WEBVTT - SAP and Carahsoft Get Probed, OpenAI Pitches the White House 0:00:02.600 --> 0:00:10.240 Bloomberg Audio Studios, podcasts, radio news. 0:00:08.760 --> 0:00:13.080 From Mahard where innovation, money and power Collie in Silicon 0:00:13.160 --> 0:00:14.040 Vallet NBN. 0:00:14.400 --> 0:00:18.920 This is Bloomberg Technology with Caroline Hyde and Ed Ludlow. 0:00:32.760 --> 0:00:34.240 Live from New York and San Francisco. 0:00:34.440 --> 0:00:38.239 This is Bloomberg Technology coming up. SAP and IT solutions 0:00:38.240 --> 0:00:42.519 provider Carasoft get probed by the US for potential price fixing. 0:00:43.560 --> 0:00:46.000 Open Ai makes a pitch to the White House. This 0:00:46.159 --> 0:00:48.680 is France gets ready for its own AI summit. 0:00:48.880 --> 0:00:51.920 In cybersecurity, startup Whiz is mulling selling shares and a 0:00:52.000 --> 0:00:53.760 twenty billion dollar valuation. 0:00:54.760 --> 0:00:56.120 First, let's get to our top story. 0:00:56.160 --> 0:01:00.680 German software giant SAP, product reseller Carasoft and other companies 0:01:00.680 --> 0:01:04.280 and being probed by the US officials potentially conspiring to 0:01:04.400 --> 0:01:08.160 overcharge government agencies over the course of a decade SAP 0:01:08.319 --> 0:01:10.200 shares as you can see currently training lower on the 0:01:10.240 --> 0:01:14.000 report bloomgs. Jake Bleiberg joins us on the story you 0:01:14.040 --> 0:01:16.880 broke it? What exactly is being investigated here? 0:01:16.959 --> 0:01:18.319 Jake, thank you for having me. 0:01:18.959 --> 0:01:23.760 The investigation is into what is a potential scheme to, 0:01:24.000 --> 0:01:28.080 over the course of ten years, overcharge the US government 0:01:28.360 --> 0:01:34.000 and primarily the Department of Defense for purchases of SAP 0:01:34.160 --> 0:01:38.920 technology through a variety of resellers, including Kerosoft, that total 0:01:39.080 --> 0:01:40.640 up to two billion dollars. 0:01:42.080 --> 0:01:46.200 Jake, what a karasof spokesperson told you is important in 0:01:46.280 --> 0:01:49.320 this story. In the context of a raid that took 0:01:49.360 --> 0:01:53.480 place at Karasoft. They gave you details, but did not 0:01:53.640 --> 0:01:56.560 go as far as to confirm that it related to 0:01:56.600 --> 0:02:01.520 a relationship with SAP. Please explain the reporting absolutely, and 0:02:01.560 --> 0:02:03.160 that is a very important detail. 0:02:04.280 --> 0:02:10.120 Yesterday, FBI and Department of Defense investigators searched Carosoft's offices 0:02:10.160 --> 0:02:15.040 in Virginia. However, we don't know whether or not that 0:02:15.720 --> 0:02:21.880 search is related to this separate civil investigation of Carosoft 0:02:22.120 --> 0:02:25.440 and SAP over the potential price fixing. 0:02:27.280 --> 0:02:29.560 Okay, let's dwell a little bit and also some other 0:02:29.720 --> 0:02:33.000 comments that have been coming. SAP says it's been cooperating 0:02:33.040 --> 0:02:35.880 with the DOJ in terms of the civil investigation since 0:02:35.960 --> 0:02:39.160 the beginning. We understand Eccentia is also involved in this 0:02:39.280 --> 0:02:41.639 in some way. A spokesmerson for that company, a subsidiary 0:02:41.680 --> 0:02:44.680 of Accentia, that's Accentia Federal Services, is responding to an 0:02:44.680 --> 0:02:48.400 administration subpoena as cooperating with the DOJ, and the Justice 0:02:48.440 --> 0:02:53.400 Department didn't respond to requests for comment. What's interesting is, 0:02:53.440 --> 0:02:56.600 of course the money you say, and the amounts that 0:02:56.639 --> 0:03:02.280 are involved here. Ultimately, how will this investigation continue and 0:03:02.320 --> 0:03:05.240 what could be the fine on the back of it. 0:03:05.880 --> 0:03:09.600 Well, so it's a great question. We don't ultimately know 0:03:10.160 --> 0:03:12.560 whether there will be a fine. But what we do 0:03:12.680 --> 0:03:16.840 know is that federal investigators are continuing their work trying 0:03:16.840 --> 0:03:20.799 to understand what happened here, and that they're framing this 0:03:21.080 --> 0:03:26.480 as a potential wrongdoing related to a very large series 0:03:26.520 --> 0:03:30.119 of purchases, hundreds of contracts totaling as much as two 0:03:30.160 --> 0:03:35.480 billion dollars over you know, since twenty fourteen. And again 0:03:35.680 --> 0:03:38.160 I should really emphasize we don't know that anyone will 0:03:38.240 --> 0:03:41.520 ultimately be accused of any wrongdoing, but this is being 0:03:41.520 --> 0:03:46.080 investigated under the False Claims Act, and that allows the 0:03:46.080 --> 0:03:49.720 government to seek or to recover as much as three 0:03:49.800 --> 0:03:53.040 times it's damages plus a penalty. So we're that to 0:03:53.240 --> 0:03:56.160 you know, we're litigation to sort of commence there. The 0:03:56.200 --> 0:03:58.120 government could be looking at a lot of money. 0:03:58.840 --> 0:04:01.840 Blomberg's take by Bay, thank you now a key interview 0:04:01.880 --> 0:04:06.160 with Bloomberg Alphabet CEO sunderpitch Ey says the company's legal 0:04:06.200 --> 0:04:09.240 battles of antitrust could drag on for many years. He 0:04:09.360 --> 0:04:12.200 was commenting on the federal anti trust trial, which the 0:04:12.280 --> 0:04:16.520 Justice Department is alleging Google manipulates the roughly six hundred 0:04:16.520 --> 0:04:20.359 and seventy five billion dollar display advertising market. Pitch I 0:04:20.400 --> 0:04:23.400 sat down with David Rubinstein for an upcoming episode of 0:04:23.400 --> 0:04:24.680 Peer to Peer. Listen to this. 0:04:26.000 --> 0:04:28.479 That case is just getting underway, So I expected to 0:04:28.520 --> 0:04:32.640 take take some time. And you know, but look where 0:04:32.680 --> 0:04:36.560 we can figure out, you know, constructive solutions. I think 0:04:36.600 --> 0:04:40.320 we will where we think it really harms our ability 0:04:40.360 --> 0:04:43.600 to in it on behalf of our users or you know, 0:04:43.640 --> 0:04:46.120 we are going to be vigorous in defending ourselves and 0:04:46.160 --> 0:04:47.680 it's going to take time for it. 0:04:47.640 --> 0:04:52.120 To play out. And just today Google is in turn 0:04:52.320 --> 0:04:55.560 ramping up a legal fight with Microsoft over the lucrative 0:04:55.600 --> 0:04:59.320 cloud market, accusing the US software giant of abusing its 0:04:59.320 --> 0:05:03.440 market dominant. In a formal complaint to the EU's antitrust watchdog, 0:05:03.720 --> 0:05:08.159 Google alleges that Microsoft's licensing terms for Azure cloud services 0:05:08.400 --> 0:05:13.240 restricts customers accessing rival platforms including Google Cloud and Amazon 0:05:13.279 --> 0:05:14.120 Web Services. 0:05:14.279 --> 0:05:16.000 Caro Let's talk open AI. 0:05:16.400 --> 0:05:18.960 It's pitched at the Biden administration on the need for 0:05:19.360 --> 0:05:21.880 massive data centers that could each use. 0:05:22.000 --> 0:05:23.720 As much power as entire cities. 0:05:23.880 --> 0:05:26.840 Framing the expansion is necessary to develop more advanced AI 0:05:26.920 --> 0:05:28.440 models and compete with China. 0:05:28.800 --> 0:05:30.400 In MEGS, Seth Feynman joins us. 0:05:30.440 --> 0:05:34.320 Now, Seth, the context of how massive, we're talking with 0:05:35.320 --> 0:05:39.320 each dentist data center as many as three nuclear plants. 0:05:38.720 --> 0:05:41.320 As many as five nuclear clients. Now, how do you 0:05:41.400 --> 0:05:43.320 power that? Where does that power come from? Is it 0:05:43.400 --> 0:05:46.280 nuclear or other clean andergy resources? Nobody can really say it, 0:05:46.320 --> 0:05:49.160 but it speaks to the tremendous scope of ambition here 0:05:49.200 --> 0:05:52.200 that Sam Moltman and have. We've reported for a lot 0:05:52.200 --> 0:05:53.880 of this year that he's been trying to build up 0:05:53.920 --> 0:05:58.160 a coalition to support infrastructure development in the US and globally. 0:05:58.440 --> 0:06:01.400 This is really the clearest, the most explicit indication we 0:06:01.440 --> 0:06:03.599 have of what he's thinking in the US and abroad. 0:06:04.520 --> 0:06:07.080 It's based on a document the Bloomberg reviewed, right, But 0:06:07.160 --> 0:06:09.479 I do think back to February when we reported and 0:06:09.520 --> 0:06:12.520 I think you edited the story about this idea that 0:06:12.560 --> 0:06:15.800 Sam Outman also went to the US government and said, well, 0:06:15.839 --> 0:06:18.400 I kind of want some green light to take money 0:06:18.720 --> 0:06:21.040 from other investors because we know seth this is going 0:06:21.080 --> 0:06:23.520 to be really capital intensive. What do we know about 0:06:23.560 --> 0:06:24.599 that side of the story. 0:06:24.720 --> 0:06:26.720 Yes, I think there's two different things happening here. At 0:06:26.800 --> 0:06:28.800 One is if they go overseas or they work with 0:06:28.839 --> 0:06:31.240 overseas and Middle Eastern investors, that brings a whole set 0:06:31.279 --> 0:06:34.919 of complications. But the document that we've seen suggests that 0:06:35.000 --> 0:06:38.400 in the US they're thinking about probably issues around permitting 0:06:38.520 --> 0:06:42.039 and streamlining and construction and approval, getting these data centers 0:06:42.080 --> 0:06:45.200 on the grid, and ultimately trying to get all the 0:06:45.200 --> 0:06:48.599 stakeholders from grid operators to data center operators to agree 0:06:48.600 --> 0:06:50.000 on how to build this stuff. And I think what 0:06:50.040 --> 0:06:52.320 they want to see from the US, in no uncertain 0:06:52.400 --> 0:06:54.880 terms as more favorable policies. And I will say that 0:06:55.279 --> 0:06:57.320 some of that seems clear in what the White House 0:06:57.360 --> 0:06:59.880 has said after a meeting earlier this month with Sam Altman, 0:07:00.240 --> 0:07:02.640 the CEO, and Vilia and other AI executives. They are 0:07:02.680 --> 0:07:06.000 putting together task forces and engagement teams and others to 0:07:06.040 --> 0:07:09.000 help begin thinking about that streamlining process. But will they 0:07:09.040 --> 0:07:11.280 move as fast as Silicon Valley likes to move. That's 0:07:11.360 --> 0:07:12.160 very much a question. 0:07:12.400 --> 0:07:15.360 Silicon Valley is already moving the Microsoft deal that was 0:07:15.400 --> 0:07:17.240 done with Constellation, but that. 0:07:17.360 --> 0:07:21.280 Is but one nuclear plant that can be refired back up. 0:07:21.840 --> 0:07:23.760 How many years is this subultimately going. 0:07:23.720 --> 0:07:24.000 To tell you? 0:07:24.320 --> 0:07:26.400 I think that's the question everyone has. Even that facility 0:07:26.440 --> 0:07:28.520 that you're describing could take years to come up online. 0:07:28.560 --> 0:07:31.560 You know, Altman himself is investing in nuclear fission and 0:07:31.600 --> 0:07:34.960 other breakthrough technologies that may never actually come to fruition, 0:07:35.040 --> 0:07:36.720 But it feels like if you can't get it through that, 0:07:37.040 --> 0:07:38.680 how else are you going to get this much power? 0:07:40.120 --> 0:07:42.600 Some Altman's talks about how every other day he meets 0:07:42.640 --> 0:07:45.160 with a government official. This is the most important AI 0:07:45.240 --> 0:07:47.440 company in the world. Probably, so we actually have a 0:07:47.480 --> 0:07:49.880 sense of the people in the White House or in 0:07:49.920 --> 0:07:53.560 federal agencies that are responsible for interacting with him. 0:07:53.880 --> 0:07:56.320 Yeah, I think frun and Center for Altman is probably 0:07:56.400 --> 0:07:59.120 Gina Raymondo. I think Doug is his key point person 0:07:59.200 --> 0:08:01.520 that we've heard and seeing him reporting here. But I 0:08:01.560 --> 0:08:04.360 think he's probably dealing with people across government, and I'm 0:08:04.360 --> 0:08:06.160 sure that he's thinking about what would meaning if there 0:08:06.200 --> 0:08:09.320 is a change an administration going forward to continue to 0:08:09.520 --> 0:08:12.280 kind of really smooth over the path towards these kind 0:08:12.280 --> 0:08:12.960 of investments. 0:08:13.000 --> 0:08:17.280 So Bloomberg Cephoeman, who leads our AI coverage, thank you 0:08:17.400 --> 0:08:20.600 very much. Now coming up, China is complicating the electric 0:08:20.680 --> 0:08:24.000 vehicle outlook for Wall Street, with Morgan Stanley changing its 0:08:24.040 --> 0:08:26.840 tune on the auto sector at large and some analysts 0:08:27.040 --> 0:08:32.200 anticipating a stellar quarter for Tesla. And later today I'll 0:08:32.200 --> 0:08:35.160 be heading to Meta Connect where the company is likely 0:08:35.240 --> 0:08:38.160 to show off some new VR and mixed reality technology. 0:08:38.200 --> 0:08:41.480 I'm going to sit down with metas CTO Andrew Bosworth 0:08:41.559 --> 0:08:45.600 or bozz for an exclusive interview. With Metas shares currently 0:08:45.679 --> 0:08:49.400 trading very near to a fresh record high, there's a 0:08:49.400 --> 0:09:03.800 lot more to come. This is Bloomberg Technology, All streets 0:09:03.880 --> 0:09:07.240 changing its tune on US electric car makers. Morgan Stanley 0:09:07.240 --> 0:09:11.679 analyst Adam Jonas cutting his overall auto outlook amid competition 0:09:11.760 --> 0:09:14.800 in China and concerns around the US consumer. He cut 0:09:14.880 --> 0:09:19.000 Rivian to equal weight today and also downgraded GM and Ford. 0:09:19.040 --> 0:09:22.559 On the other side, there's Tesla. At least four analysts 0:09:22.559 --> 0:09:26.160 have boosted their estimates for Testa's third quarter delivery numbers 0:09:26.320 --> 0:09:29.040 which are due next week, all points to signs that 0:09:29.240 --> 0:09:34.120 sales are starting to pick up for Tesla, particularly in China. Caroline, Yeah, and. 0:09:34.080 --> 0:09:38.320 We want to discuss China evs and what, of course 0:09:38.400 --> 0:09:42.640 regulation is happening in Europe with Transporter Environment Senior Director 0:09:42.720 --> 0:09:46.199 Julia chlis Conover. Julia, we hear of some of the 0:09:46.240 --> 0:09:50.800 analyst reactions to Tesla, to Rivian, to EV's. More generally, 0:09:51.040 --> 0:09:54.600 we're also seeing governments react to the worrying from their 0:09:54.679 --> 0:09:57.400 side that EV's are very cheap when they're coming from China, 0:09:57.400 --> 0:10:00.280 and it's because they're subsidized. Just set the con text 0:10:00.360 --> 0:10:01.400 where we's done right now? 0:10:03.040 --> 0:10:05.280 Right Well, I think that's a really big debate, and 0:10:05.400 --> 0:10:08.240 right now in Brussels where I'm based, basically at every 0:10:08.280 --> 0:10:11.040 minute of the day, this is what politicians and everyone 0:10:11.120 --> 0:10:14.040 is discussing. I think the bottom line is is that 0:10:14.480 --> 0:10:17.920 China had a long term vision as well as subsidies 0:10:18.160 --> 0:10:21.679 for over a decade now to build their expertise not 0:10:21.800 --> 0:10:24.800 just in electric vehicles but the entire supply chain. They 0:10:24.800 --> 0:10:27.840 had policies on demand side, on supply side, and on 0:10:27.880 --> 0:10:32.400 this just entire ecosystem. So today they simply have excellent 0:10:32.520 --> 0:10:37.720 high quality technology, dominance of the supply chain, as well 0:10:37.760 --> 0:10:42.000 as problems domestically leading to oversupply that they want to 0:10:42.120 --> 0:10:45.120 sell across the world. I'll just put one fact for 0:10:45.160 --> 0:10:48.600 you which I think really underscores why the current ev 0:10:48.840 --> 0:10:53.040 tariff negotiations in Europe are so important. Almost sixty percent 0:10:53.160 --> 0:10:57.480 of all Chinese IVY global exports are actually go into 0:10:57.559 --> 0:11:00.400 day to Europe while the US market is closed to them. 0:11:00.679 --> 0:11:03.720 That's why the current negotiations in Brussels over the higher 0:11:03.760 --> 0:11:08.080 tariffs are so tense and really really still just heating 0:11:08.160 --> 0:11:08.800 up right now. 0:11:09.320 --> 0:11:12.320 Those duties, those tariffs can be as high as about 0:11:12.360 --> 0:11:15.000 thirty five percent on addition to the ten percent tariffs 0:11:15.000 --> 0:11:15.920 that are already existing. 0:11:16.080 --> 0:11:17.319 The breaking news today. 0:11:17.080 --> 0:11:19.520 Is that we understand the EU and China will continue 0:11:19.559 --> 0:11:23.080 negotiations even as the EU looks to perhaps pass through 0:11:23.360 --> 0:11:24.960 these tariffs as. 0:11:24.800 --> 0:11:26.600 Soon as the beginning of next month. 0:11:27.240 --> 0:11:30.680 Julia, how likely do you think it is that these 0:11:30.720 --> 0:11:32.440 tariffs will go into place? 0:11:32.679 --> 0:11:33.880 And need to go into place? 0:11:35.080 --> 0:11:36.680 So, first of all, whether or not they will go 0:11:36.760 --> 0:11:37.280 into place. 0:11:37.400 --> 0:11:41.440 As we see the situation today, it still seems likely 0:11:41.840 --> 0:11:45.760 that there is no blocking majority of Member states against 0:11:45.800 --> 0:11:48.679 the tariffs. So the vote can take place any day 0:11:48.720 --> 0:11:51.720 now until the end of October, and as of today 0:11:51.800 --> 0:11:55.920 it seems that the additional tariffs would go through. However, 0:11:56.000 --> 0:11:58.760 what we do understand is that there might be a 0:11:58.840 --> 0:12:02.920 clause put into that agreement that the European Commission and 0:12:02.960 --> 0:12:06.400 the Chinese officials can continue negotiating and if they reach 0:12:06.520 --> 0:12:10.199 some sort of deal, either on prices or volumes of inputs, 0:12:10.400 --> 0:12:13.080 this might be reversed. Now that's number one, but number 0:12:13.120 --> 0:12:16.080 two is also important. Do we need those tariffs? And 0:12:16.160 --> 0:12:19.240 when we talk about tariffs on Chinese evs in Europe, 0:12:19.280 --> 0:12:21.760 I think we always forget the other big part. 0:12:21.559 --> 0:12:22.199 Of the picture. 0:12:22.520 --> 0:12:26.320 What is missing is the affordable supply of electric vehicles 0:12:26.360 --> 0:12:30.559 in Europe by European manufacturers. Some of this is planned, 0:12:30.640 --> 0:12:34.480 so about twelve models cheaper models are planned for next 0:12:34.600 --> 0:12:38.319 year that coincide with the next set of European clean 0:12:38.400 --> 0:12:42.240 car regulations, the so called CARCO two standards. So as 0:12:42.280 --> 0:12:45.480 part of the package EV tariffs make sense. If you 0:12:45.520 --> 0:12:48.000 have standards, you have locallyvs and then you have higher 0:12:48.040 --> 0:12:49.480 tariffs if you don't have. 0:12:50.200 --> 0:12:53.600 Supply, let me jump in. Let's talk about that. Package 0:12:53.640 --> 0:12:57.679 tariffs is one factor Earlier today, Bloomberg spoke to the 0:12:57.679 --> 0:13:02.400 ACA Director general basically said there's a perfect storm for 0:13:02.440 --> 0:13:04.640 the EV industry in Europe right now. Just listen to 0:13:04.679 --> 0:13:05.640 what was said. 0:13:06.400 --> 0:13:08.440 I mean, it's all coming together at the same time. 0:13:08.920 --> 0:13:12.480 We sometimes call it the almost perfect storm happening together, 0:13:12.559 --> 0:13:14.200 and we need to deal with that. And I think 0:13:14.280 --> 0:13:18.400 now also with the green transition moving from paper to reality, 0:13:18.520 --> 0:13:21.320 reality starts to hit and we see how difficult it 0:13:21.360 --> 0:13:25.840 is to not just have mandates, regulatory mandates and models 0:13:25.840 --> 0:13:29.480 on the market manufacturers making cars, but also needing a 0:13:29.559 --> 0:13:32.679 market to pull. And that's what we are missing at 0:13:32.720 --> 0:13:33.880 the moment in Europe. 0:13:35.040 --> 0:13:38.400 That's the chief of Europe's biggest autos lobby group saying 0:13:38.520 --> 0:13:41.480 things are very hard right now for our industry. A 0:13:41.520 --> 0:13:44.120 perfect storm. Do you see a perfect storm? 0:13:45.400 --> 0:13:47.600 I do agree that things are really hard. What I 0:13:47.600 --> 0:13:51.839 think we miss in Europe is a comprehensive green automotive 0:13:51.880 --> 0:13:55.640 strategy that would have the vision industrial policy to support it, 0:13:55.640 --> 0:13:59.120 be it investment support, bit trade policy. That is not happening. 0:13:59.160 --> 0:14:02.880 And that's what we're now discussing. Following the famous drug report, 0:14:03.040 --> 0:14:05.800 so that the senior report that was presented in Europe 0:14:05.840 --> 0:14:08.640 on competitiveness. But I do have to say that all 0:14:08.679 --> 0:14:12.559 of this didn't just arrive as a complete surprise. Carmakers 0:14:12.640 --> 0:14:17.800 knew about the EVY targets for years, and we always 0:14:17.840 --> 0:14:21.600 expected the market to stagnate in twenty twenty three and 0:14:21.640 --> 0:14:24.920 twenty twenty four, then to pick up in twenty twenty five. 0:14:25.640 --> 0:14:29.120 Okay, So, Julia, what is the consumer's attitude to electric 0:14:29.240 --> 0:14:30.560 vehicles in Europe? 0:14:31.480 --> 0:14:34.560 When the vehicles are brought at right prices, consumers are 0:14:34.600 --> 0:14:37.880 flocking to buy them. The French social leasing model, where 0:14:37.920 --> 0:14:41.760 smaller affordable evs were made available, was all but subscribed. 0:14:41.880 --> 0:14:44.440 So the demand is there, but it's there at the 0:14:44.640 --> 0:14:47.880 rightly priced vehicles, and that's what we're missing. We are 0:14:47.920 --> 0:14:51.680 missing affordable models. As I said, their planned for next year. 0:14:52.040 --> 0:14:55.080 So we just need to get those models, ramp those up, 0:14:55.320 --> 0:14:58.320 bring them to consumers, and actually see how the market reacts. 0:14:58.440 --> 0:15:01.840 It's too early to judge that today in twenty twenty four, 0:15:01.880 --> 0:15:03.480 when those models are not there. 0:15:05.320 --> 0:15:09.400 Transport Environment Senior Director, Julia Poliskanova, thank you very much. 0:15:18.080 --> 0:15:20.920 The global market for AI related products will hit as 0:15:20.960 --> 0:15:24.280 much as nine hundred and ninety billion dollars in twenty 0:15:24.320 --> 0:15:28.200 twenty seven. That's according to Bain's latest Global Technology Report, 0:15:28.280 --> 0:15:31.360 saying the market will grow forty to fifty five percent 0:15:31.440 --> 0:15:34.720 annually from one hundred and eighty five billion dollars last year. 0:15:34.920 --> 0:15:37.760 The consulting firm says growth will be fueled by bigger 0:15:37.800 --> 0:15:41.720 AI systems and larger data centers to train and run them, 0:15:41.960 --> 0:15:45.920 driven by companies and governments using the technology to boost efficiency. 0:15:46.120 --> 0:15:48.520 Caract, how are they using the technology? 0:15:48.960 --> 0:15:51.080 We're going to discuss that today's AI in action. We're 0:15:51.080 --> 0:15:54.920 speaking with Accelerant. So Global risk Exchange platform for speciality 0:15:55.200 --> 0:15:58.480 insurance market is booked one than four hundred million dollars 0:15:58.520 --> 0:16:00.800 in revenue over the last twelve months. So let's talk 0:16:00.840 --> 0:16:04.440 about that growth. CEO Jeff Redkey is with us. What 0:16:04.480 --> 0:16:07.600 exactly are you doing? You're trying to seek out inefficiencies 0:16:07.680 --> 0:16:08.720 in the insurance market. 0:16:08.800 --> 0:16:09.600 Just explain for us. 0:16:09.680 --> 0:16:10.360 I think that's right. 0:16:11.760 --> 0:16:16.080 So Accelerant is trying to replatform that specialty insurance market 0:16:16.520 --> 0:16:19.600 for the benefit of everyone. Let me unpack that a little. 0:16:20.720 --> 0:16:24.760 Inside that gigantic insurance industry, there are millions of policies 0:16:24.760 --> 0:16:28.360 that aren't quite standard. Something makes them unique. That's the 0:16:28.640 --> 0:16:35.200 specialty insurance market and what we do is unlike the 0:16:35.320 --> 0:16:39.480 traditional insurance industry tech, which is so inflexible that it 0:16:39.520 --> 0:16:43.400 doesn't make sense to sort of adjust and to modernize 0:16:43.400 --> 0:16:47.000 for these specialty products. We are nimble enough that with 0:16:47.160 --> 0:16:51.880 data and technology, we provide the infrastructure that replatforms the 0:16:51.920 --> 0:16:55.239 specialty insurance onto a more modern, efficient platform. 0:16:55.920 --> 0:16:58.680 Jeff. The proof is in the numbers, it always is, right. 0:16:58.720 --> 0:17:02.160 So how does the infusia as for the technology and 0:17:02.240 --> 0:17:06.240 use of the technology translate into sales for you? The 0:17:06.280 --> 0:17:08.040 proof that people are willing to pay for it? 0:17:08.760 --> 0:17:13.000 Well, we mentioned the revenue four hundred and forty four 0:17:13.040 --> 0:17:15.720 million over the trailing twelve months, and that's up over 0:17:15.760 --> 0:17:19.399 sixty percent, So certainly we're pleased with the growth. The 0:17:19.440 --> 0:17:21.840 growth of the business through the platform has been very, 0:17:21.920 --> 0:17:27.359 very strong, as both specialty underwriters as well as the 0:17:27.440 --> 0:17:31.760 risk capital partners adopt the platform and find it incredibly helpful. 0:17:32.000 --> 0:17:33.520 So I think that would be the proof. 0:17:34.080 --> 0:17:37.680 How is AI and machine learning the key to all of. 0:17:37.640 --> 0:17:41.520 This though, Well, it all starts with the data, I think, 0:17:41.680 --> 0:17:44.160 and so week by week over the past six years, 0:17:44.200 --> 0:17:48.560 the Risk Exchange has been squirreling away data, proprietary data 0:17:48.560 --> 0:17:52.840 on policies on claims on exposure and over the course 0:17:52.880 --> 0:17:56.679 of six years that's turned into tens of millions of 0:17:56.840 --> 0:18:00.600 rows of data with over eleven thousand unique attributes. And 0:18:00.640 --> 0:18:03.520 when you point the state of the art, and that's 0:18:03.520 --> 0:18:05.560 always changing. But as you point the state of the 0:18:05.640 --> 0:18:09.560 art models, whether they're machine learning or artificial intelligence models 0:18:09.600 --> 0:18:13.919 at that data, really exciting things start to happen. You 0:18:14.080 --> 0:18:17.280 know where to grow the portfolio of insurance policies, you 0:18:17.320 --> 0:18:21.560 know where to shrink the portfolio of insurance policies, You 0:18:22.280 --> 0:18:26.080 adjust the claims much more effectively. In total, we figure 0:18:26.160 --> 0:18:29.360 that twenty percent, we're picking up a twenty percent improvement 0:18:29.359 --> 0:18:32.120 in the gross margin of the business that transfers through 0:18:32.119 --> 0:18:34.760 the risk exchange because of the machine learning and AI. 0:18:35.160 --> 0:18:35.879 That's significant. 0:18:36.000 --> 0:18:38.280 The funny thing is everyone gets so caught up with 0:18:38.400 --> 0:18:41.760 generative AI, but actually you're just you're not needing that. 0:18:41.920 --> 0:18:44.400 But will that be interwoven in some way? How will 0:18:44.400 --> 0:18:46.360 you see your model progress? 0:18:46.400 --> 0:18:49.280 Do you think the generative piece, if you use that 0:18:49.320 --> 0:18:53.720 as the classic definition, I think that usually has its 0:18:53.760 --> 0:18:57.600 best use in improving other people's the customer experience, Right, 0:18:58.640 --> 0:19:01.040 I'm not sure it'll be about insight for the portfolio. 0:19:01.160 --> 0:19:04.320 So I think it'll be instead of running a query 0:19:04.400 --> 0:19:08.159 about how your portfolio is performing, you just get to 0:19:08.200 --> 0:19:10.840 ask it in natural language, and that should speed things 0:19:10.920 --> 0:19:13.920 up and probably be more enjoyable. 0:19:13.440 --> 0:19:14.080 For the users. 0:19:15.359 --> 0:19:17.240 Jeff, we talked a lot about the AI side of 0:19:17.240 --> 0:19:20.200 the story. I'm kind of actually curious about the specialty 0:19:20.280 --> 0:19:23.840 insurance side of this story, and what does that mean. 0:19:23.880 --> 0:19:29.200 Where I am in California, issues like wildfire or like 0:19:29.440 --> 0:19:32.320 mob theft in the city of San Francisco. Believe it 0:19:32.400 --> 0:19:34.200 or not, is that what we're talking about. 0:19:34.000 --> 0:19:40.760 Here, among other things. So I guess specialty insurance means 0:19:40.840 --> 0:19:44.280 that there's a risk or an insured that in some 0:19:44.400 --> 0:19:47.320 way does not fit into a standard category where a 0:19:47.320 --> 0:19:52.280 standard product with standard rates doesn't fit. So maybe that's 0:19:52.280 --> 0:19:55.600 wildfire exposure. That's a great example, but there are many others. 0:19:55.640 --> 0:19:58.000 Maybe it's a business that's too new, or maybe it's 0:19:58.040 --> 0:20:00.280 a business that's suffered a loss in the past three 0:20:00.280 --> 0:20:03.520 to five years. There are many things which can take 0:20:03.560 --> 0:20:06.760 and insured and put it into the specialty marketplace. The 0:20:07.160 --> 0:20:10.919 underlying notion of the specialty marketplace is you need incredibly 0:20:10.960 --> 0:20:14.760 tailored data sets, and there aren't millions and millions and 0:20:14.800 --> 0:20:19.760 millions of data points about each product. There are many fewer, 0:20:19.880 --> 0:20:21.840 so you have to work harder to make the data 0:20:22.240 --> 0:20:23.520 give up those insights. 0:20:23.760 --> 0:20:27.440 Does that help? Just very quick? Are more things now ensurable? 0:20:27.480 --> 0:20:28.040 Because of this? 0:20:28.400 --> 0:20:32.160 I think more things are insurable, But more importantly, products 0:20:32.200 --> 0:20:35.880 are becoming tailored to the specific needs of various insurance 0:20:36.160 --> 0:20:40.800 so it's much less of a one size fits all market. 0:20:41.680 --> 0:20:44.480 Acceleranc CEO Jeff Radke. Great to have you on the show. 0:20:44.560 --> 0:20:47.040 Thank you. Okay, more news, It's time for talking tech 0:20:47.080 --> 0:20:51.080 and first. Reddit is hiring former Google ad executive Rit 0:20:51.240 --> 0:20:54.159 Roumoff to take on the role of chief revenue Officer. 0:20:54.440 --> 0:20:56.760 This is the social media company looks to push for 0:20:56.840 --> 0:20:59.679 a wider range of ad clients. Reddit it's been investing 0:20:59.680 --> 0:21:02.520 heavily in its ad tech as it looks to compete 0:21:02.520 --> 0:21:05.920 with the likes of Alphabet and Meta Plus. Japanese memory 0:21:06.000 --> 0:21:09.560 chipmaker Kiyoshia is delaying its IPO amid a slide in 0:21:09.600 --> 0:21:13.480 semiconductor stocks. According to sources, the company is now targeting 0:21:13.480 --> 0:21:16.960 in IPO in November or December instead of October and 0:21:17.000 --> 0:21:20.439 could raise about five hundred million dollars. And Uber expands 0:21:20.480 --> 0:21:23.760 its Robotaxi fleet. The ride sharing company will partner with 0:21:23.880 --> 0:21:28.080 Chinese autonomous car company we Ride to offer robotaxis in 0:21:28.119 --> 0:21:31.600 the United Arab emorates. Uber says users in Abadabi will 0:21:31.640 --> 0:21:34.800 be able to book rides later this year. 0:21:43.040 --> 0:21:45.240 Welcome back to Blue Meg Technology. I'm Caroline Hyde in 0:21:45.240 --> 0:21:45.880 New York. 0:21:46.040 --> 0:21:48.959 And Amed Ludlow in San Francisco. All Right, a super 0:21:49.000 --> 0:21:53.119 super well read story on Bloomberg Banking Group NG is 0:21:53.160 --> 0:21:58.160 starting to use AI to price currencies, replacing jobs traditionally 0:21:58.160 --> 0:22:01.080 performed by traders and adding to its global head of 0:22:01.119 --> 0:22:05.960 electronic trading. It performed beyond their expectations. Bloomberg's Alice Watkins 0:22:06.280 --> 0:22:09.639 joins us now and has the story and the details 0:22:09.680 --> 0:22:14.960 interesting because the AI outperforms, it replaces jobs, but there's 0:22:15.000 --> 0:22:17.440 still an element of human oversight. What do you learn 0:22:17.440 --> 0:22:21.280 about what ing is actually doing well. 0:22:21.000 --> 0:22:25.040 For the FX traders at ing, adjusting currency prices for 0:22:25.119 --> 0:22:28.080 volatility was a very time consuming task, so that's why 0:22:28.119 --> 0:22:31.760 they've built this AI model to make things easier for them, 0:22:32.160 --> 0:22:35.760 and it's actually performed a lot better than they were expecting. 0:22:35.800 --> 0:22:37.800 So they were looking at getting. 0:22:37.520 --> 0:22:41.080 Grads on the desk to perform this role, but now 0:22:41.080 --> 0:22:45.200 they're going to use the model instead. It uses reinforcement learning, 0:22:45.320 --> 0:22:48.640