's people, you know, and you cut about 30% of the sales and marketing dollars, because you can cross sell into the Uber install base, you could see a scenario where you could increase Expedia's EBITDA by 75 to 100%, maybe getting it as high as $6 billion.
And while Expedia's market cap trades at $20 billion, this is off of, you know, obviously the recent news that they might get acquired. If you kind of assume a 40%, 50% price premium to the last 90-day average of the stock price, which is kind of typical or common for a deal like this, they're probably paying $26 billion for the company. and they got about 4 billion in net cash.
So you're kind of paying about 22 billion enterprise value to buy Expedia. So 22 billion of enterprise value, and if you can bump the EBITDA up to 6 billion a year, that's a pretty low multiple. I mean, you could kind of see yourself rationalizing this just from a financial basis that you're paying four times EBITDA to buy this thing.
And Dara knows this thing, and he would have great command over what needs to be done over there, and he would have a great sense of what to change and what's gone wrong. And there's a lot of interesting assets inside of Expedia. VRBO is a great one that's been under-monetized and underutilized. I don't know if you've used the UX on VRBO versus Airbnb.
There's obviously some influence Dara could have with people that he knows well that could go in and fix that interface and make it a better service.
And even as AI starts to step in and hotels maybe integrate better with agents and so on, and they show up in a more ubiquitous way, there's other things that Expedia does like build vacation packages and travel packages that are high margin products that they sell. that are a little bit different than what you're used to with just booking a flight.
Booking flights makes no money for anyone, but vacation packages is where all the money's at. And so theoretically Expedia could be smarter about how they build vacation packages and personalize them for families. And that's where they can make real margin, like 20, 30% margin. So I could see a story where this all starts to click for the board at Uber saying, maybe it makes sense.
Dara knows what he's talking about. We could buy this thing for four times pro forma EBITDA. This could be hugely accretive for us. So I think that's why this is happening, why this conversation may be happening. That's just me trying to understand what the rationale might be.
Can you just go back and explain how would they drive up EBITDA so much?
So they're spending about $8 billion a year run rate on sales and marketing at Expedia right now. And Uber has got 150 million active installed users that are using this, the Uber services every month. So the idea would be customers. Yeah.
So if Uber could cross sell some number of Expedia services to their installed base at Uber, which they could test and, you know, do a little experiment and see if it works. They may be able to reduce the marketing dollars that Expedia is spending to acquire customers through other third-party sources like Google and Bing and other places. So there's a rationale.
That's where I think the logic breaks down. I don't think Uber customers want to be cross-sold on booking a hotel. See, this is where I think MBA thinking is very different than product thinking. An MBA looking at this would say, well... You know, Expedia and Uber are both in the travel business. Their apps both involve booking trips.
So we can we can cross sell Expedia from Uber and then cut Expedia's marketing budget. I think that's how an MBA would sort of hand wave over it. I think the way like a product manager would look at this is to say, what does the user want to do? And I know that when I use the Uber app, I just want to basically make a couple of clicks, set my destination, get my car, and then move on.
And there was a product initiative a few years back at Uber where they tried to capture the user's attention during the ride. And they, you know, they added... That's right.
Yeah, they had that whole ad thing that adds... Yeah, that's making a ton of money.
It's actually printing money for them. It was like an entertainment stream or something inside the app. No, but they dialed it way back because I don't see it anymore. It was just clutter.
Would you trust Dara's judgment on this, Sax? Like, if Dara were to think about...
what the uber user would want and he could rationalize some percentage of them they could cross sell expedia services into i mean ultimately i think it's it's his decision right like well i mean what you're describing is basically a private equity play like dar is going to come in as like a private equity buyer effectively and he knows the business and will run it to reduce cost may boost some revenue and maybe there is a justification for that but
If you're trying to justify it based on cross-selling, I don't think users of the Uber app want to be cross-sold when they book a taxi. They just want to be able to affect their transaction as efficiently as possible. And just to finish the point I was making on that whole entertainment stream that they had, they dialed that product way back because it got in the way.
You'd be in the Uber app trying to figure out how to change your destination or something, and all of a sudden you're being shown some entertainment product. It's not what users wanted. And it was always kind of a banana's idea to think that just because the user books an Uber that you own their attention during that ride.
Because during that ride, you're really competing with every app on the iPhone, right? And that's the problem is you want to get in and out of the Uber app. It's about transacting efficiently.
What about not the moment when you're when you're riding in an Uber, but the moment when you say as an Uber user, hey, I need to book travel. I got to go on a vacation to Austin.
I'm never going to think to go on my Uber app for that. The only time I open. But what if they put that feature in there?
What if they had a tab that said book your travel here?
You know, when I open the Uber app, when I want to hail a taxi, that's what it's like.
There are a large number of people who maybe don't have a, you know, an assistant to book their hotels in advance. And like that would be most people, J. Calhoun.
I would not think to go into Uber to do that. It would just be clutter.
Well, no, but they already have a hotels.com partnership. And then the Uber One membership's been growing pretty nicely. And the advertising's doing a billion dollars a year. And that is just a money printing machine because you know that this person's in an Uber Black. You know that they're going to the Four Seasons. Like these users who are, you know,
They have a real ad business at Uber.
Yeah, yeah. The more Uber tries to promote some unrelated product, and what I mean by unrelated is it doesn't help you get to where you're going at that moment. It's clutter in the app. What about Uber Eats, Zags?
Uber Eats is working pretty well. It's working great. Yeah, the cross promotion is working.
That is highly related to this basically booking a car to pick up some food.
Yeah.
It's still the taxi business, basically.
I think the hotel's integration is good.
I think there's something here. We have gone through a cycle where apps and attention were highly consolidated with a few. Now the pendulum has swung the other way and apps are very narrow features that are really well described. Okay, so that's sort of where we are. That's why we have the billions and billions of apps in the App Store.
The question is, does the pendulum swing back to these super apps? And I think the big question is not whether it swings back to the super apps, but whether there's a new substrate that puts itself between the user and all of these services so that they become data-oriented services.
And this is where the question is, if you rely on an agent or you rely on a beefed-up version of search, whether that's ChatGPT or Gemini or whatever, Why would you care where all of this stuff was done? You're not going to care. And this is, I think, the big mistake in this thinking is that that real estate is actually much more fragile than I think we all think it is.
And I think a much better way to think about this is in the future, none of this UI real estate is actually worth anything. The question is, do you have a data asset that's valuable or do you do a service that's viable? Because agentically, there'll be all of these unemotional bots and workflows doing this work for you.
So I think Sachs is right in the sense that whether it's there or not, it won't matter. Could he run it like a private equity business where now Uber Corporation owns two services?
Sure, but you're probably just better off for these agents to go and cannibalize all of search because you'll be able to just get a data feed for what Expedia has to create Expedia for a few million dollars or tens of millions of dollars. You don't need to pay 20 or $30 billion for this.
Yeah, the thing that I've talked to Dara about is when they said he told me when they do something that's adjacent to what they're already doing, it explodes in terms of engagement. So like, they're doing like teens and rental cars, and then package delivery. And every time they do one of those adjacencies, it just takes off with the membership.
And to your point, Freeberg, they have those 150 customers who have their credit cards in there. And man, it's explosive. So that's what I think they're doing.
I don't think booking a vacation is an adjacency to ordering food.
I think hotels would be. I don't think flights would be because I think the flights work really well with the existing apps. But things where you have proprietary inventory like VRBO or hotels, I think those would be very powerful. And those have 20%, 30% commissions, which are in line with the commission's
That Uber's already getting and the commissions on things like flights is very small, like a couple of dollars. So I think for hotels and VRBO would be brilliant. For the other stuff, I'm not so sure. To your point, Chamath.
Well, just to finish my thought was that you'll notice that Uber eats a separate app from Uber. right? I mean, I know you can get to the eats part within Uber, but they created a separate app for a reason. It's because whether you're using Uber Eats or Uber, the goal is immediate gratification. I want to get to where I'm going. I don't book it six hours in advance. I call it right now.
And the most important thing to me is wait time. This is why Uber is beating Lyft is the wait time is lower. Same thing with food. I'm not thinking about booking my dinner right now. I'm not going to do it in advance. If you browse through the restaurants, the most important piece of data they show you in addition to the rating, is the number of minutes it takes for it to get to you.
So those apps are all about immediate gratification. And that's why you don't want other things getting in the way of them. Now, I guess the claim is somehow you're going to be able to cross-sell the booking of a vacation or a hotel that you have to think about days or weeks in advance. It's just a completely different state of mind.
I just don't think that there's much opportunity to cross-sell that. Or to use the technical jargon, I don't think the attach rate is going to be high.
What about the brand value, Zach? Because you know those people are going to another app to book their flight in their hotel. What if that other app was called Uber Travel? There might be some value in that.
I can see that. Yeah.
I think that would be the rationale where I could see the Expedia brand. Yeah.
So maybe what you could do is take VRBO, rebrand it as Uber Hotel or Uber Travel, whatever you want to call it. Exactly. And then maybe you could push people to download that app.
Well, the thing I, you know, I would, the counter I would give to this.
You could quantify the value of the installs, right? So.
Yeah, exactly. I mean, well, you could quantify because Expedia is spending on it every year right now.
I use the Bomboy app to book hotels. I use United to book my flights and I use Uber to do my rides and obviously for Eats. When you are using it, there's a tab up top. And the UI is quite nice in Uber where it's rides and eats right next to each other. I could see a third one like hotels or travel being right there. And all of a sudden, yum, yum, you just get all that inventory right in there.
And I frequently will book my hotel and I'll book my ride for the next day in advance on Uber. And I do those things. And then when I get to my hotel, I'm ordering food to my room. So I think this actually could work really well as a third tab in the app for travel. And you could actually because when you use eats in the Uber app, it's its own tab. And it's the exact same experience.
I believe in super apps. And they just launched a bus that's like a bus service in New York for 18 bucks to go to JFK. That's really awesome. I think we're a little bit disconnected because we don't book our own travel. But okay, let's keep moving here. Down the docket. All right, this big tech investing in nuclear power is off to the races.
Amazon just announced a $500 million investment in three nuclear power projects. All of these are focused on SMRs. Those are the small modular reactors. Amazon is working with dominion energy to develop a small modular nuclear reactor. near an existing nuclear power plant in Virginia. In total, Amazon plans to invest $35 billion in Virginia-based data centers by 2040.
And they want to power these by SMRs. And this is a big trend. Google is purchasing energy directly from Kairos. Power, another company building SMRs. Microsoft, as you heard, is reviving one of the Three Mile Island nuclear power plants. So this is kind of interesting, Shabab. We went from...
Nuclear not being on the table, everybody being against it, the Germans shutting down their reactors post-Fukushima. And now big tech is the customer for these with AI, and they're putting down very large deposits and payments to build them in America. And I haven't heard any opposition.
Maybe you could just speak to Chamath, what we've seen here in terms of opposition to these versus the opportunity and everybody's writing checks.
Well, they're not writing checks. So this is what I don't want to be a Debbie Downer here, but these press releases need to have an asterisk on them. So in the hierarchy of deals, right, just to unpack this for a second, there are deals where you give me X and I give you money. That's not what this is.
Then if you degrade that kind of deal structure, in a lot of heavy industry, you have deals that are called take or pay, which is there is something that's working and you need to basically take this or you need to give me the monetary equivalent of what I'm selling you. That's not what this is.
What this is, is sort of this conditional obligation where the beginning of the deal starts with a very important statement, which is if it works, and if these approvals happen, and there's a whole bunch of nested ifs, then payments can happen. So while these are important deals because they show that there are potential buyers at the finish line,
What it doesn't do is solve the two things that you need to get to the finish line, which is the actual risk capital to finish building these things and technically de-risk them, and then the regulatory approval that you need to make sure that they're allowed. So I think that these deals are good. I think it's a great signaling, but I think it's important to understand the nuances of these things.
These are not things where there's money really trading hands. And until that you see that, where irrespective of what happens, the balance sheet is investing from an Amazon or a Google, where there's corp dev folks writing $100 million or billion dollar checks into these companies, it's not yet quite there.
This is more the step before, which is sort of, can you create some marketing and some buzziness to hopefully induce somebody to then rip in billions of dollars of risk equity capital?
Freeberg, your thoughts on SMRs and these customers showing up? And then I guess you could comment on the nature of the deal structure here, because some of them are you know, contingent on the nuclear power plant turning on. Some of them do have deposits, is my understanding. We'll look that up and fact check it. There could be a range of deals here.
Yeah, I don't know the nature of the deals. I did, I think, talk about this a year ago. It was also like my prediction for the year was to buy the uranium stocks predicated on what I think is a really important point. which is as GDP per capita grows, energy consumption per capita grows.
And if you looked at the projections of GDP per capita in industrialized nations, there was no way, there is no way to meet the energy demand. And this was even pre all this crazy AI build out, which is probably part of the GDP growth. But there is no way to meet the energy demand without nuclear. There is not enough
solar geothermal or wind build-out potential that's happening, that the stopgap measure is going to have to be, and probably the right long-term solution, is to have a significant amount of baseload come from nuclear. And so what's the fastest way to do that nuclear build-out? Well, in China, they have the regulatory authority and the mandate stated.
They're going to build 300 gigawatts with 300 facilities or whatever the number is. And that's what they're doing. Very large facilities that make a gigawatt of power each.
In the US, it seems that because of the regulatory structure here and the way that utilities are regulated and the way that the states have authority on the environmental laws and all the other things, that it might be the fastest path to solving this energy gap problem is SMRs. And these things produce tens of megawatts. So again, a gigawatt is a thousand megawatts.
And we need to kind of probably grow our energy production in the United States by several terawatts over the next decade or two. So this SMR may be the fastest path. Now that could change, meaning we could end up seeing much larger facilities get built out if there's regulatory change in the US and there's more availability.
But fundamentally, we are going to need to use uranium to make electricity to meet the demand of the growing the GDP that it seems we're going to be growing it. I think this is just such a necessity. It's great to see the SMR is getting some attention. I just don't know if they're actually going to get turned on, how long it's going to take. And
you know, I don't know what this election cycle is going to bring in terms of regulatory change. I think we talked about it with several of the candidates when we were doing the interviews.
Sacks, if we are able to get a bunch of these SMRs built here in the United States, maybe if Europe follows suit, what would this do on a geopolitical basis into our relationship with the Middle East, our energy independence, and of course, the AI race to you know, general intelligence. I'll let you take it whichever direction you want to go.