WEBVTT - White House Adviseors Discuss How the US Will Export AI 0:00:02.560 --> 0:00:13.080 Bloomberg Audio Studios, Podcasts, radio news. Bloomberg Tech is live 0:00:13.119 --> 0:00:16.560 from coast to coast with Caroline Hide. 0:00:16.320 --> 0:00:19.200 In New York and Ed Lovelow in San Francisco. 0:00:22.680 --> 0:00:26.200 This is Bloomberg Tech coming up. President Trump signed executive 0:00:26.280 --> 0:00:29.160 orders to boost AI development in the US. We'll dig 0:00:29.200 --> 0:00:32.919 deeper with White House officials Michael Kratzios and Shari ram Krishnan. 0:00:33.280 --> 0:00:35.440 Plus we get back to big tech earnings, but Tesla 0:00:35.479 --> 0:00:38.800 and Alphabet's results, Tesla sliding after Musk's warning for the 0:00:38.840 --> 0:00:39.839 quarters to come. 0:00:41.640 --> 0:00:44.040 And Intel's next on the docket. We'll discuss what to 0:00:44.080 --> 0:00:46.879 expect from the ship maker's results and it's turnaround efforts. 0:00:46.960 --> 0:00:48.040 Later today the ED. 0:00:48.080 --> 0:00:50.240 First, we check in on these markets that are once 0:00:50.320 --> 0:00:53.559 again at record highs. We're focusing in on what has 0:00:53.600 --> 0:00:56.600 really been the uplift, feel good factor of the fact 0:00:56.600 --> 0:00:58.639 that the trade deal starts to roll in Japan. All 0:00:58.640 --> 0:01:00.920 eyes on Europe and we look at the naz currently 0:01:00.960 --> 0:01:02.640 up quarter of percent if you're looking at the Nazbak 0:01:02.640 --> 0:01:04.959 one hundred, but it had been a new record highs. 0:01:04.959 --> 0:01:06.680 We're seeing it on other benchmarks as well, and I 0:01:06.720 --> 0:01:09.320 know you're digging in to the key contributors to the 0:01:09.400 --> 0:01:11.720 upside and the downside in terms of points. 0:01:12.600 --> 0:01:14.759 Yeah, let's go on the downside. Tesla is the earning 0:01:14.880 --> 0:01:17.160 story on track for its biggest drop since the first 0:01:17.200 --> 0:01:19.840 week of June, a period of transition, and the core 0:01:19.920 --> 0:01:23.600 business suffering. Alphabet is AI fueled growth but at a 0:01:23.640 --> 0:01:26.440 greater cost, raising capex later in the show, going deep 0:01:26.480 --> 0:01:29.160 on both those names and Nvidia in this session is 0:01:29.200 --> 0:01:32.920 pushing fresh record highs. Jensen one front and center of 0:01:33.040 --> 0:01:37.200 yesterday's Winning the AI Race summit, but sprinkled within that 0:01:37.640 --> 0:01:40.720 a warning from the President about how closely he's looking 0:01:40.760 --> 0:01:42.800 in a video Let's get out to Bloomberg's Michael Shephard 0:01:42.840 --> 0:01:44.360 here in Washington. 0:01:44.040 --> 0:01:44.280 D C. 0:01:45.080 --> 0:01:47.800 The executive orders of the fresh bit action by this 0:01:47.880 --> 0:01:51.880 president to enact his AI action plan, a standing ovation 0:01:52.000 --> 0:01:56.560 for Jensen one, and within it a warning, so to speak. 0:01:56.280 --> 0:01:57.160 Well that's raded. 0:01:57.520 --> 0:02:01.720 In this he we saw President Donald Trump embrace artificial 0:02:01.760 --> 0:02:06.480 intelligence is a pillar of his economic and national security agendas, 0:02:06.760 --> 0:02:09.600 and with it we saw him practically throw his arms 0:02:09.639 --> 0:02:13.800 around Jensen Wang himself during the event, singling the Nvidia 0:02:14.000 --> 0:02:17.520 CEO out for praise, really for the company's prowess in 0:02:17.919 --> 0:02:23.680 making the AI chips that power this new and transformative technology. 0:02:23.840 --> 0:02:26.560 But he did sprinkle a warning in there before going 0:02:26.600 --> 0:02:29.560 on to praise Jensen. He let it slip that he 0:02:29.600 --> 0:02:32.200 had thought about trying to break up in Video, which 0:02:32.600 --> 0:02:36.640 would be a startling and very much a market shaking move, 0:02:36.840 --> 0:02:38.720 and yet he said that he was talked out of 0:02:38.760 --> 0:02:42.160 it by AIDS, who pointed to the company's importance and 0:02:42.200 --> 0:02:45.119 how it had taken years to develop that pre eminence, 0:02:45.200 --> 0:02:48.280 and if the US government had gone after Nvidia, it 0:02:48.320 --> 0:02:50.360 would be really hard to walk that back. 0:02:50.800 --> 0:02:53.280 Mike Shepherd on the latest out of that AI Action 0:02:53.400 --> 0:02:57.240 Plan event and let's get more analysis, more people in 0:02:57.240 --> 0:02:59.519 the room. Michael Kratzios is one of them, Director of 0:02:59.560 --> 0:03:02.639 the White House Officer Science and Technology Policy, and it 0:03:02.800 --> 0:03:07.240 was let's just start such an enthusiastic welcome embrace of 0:03:07.280 --> 0:03:09.560 AI written large, Michael, I got to get your take. 0:03:09.639 --> 0:03:12.880 It was perhaps jovial in nature coming from the President 0:03:12.919 --> 0:03:14.840 of the United States, But what did you make of 0:03:14.880 --> 0:03:17.119 the fact that him, once upon a time considered even 0:03:17.320 --> 0:03:18.880 thinking of breaking up and video. 0:03:20.200 --> 0:03:22.040 I think it was a tremendous day for the country 0:03:22.080 --> 0:03:25.600 and for AI more broadly. We definitively turned the page 0:03:25.720 --> 0:03:28.560 on what the last administration had tried to do to 0:03:28.639 --> 0:03:32.280 really hamper this industry and chocolate with a high level 0:03:32.280 --> 0:03:35.160 of regulation that didn't allow it to fully spread its wings. 0:03:35.400 --> 0:03:38.080 What was tremendous about the event yesterday was not only 0:03:38.160 --> 0:03:41.480 the President's speech, but the participation that the Vice President 0:03:41.600 --> 0:03:44.600 himself and five Cabinet secretaries. This was a big moment 0:03:44.600 --> 0:03:46.800 for the American I industry and it really showed how 0:03:46.800 --> 0:03:50.240 committed this administration is to ensuring continued American dominance in 0:03:50.280 --> 0:03:51.320 this critical technology. 0:03:52.520 --> 0:03:55.680 Mister Crisis, you helped to write this document and form 0:03:55.760 --> 0:04:00.640 the policy alongside the advisors David Sachs and Shri Krishnan, 0:04:00.680 --> 0:04:03.520 who will be on the program later today. What I 0:04:03.640 --> 0:04:06.160 got from Corporate America is that they feel the action 0:04:06.280 --> 0:04:12.080 plan is actionable. It's heavily rooted in deregulation and expediting permits. 0:04:12.600 --> 0:04:15.360 Of the twenty three pages, which section do you think 0:04:15.360 --> 0:04:17.040 gets enacted by this administration? 0:04:17.200 --> 0:04:17.599 First? 0:04:18.880 --> 0:04:21.080 Well, we are racing ahead on all of them. I 0:04:21.120 --> 0:04:23.480 think the one piece that you didn't mention, which I 0:04:23.480 --> 0:04:26.159 think is also very critically important, is this focus on 0:04:26.400 --> 0:04:29.960 exporting American AI. This idea that we want the entire 0:04:30.000 --> 0:04:34.279 world to be running on American artificial intelligence stack. That is, 0:04:34.360 --> 0:04:38.360 our cloud, our chips, our algorithms, all of that needs 0:04:38.400 --> 0:04:40.720 to be exported and packaged to the world so that 0:04:40.760 --> 0:04:43.839 we become the ecosystem of choice globally. And that is 0:04:43.839 --> 0:04:47.320 something that Secretary Lutnik and Secretary of State Rubio have 0:04:47.400 --> 0:04:50.039 been tasked to work on and put together for the country. 0:04:50.080 --> 0:04:53.560 So between that and our very ambitious agenda on the 0:04:53.600 --> 0:04:56.720 actual infrastructure of empower it's going to power this, there's 0:04:56.760 --> 0:04:58.040 a lot of work for us to get done in 0:04:58.040 --> 0:04:58.760 the next few months. 0:05:00.080 --> 0:05:04.839 MISSUS crisis. What's the rulemaking process from here for exporting chips? 0:05:04.880 --> 0:05:07.960 In particular, the way that your colleague, mister Lutnik put 0:05:08.000 --> 0:05:12.280 it was that the administration is comfortable with large clusters 0:05:12.640 --> 0:05:17.440 going to third parties foreign countries, but American company access 0:05:17.480 --> 0:05:20.839 to that stack would be part of the evaluation process. 0:05:21.000 --> 0:05:23.359 Do we get a successor to the diffusion rule? Have 0:05:23.480 --> 0:05:25.840 you started drawing up some guidelines? 0:05:27.120 --> 0:05:31.080 So that's being handled by BIS and Secretary Lutnik at Commerce, 0:05:31.120 --> 0:05:33.080 But generally speaking, what we're going to be doing and 0:05:33.120 --> 0:05:35.880 through the executive order of the President signed the Department 0:05:35.920 --> 0:05:38.839 of Commerce is tasked to work with American industry to 0:05:38.960 --> 0:05:42.560 create AI packages. These are packages that we then can 0:05:42.760 --> 0:05:45.320 work with the State Department to go out and export 0:05:45.320 --> 0:05:48.000 around the world. And part of that, because Commerce will 0:05:48.040 --> 0:05:50.400 be at the table when these packages are being created 0:05:50.400 --> 0:05:52.960 with our great chip companies and our algorithm companies and 0:05:53.000 --> 0:05:55.479 our model companies, we'll be able to have packages which 0:05:55.520 --> 0:05:57.599 we believe will be able to make it through the 0:05:57.720 --> 0:06:00.640 xport control restrictions of a commerce department and make them 0:06:00.640 --> 0:06:04.560 actually extraordinarily viable and very interesting and very appealing to 0:06:04.600 --> 0:06:07.000 lots of countries around the world who all are racing 0:06:07.040 --> 0:06:09.560 and very excited to have AI and deliver AI for 0:06:09.640 --> 0:06:10.120 their people. 0:06:10.760 --> 0:06:14.400 And boy are the US companies racing for AI. The 0:06:14.480 --> 0:06:17.120 investment that we hear coming from Alphabet, for example, opping 0:06:17.160 --> 0:06:20.520 by another ten billion in terms of capital expenditure. These companies, 0:06:20.560 --> 0:06:24.400 though a huge Alphabet potentially facing a breakup. We hear 0:06:24.440 --> 0:06:27.719 about Judge Meta in August. We also know that Meta 0:06:27.760 --> 0:06:31.400 itself is also being analyzed. Will these companies always be 0:06:31.440 --> 0:06:35.320 allowed to be huge if the AI innovation is the win? 0:06:36.640 --> 0:06:38.880 You know, that's not something that I particularly focus on 0:06:39.000 --> 0:06:40.080 or look at. I think you have to talk to 0:06:40.120 --> 0:06:42.640 the Department of Justice and others. But to be honest, 0:06:42.680 --> 0:06:45.159 I think we are very blessed in the United States. 0:06:45.240 --> 0:06:48.000 We have the greatest technology companies in the world that 0:06:48.040 --> 0:06:50.960 are delivering the greatest products in the world that citizens 0:06:50.960 --> 0:06:52.800 of the world all want to use. If you just 0:06:52.839 --> 0:06:55.160 look at the numbers of usage for a large language 0:06:55.160 --> 0:06:58.760 model builders, it's absolutely incredible. The entire world wants to 0:06:58.800 --> 0:07:01.200 run American rails, and we need to make sure that 0:07:01.200 --> 0:07:03.000 the regulations are out of the way to make sure 0:07:03.040 --> 0:07:05.080 that we can get it out there, not only to Americans, 0:07:05.120 --> 0:07:05.960 but also globally. 0:07:06.120 --> 0:07:08.000 And I thought what was really interesting was the comments 0:07:08.040 --> 0:07:11.760 coming from President Trump while announcing the AIX and signing 0:07:11.840 --> 0:07:14.880 executive orders, was one of copyright and almost trying to 0:07:14.920 --> 0:07:18.600 push back on this need to be compensated. If you're 0:07:18.600 --> 0:07:20.800 someone who's made the works that the ultimate models have 0:07:20.840 --> 0:07:23.000 been trained upon, what do you make of copyright and 0:07:23.040 --> 0:07:24.280 will that be in legislation? 0:07:25.520 --> 0:07:27.920 Yeah, you've burned a good point. In the President's speech, 0:07:28.000 --> 0:07:31.520 the President discussed the importance of being able to use 0:07:31.600 --> 0:07:34.280 materials in the training of these models, and this is 0:07:34.320 --> 0:07:36.680 something that is going back and forth and discussed a 0:07:36.720 --> 0:07:39.280 lot globally. And I think we're excited to kind of 0:07:39.280 --> 0:07:42.080 see where that all plays out. Ultimately speaking, what we 0:07:42.120 --> 0:07:44.000 want to do is be able to allow training to 0:07:44.040 --> 0:07:46.280 happen here in the United States, and that's something that's 0:07:46.320 --> 0:07:46.960 being sorted. 0:07:46.760 --> 0:07:47.360 Out in the courts. 0:07:48.840 --> 0:07:52.920 Mister Cratsios, why is the issue of visas and immigration 0:07:53.080 --> 0:07:55.520 of talent not included in the action plan? 0:07:56.760 --> 0:07:59.080 Well, we talk a lot about the importance of the 0:07:59.120 --> 0:08:01.600 people that are going to make the AI ecosystem thrive. 0:08:01.960 --> 0:08:03.960 And one thing that was very unique and special about 0:08:03.960 --> 0:08:06.160 this report, which actually doesn't get talked a lot about 0:08:06.200 --> 0:08:09.280 when it comes to the I ecosystem is the importance 0:08:09.360 --> 0:08:13.240 of actual builders in how we build out the infrastructure 0:08:13.240 --> 0:08:15.040 for AI. When we think of the people that we 0:08:15.120 --> 0:08:18.320 need to make AI succeed, we need HVAC people, we 0:08:18.360 --> 0:08:21.920 need electricians, we need all these amazing careers that we'll 0:08:21.920 --> 0:08:24.720 be able to actually build these massive data centers for 0:08:24.840 --> 0:08:27.480 the American people. So this report touches on that, and 0:08:27.520 --> 0:08:29.520 it also brings up the point that we need great 0:08:29.560 --> 0:08:31.320 talent and great Americans to be able to work on 0:08:31.360 --> 0:08:32.000 this technology. 0:08:34.360 --> 0:08:37.120 The way that mister Jensen Wong has put it time 0:08:37.200 --> 0:08:39.600 and time again is that more than fifty percent of 0:08:39.640 --> 0:08:43.920 the world's AI researchers are from China or currently in China. 0:08:44.320 --> 0:08:46.280 And if you look at recent newsflow, and I know 0:08:46.360 --> 0:08:50.000 that you do, mister Kratzios, all of these members of 0:08:50.160 --> 0:08:52.760 talent that are moving from Apple to Meta or Google 0:08:52.840 --> 0:08:56.560 to Meta in its Superintelligence lab our ivor first generation 0:08:56.679 --> 0:09:01.400 immigrants from somewhere around the world, but including China, do 0:09:01.480 --> 0:09:04.080 you acknowledge that the administration needs to come up for 0:09:04.160 --> 0:09:08.000 a plan to attract talent from outside of this country 0:09:08.800 --> 0:09:10.720 to fuel the AI industry. 0:09:10.320 --> 0:09:14.200 Here America is already the place where the best and 0:09:14.200 --> 0:09:16.680 brightest people from around the world want to come live, 0:09:16.800 --> 0:09:18.920 work and study. And we are blessed with that, and 0:09:18.960 --> 0:09:22.760 we continue to have amazing legal pathways for the greatest 0:09:22.800 --> 0:09:25.600 thinkers and minds and engineers to come work in the 0:09:25.640 --> 0:09:27.840 United States. Where I spend a lot of time thinking 0:09:27.840 --> 0:09:29.400 about is how we can make sure that we have 0:09:29.520 --> 0:09:33.080 more native born American talent succeeding and being employed by 0:09:33.120 --> 0:09:36.400 our great technology companies. The percentage of Americans that are 0:09:36.400 --> 0:09:39.920 now pursuing high end STEM degrees has been declined precipitously 0:09:39.960 --> 0:09:42.079 over the last twenty years. That is something that is 0:09:42.120 --> 0:09:43.840 a curve that we need to change. We need more 0:09:43.840 --> 0:09:47.719 Americans pursuing high end PhDs and things like artificial intelligence 0:09:47.760 --> 0:09:50.400 and computer science and engineering, and that's something that we're 0:09:50.400 --> 0:09:51.280 going to work very hard on. 0:09:51.800 --> 0:09:57.280 Michael. Also, what's included is well the focus on reskilling, retraining, 0:09:57.480 --> 0:10:00.920 and helping states and companies do that. There's enough there 0:10:01.080 --> 0:10:03.200 to support what is going to be a true upheaval 0:10:03.280 --> 0:10:04.120 in the labor market. 0:10:05.280 --> 0:10:07.000 I think we can get it done, and I think 0:10:07.040 --> 0:10:09.640 the Secretary of Labor and many other secretaries are very 0:10:09.640 --> 0:10:12.360 excited about driving these programs forward. As I talked to 0:10:12.400 --> 0:10:15.640 builders that are building out Stargate, they're saying that somewhere 0:10:15.640 --> 0:10:18.120 between twenty five or fifty states is where they have 0:10:18.200 --> 0:10:20.720 to bring in electricians and other folks in to actually 0:10:20.720 --> 0:10:22.719 work on these projects. And that's something that we need 0:10:22.760 --> 0:10:25.120 to fix. That's a curve that we need to bend 0:10:25.360 --> 0:10:27.679 and get more people being able to work on these 0:10:27.720 --> 0:10:31.320 actually high paying, very efficient, great jobs for Americans and 0:10:31.360 --> 0:10:33.439 be able to allow them to participate in the revolution. 0:10:34.559 --> 0:10:38.079 Miscristus very quickly. The plan includes a provision to contract 0:10:38.160 --> 0:10:42.080 only with developers using non bias models. The President brought 0:10:42.120 --> 0:10:44.960 the issue up also talked about woke in the context 0:10:45.000 --> 0:10:47.920 of models. Explain why that was included. Please, we have 0:10:48.080 --> 0:10:48.840 thirty seconds. 0:10:49.640 --> 0:10:51.839 The President has been very clear from day one about 0:10:51.880 --> 0:10:55.640 trying to remove Marxist DII ideology from the federal government, 0:10:55.920 --> 0:10:57.680 and a big part of that is making sure that 0:10:57.760 --> 0:11:01.560 the models and the technology that we actually perture and 0:11:01.600 --> 0:11:05.360 spend taxpayer dollars on does not support these types of ideologies. 0:11:05.520 --> 0:11:07.120 And that's why you're going to see changes in the 0:11:07.160 --> 0:11:09.520 way that we procure AI models to make sure that 0:11:09.559 --> 0:11:10.199 we eliminate this. 0:11:12.160 --> 0:11:14.680 Michael Kratzius, Director of the White House Office of Science 0:11:14.679 --> 0:11:16.800 and Technology Policy, Grateful to have you back on the 0:11:16.840 --> 0:11:19.199 show after a busy twenty four hours here in d C. 0:11:19.400 --> 0:11:22.600 Thank you know coming up. Elon Musk warns of rough 0:11:22.720 --> 0:11:26.480 quarters ahead, following one of Tesla's worst stretches since I 0:11:26.559 --> 0:11:28.960 first started producing EB So we're going to go to 0:11:29.040 --> 0:11:30.920 that next. This is Bloomberg Tech. 0:11:33.920 --> 0:11:37.800 We probably could have a few rough quarters and I'm 0:11:37.840 --> 0:11:42.680 not saying it well, but we could Q four, Q one, 0:11:44.320 --> 0:11:47.320 maybe Q two. But once you get to autonomy at 0:11:47.360 --> 0:11:50.439 scale in the second hand of half of next year something. 0:11:50.480 --> 0:11:54.160 By the end of next year, I think the I'll 0:11:54.200 --> 0:11:58.920 be surprised if the economics, I'm not very compelling. 0:12:00.360 --> 0:12:02.560 As a CEO. Elon Musk there on the earnings call 0:12:02.640 --> 0:12:05.920 yesterday has discussed the results. The share prices reaction is stark. 0:12:06.040 --> 0:12:08.360 William Stein is with US managing director and senior analyst 0:12:08.400 --> 0:12:12.520 at Truest Securities. Was there anything about this autonomy promise 0:12:12.880 --> 0:12:15.319 that is enough to make up for the ugly fundamental 0:12:15.360 --> 0:12:17.200 picture in the here and now? William? 0:12:17.840 --> 0:12:18.000 Oh? 0:12:18.080 --> 0:12:21.240 Sure, if you take a very long term view and 0:12:21.640 --> 0:12:24.920 look at the track record of this remarkable human being 0:12:25.360 --> 0:12:28.320 and the team that he's been able to assemble, they've 0:12:28.320 --> 0:12:33.000 done some just really amazing things with new technology in 0:12:33.120 --> 0:12:36.560 automotive and certainly in AI, and it's moving more in 0:12:36.600 --> 0:12:39.880 that direction. So one can, you know, certainly take that 0:12:40.640 --> 0:12:44.200 very much optimistic view on the long term. However, the 0:12:44.200 --> 0:12:46.000 short term clearly sounds worse. 0:12:48.320 --> 0:12:51.559 William. On the earnings call, you asked a question about 0:12:51.559 --> 0:12:55.800 the more affordable model. What happened was Lars interjected, I 0:12:55.800 --> 0:12:57.800 think it's Lars when I checked the tape and said 0:12:57.800 --> 0:12:59.959 we're not going to talk about and then Elon interjected 0:13:00.640 --> 0:13:04.320 and said, it's just a model. Why I've checked the 0:13:04.360 --> 0:13:09.080 tape so many times it's just a model. Why why 0:13:09.160 --> 0:13:11.720 is it that important that he answered you a question? 0:13:11.760 --> 0:13:14.000 And where does it fit into robotaxi? Please? 0:13:15.520 --> 0:13:20.360 So I don't think it's so relevant to Robotaxi. But 0:13:21.200 --> 0:13:24.080 this lower cost vehicle is something they've promised for a 0:13:24.160 --> 0:13:28.199 while that is supposed to be a half step between 0:13:28.640 --> 0:13:30.959 what they're doing now and what they promised at the 0:13:31.040 --> 0:13:36.000 last analyst day, which would be a significantly cost reduced car, 0:13:36.360 --> 0:13:41.680 a smaller form factor, and a different architecture entirely. What 0:13:41.800 --> 0:13:43.880 he talked about yesterday, or what I asked him to 0:13:43.880 --> 0:13:46.320 talk about, was give us more detail on this interim 0:13:46.400 --> 0:13:48.680 thing that you said you were making in the first 0:13:48.720 --> 0:13:51.720 half of this year. You reiterated that in the press 0:13:51.760 --> 0:13:54.080 release and on the call, But we're all scratching our 0:13:54.080 --> 0:13:57.080 heads where is it? And his saying that the form 0:13:57.120 --> 0:14:00.040 factor looks like a model, why, from my perspective, that 0:14:00.120 --> 0:14:03.960 is not surprising at all. I was trying to draw 0:14:04.000 --> 0:14:06.559 out from him some details as to what the architecture 0:14:06.600 --> 0:14:08.960 is going to look like and to what degree the 0:14:09.040 --> 0:14:11.840 cost is going to come down as a result of 0:14:11.960 --> 0:14:13.640 the new innovations he's bringing. 0:14:13.920 --> 0:14:15.000 Yeah, he didn't. 0:14:14.800 --> 0:14:15.360 Bite on that. 0:14:15.640 --> 0:14:20.280 We've got no clarity and ultimately the interjection of a 0:14:20.320 --> 0:14:23.080 new cheaper model is what's meant to perhaps drive sales 0:14:23.120 --> 0:14:25.960 in the next second half and the next future year. 0:14:26.800 --> 0:14:28.320 If we're not getting that clarity, are we just going 0:14:28.360 --> 0:14:30.840 to get canibalization? If it looks and feels like a. 0:14:30.800 --> 0:14:36.520 Model, why, I think that's a very smart point you're making. 0:14:36.520 --> 0:14:39.040 And the way that I'm thinking about this is, remember, 0:14:39.080 --> 0:14:44.480 there are some consumer tax credits that are going away. 0:14:45.560 --> 0:14:49.280 There are also regulatory tax credits or regulatory credits that 0:14:49.640 --> 0:14:53.600 revenue is going to decline from that. I see an 0:14:53.640 --> 0:14:57.880 incremental reduction in the cost and therefore the price that 0:14:57.920 --> 0:15:03.000 they can charge customers as perhaps maintaining volumes at this 0:15:03.080 --> 0:15:05.240 level of growing them a little bit. I don't think 0:15:05.280 --> 0:15:08.480 this is going to be even a significant evolution in 0:15:08.560 --> 0:15:10.080 the story. 0:15:11.320 --> 0:15:14.640 William Stein from Turier Securities, thank you very much, recapping 0:15:14.720 --> 0:15:17.600 One of the questions posed on the call. Google parent 0:15:17.680 --> 0:15:19.840 Alphabet was also about with this earnings last night, the 0:15:19.840 --> 0:15:22.840 company said demand for AI products boost sales and now 0:15:22.920 --> 0:15:27.800 requires an extreme increase in capital spending. Schweeder Cajeria, managing 0:15:27.840 --> 0:15:31.200 director of Global Internet at Wolf Research, joins us I 0:15:31.240 --> 0:15:35.080 put the earnings transcript through various generative AI tools. And 0:15:35.120 --> 0:15:38.040 the answer I got out Schwetter was AI fueled growth 0:15:38.400 --> 0:15:40.800 at a higher cost. Does that summit for you? 0:15:42.880 --> 0:15:43.280 I think so. 0:15:43.360 --> 0:15:45.600 First of all, thanks for having me ed. I think so. 0:15:45.960 --> 0:15:50.480 There's just a tremendous amount of focus on AI, and 0:15:50.760 --> 0:15:56.040 the conclusion here from the earnings is that AI related 0:15:56.080 --> 0:16:01.440 investments are showing are bearing fruit, not only in search, 0:16:01.520 --> 0:16:04.200 which is their bread and butter, but also in cloud, 0:16:04.360 --> 0:16:10.160 where revenue grew significantly higher than where expectations were. So 0:16:10.520 --> 0:16:14.440 fundamentals are getting stronger for Google, They're showing it, and 0:16:14.480 --> 0:16:19.320 they're justifying with greater gap xpend so that they can continue. 0:16:20.160 --> 0:16:22.400 The durability is justified. 0:16:22.880 --> 0:16:25.080 We'll see whether Ed was putting into notebook ILM and 0:16:25.120 --> 0:16:27.479 making it into a podcast shreutter. But I am interested 0:16:27.720 --> 0:16:30.000 as to why this is negative for Amazon. Your note 0:16:30.040 --> 0:16:32.120 is so interesting. Ultimately, think that this is a beat 0:16:32.120 --> 0:16:33.800 across the board. You think this is a good news 0:16:33.840 --> 0:16:35.320 story for METSA but not for Amazon. 0:16:36.520 --> 0:16:38.240 Well, I think it is a positive. I should have 0:16:38.240 --> 0:16:41.440 been a little bit clearer in my note, just to 0:16:41.480 --> 0:16:43.680 be you know, a little bit more upfront. In the 0:16:43.800 --> 0:16:46.880 second quarter, I think it would be positive for Amazon. 0:16:47.240 --> 0:16:50.560 With the cloud numbers, we're pretty strong for Google, and 0:16:51.120 --> 0:16:53.280 there's not a ton of reason to think Amazon also 0:16:53.320 --> 0:16:57.520 wouldn't have seen benefits of demand. The incremental negative pieces 0:16:57.600 --> 0:17:01.920 that Google also commented on applied demand remaining tight and 0:17:02.000 --> 0:17:04.800 so the capacity constraints. They are not out of the woods. 0:17:05.119 --> 0:17:07.760 And one of the key focus areas for investors was 0:17:07.800 --> 0:17:11.040 whether we'd see acceleration in AWS revenue growth in the 0:17:11.080 --> 0:17:13.560 back half, and as early as the third quarter. We 0:17:13.640 --> 0:17:16.159 may not see that just yet. It could be pushed 0:17:16.160 --> 0:17:17.320 into Q four or beyond. 0:17:19.160 --> 0:17:21.080 One of your self psych colleagues had a know oubt 0:17:21.119 --> 0:17:24.240 that basically said Google's hand has been forced by open 0:17:24.280 --> 0:17:27.359 AI and open hours commitment to spend also in the 0:17:27.440 --> 0:17:30.840 talent domain. How honest do you feel Alphabet or Google 0:17:30.920 --> 0:17:34.399 were about the pressures they're facing to spend their way 0:17:34.760 --> 0:17:36.920 into a situation or out of a situation. 0:17:38.359 --> 0:17:41.960 Well, I think that there's a growing amount of pressure 0:17:41.960 --> 0:17:45.080 and that is pretty well documented across these you know, 0:17:45.440 --> 0:17:49.000 across what Mark Zuckerberg is saying and what is actually happening. 0:17:49.520 --> 0:17:52.640 But the benefit with Google is also that they do 0:17:52.840 --> 0:17:57.359 also have the advantage of the infrastructure of the Deep 0:17:57.680 --> 0:18:00.800 Deep Mind Research group of some of the best engineers 0:18:00.800 --> 0:18:04.600 that are out there. So not only a high bay package, 0:18:04.680 --> 0:18:08.960 but also who you work with, how how innovative you 0:18:09.000 --> 0:18:11.240 work is all that also matters. I do think that 0:18:11.280 --> 0:18:14.200 there was some truth behind what Sundar was saying that, 0:18:14.600 --> 0:18:17.560 you know, the retention levels are kind of where they 0:18:17.640 --> 0:18:18.960 want want to be. 0:18:19.600 --> 0:18:22.160 Schwider Cojuria of Wolf Research is always great to get 0:18:22.160 --> 0:18:24.879 your notes and your expertise on the show. Thank You Now. 0:18:24.960 --> 0:18:28.280 Another company out with earnings, Service Now. The company says 0:18:28.320 --> 0:18:30.800 it's these strong outlook for revenue growth in the third 0:18:30.840 --> 0:18:33.919 quarture as customers adopts AI software tools. I spoke with 0:18:33.960 --> 0:18:36.800 Service Now CEO Bill McDermott. You're out to the company's 0:18:36.800 --> 0:18:39.320 savings on labor with the help of its own AI. 0:18:40.800 --> 0:18:45.040 We're slowing down the hiring in jobs that are quite 0:18:45.080 --> 0:18:48.480 frankly soul crushing jobs. I mean, just think about like 0:18:48.640 --> 0:18:52.960 it support. You know, ninety seven percent of standard software 0:18:53.240 --> 0:18:59.040 automatically generated by AI now, and the pressure on supporting 0:18:59.400 --> 0:19:03.960 function and people has been cut in half. Customer support 0:19:04.080 --> 0:19:07.680 eighty percent of the customer inquiries in cases now are 0:19:07.680 --> 0:19:11.600 fully managed by agentic AI. And think about security and 0:19:11.720 --> 0:19:14.480 risk management in real time. All the patchwork and the 0:19:14.560 --> 0:19:17.960 change management is done by agents now. And the same 0:19:18.040 --> 0:19:20.320 is true go to market. You know, think about people 0:19:20.359 --> 0:19:23.480 that talk to customers. There's no prep work anymore. Their 0:19:23.520 --> 0:19:27.520 AI teammate is doing that for them at service now, 0:19:27.880 --> 0:19:31.399 So the speed of the company has accelerated dramatically. The 0:19:31.480 --> 0:19:35.040 supporting cast of the sole crushing work is now being 0:19:35.080 --> 0:19:37.600 done by agents. They work hard twenty four by seven. 0:19:37.640 --> 0:19:39.840 You don't have to pay them, and they don't need 0:19:39.840 --> 0:19:43.040 any lunch, and they don't have any healthcare benefits, so 0:19:43.080 --> 0:19:47.199 they're very affordable. And that really complements all workforce. So 0:19:47.240 --> 0:19:51.359 we're still hiring, but we're hiring less in those supporting functions. 0:19:51.640 --> 0:19:54.160 And I suppose that's going to happen to all well 0:19:54.240 --> 0:19:57.480 run companies. And I say that because we now need 0:19:57.520 --> 0:20:01.760 to think about how you've run a core operation on service. Now, 0:20:02.040 --> 0:20:05.000 how you change the way you run your business processes. 0:20:05.320 --> 0:20:07.720 It could be ordered a cash, procure to pay. 0:20:07.600 --> 0:20:11.680 Hire to retire, designed to product. These are all processes 0:20:11.720 --> 0:20:15.000 that go across functions and we need to have those 0:20:15.000 --> 0:20:19.280 twentieth century ORG charts obliterated. We now need to work 0:20:19.280 --> 0:20:22.919 in teams. Teams need to work across processes and the 0:20:23.040 --> 0:20:25.840 agents need to be the best friend to humans. And 0:20:25.880 --> 0:20:28.160 if we do this right, the GDP of the world 0:20:28.200 --> 0:20:31.639 will go up. Companies will outperform not just on the 0:20:31.640 --> 0:20:35.080 op X costs, takeout and productivity, but they will be 0:20:35.600 --> 0:20:39.359 building new horizons, new business models, and new dreams on 0:20:39.440 --> 0:20:41.480 the service now platform. And we're excited. 0:20:42.040 --> 0:20:44.639 Service Now CEO Bill McDermott. Now, let's just shift to 0:20:44.680 --> 0:20:47.920 IBM shares briefly because they are down pretty significantly day 0:20:48.280 --> 0:20:51.480 of their lows. But actually spoke with Avin Krishna yesterday, 0:20:51.520 --> 0:20:54.800 the CEO of IBM. He had been pretty upbeat about 0:20:54.800 --> 0:20:57.160 his revenues, thinking it was of high quality, felt good 0:20:57.160 --> 0:20:59.800