WEBVTT - AWS CEO Matt Garman Lays Out Amazon’s AI Plans 0:00:02.520 --> 0:00:09.840 Bloomberg Audio Studios, podcasts, radio news from the heart of 0:00:09.920 --> 0:00:15.640 where innovation, money and power collide in Silicon Valley and beyond. 0:00:16.040 --> 0:00:21.320 This is Bloomberg Technology with Caroline Hyde and Ed Ludlow. 0:00:35.640 --> 0:00:39.080 Live from San Francisco. This is Bloomberg Technology coming up. 0:00:39.159 --> 0:00:43.000 US President Trump accuses China of violating an agreement with 0:00:43.080 --> 0:00:47.200 the US to east tariffs, escalating tensions between the two countries. 0:00:47.440 --> 0:00:48.839 Plus sales of Apple's. 0:00:48.479 --> 0:00:51.400 iPhone are set to take a big hit from Trump's 0:00:51.440 --> 0:00:54.240 new tariff policy. We get into the numbers and we 0:00:54.280 --> 0:00:57.760 speak with the CEO of AWS as the leading Hyperscala 0:00:57.840 --> 0:01:01.600 runs faster on a What a week it has been 0:01:01.640 --> 0:01:02.760 in the world of technology. 0:01:02.760 --> 0:01:04.120 I'm looking at the Nazak one hundred. 0:01:04.160 --> 0:01:07.759 It is a megacap and technology heavy index, and over 0:01:07.800 --> 0:01:10.280 the course of the week we're up a percentage point. 0:01:10.280 --> 0:01:13.040 We're actually lower in Friday session, but this is a 0:01:13.080 --> 0:01:16.920 rebound from last week where earnings and that tariff story 0:01:17.160 --> 0:01:19.480 have been the driver of the markets and the headlines. 0:01:19.520 --> 0:01:21.759 There are two names in particular that we have been 0:01:21.800 --> 0:01:24.520 so focused on. Those names are in video. Of course, 0:01:24.560 --> 0:01:28.600 we had earnings twenty four hours ago, thirty eight hours ago, 0:01:28.760 --> 0:01:31.400 forty eight hours ago, and then we spoke to Gensen one. 0:01:31.560 --> 0:01:34.399 It's giving some of the gains that it had seen back, 0:01:34.720 --> 0:01:37.560 but there is a big question still about China and 0:01:37.600 --> 0:01:40.479 cross border trade of the lead edge technology. And then 0:01:40.520 --> 0:01:43.880 Apple again softer four tens and one percent just twenty 0:01:43.880 --> 0:01:46.880 four hours ago, and Video was the world's most valuable company. 0:01:47.040 --> 0:01:49.960 Apple had regained some status as a three trillion dollar club. 0:01:50.120 --> 0:01:52.320 We're under pressure. Now. Here's the top story. 0:01:52.640 --> 0:01:56.080 China's planning to allocate seventy billion dollars of capital in 0:01:56.160 --> 0:01:59.680 a new effort to fast track new infrastructure projects and 0:02:00.160 --> 0:02:02.520 up its own economy from ongoing US tariffs. 0:02:02.520 --> 0:02:03.520 That's according to sources. 0:02:03.560 --> 0:02:07.960 Meanwhile, President Trump reigniting those tariff war concerns, posting on 0:02:08.000 --> 0:02:12.079 True Social that China has quote totally violated its agreement 0:02:12.200 --> 0:02:16.520 with US or US, offering no details of what agreements 0:02:16.600 --> 0:02:20.240 China could have violated. Let's get more from Bloomberg's Mike Shephard. 0:02:20.520 --> 0:02:23.160 Do we know what the president is talking about here? 0:02:24.440 --> 0:02:27.280 Well, we haven't gotten a clear signal from him directly 0:02:27.360 --> 0:02:30.080 about what had set him off this morning, but we 0:02:30.200 --> 0:02:32.880 have seen ed over the past few weeks since that 0:02:33.160 --> 0:02:36.880 sit down in Geneva, between Treasury Secretary Scott Besson and 0:02:36.919 --> 0:02:40.360 his Chinese counterparts signed that that little agreement that that 0:02:40.560 --> 0:02:44.320 daytime that they had forged there in Switzerland was starting 0:02:44.360 --> 0:02:47.520 to come apart. The Chinese government had objected to some 0:02:47.600 --> 0:02:52.440 export controls that the US has increasingly been applying to China, 0:02:52.800 --> 0:02:55.560 keeping advanced technology from getting in there, And of course 0:02:55.600 --> 0:02:57.920 this has been a sore spot for Beijing for years, 0:02:58.160 --> 0:03:01.800 with limits on exports of semikins, but they have been escalating, 0:03:01.840 --> 0:03:04.560 and in more recent weeks we even saw the US 0:03:04.600 --> 0:03:08.919 go after makers of chip design software, telling them essentially, 0:03:09.000 --> 0:03:12.920 you can't sell to China without violating our export restrictions. 0:03:13.240 --> 0:03:16.360 So that is something that already Beijing was seeing as 0:03:16.400 --> 0:03:19.320 a violation of the spirit of those conversations. We had 0:03:19.360 --> 0:03:22.480 heard that a week ago, but now today we hear 0:03:22.520 --> 0:03:26.560 from the top US trade negotiator Jamison Greer talking about 0:03:26.760 --> 0:03:31.240 rare earths and critical minerals. And during that discussion in Geneva, 0:03:31.280 --> 0:03:33.960 the two sides had agreed that Beijing might lift some 0:03:34.040 --> 0:03:36.920 of the barriers that it had placed in retaliation for 0:03:37.080 --> 0:03:41.600 US export curbs on exports to the US of critical minerals. 0:03:41.680 --> 0:03:42.520 So there is a bit of. 0:03:42.560 --> 0:03:44.480 Tit for tat that we are seeing here, and it 0:03:44.520 --> 0:03:47.480 may be bubbling over in the form of Donald Trump's 0:03:47.480 --> 0:03:49.000 truth social feed here today. 0:03:49.080 --> 0:03:51.840 Yet Yeah, I'm reading the news on the Bloomberg terminal, 0:03:51.840 --> 0:03:54.280 and I go back to what Michael Kratsios, who runs 0:03:54.320 --> 0:03:57.160 the White House Office of Science and Technology, right told us. 0:03:57.400 --> 0:04:02.520 The policy from this administration is promote and protect. Protect 0:04:02.600 --> 0:04:05.080 is the tariffs bit right. Promote is on shoring and 0:04:05.160 --> 0:04:09.640 stimulus for the tech industry. China is doing the same thing, right, 0:04:09.680 --> 0:04:12.160 It's trying to drum up some money seventy billion dollars 0:04:12.200 --> 0:04:15.320 we're reporting for its own industries in China. 0:04:15.360 --> 0:04:16.320 What do we know about that? 0:04:17.760 --> 0:04:20.359 Well, And that's a good question, and I'm glad you 0:04:20.360 --> 0:04:24.520 brought that up, because it's unclear whether the administration here 0:04:24.600 --> 0:04:28.760 is responding to that report of additional Chinese stimulus toward 0:04:28.839 --> 0:04:33.200 artificial intelligence and other high tech projects that they have 0:04:33.320 --> 0:04:36.880 planned as a way to spark their economy and make 0:04:36.920 --> 0:04:40.479 sure that they have some domestic demand to satisfy the 0:04:40.560 --> 0:04:43.679 industry that is trying to cultivate. But nonetheless it would 0:04:43.680 --> 0:04:47.000 not sit well with the US administration here because they 0:04:47.000 --> 0:04:49.960 are trying to bring some of that manufacturing to American 0:04:50.040 --> 0:04:54.040 shores to try to build up the American tech manufacturing base. 0:04:54.160 --> 0:04:57.520 We've seen so much of it, especially in semiconductors over decades, 0:04:57.920 --> 0:05:01.239 move to Asia, and Donald Trump there's been applying pressure 0:05:01.320 --> 0:05:03.839 to none other than Tim Cook, the Apple CEO, to 0:05:03.920 --> 0:05:07.279 try to get the iPhone production moved here. Of course, 0:05:07.640 --> 0:05:10.640 in our conversations with Mark German and so many others, 0:05:10.880 --> 0:05:12.919 we know that that is a bit of a pipe dream, 0:05:12.920 --> 0:05:16.279 but nonetheless it is a priority for the administration, and 0:05:16.360 --> 0:05:18.680 so you do see, as you said from your interview 0:05:18.680 --> 0:05:22.440 with Michael Kratzios, this desire to promote American tech here 0:05:22.480 --> 0:05:25.640 at home, but also protect it against the challenge it 0:05:25.720 --> 0:05:26.440 sees coming. 0:05:26.240 --> 0:05:29.680 From China, Bluembos, Mike Shephard, do not go far, how 0:05:29.760 --> 0:05:32.880 a sense, more news is coming on those topics. Meanwhile, 0:05:33.160 --> 0:05:35.960 sales of Apple's iPhone and those of its rivals are 0:05:36.040 --> 0:05:39.640 set to take a significant blow from President Trump's tariffs. 0:05:39.680 --> 0:05:44.039 IDC research showing smartphone growth of just zero point six 0:05:44.160 --> 0:05:49.880 percent post tariffs. For more, Nabila popeu IDC senior research director, 0:05:50.000 --> 0:05:53.279 joins us, I always track your data, not just the 0:05:53.320 --> 0:05:57.640 backward looking data, but extrapolating out and looking forward, you've 0:05:58.160 --> 0:06:01.240 kind of now got the post tariffscenario. Just explain what 0:06:01.279 --> 0:06:05.720 you published overnight and how this impacts in particular Apple. 0:06:07.240 --> 0:06:10.640 Hi, thank you and nice to be here as always. Yeah, 0:06:10.680 --> 0:06:13.360 so exactly, And whenever we publish our forecast, we always 0:06:13.400 --> 0:06:15.680 have to show what was the scenario before, right, So 0:06:16.520 --> 0:06:19.520 in our latest forecast, which we just published, we're showing 0:06:19.600 --> 0:06:22.040 a flat growth and it's really important to see how 0:06:22.080 --> 0:06:25.040 that's changed, right, because twenty four was. 0:06:25.640 --> 0:06:28.720 You know, the year of we had like twenty three. 0:06:29.000 --> 0:06:31.400 We basically we've been in a few years of decline. 0:06:31.440 --> 0:06:34.039 In the twenty four exactly, thank you for showing that 0:06:34.160 --> 0:06:36.760 up was a big year of growth. And then twenty 0:06:36.760 --> 0:06:39.279 five was supposed to continue this growth. But we were 0:06:39.640 --> 0:06:42.719 supposed to have about two percent two point three percent, 0:06:43.720 --> 0:06:48.280 you know, continued recovery. But since February, when was our 0:06:48.360 --> 0:06:52.600 last forecast was published, we've been going through such tumulus times. Right, 0:06:52.640 --> 0:06:55.560 there's been a world ruin of uncertainty. You've seen that 0:06:55.600 --> 0:06:58.719 we're thrown around many times. We've had to pull down 0:06:58.720 --> 0:07:01.839 the forecast because of our we see high uncertainty, so 0:07:01.960 --> 0:07:06.400 much terrrift supply chain turmoil, and there's a lot of 0:07:06.520 --> 0:07:11.240 you know, just soft demand across the board in large 0:07:11.280 --> 0:07:12.320 global markets. 0:07:13.960 --> 0:07:18.120 I'm trying to understand policy and how you factor policy 0:07:18.160 --> 0:07:21.360 into that model. Right, So, the situation was that smartphones 0:07:21.360 --> 0:07:22.760 were given a tariff exemption. 0:07:23.240 --> 0:07:24.320 That was mid April. 0:07:24.760 --> 0:07:29.200 Then President Donald Trump just this month May said that 0:07:29.520 --> 0:07:34.040 for smartphones made outside of America there should be a well, 0:07:34.120 --> 0:07:37.640 let's be honest, not just smartphones, the iPhone made outside 0:07:37.680 --> 0:07:41.000 of America, there should be a levy of twenty five percent. 0:07:41.600 --> 0:07:42.800 How have you factored that in? 0:07:44.280 --> 0:07:47.240 So you know, that came out right actually after we 0:07:47.280 --> 0:07:50.720 had finalized our forecast. So we you know, what we 0:07:50.760 --> 0:07:53.200 had taken into consideration, right, because we knew that even 0:07:53.240 --> 0:07:56.760 when the exemptions were announced or whatever wherever the tariff stood, 0:07:56.880 --> 0:08:00.800 that it was all temporary. That things could always change. 0:08:00.840 --> 0:08:02.840 So we had to draw a line under stand at 0:08:02.840 --> 0:08:03.320 some point. 0:08:03.800 --> 0:08:05.600 But there was you know. 0:08:05.640 --> 0:08:08.200 We had we did bacon that there is I think 0:08:08.280 --> 0:08:11.960 things could always change. So there the what we're you know, 0:08:12.000 --> 0:08:14.480 saying to our clients or anyone that you know essentially 0:08:14.520 --> 0:08:16.760 asked us this question is that that there is a 0:08:16.880 --> 0:08:20.960 huge downside risk to the US forecast and the tariffs 0:08:21.000 --> 0:08:25.360 that twenty five percent essentially brings, you know, would impose 0:08:26.640 --> 0:08:30.920 the US market, right, It would impact increased prices, It 0:08:30.920 --> 0:08:33.439 would change the demand right to. 0:08:33.440 --> 0:08:34.160 The US market. 0:08:34.240 --> 0:08:37.120 Right now, we're expecting the US market to grow about 0:08:37.120 --> 0:08:42.120 one point nine percent, and what which is actually US 0:08:42.120 --> 0:08:44.200 and China are what are holding up that even that 0:08:44.240 --> 0:08:47.040 flat zero point six percent growth that we're seeing globally, 0:08:47.320 --> 0:08:51.080 but it was pulled down Both those large markets were 0:08:51.080 --> 0:08:54.079 previously expected to grow higher. And if this twenty five 0:08:54.120 --> 0:08:57.959 percent were to grow go into effect, you know, US 0:08:57.960 --> 0:09:00.000 market would potentially even go into a decline. 0:09:02.400 --> 0:09:05.480 We should say that that this data from you and 0:09:05.559 --> 0:09:08.800 the newsflow, it doesn't just apply to the iPhone, right, 0:09:08.840 --> 0:09:12.960 it also applies to other smartphone makers like on Android 0:09:12.960 --> 0:09:15.800 os as well. But this has been our focus, which 0:09:15.840 --> 0:09:19.040 is Apple and the pressure that Tim Cook has been 0:09:19.120 --> 0:09:24.440 under from this administration. The supply chain is a question mark. 0:09:24.800 --> 0:09:28.560 But right now, do you envisage any impact to pricing 0:09:28.600 --> 0:09:32.679 of iPhone or any giant shift to where that company 0:09:32.720 --> 0:09:34.000 manufactures that handset. 0:09:35.480 --> 0:09:38.400 So I'll address the second part first, right even you know, 0:09:38.480 --> 0:09:43.079 when when you know President Trump announced this potential twenty 0:09:43.080 --> 0:09:45.880 five percent for anywhere he even said that it's not 0:09:45.920 --> 0:09:47.880 just Apple, it would be to any brand that as 0:09:47.920 --> 0:09:50.880 long as phones coming into the US were not made 0:09:50.920 --> 0:09:54.160 in the US, it would they would face potentially twenty 0:09:54.200 --> 0:09:57.360 five percent or anything wherever they lack, Right, We do 0:09:57.559 --> 0:10:01.400 think that despite wherever, whatever, if you wherever the tears 0:10:02.080 --> 0:10:06.320 happen to land eventually if they do, regardless of that, 0:10:06.720 --> 0:10:10.160 we think smartphone OEMs will continue to diversify outside of 0:10:10.280 --> 0:10:13.320 China and Vietnam and India. We said this in our 0:10:13.360 --> 0:10:17.080 press release, will continue to be the two primary, you know, 0:10:17.800 --> 0:10:22.000 global hubs for smartphone manufacturing for reasons that everyone on 0:10:22.040 --> 0:10:24.080 your show that has come on to say, right, for 0:10:24.400 --> 0:10:26.920 because of the supply chain ecosystem that has already been 0:10:27.000 --> 0:10:32.720 established there prior to all this tariffs madness that began 0:10:33.120 --> 0:10:36.240 earlier this year. Right, it's really hard. I think anyone 0:10:36.280 --> 0:10:40.320 that says smartphone manufacturing, I mean, you understand that, I 0:10:40.400 --> 0:10:45.280 understand the ideal ideology behind bringing manufacturing to the US, 0:10:45.320 --> 0:10:47.960 And the question has also been asked, it's a dream, 0:10:48.880 --> 0:10:51.320 but I you know, I don't see any reality where 0:10:51.480 --> 0:10:55.440 that is possible. So it's just yeah, So that's the 0:10:55.440 --> 0:10:57.680 reason why we do feel that it's continued, those two 0:10:57.679 --> 0:11:00.200 regions are going to continue to be the focus of 0:11:00.240 --> 0:11:02.120 oh m's diversification plans. 0:11:03.120 --> 0:11:06.320 IDC Senior Research Director Nabila Popad's great to have you 0:11:06.360 --> 0:11:06.760 on the show. 0:11:06.800 --> 0:11:07.480 Thank you so much. 0:11:07.520 --> 0:11:11.400 Now coming up, UiPath raised its outlook, but competition in 0:11:11.520 --> 0:11:15.520 agentic WORKFOW solutions it's intensifying. We'll talk to CEO Daniel 0:11:15.600 --> 0:11:17.520 Dines about what that means for his company. 0:11:17.679 --> 0:11:24.040 That's next. This is Bloomberg Technology. 0:11:31.679 --> 0:11:35.600 Automation software company UiPath raised its full years sales outlook 0:11:35.679 --> 0:11:38.720 yesterday when it released its earnings. The company also stressed 0:11:39.000 --> 0:11:43.720 that it's public sector clients renewed contracts. Let's get more 0:11:43.760 --> 0:11:46.480 on this with UIPAR CEO Daniel Dines. Daniel, it's great 0:11:46.480 --> 0:11:48.800 to have you back here on Bloomberg Technology. You know, 0:11:48.840 --> 0:11:52.200 a point of discussion for most of the year so 0:11:52.360 --> 0:11:57.800 far has been the impact on of joje on companies 0:11:57.880 --> 0:12:03.480 just like yours that have government federal level contracts. I'm 0:12:03.520 --> 0:12:05.280 trying to read between the lines and the transfer to 0:12:05.320 --> 0:12:07.680 the earnest call. Was there any impact to you, positive 0:12:07.760 --> 0:12:09.760 or negative from that doge activity. 0:12:11.720 --> 0:12:16.360 Thank you for having me. Look, we had a great 0:12:16.400 --> 0:12:20.080 event in dc A month ago and we talked to 0:12:20.160 --> 0:12:23.440 it a lot of agency. There is a renewed interest 0:12:23.600 --> 0:12:32.200 in agentic automation as a space, and we basically renewed 0:12:32.280 --> 0:12:35.480 our contract. We have a great deal with the Air Force. 0:12:35.640 --> 0:12:40.240 They are coming with a great initiative called Airmen which 0:12:40.360 --> 0:12:44.959 is aiming at a very transformative of their business processes. 0:12:45.760 --> 0:12:46.000 So on. 0:12:46.160 --> 0:12:52.840 In one, I think it's a more positive environment in DC, 0:12:53.000 --> 0:12:57.880 but still there is a lot of moving pieces in 0:12:57.960 --> 0:13:04.280 transition there. From our perspective and our guidance, it's not 0:13:04.520 --> 0:13:05.360 a big difference. 0:13:06.640 --> 0:13:11.199 Many points to new ar R R additions and then 0:13:11.240 --> 0:13:12.439 being low. 0:13:12.520 --> 0:13:13.839 What then accounts for that. 0:13:17.120 --> 0:13:22.160 Well, we continue to execute. Well, we just finish a 0:13:22.200 --> 0:13:26.560 big transition. We had an entire year of changes in 0:13:26.600 --> 0:13:31.800 the company. We accommodate everything to go really big in 0:13:31.920 --> 0:13:33.599 agentic automation. 0:13:33.760 --> 0:13:37.960 Basically, what is the growth picture? 0:13:38.360 --> 0:13:42.400 Our analyst right that growth indicators a mix for UiPath. 0:13:42.880 --> 0:13:45.240 Would you agree with that analysis of it being mixed? 0:13:45.280 --> 0:13:49.559 And if you do see growth, where's it coming from? 0:13:49.679 --> 0:13:54.240 Well, I think that agentic automation is a tremendous opportunity 0:13:54.320 --> 0:13:58.959 and it's I think everybody agrees that this is gonna 0:13:59.000 --> 0:14:02.439 be very trans formative on the business landscape. 0:14:02.720 --> 0:14:03.800 We believe this is. 0:14:03.720 --> 0:14:06.840 Going to be a bigger opportunity than RPA, and we 0:14:06.920 --> 0:14:11.600 are really well positioned to capture a significant share of it. 0:14:11.880 --> 0:14:15.640 So all in once, I think this is really our 0:14:15.679 --> 0:14:17.360 biggest opportunity for growth. 0:14:18.720 --> 0:14:21.080 I'm looking at the cost of inference right now, you know, 0:14:21.120 --> 0:14:24.760 particularly in the context of companies just like yours trickling down. 0:14:26.320 --> 0:14:30.440 A lot of companies are talking about the momentum that 0:14:30.520 --> 0:14:33.760 agent KI is giving them. What's your point of difference. 0:14:33.880 --> 0:14:36.200 What is it that you're able to offer that others 0:14:36.240 --> 0:14:36.480 are not. 0:14:39.000 --> 0:14:41.200 I think a lot of agents that we are seeing 0:14:41.200 --> 0:14:46.120 today are conversational agents, but customers would like agents that 0:14:46.240 --> 0:14:51.000 are deployed in the context of their existing business processes. 0:14:51.360 --> 0:14:55.520 And how can you trust the agents? It's the main 0:14:55.560 --> 0:14:59.480 point of the conversations. And we offer a technology to 0:14:59.800 --> 0:15:05.200 ork straight between RPA bots and agents and humans that 0:15:05.520 --> 0:15:10.160 really enhance the reliability of agents and the trust of customers. 0:15:12.680 --> 0:15:15.760 Right now, what is the biggest cost consideration for you? 0:15:16.120 --> 0:15:20.800 I talked about inference costs coming down. Even so, investors 0:15:20.800 --> 0:15:25.080 in particular love to see a commitment to access to 0:15:25.120 --> 0:15:29.400 compute and also just some like long term commitment that 0:15:29.440 --> 0:15:30.720 the investment's worthwhile. 0:15:32.920 --> 0:15:34.920 Yeah, I would not say that. In our business, the 0:15:34.960 --> 0:15:41.840 infant cost is significant. The real cost comes with the 0:15:41.880 --> 0:15:47.720 implementing of the agents. Still, the capability is required to 0:15:47.920 --> 0:15:55.360 create really good agents that mimic the capability of humans 0:15:55.600 --> 0:15:58.720 are kind of rare, and we are working a lot 0:15:58.760 --> 0:16:04.120 to build technology that helped to democratize the ability to 0:16:04.160 --> 0:16:07.520 build agents. To me, this is gonna be the real 0:16:07.600 --> 0:16:10.160 cost of deploying a gentic automation. 0:16:10.320 --> 0:16:10.760 That's key. 0:16:12.000 --> 0:16:15.520 Daniel dines UiPath CEO, thank you so much for joining 0:16:15.600 --> 0:16:18.520 us here on Bloomberg Technology. Now coming up, Microsoft's push 0:16:18.600 --> 0:16:22.120 to get corporate customers on board with co Pilot seems 0:16:22.120 --> 0:16:23.840 to be taking off. We've got the details on its 0:16:23.840 --> 0:16:25.520 AI sales coming up next. 0:16:25.720 --> 0:16:26.960 This is Bloomberg Technology. 0:16:42.200 --> 0:16:45.120 A growing push for sovereign AI has led to a 0:16:45.120 --> 0:16:48.280 wave of data center announcements, including one this week from 0:16:48.320 --> 0:16:51.280 Bell Canada. It plans to invest hundreds of millions of 0:16:51.280 --> 0:16:54.560 dollars to build AI data centers across the country, and 0:16:54.640 --> 0:16:58.040 it is named chip startup Grock as the project's exclusive 0:16:58.160 --> 0:17:01.960 inference partner. Grock CEO Jonathan Ross spoke with Caroline Hyde 0:17:02.000 --> 0:17:05.199 about sovereign AI, but also US chip curbs on China. 0:17:05.800 --> 0:17:08.720 From grox's beginning, we have done no business in China, 0:17:08.880 --> 0:17:10.920 and this is not a geopolitical thing. This is just 0:17:10.960 --> 0:17:13.840 a shrewd business thing. We saw all of these very 0:17:13.880 --> 0:17:17.119 large tech companies go into China and lose over and 0:17:17.119 --> 0:17:18.879 over and again, and there's just a thumb on the 0:17:18.920 --> 0:17:23.560 scale for any Western company. I hope that someday China 0:17:23.640 --> 0:17:27.240 allows real competition because real competition makes people stronger, and 0:17:27.280 --> 0:17:30.240 I want to compete against those Chinese companies. But until 0:17:30.240 --> 0:17:32.760 they allow real competition, we can't do that. So we've 0:17:32.800 --> 0:17:36.080 blocked all Chinese companies from having access to our API. 0:17:36.400 --> 0:17:38.280 We will not sell chips in China because they're just 0:17:38.280 --> 0:17:40.399 going to reverse engineer them and try and build them themselves, 0:17:40.400 --> 0:17:43.040 and they're not going to compete fairly. As a result, 0:17:43.200 --> 0:17:46.320 when we do these deals, we're not allowing Chinese companies 0:17:46.320 --> 0:17:49.240 to run on them, and we do build operate transfer. 0:17:49.280 --> 0:17:51.960 In these sovereign cases, we're doing build operate, so we're 0:17:52.000 --> 0:17:55.159 the ones operating that, and that actually gives commerce and 0:17:55.200 --> 0:17:57.879 others a lot of comfort that Chinese companies aren't going 0:17:57.880 --> 0:17:58.560 to be getting. 0:17:58.280 --> 0:18:01.520 Access, But does the strictions on say in video selling 0:18:01.640 --> 0:18:04.240 H twenties, as Jensen would say, mean that China just 0:18:04.280 --> 0:18:04.960 gets better at. 0:18:04.840 --> 0:18:05.439 Doing it themselves. 0:18:05.480 --> 0:18:07.920 In a Huawei will be able to compete with growth in. 0:18:07.840 --> 0:18:12.440 The future potentially. However, the reality is if you enable 0:18:12.480 --> 0:18:14.720 people now, then they're going to be able to use 0:18:14.720 --> 0:18:17.040 that advantage to keep growing quicker. We're going to be 0:18:17.080 --> 0:18:19.639 using AI to design our chips, We're going to be 0:18:19.720 --> 0:18:22.800 using AI to design our software, So why would you 0:18:22.800 --> 0:18:25.359 give that advantage? I think it's different going back to 0:18:25.400 --> 0:18:29.040 information age technology, it's about replicating and distributing. Compute is 0:18:29.080 --> 0:18:32.000 about doing something specific. If you don't have the compute, 0:18:32.000 --> 0:18:36.000 you can't do it, whereas information knowledge sort of diffuses 0:18:36.119 --> 0:18:39.239 and you'll catch up. With generative age technology you can 0:18:39.280 --> 0:18:39.840 pull ahead. 0:18:40.800 --> 0:18:44.199 You talk about wand in competition and to be competitive 0:18:44.600 --> 0:18:47.359 in China, for example, what about the competition within VideA. 0:18:47.640 --> 0:18:49.720 How do you stand compared to AMD and the other 0:18:49.800 --> 0:18:51.680 offerings Here in the United States. 0:18:51.440 --> 0:18:53.040 We actually think we're one of the best things that's 0:18:53.040 --> 0:18:57.359 ever happened to in Vidia shareholders. So in Vidia sells 0:18:57.400 --> 0:18:59.960 a product that is a premium product. It's a luxury good, 0:19:00.280 --> 0:19:02.919 it's for training the models. When you train a model, 0:19:03.119 --> 0:19:05.399 you're willing to spend more money because you get to 0:19:05.400 --> 0:19:08.920 amortize that across an enormous number of users. If in 0:19:09.040 --> 0:19:11.639 Vidia continues focusing on inference, which they have to do 0:19:11.760 --> 0:19:14.479 right now because there isn't enough inference capacity, they're going 0:19:14.520 --> 0:19:16.280 to have to bring their margins down. They're going to 0:19:16.359 --> 0:19:18.360 have to sell their chips for a lot less. Remember, 0:19:18.640 --> 0:19:22.119 our cost is much less than in Vidia's cost. On 0:19:22.200 --> 0:19:26.320 top of that, higher volume, lower margin. And so the 0:19:26.359 --> 0:19:29.479 more inference compute we deploy, the more demand there is 0:19:29.480 --> 0:19:31.760 for training compute. The more demand for training compute, the 0:19:31.800 --> 0:19:34.240 more GPUs and VideA sells. And the one thing that's 0:19:34.320 --> 0:19:36.840 really important here in Video wants to do between three 0:19:36.840 --> 0:19:39.880 and five million GPUs for AI this year. But that's 0:19:39.960 --> 0:19:43.080 based on what's called HPM high bandwidth memory. It's this 0:19:43.240 --> 0:19:45.960 very scarce component. They can make as many GPU chips 0:19:46.000 --> 0:19:47.199 as they want, they can't make as much of the 0:19:47.200 --> 0:19:49.359 memories they want. Those chips are going to be built 0:19:49.359 --> 0:19:51.119 and sold, and the only question is what are they 0:19:51.119 --> 0:19:53.320 going to be sold to do when we come in 0:19:53.640 --> 0:19:56.639 and we sell inference chips than what we're doing is 0:19:56.640 --> 0:19:58.520 we're allowing those GPUs to be sold into training, which 0:19:58.560 --> 0:20:00.639 is a higher margin business. The video is going to 0:20:00.720 --> 0:20:03.320 make more profit thanks to us. 0:20:05.520 --> 0:20:08.480 All Right, we have some breaking news crossing the Bloomberg terminal. 0:20:08.720 --> 0:20:13.080 Utel Sat satellite based internet providers in talks to Rays 0:20:13.480 --> 0:20:18.240 one point five billion euros. This is interesting because in Europe, 0:20:18.280 --> 0:20:21.280 the continent's hoping that you tel Sat will be able 0:20:21.720 --> 0:20:25.720 to displace and compete against the likes of Elon, Musk 0:20:25.720 --> 0:20:27.000 Starlink in Europe. 0:20:27.119 --> 0:20:28.040 That's the red headline. 0:20:28.080 --> 0:20:31.480 You tail Satin talks to Rays one point five billion 0:20:31.920 --> 0:20:35.840 euros with a doubling of the steak from French entities. 0:20:35.880 --> 0:20:37.840 We get more on that story as we get more 0:20:37.840 --> 0:20:38.119 on it. 0:20:38.200 --> 0:20:39.760 All right, let's get back to what's happening in the 0:20:39.760 --> 0:20:44.200 world of AI, and Microsoft is focused in driving adoption 0:20:44.400 --> 0:20:46.720 of its AI tool co Pilot, and at a town 0:20:46.720 --> 0:20:49.240 hall this week, it touted its progress. 0:20:49.280 --> 0:20:51.840 Who broke the story, Bloomberg's Brody Ford. 0:20:51.840 --> 0:20:53.920 I think this is really interesting, right because it's quick 0:20:53.960 --> 0:20:56.320 on the heels of this deep dive we did in 0:20:56.400 --> 0:21:00.000 Business Week about co Pilot, how it happened, how it launched, 0:21:00.119 --> 0:21:03.800 What Microsoft shypes are for it, what you reported are 0:21:04.119 --> 0:21:09.360 names across various industries, big corporate names that are doing deals. 0:21:10.640 --> 0:21:10.840 Right. 0:21:11.000 --> 0:21:14.840 Yeah, The background here is that Microsoft salesforce companies like 0:21:14.920 --> 0:21:17.879 them they have these AI tools, right. I mean, in 0:21:17.880 --> 0:21:20.720 the consumer space, Chat, GPT and tools like it have 0:21:20.880 --> 0:21:24.359 ramped incredibly. In the enterprise, it hasn't been quite as 0:21:24.400 --> 0:21:28.320 easy of an adoption path, and so with tools like Copilot, 0:21:28.400 --> 0:21:32.720 Microsoft's AI assistance, we don't have great intel over how 0:21:32.760 --> 0:21:35.000 many folks are really using it, how many folks are 0:21:35.040 --> 0:21:38.840 really paying for it. What we learned from the meeting yesterday, 0:21:38.920 --> 0:21:42.359 which sources we're telling us about, is that you know, 0:21:42.640 --> 0:21:45.920 there are dozens of customers who are having over one 0:21:45.960 --> 0:21:49.720 hundred thousand paid users. That's a pretty significant amount. I mean, 0:21:49.760 --> 0:21:52.199 of course, you know there's some discounting in that, but 0:21:52.720 --> 0:21:56.480 this is a sign that adoption is ramping, maybe more 0:21:56.640 --> 0:21:58.840 than investors had been appreciating. 0:22:00.160 --> 0:22:00.840 Your reporting. 0:22:00.880 --> 0:22:05.120 We note that Microsoft CEO Sati Endella is very closely 0:22:05.200 --> 0:22:09.680 tracking all kinds of data, including who is using co 0:22:09.800 --> 0:22:11.159 pilot and how they're using it. 0:22:11.200 --> 0:22:12.880 Why is that important, right? 0:22:12.920 --> 0:22:15.760 I Mean, what's really important is that you don't have shelfware, 0:22:15.840 --> 0:22:18.840 but you don't buy one hundred thousand licenses, but you know, 0:22:19.040 --> 0:22:22.080 fifty people use it in finance, right. I mean, this 0:22:22.160 --> 0:22:24.560 is the classic problem with software where maybe you can 0:22:24.600 --> 0:22:26.520 sell a lot of things, but you have to really 0:22:26.520 --> 0:22:28.680 make sure they're using it. 0:22:28.760 --> 0:22:31.560 Bloomberg's Brodie Ford another excellent piece of reporting. 0:22:31.600 --> 0:22:32.360 Thank you very much. 0:22:32.440 --> 0:22:36.560 Okay, coming up, Patrick McGoldrick from JP Morgan Private Capital 0:22:36.680 --> 0:22:40.440 joins us to talk about its approach to AI investment opportunities. 0:22:40.720 --> 0:22:55.119 Their conversation's next, This is Bloomberg Technology. Welcomes back to 0:22:55.160 --> 0:22:58.119 Bloomberg Technology, Amed Ludlow in San Francisco. It's kind of 0:22:58.160 --> 0:23:00.879