WEBVTT - Bloomberg Businessweek Weekend - August 7th, 2026 0:00:00.280 --> 0:00:07.240 Bloomberg Audio Studios, Podcasts, Radio News. 0:00:08.640 --> 0:00:13.000 This is Bloomberg Business Week Daily reporting from the magazine 0:00:13.000 --> 0:00:17.599 that helps global leaders stay ahead with insight on the people, companies, 0:00:17.640 --> 0:00:22.520 and trends shaping today's complex economy. Plus global business finance 0:00:22.560 --> 0:00:25.880 and tech news as it happens. The Bloomberg Business Week 0:00:25.960 --> 0:00:31.320 Daily Podcast with Carol Masser and Tim Steneveek on Bloomberg Radio. 0:00:32.120 --> 0:00:35.040 Hi, everyone, Welcome to the Bloomberg Business Week Weekend Podcast. 0:00:35.440 --> 0:00:38.440 A big focus for everyone this week. I gotta be honest. 0:00:38.840 --> 0:00:41.320 Since the IPO back in June, this has been a countdown. 0:00:41.479 --> 0:00:45.839 We are talking about SpaceX and its earnings. Investors got 0:00:45.840 --> 0:00:50.000 their first look at SpaceX's financials following its landmark IPO. 0:00:50.200 --> 0:00:52.720 For all the details on the results, head to the 0:00:52.760 --> 0:00:55.520 Bloomberg End at Bloomberg dot com. But Tim, like you 0:00:55.640 --> 0:00:58.280 kept saying it wasn't really about earnings because the company 0:00:58.320 --> 0:00:59.600 isn't profitable overall. 0:00:59.680 --> 0:01:02.760 Yeah, that always reminds me. Earnings mean you have earnings, 0:01:02.760 --> 0:01:06.960 and SpaceX didn't have any earnings. Bottom line For investors' 0:01:07.000 --> 0:01:11.160 top line, revenue surged staggering. Capex sent the stock tumbling 0:01:11.240 --> 0:01:14.759 in the release. That theme of the iwatering AI spend. 0:01:15.160 --> 0:01:17.399 It's something we dig into this hour with a noted 0:01:17.400 --> 0:01:19.400 critic of the AI build and spend. 0:01:19.480 --> 0:01:22.039 That tech critic he's also the publisher of Where's Your 0:01:22.160 --> 0:01:25.000 Ed At We're talking about ed Zitron. He's stopped by 0:01:25.040 --> 0:01:27.600 to dissect the hundreds of billions of dollars going into 0:01:27.640 --> 0:01:31.959 the AI capex movement and spend overall, and why the 0:01:32.040 --> 0:01:37.039 massive gap between data center spending and actual AI revenue 0:01:37.480 --> 0:01:41.560 is creating what Ed calls quote an unsustainable circular economy. 0:01:41.640 --> 0:01:43.520 He's kind of not alone in that thinking. 0:01:43.920 --> 0:01:46.600 No, he's definitely not, and I think what he says 0:01:46.680 --> 0:01:50.400 really resonates with an audience, and we see that when 0:01:50.440 --> 0:01:53.160 he comes on the program. AI's talked a lot about 0:01:53.400 --> 0:01:55.840 by us here at Bloomberg. You know that at this point, 0:01:56.000 --> 0:01:58.960 so too increasingly our prediction markets and the battles they 0:01:58.960 --> 0:02:00.920 are dealing with when it comes to their role in 0:02:00.960 --> 0:02:03.720 the financial world, and the growing legal challenges that are 0:02:03.800 --> 0:02:04.520 questioning them. 0:02:04.680 --> 0:02:07.200 That's right. We wanted to hear how calshe is keeping 0:02:07.280 --> 0:02:10.240 up with all the lawsuits filed against them. We do 0:02:10.320 --> 0:02:13.640 that with Bobby Denault, head of Enforcement and Legal Counsel 0:02:13.720 --> 0:02:16.680 at calshe needless to say, there's a lot. 0:02:16.560 --> 0:02:19.320 At stake and speaking of high stakes, and back to AI. 0:02:19.680 --> 0:02:23.200 Aaron Brown, Bloomberg opinion columnist and former chief risk manager 0:02:23.240 --> 0:02:27.640 at AQR Capital Management. That's Cliff Asns's fund weighs in 0:02:27.720 --> 0:02:30.679 on the AI hedge fund situational awareness and why is 0:02:30.760 --> 0:02:33.799 staggering four hundred and thirty nine percent first half return 0:02:34.400 --> 0:02:36.919 was a glaring warning sign about market mania. 0:02:37.000 --> 0:02:39.600 All of that to come this hour. We begin with 0:02:39.800 --> 0:02:43.720 legal battles surrounding prediction markets. The recent surge and volume 0:02:43.720 --> 0:02:47.800 across platforms like Calshie and Polymarket has made event based 0:02:47.840 --> 0:02:51.079 trading one of the fastest growing corners of finance. 0:02:51.080 --> 0:02:55.079 Yes, but also growing regulatory scrutiny of these markets. Just 0:02:55.200 --> 0:02:57.960 last week, New York State authorities sued call she for 0:02:58.040 --> 0:03:01.360 allegedly running an illegal on the un licensed gambling operation 0:03:01.480 --> 0:03:04.320 in the state, marking another legal hurdle for an industry 0:03:04.360 --> 0:03:06.359 that is one support from the Trump administration. 0:03:06.680 --> 0:03:08.720 We needed to learn more, and so for that we 0:03:08.760 --> 0:03:11.640 caught up with Robert Denault, head of Enforcement and legal 0:03:11.680 --> 0:03:12.800 counsel at Calshie. 0:03:13.160 --> 0:03:15.560 I mean, as both a New Yorker and a lawyer, 0:03:15.800 --> 0:03:19.200 I'm alarmed by the overreaching sentiment that's coming from the 0:03:19.200 --> 0:03:24.720 Attorney General's office. So calshe is a licensed federally regulated exchange. 0:03:25.200 --> 0:03:28.800 By her logic, any federally regulated exchange that's operating with 0:03:28.880 --> 0:03:31.639 a federal license and overseen by a federal regulator can 0:03:31.680 --> 0:03:35.560 suddenly be subject to the whims of state criminal enforcement 0:03:35.720 --> 0:03:39.360 if the Attorney General of a particular state wakes up 0:03:39.360 --> 0:03:42.960 and decides one day that these contracts actually come within 0:03:43.040 --> 0:03:46.680 New York state gambling law. That's not how any exchange 0:03:46.720 --> 0:03:49.240 in US history has ever operated, right, So we have 0:03:49.720 --> 0:03:51.440 we're here at Bloomberg. You guys talk about the New 0:03:51.520 --> 0:03:55.280 York Stock Exchange NASDAQ, other exchanges, all of them could 0:03:55.320 --> 0:03:58.000 be potentially affected by the breath and scope of the 0:03:58.040 --> 0:04:02.960 New York Attorney generals approach and legal theory in terms 0:04:02.960 --> 0:04:05.400 of how she believes she can regulate and bring to 0:04:05.520 --> 0:04:08.120 heal federally licensed exchanges here in New York State. 0:04:08.160 --> 0:04:10.240 Bobby, are you saying you're the exact same things as 0:04:10.280 --> 0:04:12.440 the New York Stock Exchange or the NASDAK markets. That 0:04:12.440 --> 0:04:16.400 you're saying you're the exact same thing apples to apples. 0:04:16.000 --> 0:04:19.480 Then what I am saying is that federal law dictates that. 0:04:19.480 --> 0:04:22.159 That is the case when a federal law like the 0:04:22.160 --> 0:04:25.279 Commodity Exchange Act exists and provides for a way for 0:04:25.360 --> 0:04:28.039 an entity to get licensed by something like the Commodity 0:04:28.080 --> 0:04:31.880 Futures Trading Commission or the Securities Exchange Commission, that licensure 0:04:32.000 --> 0:04:35.720 and that federal regulation is what governs that marketplace. Now, 0:04:35.920 --> 0:04:39.359 if states want to litigate with that regulator on a 0:04:39.440 --> 0:04:41.800 case by case basis about what types of contracts might 0:04:41.839 --> 0:04:44.200 implicate some state laws, that's one question, and that's some 0:04:44.200 --> 0:04:47.080 of the lawsuits we've seen over the last year on sports. 0:04:47.279 --> 0:04:50.760 But this is much more far reaching. This is claiming 0:04:50.800 --> 0:04:53.800 that that license means nothing, and if you don't hold 0:04:53.839 --> 0:04:56.680 a New York license, you're running a criminal operation and 0:04:56.720 --> 0:04:58.120 you need to be run out of the state. And 0:04:58.560 --> 0:05:01.080 I think it's important to ground this sort of in 0:05:01.240 --> 0:05:04.320 history of disruptive sort of new players in marketplaces. Right, 0:05:04.520 --> 0:05:08.080 call she's new, but this sort of licensed regime has 0:05:08.120 --> 0:05:11.839 existed for many decades. We've seen similar playbooks used against 0:05:11.839 --> 0:05:14.719 companies like Uber and Airbnb, where states try to throw 0:05:14.760 --> 0:05:17.760 their weight around and bring crazy cases to block what 0:05:17.920 --> 0:05:21.200 customers want. As a reasonable alternative, we think that that's 0:05:21.279 --> 0:05:24.279 a pretty similar playbook to what the Attorney General's following here. 0:05:24.320 --> 0:05:26.880 So why shouldn't cal she seek a license from the 0:05:26.920 --> 0:05:28.960 New York State Gaming Commission? Why wouldn't you do that? 0:05:29.240 --> 0:05:32.000 So it really goes to the way that the business operates. 0:05:32.400 --> 0:05:35.560 We are a federally licensed exchange that requires us to 0:05:35.640 --> 0:05:39.680 run open markets that are available nationwide. Our users set 0:05:39.720 --> 0:05:42.479 the price, traders set the price. We match traders in 0:05:42.520 --> 0:05:45.479 an open marketplace with one another. We are not on 0:05:45.520 --> 0:05:48.320 the other side of individuals trading. We don't run a 0:05:48.320 --> 0:05:50.719 casino where people can come in and drink and play 0:05:50.720 --> 0:05:52.919 card games. We don't run a sports book where we 0:05:52.920 --> 0:05:55.520 profit when people lose. What we do is run open 0:05:55.560 --> 0:05:59.120 marketplaces where users define the price point and that type 0:05:59.120 --> 0:06:01.400 of financial product, even if it touches on topics that 0:06:01.480 --> 0:06:04.479 is similar to a topic touched on by a sportsbook, 0:06:04.640 --> 0:06:08.159 the way that that product operates is typically what governs 0:06:08.160 --> 0:06:09.520 what regulations apply. 0:06:10.520 --> 0:06:15.359 So even though people keep coming at you and say gambling, gambling, gambling, 0:06:16.360 --> 0:06:17.719 that's why you're not gambling. 0:06:18.320 --> 0:06:21.160 So I think it's important to define exactly what they 0:06:21.200 --> 0:06:23.720 mean when they say gambling. To me, gambling is when 0:06:23.760 --> 0:06:25.599 you go up against the house. When you go to 0:06:25.640 --> 0:06:27.720 a place that controls whether you're going to win or lose, 0:06:27.760 --> 0:06:29.880 they're going to chase losers, They're going to maximize their 0:06:29.920 --> 0:06:32.640 ability to limit winners. They're not going to run like 0:06:32.680 --> 0:06:36.760 a true business. But where an exchange exists and individuals 0:06:36.800 --> 0:06:39.000 are setting price with one another, I think that it 0:06:39.040 --> 0:06:42.240 operates under a different regulatory framework. And it's not to 0:06:42.279 --> 0:06:45.479 say that a whole panoply of different topics can be 0:06:45.560 --> 0:06:49.560 touched on by different regulated products. So we see concerns 0:06:49.560 --> 0:06:52.760 in options trading or leverage trading, retail traders moving into 0:06:52.800 --> 0:06:55.960 markets that are traditionally regulated markets at the federal level, 0:06:56.160 --> 0:06:58.520 I think some of the same concerns exist for those markets. 0:06:58.560 --> 0:07:01.080 And if you read the lawsuit closely, a ton of 0:07:01.120 --> 0:07:03.640 the definitions and the language and the descriptions that Attorney 0:07:03.640 --> 0:07:08.520 General James's office uses could easily be transposed onto robinhood, 0:07:08.800 --> 0:07:12.920 crypto trading, derivatives trading, any sort of trading activity that 0:07:12.920 --> 0:07:16.120 the Attorney General suddenly decides poses a customer threat to 0:07:16.160 --> 0:07:18.040 individuals who want to participate in I mean. 0:07:17.880 --> 0:07:20.800 There is In April, she did sue coinbase in Gemini 0:07:20.800 --> 0:07:24.480 for running illegal gambling platforms. There were some like why 0:07:24.520 --> 0:07:26.160 do you think call she was signaled out in this 0:07:26.560 --> 0:07:29.320 typical or this this iteration of the lawsuit. 0:07:29.720 --> 0:07:32.280 You'd have to ask the Governor's office, why you know 0:07:32.360 --> 0:07:35.680 call She's was signal signaled out. I think the truth 0:07:35.760 --> 0:07:37.840 is there are a number of prediction markets that would 0:07:37.840 --> 0:07:40.040 be affected by this approach. There's a number of other 0:07:40.120 --> 0:07:43.080 exchanges that would be affected. There are number of like 0:07:43.160 --> 0:07:47.520 robin Hood's based in New York, right, novig Polymarket, They're 0:07:47.560 --> 0:07:49.320 all based in New York, none of whom have been 0:07:49.360 --> 0:07:51.760 subject to the Attorney General's focus. So I don't know 0:07:51.760 --> 0:07:53.400 if that's coming. I don't know if you'd have to 0:07:53.400 --> 0:07:56.040 ask her office, but certainly, you know, we're a bit 0:07:56.080 --> 0:07:59.560 alarmed that the Attorney General is seeing the scope of 0:07:59.600 --> 0:08:04.440 New York gambling law applying to potentially federal derivatives exchanges. 0:08:04.600 --> 0:08:07.960 But I do think about exchanges. Do you think about 0:08:08.000 --> 0:08:10.560 the composition of the bets that are on Like if 0:08:10.560 --> 0:08:13.920 we're going to use the gambling analogy that keeps getting 0:08:14.160 --> 0:08:17.040 kind of lobbied at you. I mean, if I think 0:08:17.080 --> 0:08:20.840 about exchange, there is incredible oversight, you know, and there's 0:08:21.000 --> 0:08:24.360 rules and so that there isn't you know, trades that 0:08:24.400 --> 0:08:27.400 are happening that are not legit or fair or right. 0:08:28.040 --> 0:08:31.280 You know, if if gambling is as you say, there 0:08:31.360 --> 0:08:35.600 is no house, right, it's just two parties figuring out something. 0:08:35.960 --> 0:08:39.320 Where is the oversight and concern about the composition of 0:08:39.360 --> 0:08:42.440 the bets being made and making sure that they are true? 0:08:42.480 --> 0:08:44.760 Because I think that's where you're get into gambling. If 0:08:44.800 --> 0:08:46.840 anybody can kind of put up a bet and another 0:08:46.880 --> 0:08:49.080 person can take the side, and who knows if those 0:08:49.120 --> 0:08:51.360 two sides didn't get together to figure out this bet, 0:08:51.400 --> 0:08:53.880 like you know what I'm saying, So that to me 0:08:54.400 --> 0:08:57.000 is where you get into the gambling. It's a little 0:08:57.320 --> 0:08:58.480 it feels a little loose. 0:08:58.720 --> 0:09:01.600 Sure, I don't think it's loose at all. So I 0:09:01.679 --> 0:09:04.920 think that when you're just to sort of zero in 0:09:04.920 --> 0:09:07.480 on some of the nuance that you just said, gambling 0:09:07.640 --> 0:09:09.760 is when there is the house. What you do on 0:09:09.800 --> 0:09:12.120 an exchange is where there is no house, there are 0:09:12.120 --> 0:09:15.240 two counterparties in a particular market agreeing on a price 0:09:15.320 --> 0:09:18.040 together and executing what we characterize as a swap, what 0:09:18.120 --> 0:09:21.640 the CEA qualifies as a swap. Those types of trades 0:09:21.800 --> 0:09:25.400 are heavily regulated under CFTC regulation. There are hundreds of regulations. 0:09:25.480 --> 0:09:27.400 You're saying all the trades that are happening on Kowshi 0:09:27.440 --> 0:09:29.880 will be heavily regulated and overset, And we understand the 0:09:29.920 --> 0:09:30.480 two sides. 0:09:30.559 --> 0:09:33.720 They are currently heavily regulated and being overseen. I meet 0:09:33.720 --> 0:09:36.240 with the CFTC's Enforcement Division. That's my line of work 0:09:36.280 --> 0:09:38.760 at Kalshi. I lead our exchange enforcement. I meet with 0:09:38.800 --> 0:09:41.960 their Enforcement division multiple times a week. But separately, they 0:09:41.960 --> 0:09:45.920 have other divisions division Market Oversight, other divisions that oversee 0:09:45.960 --> 0:09:48.920 the implementation of markets so that they are structurally fair 0:09:49.320 --> 0:09:53.480 for the participants operating in them, and that oversight requires 0:09:53.559 --> 0:09:56.120 Kalshi as an exchange, but all prediction market exchanges to 0:09:56.160 --> 0:09:58.600 work very closely with the CFTC on the products that 0:09:58.600 --> 0:10:01.360 they offer. They have to sell certify them to the CFTC, 0:10:01.720 --> 0:10:05.200 who can revoke approval for those products. And the CFTC 0:10:05.280 --> 0:10:07.600 recently came out with a two hundred and sixty seven 0:10:07.600 --> 0:10:10.640 page rule making that was specific toward prediction market contract 0:10:10.720 --> 0:10:13.000 So I think you know it isn't correct to say 0:10:13.000 --> 0:10:14.760 these aren't currently heavily regulated. 0:10:15.000 --> 0:10:17.440 We're speaking with Bobby Denald, he's head of Enforcement and 0:10:17.480 --> 0:10:19.640 Legal counsel at Kalshe He joins us here in the 0:10:19.640 --> 0:10:22.040 Bloomberg BusinessWeek studio, I want to shift gears and talk 0:10:22.040 --> 0:10:24.439 about some other recent news. It seems like you're really 0:10:24.440 --> 0:10:26.640 playing whack a mole at this point with these lawsuits. 0:10:26.640 --> 0:10:29.640 Like obviously it's keeping you busy. Also keeping you busy 0:10:29.640 --> 0:10:32.720 as enforcement and former Congressman George Santos agreed to pay 0:10:32.760 --> 0:10:35.280 more than thirty five thousand dollars to settle allegations that 0:10:35.320 --> 0:10:38.199 he manipulated a wager on your platform about whether he'd 0:10:38.200 --> 0:10:40.400 attend the twenty twenty six State of the Union. Is 0:10:40.400 --> 0:10:42.640 it your responsibility to police that. Are you the one 0:10:42.679 --> 0:10:45.120 who recognized that and flagged it? 0:10:45.400 --> 0:10:47.320 Yes, it is our responsibility to police that. 0:10:47.400 --> 0:10:49.959 Do you have the resources like this is one example, 0:10:49.960 --> 0:10:52.760 But do you have the resources to find every single 0:10:52.800 --> 0:10:55.680 one of them that allegedly has insider information? 0:10:55.960 --> 0:10:59.240 So I'd pause there and say, that's not our standard 0:10:59.280 --> 0:11:01.240 at the New York stylelock exchange, and that's not our 0:11:01.280 --> 0:11:03.800 standard in the securities and equities markets. What we expect 0:11:04.120 --> 0:11:07.640 our federally regulated exchanges to do is to have reasonable 0:11:07.640 --> 0:11:12.040 procedures to detect insider trading, market manipulation, etc. Calhi has 0:11:12.320 --> 0:11:15.720 very robust and in fact more robust than stock exchanges 0:11:16.200 --> 0:11:19.640 policing measures to prevent and detect insider trade. We do 0:11:19.760 --> 0:11:23.040 have twenty four to seven market surveillance. We use a 0:11:23.080 --> 0:11:26.040 surveillance vendor, but also in the coming days are going 0:11:26.040 --> 0:11:29.600 to announce an expansion of our surveillance systems and we 0:11:29.679 --> 0:11:32.640 police these markets both in real time and in retrospect, 0:11:32.800 --> 0:11:35.560 and we work with the CFTC to investigate specific markets 0:11:35.600 --> 0:11:38.640 and detect anomalous activity in those markets. We did detect 0:11:38.760 --> 0:11:42.160 the trading activity by mister Santos. We conducted an investigation 0:11:42.280 --> 0:11:44.679 that involved an interview with mister Santos, and we referred 0:11:45.200 --> 0:11:48.200 the entire matter and the evidence that we collected the CFTC, 0:11:48.240 --> 0:11:51.560 which allowed them to pursue enforcement. But separately, in the 0:11:51.559 --> 0:11:55.840 derivatives and commodities space, these exchanges, like ours, have a 0:11:55.880 --> 0:11:58.960 responsibility to also bring enforcement actions directly against the users 0:11:59.000 --> 0:12:02.400 who participate on them, and so we pursue direct exchange 0:12:02.480 --> 0:12:06.000 enforcement in areas where individuals come on and violate our 0:12:06.200 --> 0:12:07.720 CFTC approved exchange rules. 0:12:07.720 --> 0:12:11.040 And since that's happening, and how many investigations are you triggering, 0:12:11.080 --> 0:12:12.760 like on a daily basis, or a weekly. 0:12:12.520 --> 0:12:17.120 Basis, so we sort of estimate quarterly, and volumes upticked 0:12:17.160 --> 0:12:19.720 quite a bit, So the investigations have upticked quite a bit, 0:12:19.760 --> 0:12:21.560 but somewhere between one hundred and fifty and two hundred 0:12:21.559 --> 0:12:24.760 and fifty a quarter become material investigations. We make a 0:12:24.840 --> 0:12:27.000 number of referrals. I think year to date we've probably 0:12:27.040 --> 0:12:30.120 made about forty or fifty referrals to the CFTC. These 0:12:30.160 --> 0:12:32.920 matters take time, though the legal process. Everybody deserves due 0:12:32.920 --> 0:12:35.600 process rights, yep, So we afford people do process. We've 0:12:35.640 --> 0:12:38.920 settled or brought disciplinary actions in a number of cases. 0:12:38.960 --> 0:12:40.840 We're going to continue to do so. By the end 0:12:40.840 --> 0:12:42.760 of the year, you know, I'm sure you'll see a 0:12:42.840 --> 0:12:46.560 meaningful number of actions. I put for context, the SEC 0:12:46.600 --> 0:12:48.640 in the last year of the Biden administration bought thirty 0:12:48.640 --> 0:12:52.280 five insider trading actions, So you know, I think the 0:12:52.320 --> 0:12:54.800 expectation somewhere in the ballpark of that number. 0:12:54.920 --> 0:12:56.840 Do you think that there's some markets that are listed 0:12:56.840 --> 0:13:00.120 on KYSHI that are more susceptible to manipulation than others? 0:13:00.840 --> 0:13:04.640 I think, much like insider trading in the traditional securities 0:13:04.640 --> 0:13:07.400 and equities markets, there are certain paradigms that exist that 0:13:07.440 --> 0:13:10.400 create more likelihood for insider or manipulation. 0:13:10.000 --> 0:13:11.800 Risk, like mentioned markets. 0:13:11.840 --> 0:13:15.319 Perhaps you know, I think mention markets are a unique category. 0:13:16.040 --> 0:13:19.800 I don't know that they're necessarily more susceptible to insider risk, 0:13:19.880 --> 0:13:22.000 but I can see the context or argument for saying 0:13:22.040 --> 0:13:25.000 they maybe are more susceptible to manipulation where one person 0:13:25.040 --> 0:13:27.280 controls what word they say. On the counter of that, 0:13:27.800 --> 0:13:30.360 there's a small pool of people who could possibly be 0:13:30.520 --> 0:13:33.200 capable of manipulating that market. Because we're in exchange that 0:13:33.240 --> 0:13:37.319 collects user information, we follow KYC know your customer rules 0:13:37.320 --> 0:13:39.720 for every single individual trading on the exchange, we have 0:13:39.720 --> 0:13:42.079 a pretty good idea of who's taking positions in certain 0:13:42.080 --> 0:13:44.760 markets that allows us to police mention markets, just like 0:13:44.800 --> 0:13:45.559 we police all market. 0:13:45.600 --> 0:13:49.360 You're talking about Calsey public companies, right, I just are no. 0:13:49.480 --> 0:13:50.480 I mean, yeah, that's part of it. 0:13:50.520 --> 0:13:51.760 I mean, that's the new Just. 0:13:52.080 --> 0:13:55.040 Sixty seconds, I know, we've got to run it's interesting 0:13:55.120 --> 0:14:00.240 earnings updates, KPI forecasts, earnings I do, and I'm thinking, God, 0:14:00.320 --> 0:14:03.120 does this replace ultimately the earnings estimates that we follow? 0:14:03.400 --> 0:14:04.000 Is that the goal? 0:14:04.160 --> 0:14:06.400 Just quickly, I think there's two goals. I think the 0:14:06.480 --> 0:14:08.720 first goal is what we find is so many people 0:14:08.760 --> 0:14:11.600 are using Kalshi as a resource just to obtain information. 0:14:11.720 --> 0:14:13.959 Seventy five percent of people who visit the platform are 0:14:13.960 --> 0:14:17.160 there just to look to learn, They're not there to trade. 0:14:17.240 --> 0:14:19.160 And so this is a resource for people who might 0:14:19.200 --> 0:14:20.880 be on a trading desk, people who might be in 0:14:20.920 --> 0:14:23.520 a financial position where they want to analyze these really 0:14:23.720 --> 0:14:28.160 subtopic financial questions about a particular KPI. But you know, 0:14:28.240 --> 0:14:31.160 of course, we are seeing academic research that shows these 0:14:31.160 --> 0:14:34.520 are somehow and sometimes more accurate than our traditional metrics 0:14:34.720 --> 0:14:37.560 and financial markets, and so it's exciting to develop further 0:14:37.560 --> 0:14:38.720 in that Spaceically, we've come back and. 0:14:38.720 --> 0:14:41.360 Talk more about this, so I am fascinated about kind 0:14:41.360 --> 0:14:44.520 of where this goes and who will all be on it. Bobby, 0:14:44.560 --> 0:14:45.240 thank you so much. 0:14:49.000 --> 0:14:53.000 You're listening to the Bloomberg Business Weekdaily podcast. Catch us 0:14:53.080 --> 0:14:56.520 live weekday afternoons from two to five pm Eastern Listen 0:14:56.560 --> 0:15:00.120 on Apple CarPlay and Android Auto with the Bloomberg Business app, 0:15:00.240 --> 0:15:02.040 or watch us live on YouTube. 0:15:02.840 --> 0:15:05.280 This past week, we got further information that points to 0:15:05.320 --> 0:15:09.160 Microsoft generating most of its AI revenues and likely about 0:15:09.200 --> 0:15:13.160 seventy percent from one customer open AI. This is all 0:15:13.160 --> 0:15:17.640 according to new Microsoft company disclosures. Under an agreement between 0:15:17.680 --> 0:15:21.320 the two companies, Open Ai pays Microsoft our computing power, 0:15:21.480 --> 0:15:24.640 costs associated with building AI models, and a share of 0:15:24.680 --> 0:15:25.400 its revenue. 0:15:25.520 --> 0:15:28.280 Our next guest says that is worth watching here to 0:15:28.280 --> 0:15:31.640 pull apart big text catpecs obsession and why he believes 0:15:31.640 --> 0:15:35.120 this circular ecosystem is nearing a tipping point is ed Zitron. 0:15:35.280 --> 0:15:37.960 He's CEO of Easy Primary Research. He's the host of 0:15:38.000 --> 0:15:40.880 the Better Offline podcast And to note, we caught up 0:15:40.920 --> 0:15:44.080 with that before the news broke of Microsoft's AI sales. 0:15:44.560 --> 0:15:48.040 Where are we in terms of the AI narrative in 0:15:48.080 --> 0:15:49.440 your view and what's the reality? 0:15:49.840 --> 0:15:52.520 Well, I think investors have to ask a question right now, 0:15:52.640 --> 0:15:55.200 what am I getting into when I invest in Microsoft, Google, 0:15:55.240 --> 0:15:58.840 and Amazon. So UBS estimates that twenty seven percent of 0:15:58.880 --> 0:16:02.200 Google Clouds revenue this year will be open ai and anthropic, 0:16:02.360 --> 0:16:05.720 increasing to over forty eight percent next year. That is 0:16:05.720 --> 0:16:07.560 a remarkable amount of money. It's going to be over 0:16:07.600 --> 0:16:11.000 one hundred and twenty four billion dollars next year. Everyone 0:16:11.080 --> 0:16:13.240 is buying into these stocks because they believe all of 0:16:13.280 --> 0:16:17.160 that capex is going towards diverse and spread out AI demand, 0:16:17.320 --> 0:16:20.120 when in fact, what it's actually doing is helping create 0:16:20.200 --> 0:16:23.960 infrastructure for two unprofitable, unsustainable companies. 0:16:24.560 --> 0:16:27.520 So those the other one would be anthropic. Yes, you argue, 0:16:27.600 --> 0:16:29.280 so give us more data because you have the micro 0:16:29.360 --> 0:16:32.480 you're citing Microsoft, But what about AWL, Well. 0:16:32.320 --> 0:16:34.440 That was what I was saying. So Barclay's actually says 0:16:34.480 --> 0:16:37.120 that this year thirteen percent of AWS revenue will be 0:16:37.120 --> 0:16:40.040 both open air and anthropic, and next year will be 0:16:40.080 --> 0:16:43.440 eighteen percent. AWS much bigger business than Google Cloud. Now, 0:16:43.520 --> 0:16:45.240 just to be clear, when I was saying that twenty 0:16:45.320 --> 0:16:48.360 seven percent this year and forty eight percent next year 0:16:48.400 --> 0:16:51.480 for Google Cloud, I meant both anthropic and open al Okay, 0:16:51.560 --> 0:16:53.600 most people don't know that open ai is a large 0:16:53.640 --> 0:16:56.200 customer of Google Cloud. It's not a well, it's not 0:16:56.240 --> 0:16:58.720 a well known fact. But this was this was actually 0:16:58.760 --> 0:16:59.840 mentioned by you. 0:17:00.080 --> 0:17:02.320 Yes is Stephen jew So where would those companies be 0:17:02.480 --> 0:17:05.000 right now without anthropic and without open ai? 0:17:05.160 --> 0:17:07.560 Well, I have serious questions about that. So in counter 0:17:07.680 --> 0:17:10.280 year twenty twenty five, according to my own reporting about 0:17:10.280 --> 0:17:13.680 open aiy's numbers, sixty nine percent if the year over 0:17:13.760 --> 0:17:17.520 year growth of Microsoft Intelligent Cloud segment was actually from 0:17:17.520 --> 0:17:20.320 open Ai. Without that, it would have only grown eight 0:17:20.320 --> 0:17:23.119 percent year of view, which is barely beating inflation. And 0:17:23.200 --> 0:17:26.199 so everyone is being sold what I consider kind of 0:17:26.200 --> 0:17:28.240 a lie. It's honestly kind of a scandal. 0:17:28.560 --> 0:17:30.800 So this goes back to I feel like we have 0:17:30.880 --> 0:17:35.479 companies the circular financing the circularity of it all, and 0:17:35.560 --> 0:17:40.720 kind of creating demand for their products. So when does 0:17:40.760 --> 0:17:42.200 the party end? In your view? 0:17:42.560 --> 0:17:44.800 So with open Aiy's IPO, I think that could be 0:17:44.800 --> 0:17:46.919 one of the flashpoints. Remember, this company was meant to 0:17:46.920 --> 0:17:49.080 go public this year. They failed about a month or 0:17:49.119 --> 0:17:51.320 two ago, and now the New York Times has reported 0:17:51.320 --> 0:17:54.680 that they're considering they are delaying until twenty twenty seven. 0:17:55.000 --> 0:17:57.640 That's lethal for a number of people. But open Ai 0:17:57.680 --> 0:18:01.199 and Anthropic need continual flow of capital. They do not 0:18:01.400 --> 0:18:04.240 pay their bills out of existing cash flow, so when 0:18:04.280 --> 0:18:06.760 anything happens to that cash, I think that's the first 0:18:07.600 --> 0:18:10.160 kind of domino to four. But then again, there's also 0:18:10.200 --> 0:18:12.680 the overall problem of data center's just not getting built 0:18:12.800 --> 0:18:16.080 very fast, taking about twelve to thirty six months, depending 0:18:16.119 --> 0:18:18.880 on how small or larger data center is actually being 0:18:18.920 --> 0:18:22.280 built at And the problem is is that everyone believes 0:18:22.280 --> 0:18:24.720 that AI is coming out of cash flow, that AI 0:18:24.880 --> 0:18:27.199 is coming out of just this diverse revenue base, when 0:18:27.240 --> 0:18:30.119 it's really not. It's extremely narrow. The Information reported a 0:18:30.119 --> 0:18:32.800 few months ago that eighty nine percent of the largest 0:18:32.800 --> 0:18:36.000 AI companies, well, their revenue comes just from open AI 0:18:36.040 --> 0:18:38.240 and anthropic it's heavily centralized. 0:18:38.320 --> 0:18:40.760 Doesn't it have to be centralized to some extent? This 0:18:40.800 --> 0:18:43.280 is expensive to do or no in terms of data 0:18:43.280 --> 0:18:45.000 center build out and so on and so forth. And 0:18:45.080 --> 0:18:48.160 what's going to make AI generative? 0:18:48.200 --> 0:18:48.440 AI? 0:18:49.320 --> 0:18:51.760 The ability for it to be really, really good is 0:18:51.840 --> 0:18:54.199 having access to lots of information. So doesn't it have 0:18:54.320 --> 0:18:56.640 to be to some extent ed concentrated. 0:18:56.840 --> 0:18:59.720 Well, when I say concentration, I mean concentration of revenue 0:18:59.840 --> 0:19:00.840 in these two companies. 0:19:00.880 --> 0:19:02.879 No, I understand, but to make it good. So doesn't 0:19:02.880 --> 0:19:05.720 it make sense that those who are expose the most 0:19:05.840 --> 0:19:07.639 it's going to be concentrated to some extent. 0:19:07.680 --> 0:19:10.360 Well, I mean when we're talking about so Siteline Climates 0:19:10.400 --> 0:19:12.879 said that they saw back in February about one hundred 0:19:12.880 --> 0:19:15.600 and ninety gig what's worth of data center capacity being 0:19:15.640 --> 0:19:17.840 built in the next few years. It was built or 0:19:17.920 --> 0:19:20.159 under planning. Now, if you work that out. 0:19:20.000 --> 0:19:20.520 With a PUE. 0:19:20.640 --> 0:19:23.160 So just the efficiency rating of one point three, you're 0:19:23.200 --> 0:19:25.920 coming out to twelve million O megawatt over one point 0:19:25.960 --> 0:19:29.760 six trillion dollars of annual revenue needed to satiate those 0:19:29.840 --> 0:19:32.919 data centers. Having two customers is not going to do that. 0:19:33.280 --> 0:19:36.800 Even that, they're most spendy and rober and open Ai 0:19:37.359 --> 0:19:39.520 they can't afford anything they need venture Cappital, but they're 0:19:39.520 --> 0:19:41.879 only going to spend four hundred billion a year, and 0:19:41.880 --> 0:19:43.600 that's if they get that far, which I don't believe 0:19:43.600 --> 0:19:43.960 they will. 0:19:44.160 --> 0:19:45.840