WEBVTT - Nvidia’s New AI Offerings, Musk Buys More X Shares 0:00:00.960 --> 0:00:05.279 From the heart of where innovation, money and power collide 0:00:05.480 --> 0:00:10.760 in Silicon Valley and beyond. This is Bloomberg Technology with 0:00:10.920 --> 0:00:32.280 Caroline Hide and Ed Ludlow. 0:00:27.280 --> 0:00:30.080 Live from New York and Caroline Hyde and I'm Tim 0:00:30.120 --> 0:00:31.360 Steinbeck in San Francisco. 0:00:31.560 --> 0:00:33.000 This is Bloomberg. 0:00:32.600 --> 0:00:34.120 Technology coming up in video. 0:00:34.159 --> 0:00:37.040 CEO Jensen Wang unveils new AI chips and technology at 0:00:37.080 --> 0:00:40.840 the company's annual Developer conference, but fails to wow investors 0:00:41.120 --> 0:00:44.000 will have all the details. Plus China's ten Cent posts 0:00:44.040 --> 0:00:46.960 its fastest revenue growth since twenty twenty three for the 0:00:47.000 --> 0:00:50.600 company planning to ramp up AI spending and cloud flare 0:00:50.920 --> 0:00:54.520 out with new cybersecurity products aimed to help safeguard against 0:00:54.560 --> 0:00:58.560 AI vulnerabilities. Who speak with the CEO Matthew Prince a 0:00:58.640 --> 0:01:01.240 first rechecking on these markets which are not all consumed 0:01:01.240 --> 0:01:03.680 with AI but instead all consumed with the FED. 0:01:03.920 --> 0:01:05.120 We're currently up eight. 0:01:05.040 --> 0:01:07.479 Five ten percent on the Nasdaq one hundred, bouncing back 0:01:07.520 --> 0:01:10.720 off what had been a significant sell off yesterday, and 0:01:10.920 --> 0:01:14.640 overall we still contend with the macro forces versus the 0:01:14.720 --> 0:01:17.560 micro on the points higher, though Apple doing well to 0:01:17.560 --> 0:01:19.480 push the NASDAK one hundred higher so to is in 0:01:19.560 --> 0:01:21.120 video when we go into that particular name. 0:01:21.160 --> 0:01:24.160 As GtC continues, we're up nine tens percent. 0:01:24.280 --> 0:01:27.959 Remember it fell three point four percent yesterday as Jenson 0:01:28.000 --> 0:01:30.959 Wang took to the stage unveiling our whole host of 0:01:31.080 --> 0:01:36.039 new AI improvements chips that move from the Blackwell Ultra 0:01:36.200 --> 0:01:38.959 onto vira Ubin. We used to also get a new 0:01:39.040 --> 0:01:42.640 scientist name for the next iteration, but tim it didn't 0:01:42.640 --> 0:01:43.319 wow investors. 0:01:43.319 --> 0:01:45.080 It feels it didn't. 0:01:45.120 --> 0:01:47.800 There's always today and there's always the rest of the week. 0:01:47.840 --> 0:01:49.720 From more on in video, Man Deep saying a Bloomberg 0:01:49.720 --> 0:01:54.280 Intelligence joins us. Now Mandeep, there's this narrative emerging that 0:01:54.440 --> 0:01:57.280 AI costs are just getting out of control. What did 0:01:57.320 --> 0:01:59.840 we see yesterday? What did you see yesterday from Nvidia 0:02:00.080 --> 0:02:03.200 said to customers, you're getting more value out of what 0:02:03.240 --> 0:02:03.840 we're bringing you. 0:02:04.560 --> 0:02:08.600 Well, one, there was a clear shift towards reasoning, and 0:02:08.680 --> 0:02:11.080 you know the fact that he said reasoning requires one 0:02:11.160 --> 0:02:15.880 hundred times more compute. So from that perspective, you know, 0:02:16.000 --> 0:02:19.560 the demand trends seem to be reassuring coming from Nvidia, 0:02:19.639 --> 0:02:22.440 and clearly they are ahead when it comes to you know, 0:02:22.480 --> 0:02:25.360 the performance per what that everyone seems to be focused 0:02:25.400 --> 0:02:28.440 on with these chips, and look, you could argue, you know, 0:02:29.040 --> 0:02:32.000 there isn't a new end market on the horizon, even 0:02:32.040 --> 0:02:34.760 though they announced, you know, the partnership with GM or 0:02:35.600 --> 0:02:38.840 the Humanoids was part of the keynote as well. But 0:02:39.560 --> 0:02:42.680 to me, it's still all about data centers and servers, 0:02:43.160 --> 0:02:46.079 and they seem to have a very clear roadmap till 0:02:46.120 --> 0:02:48.720 twenty twenty seven in terms of what kind of performance 0:02:48.760 --> 0:02:52.040 improvements they're looking with the next version of their chips. 0:02:52.280 --> 0:02:55.400 I think via Rubin expected to three x the improvements 0:02:55.400 --> 0:02:58.800 on Blackwell Ultra. We get Feynman being the next name 0:02:58.880 --> 0:03:03.480 for the future platform a focus. But look, analysts overall 0:03:03.720 --> 0:03:06.280 once again declaring that the moat is strong, that maybe 0:03:06.320 --> 0:03:08.320 eight years ahead of the competition. If you're talking to 0:03:08.320 --> 0:03:11.760 certain hyperscalers, what more can you do to douce what's 0:03:11.840 --> 0:03:13.560 currently in the valuation right now? 0:03:13.639 --> 0:03:16.880 Well, so, if you look at the trajectory of Nvidia's numbers, 0:03:17.000 --> 0:03:20.000 they've always moved to the upside because they have come 0:03:20.080 --> 0:03:23.600 up with something new or you know, an additional customer 0:03:23.720 --> 0:03:25.679 that is of the same ramp. 0:03:25.480 --> 0:03:26.040 Up all time. 0:03:26.080 --> 0:03:27.040 It's not new enough for that. 0:03:27.200 --> 0:03:30.200 Well, so you have to look at the hyper scale investment, 0:03:30.320 --> 0:03:33.919 and they are all developing their own A six. So 0:03:34.080 --> 0:03:37.680 the fact that you have competition on the horizon, even 0:03:37.680 --> 0:03:40.440 though Nvidia is way ahead of everyone else in terms 0:03:40.520 --> 0:03:43.920 of performance, but you just can't rule out the fact 0:03:43.960 --> 0:03:47.160 that A six are a real possibility. That's what Broadcom 0:03:47.240 --> 0:03:51.160 called out, and to me, that competitive environment will get challenging. 0:03:51.200 --> 0:03:54.800 And look, they made the pivot to reasoning after Deep Seek. 0:03:55.120 --> 0:03:57.680 This wasn't six they're six months back. It was all 0:03:57.680 --> 0:04:00.840 about pre training. The keynote this time around was all 0:04:00.880 --> 0:04:04.840 about inferencing and reasoning. So they're very good at making 0:04:04.880 --> 0:04:08.040 those pivots. But the fact is the market keeps evolving 0:04:08.160 --> 0:04:11.280 and I think A six are still a real threat 0:04:11.320 --> 0:04:13.960 when it comes to in video sales at the data 0:04:13.960 --> 0:04:17.400 center level. Yes, it's a one trillion dollar market, but 0:04:17.800 --> 0:04:20.120 who can give the assurance that it will be in 0:04:20.240 --> 0:04:24.479 Vidia who captures most of that in market because your 0:04:24.520 --> 0:04:27.840 competition is the four or five main hyperscalers that are 0:04:27.920 --> 0:04:29.880 still looking to develop that ac. 0:04:29.600 --> 0:04:31.120 Option frenemies continue. 0:04:31.120 --> 0:04:33.919 Man Deep seeing of Blue meg Intelligence love having him. Meanwhile, 0:04:33.960 --> 0:04:36.479 a wider investor look now at in Nvidia and some 0:04:36.520 --> 0:04:38.640 of the others in the chip sector, Kim forrest A 0:04:38.720 --> 0:04:42.800 Capital Partners CIO joins us and Kim. We mentioned how 0:04:42.920 --> 0:04:45.680 their own clients are becoming their competitors. Not to mention 0:04:46.040 --> 0:04:48.599 AMD still trying to be and also Ran in this space, 0:04:48.640 --> 0:04:52.480 not to mention Intel really languishing. But can Invidia remain 0:04:52.880 --> 0:04:54.360 the king when it comes to AI? 0:04:56.320 --> 0:04:59.480 Well maybe I think it's a strong maybe there, and 0:04:59.520 --> 0:05:03.320 maybe they do rule the hardware world or a lot 0:05:03.320 --> 0:05:07.279 of it, because you know, there really is no substitute 0:05:07.360 --> 0:05:12.320 right now. Here's the thing, though, human beings are incredibly inventive, 0:05:12.920 --> 0:05:16.919 and in Nvidia, although they always trot out the fact 0:05:16.920 --> 0:05:22.000 that they have software that developers may use, I think 0:05:22.040 --> 0:05:26.000 the very largest developers that are using these chips are 0:05:26.040 --> 0:05:29.040 writing their own software, not depending on KUDA. 0:05:29.720 --> 0:05:30.680 So keep that in the. 0:05:30.640 --> 0:05:31.640 Back of your mind. 0:05:31.760 --> 0:05:35.440 But I believe what deep Seek has shown the investing 0:05:35.520 --> 0:05:38.320 world is there's more than one way to skin a cat, 0:05:38.720 --> 0:05:42.520 and just throwing really expensive chips and tons of chips 0:05:42.520 --> 0:05:45.920 at a problem is not always the answer. So I'm 0:05:45.960 --> 0:05:49.240 looking out there for innovation in the software world to 0:05:49.400 --> 0:05:53.720 make the build out of a data center a little 0:05:53.839 --> 0:06:00.680 less like key into developing AI in the future, software 0:06:00.920 --> 0:06:02.839 less hardware Kim. 0:06:02.680 --> 0:06:05.360 Do you think that the market has fully priced in 0:06:05.480 --> 0:06:06.600 the impact of deep seek? 0:06:09.640 --> 0:06:09.680 No? 0:06:10.240 --> 0:06:14.640 Actually, and I think this is a really difficult area 0:06:14.680 --> 0:06:18.240 to talk about AI without really nerding out. So let's 0:06:18.320 --> 0:06:20.279 keep it at the fifty thousand foot level. 0:06:21.200 --> 0:06:22.919 This is a show to not out on, Kim. This 0:06:23.040 --> 0:06:24.080 is a show to net out on. 0:06:24.080 --> 0:06:27.880 I'm just saying, okay, yeah, I'm yeah, okay, I'm trying 0:06:27.880 --> 0:06:31.440 to restrain myself. But here's here's what I see is 0:06:31.560 --> 0:06:36.840 the essential problem is there are limitations in the physical 0:06:36.839 --> 0:06:40.240 world power. You know, how quickly you can build a 0:06:40.320 --> 0:06:44.560 data center, how quickly you can build chips, the machines 0:06:44.600 --> 0:06:48.160 that create these exotic chips, how quickly you can build that. 0:06:48.560 --> 0:06:52.760 It's all, you know, a physical world problem. In many ways, 0:06:52.800 --> 0:06:55.640 software is exactly what we've used in the past to 0:06:55.680 --> 0:06:59.800 become more productive, and why not use it on AI itself, 0:07:00.320 --> 0:07:05.480 to use these data centers that we already have and 0:07:05.680 --> 0:07:10.000 rethink the software and the training, testing and validation kind 0:07:10.040 --> 0:07:14.120 of exercises we do to build large language models. Why 0:07:14.160 --> 0:07:16.840 not make that better by better software? 0:07:17.280 --> 0:07:19.560 Crazy, Well, you have. 0:07:19.560 --> 0:07:21.880 The background to nerd out on this, because before finance, 0:07:21.920 --> 0:07:23.679 you spend more than a decade working on this stuff. 0:07:23.680 --> 0:07:25.560 I mean this was twenty years ago, so you've been 0:07:25.560 --> 0:07:26.640 thinking about this stuff for. 0:07:26.560 --> 0:07:27.239 A long time. 0:07:27.680 --> 0:07:31.240 But now, do the economics of inference work? Do the 0:07:31.560 --> 0:07:35.400 economics of agentic models work right now? 0:07:36.600 --> 0:07:41.520 I don't think so, because they're so expensive to build 0:07:42.160 --> 0:07:44.360 and it doesn't look like people are lining up to 0:07:44.400 --> 0:07:47.520 buy it. I mean, is it really that simple? I 0:07:47.560 --> 0:07:50.960 think the answer is yes. I have not been convinced 0:07:51.000 --> 0:07:54.680 that there is a strong enough problem that can be 0:07:54.760 --> 0:08:00.360 easily identified and thus monetized in AI. I think it's 0:08:00.360 --> 0:08:04.520 the future, absolutely, but I think that businesses who have 0:08:04.600 --> 0:08:09.320 to put their assets at risk buying this stuff. 0:08:09.600 --> 0:08:11.080 And then using it. 0:08:11.120 --> 0:08:14.240 And remember a lot of language large language models are 0:08:14.320 --> 0:08:18.840 eighty five to ninety five percent correct, and that is underwhelming. 0:08:19.080 --> 0:08:20.920 You know, if you're going to build your business on 0:08:21.000 --> 0:08:25.480 something that's incorrect, you know, fifteen to ten to fifteen 0:08:25.480 --> 0:08:30.280 percent of the time. Yeah, these are real world problems 0:08:30.320 --> 0:08:34.360 that we have to get unexcited about and fix. And 0:08:34.400 --> 0:08:37.920 again my answer is software, because that's what I what 0:08:38.000 --> 0:08:41.000 about my book. I know that that area, as opposed 0:08:41.040 --> 0:08:43.120 to building better hardware, we'll. 0:08:43.040 --> 0:08:43.800 Go to the software. 0:08:43.840 --> 0:08:46.920 Go to the newly released open source software dynamo and look, 0:08:46.960 --> 0:08:49.920 there's growing that that makes deep sinks are one thirty 0:08:50.400 --> 0:08:54.000 x faster when you run this new open source software 0:08:54.000 --> 0:08:56.520 being offered by n videos. So are they not adapting 0:08:56.600 --> 0:08:58.480 quickly enough because it centainly feels as though they've gone 0:08:58.520 --> 0:08:59.240 answer for everything. 0:09:00.360 --> 0:09:02.559 Well, I think we really have to do is attack. 0:09:02.760 --> 0:09:05.520 I mean, faster is better, right, But then how do 0:09:05.600 --> 0:09:08.440 we stop wrong answers? That's where the humans have to 0:09:08.480 --> 0:09:11.920 come in and kind of build stuff around that to 0:09:12.040 --> 0:09:15.840 wall off the bad answers. And you know, whenever I 0:09:16.080 --> 0:09:19.200 was doing this back in the late nineties, that's where 0:09:19.240 --> 0:09:22.520 a lot of our time was spent on refining the models, 0:09:22.960 --> 0:09:25.880 was to make sure that we weren't returning bad answers. 0:09:26.120 --> 0:09:28.679 And I don't really see that as a focus. And 0:09:29.240 --> 0:09:34.080 again I'm talking my book what I know, And it's 0:09:34.120 --> 0:09:37.160 still a problem. That's the crazy thing, you know, thirty 0:09:37.240 --> 0:09:40.320 years later, wrong answers are still a problem. 0:09:41.080 --> 0:09:43.760 Kim forrest A Bocha Capital Partners is Kim always great 0:09:43.760 --> 0:09:44.160 to check in with. 0:09:44.240 --> 0:09:45.880 You appreciate you joining us today. 0:09:46.400 --> 0:09:49.040 Meanwhile, as many as four hundred thousand in Vidio chips 0:09:49.040 --> 0:09:51.160 are set to have a new home in the small 0:09:51.240 --> 0:09:54.320 city of Abilene, Texas. It's the site of the first 0:09:54.440 --> 0:09:58.760 data center complex, or open AI's Stargate infrastructure venture. According 0:09:58.760 --> 0:10:01.880 to the developer the prom project, it will be completed 0:10:01.920 --> 0:10:04.760 by mid twenty twenty six with support for hundreds of 0:10:04.760 --> 0:10:07.679 thousands of AI chips, though no word yet on how 0:10:07.720 --> 0:10:10.439 many chips have been committed to the project so far. 0:10:10.720 --> 0:10:15.480 Caroline, coming up, we're talking connectivity and connectivity firm cloud 0:10:15.480 --> 0:10:19.560 Flair introducing new AI tools to take on vulnerabilities. When 0:10:19.600 --> 0:10:21.200 we're speaking with a CEO, Matthew Prince. 0:10:21.240 --> 0:10:22.800 That's next. This is room meg Technology. 0:10:34.440 --> 0:10:37.120 Cloud connectivity firm cloud Flare is in the midst of 0:10:37.160 --> 0:10:40.160 its annual security Week, announcing a slew of new products, 0:10:40.160 --> 0:10:43.920 including one to help safeguard against AI vulnerabilities. Here to 0:10:43.920 --> 0:10:48.240 discusses cloud Flare CEO Matthew Prince. Matthew just announced cloud 0:10:48.280 --> 0:10:51.560 Flare for AI this week. It essentially allows customers to 0:10:51.679 --> 0:10:54.920 safely use and develop AI tech. Up to now, though, 0:10:55.320 --> 0:10:57.320 what has been the problem, what has been the challenge? 0:10:57.360 --> 0:10:58.560 What has been the vulnerability? 0:10:59.800 --> 0:11:02.560 Thing that's really hard in the AI spaces, as your 0:11:02.600 --> 0:11:05.280 last segment was talking about, is everyone's just rushing as 0:11:05.320 --> 0:11:07.760 fast as possible everyone's bored is saying what are we 0:11:07.840 --> 0:11:10.120 doing in AI? And everyone is trying to go as 0:11:10.160 --> 0:11:12.840 fast as possible and what that means is mistakes get made, 0:11:12.920 --> 0:11:16.120 Data that shouldn't leave your enterprise gets sucked. 0:11:15.840 --> 0:11:17.040 Up into an AI model. 0:11:17.480 --> 0:11:20.439 People, maybe you're using AI in a way which might 0:11:20.480 --> 0:11:22.839 embarrass you or your firm. And what we've heard from 0:11:22.840 --> 0:11:25.679 our customers is that they want guardrails, they want controls 0:11:25.720 --> 0:11:27.560 to be able to make sure that even as they 0:11:27.600 --> 0:11:30.360 go as fast as possible, they're not going to break anything, 0:11:30.400 --> 0:11:31.679 which is really really important. 0:11:31.720 --> 0:11:34.360 And so that's what we're launching with cloud Flare for AI, and. 0:11:34.240 --> 0:11:37.560 We're really excited to see already customers using it to 0:11:37.600 --> 0:11:39.600 make sure they can go as fast as possible but 0:11:39.720 --> 0:11:41.080 not make those critical mistakes. 0:11:41.120 --> 0:11:42.239 In the space of AI. 0:11:42.760 --> 0:11:46.160 MAV you've got so many current security offerings already when 0:11:46.160 --> 0:11:49.640 you think about web applications, firewalls of ability, API protection. 0:11:49.960 --> 0:11:52.319 What's new here in terms of cloud Flare for AI. 0:11:52.400 --> 0:11:55.000 Is it a new bundling thematic or is there really 0:11:55.000 --> 0:11:55.880 new offerings coming? 0:11:56.720 --> 0:11:59.080 You know, I think that it is a combination of 0:11:59.200 --> 0:12:02.120 building on the foundation that we have where we can 0:12:02.200 --> 0:12:05.520 process just vast amounts of data flowing through our systems, 0:12:05.559 --> 0:12:08.640 do analysis do those controls? But then adding on top 0:12:08.720 --> 0:12:11.040 of it, not just how are we looking for hackers, 0:12:11.040 --> 0:12:12.960 how are we looking for threats? 0:12:13.000 --> 0:12:15.280 But how are we actually looking for the ways. 0:12:15.000 --> 0:12:17.640 That people might be using AI in a way that 0:12:17.720 --> 0:12:19.480 might be embarrassing to a company, in a way that 0:12:19.600 --> 0:12:22.280 might be revealing secrets that are there. And I think 0:12:22.440 --> 0:12:24.280 part of what's powerful here is, yes, we have a 0:12:24.280 --> 0:12:27.640 whole bunch of customers that care about keeping their use 0:12:27.640 --> 0:12:29.880 of AI secure, but a lot of the AI companies, 0:12:29.920 --> 0:12:32.920 in fact, most of the major AI companies also use Cloubler. 0:12:33.000 --> 0:12:34.560 So we've been able to actually. 0:12:34.240 --> 0:12:36.800 Integrate those two sides in a way that give you 0:12:36.880 --> 0:12:39.520 a holistic offering and make sure that you can again 0:12:39.640 --> 0:12:41.960 go as fast as possible with AI, but do it 0:12:41.960 --> 0:12:44.920 in a way which is secure and has real guardrails. 0:12:44.840 --> 0:12:47.760 That speed, that desire to get out in front. When 0:12:47.760 --> 0:12:50.559 it comes to AI, many would cass as hype. Can 0:12:50.600 --> 0:12:53.040 you just set the broader context here of whether you 0:12:53.080 --> 0:12:56.440 think the AI hype is turning into AI productivity reality? 0:12:57.440 --> 0:12:57.640 Yeah? 0:12:57.640 --> 0:13:00.880 You know, I think there's anytime you have something which 0:13:00.920 --> 0:13:04.640 is as disruptive a new technology is AI. There are 0:13:04.679 --> 0:13:06.959 going to be a certain amount of money which is 0:13:07.000 --> 0:13:08.800 spent on it, which is just lit on fire. And 0:13:08.840 --> 0:13:11.040 I think that's just the reality that's there. But what 0:13:11.080 --> 0:13:13.400 we're seeing is that there's real value which is being 0:13:13.480 --> 0:13:16.120 created as well. And so even if we assume that 0:13:16.200 --> 0:13:19.880 only one percent of what's being done today turns into value, 0:13:19.920 --> 0:13:22.520 what we see from our customers, what we see from 0:13:22.520 --> 0:13:24.520 the AI companies that are out there, is that that 0:13:24.559 --> 0:13:27.520 one percent has the opportunity to have enormous returns. And 0:13:27.559 --> 0:13:29.959 so again, I think anyone who's not doing at least 0:13:29.960 --> 0:13:33.240 something experimenting in AI is probably missing the boat here. 0:13:33.640 --> 0:13:35.520 And again I don't hear a lot of customers that 0:13:35.559 --> 0:13:38.760 are saying, Gosh, we're going to shut down our AI efforts. Instead, 0:13:38.800 --> 0:13:41.400 I think people are stepping on the accelerator, finding more 0:13:41.440 --> 0:13:43.439 ways to use this. And again I think we have 0:13:43.480 --> 0:13:45.199 to acknowledge some of this is going to be just 0:13:45.280 --> 0:13:47.240 lighting money on fire, but a lot of it is 0:13:47.240 --> 0:13:49.600 going to turn into real innovation over the long term. 0:13:50.640 --> 0:13:54.080 Matthew and investors watching right now probably are wondering how 0:13:54.120 --> 0:13:56.880 big of a business does this turn into. For cloud Flair, 0:13:56.960 --> 0:13:58.200 what are you modeling internally? 0:14:00.000 --> 0:14:02.599 Again, we don't forecast anything in particular, but what I 0:14:02.600 --> 0:14:04.880 will say is customers are coming to us, and this 0:14:05.000 --> 0:14:08.360 was a real demand that was driven by our customers 0:14:08.400 --> 0:14:11.200 saying you are extremely well positioned to be able to 0:14:11.280 --> 0:14:12.360 solve this problem. 0:14:12.440 --> 0:14:14.800 This is something that we want and that we want 0:14:14.840 --> 0:14:15.400 you to build. 0:14:15.640 --> 0:14:17.079 And so I think that that gives us a lot 0:14:17.080 --> 0:14:19.520 of confidence that as we invest in this space, that 0:14:19.680 --> 0:14:22.080 is something where there will be real returns. It's something 0:14:22.080 --> 0:14:24.680 that customers want, it's something that the largest enterprises are 0:14:24.720 --> 0:14:25.360 realizing that. 0:14:25.320 --> 0:14:27.120 They need that they can't just. 0:14:27.120 --> 0:14:31.240 Go at the full throttle without those guardrails in place. 0:14:31.240 --> 0:14:33.320 And because of the unique position that cloud Flairer is in, 0:14:33.560 --> 0:14:35.560 we can deliver those guardrails and make sure that you 0:14:35.600 --> 0:14:36.640 can use AI safely. 0:14:36.840 --> 0:14:39.800 Because you're offering the platform to build to deploy the apps, 0:14:39.800 --> 0:14:42.440 you're also offering the security. But paint a picture of 0:14:42.440 --> 0:14:44.600 the ecosystem right now because people might be thinking, Okay, 0:14:44.640 --> 0:14:47.560 your competitors are like AWS and fastly. But then there's 0:14:47.560 --> 0:14:50.240 this deal with Google buying wiz and what that does 0:14:50.320 --> 0:14:53.800 for overall security of infrastructure, cloud infrastructure. 0:14:53.840 --> 0:14:56.400 Where do you fit in with cloud flair Matthew, I. 0:14:56.400 --> 0:14:58.320 Mean those are all very different companies, and I think 0:14:58.320 --> 0:15:00.880 the first thing to say is congratulate to the Whiz team. 0:15:00.880 --> 0:15:02.960 It's incredibly hard to build companies, and the fact that 0:15:03.000 --> 0:15:06.400 they so quickly generate build something that had enormous value 0:15:06.600 --> 0:15:08.440 and we're able to sell it to Google, I think 0:15:08.720 --> 0:15:12.280 is something that again deserves an incredible amount of tip 0:15:12.320 --> 0:15:14.360 of the hat to them and what they're doing. 0:15:14.520 --> 0:15:17.000 I think the fact that Google had to build Whiz. 0:15:16.920 --> 0:15:20.000 Just shows how security has to be part of any 0:15:20.040 --> 0:15:22.560 cloud platform, and cloud Flair really started with security, and 0:15:22.600 --> 0:15:25.240 I think that's given us a huge advantage over over time. 0:15:25.520 --> 0:15:29.360 Wiz is really great at identifying what an enterprise's security 0:15:29.400 --> 0:15:32.240 vulnerabilities are. And what we're proud of at cloud Flair 0:15:32.320 --> 0:15:34.960 is that we have actually been the solution once Wiz 0:15:34.960 --> 0:15:38.200 that identified the problem to actually solving the security problem 0:15:38.240 --> 0:15:41.280 behind the scenes, where they're recommending us as that solution. 0:15:41.360 --> 0:15:43.400 Google's been a long term partner of cloud Flair. We 0:15:43.480 --> 0:15:44.480 hope that will continue. 0:15:44.720 --> 0:15:46.840 I expect that will continue over over time, and I 0:15:46.840 --> 0:15:49.720 think integrating Whiz actually makes it make more sense for 0:15:49.760 --> 0:15:52.640 them to be integrating closely with cloud Flair and being 0:15:52.680 --> 0:15:53.520 able to deliver our. 0:15:53.400 --> 0:15:55.920 Services as one seamless, offering some of. 0:15:55.840 --> 0:15:58.320 The best as I perspective of the internet. Writ large, 0:15:58.320 --> 0:16:00.400 we love having you on Matthew prints cloud flairer CEO. 0:16:00.560 --> 0:16:03.800 Thanks so much, Tim I forgot well. 0:16:03.680 --> 0:16:05.280 So I'm now for talking tech. 0:16:05.560 --> 0:16:09.240 First up, Google launch Date redesigned version of its budget 0:16:09.280 --> 0:16:12.600 Pixel phone, upgrading the chip and battery life weeks after 0:16:12.640 --> 0:16:15.880 Apple's lower end iPhone sixteen went on sale. The new 0:16:15.880 --> 0:16:18.200 Pixel nine A costs four hundred and ninety nine dollars, 0:16:18.320 --> 0:16:21.120 in line with the price of last year's Pixel eight 0:16:21.200 --> 0:16:25.920 A plus. Tesla was granted approval to begin carrying passengers 0:16:25.960 --> 0:16:29.560 in California. The approval does not allow Tesla to offer 0:16:29.680 --> 0:16:33.040 rides in autonomous vehicles, but a ride hailing business with 0:16:33.160 --> 0:16:36.360 human drivers could pave the way for Tesla to eventually 0:16:36.400 --> 0:16:40.640 introduce robotaxi service in the state where Weimo already operates 0:16:40.640 --> 0:16:44.960 here in San Francisco, and Shaomi is working to expand 0:16:45.040 --> 0:16:48.640 its production capabilities to meet demand. This comes after it 0:16:48.760 --> 0:16:51.800 raised its twenty twenty five delivery target for evs to 0:16:51.840 --> 0:16:55.480 three hundred and fifty thousand units, reflecting inroads into the 0:16:55.560 --> 0:16:59.720 Chinese EV market. Shami Group Vice president and CFO A 0:17:00.120 --> 0:17:03.520 Lamb joined a Bloomberg Television in an exclusive interview to 0:17:03.560 --> 0:17:04.199 weigh in on this. 0:17:05.160 --> 0:17:08.320 You see a couple of our evs right that, the 0:17:08.359 --> 0:17:11.200 red one being the Strew seven, and then the yellow one, 0:17:11.200 --> 0:17:13.520 which is something that we just launched, is our s 0:17:13.600 --> 0:17:16.080 U seven Ultra, which we are start selling at a 0:17:16.119 --> 0:17:19.080 starting price of over five hundred thousand RMP. We are 0:17:19.160 --> 0:17:23.159 running one factory right now in Beijing which is fully 0:17:23.240 --> 0:17:25.879 rammed up. We are working very hard to squeeze the 0:17:25.920 --> 0:17:29.080 extra production volume out of that out of that factory. 0:17:29.440 --> 0:17:32.000 But at the same time, obviously we need to continue 0:17:32.280 --> 0:17:43.960 to expand our production capabilities by adding new production sites. 0:17:45.200 --> 0:17:48.000 Ten Cent posted its fastest pace of revenue grows is 0:17:48.000 --> 0:17:51.040 twenty twenty three revenue rising eleven percent for the three months. 0:17:50.960 --> 0:17:53.720 Ending in December, and it also shed some light on 0:17:53.760 --> 0:17:54.840 its AI ambitions. 0:17:55.240 --> 0:17:57.199 Makes Henry ren joys us now for a breakdown and 0:17:57.240 --> 0:17:58.959 ponymar are really setting out that they're kind of going 0:17:59.000 --> 0:18:00.800 to do in AI what they did for gaming with 0:18:01.000 --> 0:18:02.800 their own but also of the third parties. 0:18:04.640 --> 0:18:07.960 Indeed, so so breakdown for results shows that the company 0:18:08.000 --> 0:18:10.800 is doing pretty strong across it's all its business lines, 0:18:10.800 --> 0:18:14.200 from gamings to social network of free chat as well 0:18:14.240 --> 0:18:17.159 as the payment and fintech, but of course AI is 0:18:17.200 --> 0:18:21.240 the key business focused for today, and the company said 0:18:21.240 --> 0:18:24.679 that it made it so called emergency purchase of GPUs 0:18:24.760 --> 0:18:27.200 in the fourth quarter because it was seeing a search 0:18:27.400 --> 0:18:30.440 in AI demound. And indeed we're seeing this coming through 0:18:30.480 --> 0:18:32.280 on the revenue side of things as well, because the 0:18:32.280 --> 0:18:36.520 company said that it's AI cloud business double this revenue 0:18:36.560 --> 0:18:40.080 in the fourth quarter. And just to give you a 0:18:40.160 --> 0:18:42.600 sense of things in terms of how much the company 0:18:42.640 --> 0:18:44.879 is spending on AI these days, the company said that 0:18:44.920 --> 0:18:49.240 it's fourth quarter capital spending quadrupled from the same quarter 0:18:49.320 --> 0:18:52.000 in twenty twenty three. So it really shows you the 0:18:52.200 --> 0:18:55.879 urgency that the company is having in terms of catching 0:18:55.960 --> 0:18:59.000 up on other Chinese big tech companies in terms of 0:18:59.280 --> 0:19:02.080 you know, enhancing its AI capabilities. 0:19:03.400 --> 0:19:06.240 Whomberg's Henry Wrenn joining us from London. Henry, good to 0:19:06.240 --> 0:19:09.159 see you, Thanks so much. Meanwhile, Elon Musk invested a 0:19:09.160 --> 0:19:12.159 one hundred and fifty million dollars to acquire more shares 0:19:12.240 --> 0:19:14.800 in X last year at evaluation, approaching the price he 0:19:14.880 --> 0:19:18.000 paid for the company's equity back in twenty twenty two. 0:19:18.359 --> 0:19:21.760 Woomberg's Kurt Wagner joins us now with more Kurt, what's 0:19:21.800 --> 0:19:22.640 the endgame here? 0:19:22.720 --> 0:19:25.000 Does he just want to own the entire thing? 0:19:26.440 --> 0:19:29.439 Yeah, I mean he already owns around seventy five percent 0:19:29.480 --> 0:19:32.240 of X, So I'm not sure how much of this is. 0:19:32.280 --> 0:19:35.040 You know, buying one hundred and fifty million more in 0:19:35.080 --> 0:19:37.679 shares is not going to dramatically change his ownership. It 0:19:37.720 --> 0:19:39.760 could be he was, you know, buying out somebody who 0:19:39.800 --> 0:19:42.600 wanted out. I think more than that, it's it's perhaps 0:19:42.800 --> 0:19:47.480 a signal, right that this company is still valued at 0:19:47.560 --> 0:19:50.520 where he or close to where he purchased it. We 0:19:50.600 --> 0:19:53.639 know they're out raising money right now from other investors, 0:19:53.640 --> 0:19:55.800 and so maybe having that sale at the end of 0:19:55.840 --> 0:19:59.080 last year sort of setting the benchmark, right, Like, if 0:19:59.119 --> 0:20:01.400 Elon's willing to buy in close to what he paid 0:20:01.440 --> 0:20:04.040 for it, maybe others should as well. So perhaps it 0:20:04.119 --> 0:20:06.240 was a little bit more of a signal than anything. 0:20:07.160 --> 0:20:09.960 The signal was an amazing deep dive by you and 0:20:10.000 --> 0:20:13.600 other reporters into Kingdom Holdings and your report. That's a 0:20:13.640 --> 0:20:16.080 Saudi investment firm that seems to where we got the 0:20:16.080 --> 0:20:18.720 tip off that he was offering these sorts of prices 0:20:18.720 --> 0:20:19.440