WEBVTT - Kinda Spooky: META, AI, TSLA 0:00:02.720 --> 0:00:08.560 Bloomberg Audio Studios, podcasts, radio news. I feel like we 0:00:08.600 --> 0:00:11.320 need your address things. Yes, go One is that I 0:00:11.360 --> 0:00:11.920 sound like this. 0:00:12.119 --> 0:00:14.080 You sound great, I know, but it's different. 0:00:14.120 --> 0:00:16.319 I have a cold, so it's going to be like 0:00:16.400 --> 0:00:19.480 the best, the best voice you've ever heard on a 0:00:19.520 --> 0:00:21.279 podcast for like ten minutes, and then I'm going to 0:00:21.320 --> 0:00:23.279 lose my voice and it's going to be over. So 0:00:23.280 --> 0:00:24.599 it'll be great on both accounts. 0:00:24.800 --> 0:00:25.000 Yeah. 0:00:25.120 --> 0:00:26.400 The other thing I want to address is what are 0:00:26.400 --> 0:00:27.200 you going to be for Halloween? 0:00:27.320 --> 0:00:30.040 Oh my god, man, I'm so thrilled you asked. I'll 0:00:30.040 --> 0:00:32.919 probably just dress in all black and put on caddiers. Okay, 0:00:33.000 --> 0:00:35.800 but that's what I do most Fridays. Yeah, right, what 0:00:35.800 --> 0:00:36.239 about you? 0:00:36.720 --> 0:00:38.159 I think I've dies this bullet? 0:00:38.440 --> 0:00:40.239 Is it a bullet? Isn't it just joyful? 0:00:40.640 --> 0:00:44.320 Is joyful to take your kids around in their costumes 0:00:44.440 --> 0:00:47.239 when you're not in a costume? I shouldn't it? Like 0:00:47.360 --> 0:00:49.680 I see like the people with family costers, like, Okay, 0:00:49.680 --> 0:00:50.960 that's cool, and I kind of want to do it, 0:00:51.000 --> 0:00:54.200 but like not enough to Yeah, for adults, it's got 0:00:54.240 --> 0:00:54.960 to be creative. 0:00:55.120 --> 0:00:57.960 There are onesies out there, but I hear, yeah, I 0:00:58.000 --> 0:01:01.520 don't really want to cut around really psyched. I mean 0:01:01.520 --> 0:01:03.960 not to go running a one sie. I spoke too soon, 0:01:04.480 --> 0:01:07.520 but I'm just I love Halloween so much. I'm really excited. 0:01:08.360 --> 0:01:10.320 I'm an adult and I don't have a child, so 0:01:10.360 --> 0:01:12.440 I feel like I can't. But we're going to walk 0:01:12.480 --> 0:01:18.000 around Hoboken because Hoboken goes hard on Halloween. A lot 0:01:18.040 --> 0:01:18.280 of these. 0:01:19.319 --> 0:01:20.399 Yes, but okay, I. 0:01:20.319 --> 0:01:22.039 Mean I'm not going to like knock on doors and 0:01:22.080 --> 0:01:25.360 take candy. If someone with a bucket of candy comes 0:01:25.440 --> 0:01:26.679 up to me and says, do you want a piece, 0:01:27.280 --> 0:01:29.680 I'm not going to turn them down. That'd be crazy. 0:01:30.200 --> 0:01:33.520 Are you going to wear your cat ears on air? 0:01:33.840 --> 0:01:35.479 No? No, I won't. 0:01:35.680 --> 0:01:37.280 I feel like one year, I want you to wear 0:01:37.360 --> 0:01:39.720 like a children's ghost costume, you know, just like a 0:01:39.840 --> 0:01:41.600 sheet of holes for eyes on air. 0:01:41.760 --> 0:01:44.399 I usually subtly try to dress up on air. 0:01:44.520 --> 0:01:46.000 Like you've worn the cat I feel like we talked 0:01:46.000 --> 0:01:48.280 about this last year, but you really did. 0:01:48.680 --> 0:01:50.960 Some dedicated listener can tell us. I mean, I always 0:01:51.080 --> 0:01:53.560 I have this like Christmas shirt that I try to 0:01:53.560 --> 0:01:56.040 sneak in once a year. Tomorrow, I've I think I'm 0:01:56.040 --> 0:01:58.520 going to wear all black, like a black turtleneck and 0:01:58.640 --> 0:02:00.400 just be kind of spooky. 0:02:00.680 --> 0:02:03.000 Okay, just subtle, just a. 0:02:03.000 --> 0:02:06.960 Little maybe like a cat eye eyeliner. But I will 0:02:07.000 --> 0:02:10.640 mention over and over again that it's Halloween. I love Halloween. Yeah, 0:02:10.760 --> 0:02:12.440 I want to put someone in a ghost costume and 0:02:12.520 --> 0:02:15.400 just have them walk behind me at some point, just once. 0:02:15.840 --> 0:02:17.760 I think you should actually have a ghost cut host 0:02:17.800 --> 0:02:18.720 who doesn't say anything. 0:02:18.800 --> 0:02:22.040 That would be kind of funny. They just like a 0:02:22.160 --> 0:02:25.240 human and a ghost, like yeah, like. 0:02:25.200 --> 0:02:27.000 Many glasses over the ghost pace, you know. 0:02:27.200 --> 0:02:30.360 Yeah. I turned to them and like say something, and 0:02:30.360 --> 0:02:35.800 then I turned back like this sounds great. 0:02:36.200 --> 0:02:39.000 I've got your TV shows all right, podcast. 0:02:38.720 --> 0:02:40.919 Now, we're just trying to run out the clock here. 0:02:42.200 --> 0:02:43.679 To the podcast like that. 0:02:43.880 --> 0:02:47.880 Yeah scary. 0:02:49.600 --> 0:02:53.040 Hello and welcome to the Money Stuff podcast, your weekly podcast. 0:02:53.080 --> 0:02:57.120 Where are you talking about stuff related to money? I'm 0:02:57.120 --> 0:02:59.079 Matt Levine and I write the Money Stuff column for 0:02:59.160 --> 0:02:59.960 Bloomberg Get Thing. 0:03:00.680 --> 0:03:03.160 And I'm Katie Greifeld, a reporter for Bloomberg News and 0:03:03.200 --> 0:03:18.920 an anchor for Bloomberg Television. We're in We are in 0:03:19.000 --> 0:03:23.120 the heart of magnificent seven tech earnings, which means we 0:03:23.160 --> 0:03:26.400 are in the heart of those tech companies just saying 0:03:26.440 --> 0:03:29.840 how much they're going to spend on AI infrastructure, and 0:03:29.919 --> 0:03:31.040 it is so much. 0:03:31.639 --> 0:03:34.320 It's really like either you do or you don't believe that, 0:03:34.360 --> 0:03:36.880 Like the entire world is going to change with AI. 0:03:37.080 --> 0:03:38.640 And if you believe it, you're like, yeah, sure, ten 0:03:38.680 --> 0:03:41.360 trillion dollars, no problem, right, But it is like it's 0:03:41.400 --> 0:03:44.960 easy to be nervous or skeptical or make fun of 0:03:44.960 --> 0:03:46.480 those numbers because it's a lot of money. 0:03:46.640 --> 0:03:49.520 Yeah. Well, something that I've been talking about on television 0:03:49.560 --> 0:03:52.080 and casually with my friends in the newsroom is how 0:03:53.000 --> 0:03:55.440 a lot of this these hopes and dreams and these 0:03:55.440 --> 0:03:58.520 ambitions of these big tech companies to spend all this money, 0:03:59.160 --> 0:04:01.200 A lot of it has been able to be funded 0:04:01.240 --> 0:04:04.000 from free cash flow, but increasingly they're having to tap 0:04:04.040 --> 0:04:06.280 the bond markets. But as you laid out in a 0:04:06.280 --> 0:04:08.760 column this week, they have stellar credit ratings, so they 0:04:08.800 --> 0:04:12.360 don't necessarily want to rely too much on the bond market. 0:04:12.720 --> 0:04:14.080 Yeah, it's a mixed bag, right, Like you have a 0:04:14.120 --> 0:04:16.200 stellar credit rating, so you can borrow in the bond market, 0:04:16.200 --> 0:04:17.680 but like you can't do too much because then you 0:04:17.720 --> 0:04:19.880 lose the credit iting. So Meta is out this week 0:04:19.920 --> 0:04:22.919 with a twenty five billion dollar investment grade deal. Yes, 0:04:22.960 --> 0:04:25.840 but they want more than twenty five billion dollars and 0:04:26.520 --> 0:04:31.520 to do that they have tapped the woo like I guess, 0:04:31.560 --> 0:04:33.840 I don't know. You call it like people sometimes called 0:04:33.880 --> 0:04:36.520 the hybrid's market, but also like it's the kind of 0:04:36.520 --> 0:04:40.480 project finance. Yeah, it's like off balance sheet financing. Basically, 0:04:40.560 --> 0:04:43.920 like they set up a joint venture where the joint 0:04:44.000 --> 0:04:47.040 venture will operate their data centers, and the joint venture 0:04:47.120 --> 0:04:50.159 will be owned by in this case Blue Owl, but 0:04:50.279 --> 0:04:56.640 in the general case someone who is uh, basically credit 0:04:56.680 --> 0:04:59.000 firm but is willing to do weird stuff, and so 0:04:59.160 --> 0:05:01.000 Blue Oul is the equ the owner of the data center, 0:05:01.400 --> 0:05:04.800 and Meta signs a series of contracts that in the 0:05:04.839 --> 0:05:06.840 aggregate kind of look like that. Like they sign like 0:05:06.960 --> 0:05:10.480 leases and a residual value guarantee, where like at the 0:05:10.560 --> 0:05:11.919 end of the term of the lease they have to 0:05:11.920 --> 0:05:15.480 buy the chips in the data center. But those contracts 0:05:15.560 --> 0:05:19.039 look enough like debt that Blue Owl can package it 0:05:19.040 --> 0:05:21.520 into bonds and sell it to Pimcar or whatever, but 0:05:21.680 --> 0:05:26.600 also enough not like debt that it doesn't affect Meta's 0:05:27.000 --> 0:05:30.320 accounting or its credit ratings. So it's like not debt 0:05:30.320 --> 0:05:32.800 of Meta. For the purposes of like Meta, but it 0:05:32.839 --> 0:05:35.280 is kind of debt of Meta for the purposes of 0:05:35.320 --> 0:05:36.400 the people buying the bonds. 0:05:36.839 --> 0:05:41.599 So you talked about how this is a fantastic trick 0:05:41.839 --> 0:05:44.040 that if you want to call it that that has connotations, but. 0:05:44.080 --> 0:05:46.040 It's just like it's just the thing people do. Right, 0:05:46.160 --> 0:05:48.120 It's like it turns out there are things that are 0:05:48.160 --> 0:05:49.760 debt and there are things that are not debt, but 0:05:49.800 --> 0:05:52.640 then there's like this huge middle ground of things that 0:05:52.760 --> 0:05:56.320 they kind of that that and like, for the most part, 0:05:57.200 --> 0:05:59.880 things are pretty binary for like ratings and balance sheet. 0:06:00.160 --> 0:06:02.360 Either it's on your balance sheet or it's not. Either 0:06:02.440 --> 0:06:05.040 it effects your credit anying credit atings are more not 0:06:05.120 --> 0:06:08.520 quite binary, like the ratings agency sometimes gives things partial credit. 0:06:09.080 --> 0:06:11.120 But you know, if you like check the right boxes, 0:06:11.160 --> 0:06:14.400 then it's not quite debt. But you can convince people 0:06:14.920 --> 0:06:17.520 to buy the bonds because they look pretty much like 0:06:17.560 --> 0:06:18.000 your debt. 0:06:18.120 --> 0:06:20.839 Yeah, give or take you do, right though. That The 0:06:20.880 --> 0:06:25.039 bonds for the Hyperion Hyperion the name of the data center. 0:06:25.120 --> 0:06:27.560 They price with a coupon of almost six point six percent, 0:06:27.600 --> 0:06:31.360 which was roughly a percentage point higher than Meta's outstounding 0:06:31.400 --> 0:06:34.239 corporate bonds and in line with the average jump on yields. 0:06:34.360 --> 0:06:37.920 So hear what you're saying. But at a certain point, 0:06:37.920 --> 0:06:40.840 I feel like people and investors see through this. 0:06:41.640 --> 0:06:43.440 They do, and they don't, right, I mean, they don't 0:06:43.640 --> 0:06:45.360 treat it like Meta debt exactly. 0:06:45.560 --> 0:06:45.760 Yeah. 0:06:45.760 --> 0:06:47.600 They you know, they're happy to buy it at a 0:06:47.839 --> 0:06:50.480 yield and they're like, oh, yeah, it's great. It's like 0:06:50.560 --> 0:06:53.359 it's backed by Meta. You know, it has structural features 0:06:53.360 --> 0:06:55.240 that make it not quite as good as Meta debt, 0:06:55.240 --> 0:06:58.279 but yeah, so it trades that a yield that is 0:06:59.200 --> 0:07:01.640 a reasonable cut the capital for building a data center 0:07:01.680 --> 0:07:05.080 and kind of keeps it off of Meta's pristine balance sheet. 0:07:05.320 --> 0:07:10.320 Yeah, we're talking about this like weird structure, this SPV 0:07:10.520 --> 0:07:11.280 that Meta is doing. 0:07:12.760 --> 0:07:15.360 It's weird ish a little, but it's also like, you know, 0:07:15.480 --> 0:07:19.240 the basic format of doing some sort of project finance JV. 0:07:20.240 --> 0:07:22.080 They didn't invent that for data centers. 0:07:22.280 --> 0:07:22.840 That's true. 0:07:22.920 --> 0:07:24.760 It's been done for a while and a lot of 0:07:24.760 --> 0:07:27.320 industries in different forms. And you know, I wrote this 0:07:27.360 --> 0:07:30.760 week about it's very general purpose. And so Carrig Doctor 0:07:30.760 --> 0:07:35.520 Pepper is putting its k cup manufacturing in JV so 0:07:35.560 --> 0:07:39.040 that it can borrow against the k cups without tarnishing 0:07:39.080 --> 0:07:40.760 its credit rating to do an acquisition. 0:07:40.960 --> 0:07:42.640 Okay, so it's not weird. I do feel like a 0:07:42.680 --> 0:07:45.360 lot of folks, myself included, have thought more about this 0:07:45.440 --> 0:07:48.680 in the past two weeks, and maybe I have. Ever, 0:07:49.320 --> 0:07:53.000 it's a concept that's being more widely socialized than perhaps 0:07:53.000 --> 0:07:53.240 in the. 0:07:53.200 --> 0:07:54.520 Past and higher profile. 0:07:54.760 --> 0:07:55.960 Yes, that's the thing. 0:07:56.160 --> 0:07:57.800 And the other thing about it that is interesting is 0:07:57.800 --> 0:08:00.000 that it's really interacting with the private credit right. Yeah, 0:08:00.200 --> 0:08:03.040 you look at all the private credit firms and stereotypically 0:08:03.080 --> 0:08:07.920 private credit means leveraged acquisition finance. And the reason for 0:08:08.000 --> 0:08:11.040 that is because you know, private equity sponsors are willing 0:08:11.080 --> 0:08:15.880 to pay up for flexible, fast, guaranteed debt, so you 0:08:15.920 --> 0:08:17.800 can make money doing that, Like you can charge more 0:08:17.800 --> 0:08:20.400 than the public markets would if you can offer sponsors 0:08:20.440 --> 0:08:23.720 like user friendly product financing their acquisitions. But then, like 0:08:23.800 --> 0:08:25.200 you know, these private credit firms, like they have a 0:08:25.240 --> 0:08:27.600 lot of insurance clients, they want to deploy a lot 0:08:27.640 --> 0:08:30.240 of money. They need to eventually get into investment grade debt, 0:08:30.400 --> 0:08:33.360 and it is harder to get into investment grade debt 0:08:33.360 --> 0:08:34.920 because you go to like Meta and you say I'll 0:08:34.960 --> 0:08:36.840 buy your bonds and metas like, sure, our bond's price 0:08:36.880 --> 0:08:40.120 that you know, that's right, and it's not super exciting 0:08:40.160 --> 0:08:43.800 for private credit. But then you do this project financing, 0:08:43.840 --> 0:08:45.200 and you can do a thing that gets an A 0:08:45.280 --> 0:08:50.000 plus rating and also has a yield. And also you're 0:08:50.040 --> 0:08:52.160 not competing in the same way with the public markets 0:08:52.160 --> 0:08:56.720 because regular public market bond investors can't necessarily buy two 0:08:56.760 --> 0:08:59.000 point five billion dollars of equity in a data center, right, 0:08:59.040 --> 0:09:01.480 Like the private credit firms are a little bit more 0:09:01.480 --> 0:09:04.040 flexible and like what kind of financing arrangements they can do, 0:09:04.080 --> 0:09:05.760 and a little bit more creative, and so they can 0:09:06.000 --> 0:09:08.439 do this to get tens of billions of dollars of 0:09:08.520 --> 0:09:10.440 investment grade financing, which is what they want. 0:09:10.679 --> 0:09:14.760 Yeah, I do want to talk more broadly about META, Yes, 0:09:14.840 --> 0:09:17.280 but before we get there, so you made the point that, okay, 0:09:17.320 --> 0:09:20.920 even with this what they're working out, they also tap 0:09:20.960 --> 0:09:23.160 the market, the public market for that twenty five billion 0:09:23.200 --> 0:09:27.160 dollar bond. Apparently it was super oversubscribed according to our reporting. 0:09:27.200 --> 0:09:28.720 But you think about. 0:09:30.480 --> 0:09:33.040 You know, META has capacity to do hundreds of billions 0:09:33.040 --> 0:09:34.360 of dollars then, and they don't do it because they 0:09:34.400 --> 0:09:36.400 want to keep good rating. So they Yeah, they want 0:09:36.440 --> 0:09:37.000 to met a bond. 0:09:37.080 --> 0:09:39.440 Yeah, one hundred and twenty five billion dollars of bids 0:09:39.520 --> 0:09:40.439 for this twenty. 0:09:40.160 --> 0:09:42.040 Five So like over subscribe never means anything. 0:09:42.120 --> 0:09:45.679 Yeah, well nothing truly means anything, Matt. But something I've 0:09:45.679 --> 0:09:47.800 been thinking about, not just as it relates to Meta, 0:09:47.840 --> 0:09:50.760 but all the spending that all these big tech companies 0:09:50.800 --> 0:09:55.680 are doing. You think about the typical profile of investing 0:09:55.800 --> 0:09:59.959 in tech from the equity market perspective, It's like, Okay, 0:10:00.160 --> 0:10:04.680 these companies have pristine, tidy balance sheets, they're very capital 0:10:04.760 --> 0:10:07.520 light businesses, et cetera, et cetera. And I do wonder, 0:10:07.760 --> 0:10:11.920 you know, if the nature of investing in the tech 0:10:12.120 --> 0:10:15.719 space is starting to change, And I keep asking people. 0:10:15.440 --> 0:10:18.120 That just not capital light these days? I know, I 0:10:18.160 --> 0:10:19.440 know it's the opposite. 0:10:19.640 --> 0:10:21.880 It's just you think, like five years ago, the conversation 0:10:22.080 --> 0:10:26.600 was about you know, intangible assets, and that week. 0:10:26.600 --> 0:10:29.840 Meta right like five years ago was Meta called Facebook. Yeah, 0:10:30.640 --> 0:10:33.120 but like you think, like the stereotypical business model of 0:10:33.160 --> 0:10:35.360 Facebook is like it serves a web page, right, It's 0:10:35.360 --> 0:10:38.000 like marginal cost is zero, right, so it's scales like 0:10:38.160 --> 0:10:41.120 really really effectively. So Facebook is this classic investment where 0:10:41.160 --> 0:10:44.319 you know, venture capitalists put money in and it goes 0:10:44.360 --> 0:10:47.600 from being a dorm room project to being a one 0:10:47.679 --> 0:10:51.840 hundred billion dollar company with kind of no additional spending. Yeah, 0:10:52.120 --> 0:10:55.880 and that switch just flipped and now it's just everything 0:10:55.960 --> 0:10:57.720 is a trillion dollars of data centers. 0:10:57.800 --> 0:11:02.000 Well, I love talking about meta because you remember the 0:11:02.040 --> 0:11:05.840 reason why Meta is called meta because they do. Hey, 0:11:05.840 --> 0:11:09.280 one's so big on the metaverse, which I never talked 0:11:09.320 --> 0:11:12.240 about before Mark Zuckerberg was lighting money on fire. 0:11:12.440 --> 0:11:15.160 I don't really know what that entel, but it would 0:11:15.160 --> 0:11:17.439 be funny if there were like data centers out there 0:11:17.520 --> 0:11:20.080 that were optimized only for the metaverse and are now 0:11:20.120 --> 0:11:21.760 just like tumbleweed is blowing through them. 0:11:21.840 --> 0:11:24.360 I love it. I love it. But that led to 0:11:24.480 --> 0:11:26.600 Meta's year of efficiency, et cetera. 0:11:26.679 --> 0:11:29.000 It was this whole thing and now the opposite. 0:11:28.720 --> 0:11:31.880 Alphabet and Microsoft. You can make a case for why 0:11:31.880 --> 0:11:35.240 they're going so hard on AI when it comes to Meta. 0:11:36.160 --> 0:11:38.920 I'm not sure what the payoff is. They said that 0:11:39.120 --> 0:11:42.280 they're investing so hard because it's going to help better 0:11:42.320 --> 0:11:48.760 target their advertising. But is that is that? Is that 0:11:48.840 --> 0:11:53.560 a trillion dollar endeavors? It's wild, it's wild, and I've. 0:11:53.480 --> 0:11:56.160 Learned is that the purpose of the economy is to 0:11:56.200 --> 0:11:57.640 serve customized ads. 0:11:57.960 --> 0:12:01.440 For sure, for sure, all the water sources be damned. 0:12:01.440 --> 0:12:04.040 But I don't know. I'm curious to see how this 0:12:04.080 --> 0:12:07.520 plays out for Meta Apparently, Mark Zuckerberg did say on 0:12:07.520 --> 0:12:10.040 the earnings call they reported on Wednesday this week that 0:12:10.080 --> 0:12:12.400 the company has options if it ends up spending too 0:12:12.440 --> 0:12:15.720 much on infrastructure. In one scenario, he said, the company 0:12:15.720 --> 0:12:18.480 could use the extra computing capacity for its core business. 0:12:18.720 --> 0:12:21.640 In another, it could sell the power to other companies. 0:12:21.400 --> 0:12:24.400 Or or really good metaverse. 0:12:24.920 --> 0:12:27.080 That metaverse, really good metaverse. 0:12:27.360 --> 0:12:30.760 Yeah, there's the centers full of legless Mark Zuckerberg. 0:12:31.240 --> 0:12:33.520 I mean, they're still committed to that in there. I 0:12:33.520 --> 0:12:37.199 believe that falls under the Reality Labs division, which reported 0:12:37.240 --> 0:12:40.120 a loss of four point four billion dollars for the 0:12:40.160 --> 0:12:41.000 third quarter. 0:12:40.880 --> 0:12:43.319 On revenue of twelve cents revenue. 0:12:42.960 --> 0:12:46.600 Of four hundred and seventy million. I'll have you know. Anyway, 0:12:46.960 --> 0:12:50.800 I think it's super interesting. At least Mark Zuckerberg's thinking 0:12:50.840 --> 0:12:54.080 about the prospect of maybe they're overbuilding. 0:13:08.840 --> 0:13:10.720 Also, we're in hundreds of billions of dollars of a 0:13:10.880 --> 0:13:13.120 spending open AI is a regular company. 0:13:12.880 --> 0:13:16.920 Now I know it's exciting so regularly anti climactically. 0:13:17.280 --> 0:13:20.000 Yeah, I was assuming there'd be like years of litigation, 0:13:20.240 --> 0:13:24.320 and like all the state attorney generals and Microsofts were like, okay, good, 0:13:24.360 --> 0:13:26.400 it's fine, So cool. They did their. 0:13:26.320 --> 0:13:29.319 Conversial at this moment too. Elon Musk is a bit distraction, 0:13:29.679 --> 0:13:32.520 which will brew we'll get to I'm sure he'll keep suing, 0:13:32.600 --> 0:13:35.560 but yeah, we'll get to that anyway. Open Ai apparently 0:13:35.600 --> 0:13:39.240 is so normal that maybe they're going to IPO in 0:13:39.480 --> 0:13:41.320 next two years or so. 0:13:41.559 --> 0:13:43.400 Yeah, sure, maybe, right. 0:13:43.280 --> 0:13:44.600 That's what Reuter's told us. 0:13:44.840 --> 0:13:48.000 Sure. I just feel like any large private tech company 0:13:48.080 --> 0:13:50.080 might IPO in the next two years. 0:13:49.920 --> 0:13:52.760 I know, but this one I feel like it urgently 0:13:52.800 --> 0:13:55.760 needs to really. Okay, Yeah, they've committed to like one 0:13:55.800 --> 0:13:58.520 point four trillion dollars of spending. Where are they going 0:13:58.559 --> 0:13:59.640 to get the funds. 0:14:00.040 --> 0:14:02.200 From, like an IPO? Maybe a little from an IPO? 0:14:02.400 --> 0:14:06.120 Sam Almons, if open Ai and Fanny and Freddie I'll 0:14:06.120 --> 0:14:09.800 go public like next year, God, what IPO? 0:14:10.440 --> 0:14:14.160 So many good podcasts? No. Sam Almons said that I 0:14:14.200 --> 0:14:16.280 think it's fair to say that it's the most likely 0:14:16.320 --> 0:14:18.920 path for us, given the capital needs that we'll have. Yeah, 0:14:19.520 --> 0:14:20.960 during a live stream on Tuesday. 0:14:20.960 --> 0:14:25.440 Matt find the the public equity markets and the public 0:14:25.480 --> 0:14:29.600 bond markets and the project finance markets. It's amazing, so 0:14:29.720 --> 0:14:32.080 much money for AI. Yeah, so they're a normal company. 0:14:32.080 --> 0:14:33.880 I don't know, a normal company. 0:14:33.680 --> 0:14:35.720 Owns twenty seven percent. 0:14:35.960 --> 0:14:37.280 Twenty seven percent. 0:14:37.480 --> 0:14:41.920 The nonprofit owns twenty six percent but has like super 0:14:42.000 --> 0:14:46.840 voting board control. It's very hard. Yeah, you be like, 0:14:47.440 --> 0:14:49.520 we started as a non profit. We shore donations. Theydn't 0:14:49.520 --> 0:14:52.320 think that much in donations, but in donations, and now 0:14:52.480 --> 0:14:55.400 we're a five hundred million dollars private company for the 0:14:55.400 --> 0:14:58.040 benefit of our investors. And one way to sort of 0:14:58.440 --> 0:15:01.560 smooth that transition is that the nonprofit will continue to 0:15:01.600 --> 0:15:05.280 control the company and have like supervoting rights. I don't 0:15:05.280 --> 0:15:07.280 know how that will go for an IPA, but it'll 0:15:07.320 --> 0:15:10.480 go fine. Like no one cares. You could imagine being like, 0:15:11.320 --> 0:15:13.760 if this company is not run for profit, maybe I 0:15:13.800 --> 0:15:18.480 shouldn't invest in it. Yeah, seeking an enormous financial return, 0:15:18.960 --> 0:15:20.400 But no one has ever had that thought. 0:15:20.640 --> 0:15:22.720 Yeah, or ever will. Yeah, that's so clean. 0:15:22.920 --> 0:15:25.520 It's like I is definitely looking to give you a 0:15:25.600 --> 0:15:28.800 high financial return on your investment, which is nice. 0:15:29.200 --> 0:15:32.200 So do you not take Samulmon at his words, which 0:15:32.240 --> 0:15:35.400 words that they need to do this to get the 0:15:35.400 --> 0:15:36.400 capital that they need to know. 0:15:36.920 --> 0:15:38.920 I believe that they need to do this to get 0:15:38.920 --> 0:15:42.160 the capital. But like even now that they're a for 0:15:42.280 --> 0:15:46.080 profit company, they're a public benefit company, which is slightly 0:15:46.160 --> 0:15:48.320 less than for profit but still pretty for profit. But 0:15:48.320 --> 0:15:51.440 they're controlled by a nonprofit board, right, Like the nonprofit 0:15:51.480 --> 0:15:56.040 boards to put the interest of humanity first. But no 0:15:56.040 --> 0:15:58.000 one cares about that. No one's worried about that. 0:15:58.080 --> 0:16:01.720 I thought that a nonprofit board couldn't fire members. 0:16:02.400 --> 0:16:04.800 Well, they're not going to fire him like that, lesson. 0:16:05.200 --> 0:16:08.080 Well I know that, but even other members, like other 0:16:08.200 --> 0:16:10.680 open AI executives. So when when you say they're in control, 0:16:10.720 --> 0:16:12.040 I mean, what does that mean if they're not going 0:16:12.120 --> 0:16:14.320 to fire the people running it? 0:16:14.600 --> 0:16:17.160 That's a good question the board of directors, right, So, 0:16:17.240 --> 0:16:19.880 like typically one of the main things the board does 0:16:19.960 --> 0:16:22.480 is decide whether or not to do a merger. And 0:16:22.520 --> 0:16:25.200 they're too big to do a merger, you know, yeah, right, Like, 0:16:25.200 --> 0:16:27.760 I don't know how much operational control they have. One 0:16:27.760 --> 0:16:30.360 thing the board of directors does is fire the CEO 0:16:30.440 --> 0:16:33.680 if he does a bad job. But there's a constraint 0:16:33.760 --> 0:16:36.920 on that job number one, Right, Like the board of 0:16:36.920 --> 0:16:42.360 open Ai within recent memory has fired Sam Altman, not 0:16:42.400 --> 0:16:44.880 for doing a bad job of developing products or leading 0:16:45.040 --> 0:16:48.880 the people or achieving shareholder value, but for the crime of, 0:16:50.400 --> 0:16:52.440 I don't know, something else. No one really understood it. 0:16:52.480 --> 0:16:54.720 But it was kind of in the ballpark of like 0:16:54.920 --> 0:16:57.760 not being for the benefit of humanity or like deceiving 0:16:57.760 --> 0:16:59.760 the board. And I'm like, no one really knows, and 0:16:59.800 --> 0:17:02.880 they quickly unfired him and then fired themselves. Yeah, so 0:17:03.640 --> 0:17:04.600 that won't happen again. 0:17:04.680 --> 0:17:06.520 Yeah, that was just about two years ago. 0:17:06.720 --> 0:17:09.440 Yeah, it's amazing in theory, Like you can look around 0:17:09.480 --> 0:17:13.760 a lot of like boards of big founder led tech startups. 0:17:14.440 --> 0:17:17.400 Sometimes those founders do weird stuff and you can imagine 0:17:17.400 --> 0:17:19.600 a board firing them, but you can't really because they 0:17:19.640 --> 0:17:20.359 never would. You know. 0:17:20.520 --> 0:17:23.760 There's been so many natural segues to yes. 0:17:23.720 --> 0:17:25.200 Right, let's move straight along. 0:17:25.440 --> 0:17:28.160 Okay, Well, the only thing I was going to say 0:17:28.240 --> 0:17:30.320 is that when it comes to the capital needs, obviously 0:17:30.359 --> 0:17:32.439 we've talked a lot about the private markets and the 0:17:32.440 --> 0:17:35.240 public markets melding that if you're a private company, you 0:17:35.280 --> 0:17:36.960 can get most of the funding that you need in 0:17:37.000 --> 0:17:41.440 the private markets. This obviously is such an extreme example 0:17:41.960 --> 0:17:44.720 that maybe it's not even worth mentioning, but it is 0:17:44.760 --> 0:17:47.000 a limit perhaps of the private markets. 0:17:47.280 --> 0:17:50.040 Yeah, sure, right, I tak that point, although we're not 0:17:50.160 --> 0:17:52.520 there yet, right, No, this is like they're rumored to 0:17:52.560 --> 0:17:54.200 be interested in going public in twenty twenty. 0:17:54.320 --> 0:17:57.240 My mind, they're already public for now. They have floated 0:17:57.240 --> 0:17:58.680 a trial balloon with Reuters. 0:17:58.880 --> 0:18:01.000 No, Like, what do they need capital for? They need 0:18:01.040 --> 0:18:03.760 to like get chips and data centers and compute like 0:18:03.760 --> 0:18:06.119 they're getting all of that, right, but they're funding it 0:18:06.160 --> 0:18:10.440 with vague promises of future trillions of dollars of capital raising, right, 0:18:10.480 --> 0:18:14.560 Like they have not needed to go public to meet 0:18:14.600 --> 0:18:18.479 their enormous capital spending needs yet. Yeah, right, they're incurring 0:18:18.520 --> 0:18:21.480 obligations that might require going public. But loo, I agree 0:18:21.480 --> 0:18:23.200 with you that, like, if Sam Altman is feeling the 0:18:23.240 --> 0:18:26.240 pressure to go public, that does show some limit on 0:18:26.920 --> 0:18:30.480 the ability of private markets to fund absolutely everything. Yeah, 0:18:30.640 --> 0:18:34.119 that he's got a five hundred billion dollar company that 0:18:34.200 --> 0:18:38.400 keeps signing trillion dollar deals and is not public. So true, 0:18:38.480 --> 0:18:40.040 the limit is we haven't hit it yet. 0:18:56.440 --> 0:19:00.359 If we're looking for another segue, SpaceX obviously is private 0:19:00.400 --> 0:19:01.960 company that has a lot of capital needs. 0:19:02.040 --> 0:19:04.400 Yes, so it turns out that shooting rockets to Mars 0:19:04.920 --> 0:19:08.680 costs much, much, much less than training chatbots. 0:19:08.960 --> 0:19:11.399 What a world. It's what a world, right, Elon Musk, 0:19:12.160 --> 0:19:15.639 he's the whole company, the whole board of Tesla. Well really, 0:19:15.880 --> 0:19:18.680 Robin Denholm has been on a media tour. 0:19:18.960 --> 0:19:21.920 Yeah, Robin Denholm and the board of Tesla are going 0:19:21.960 --> 0:19:27.480 around to big shareholders and also frankly media, Yeah, to 0:19:27.680 --> 0:19:30.560 drum up support for Elon Musk's big pay package because 0:19:30.760 --> 0:19:33.560 the sharelders have to vote on it, and the big 0:19:34.080 --> 0:19:37.200 shaholder advisory firms Iss and Glass Lewis have recommended that 0:19:37.240 --> 0:19:40.360 the sharelders vote against giving Elon Musk a pay package 0:19:40.440 --> 0:19:45.120 that has a nominal value of one trillion dollars. Yeah. 0:19:45.160 --> 0:19:47.720