1 00:00:02,520 --> 00:00:09,959 Speaker 1: Bloomberg Audio Studios, podcasts, radio news from the heart of 2 00:00:10,000 --> 00:00:15,720 Speaker 1: where innovation, money and power collide in Silicon Valley and beyond. 3 00:00:16,160 --> 00:00:40,360 Speaker 1: This is Bloomberg Technology with Caroline Hyde and Ed Ludlow. 4 00:00:35,040 --> 00:00:36,280 Speaker 2: Live from San Francisco. 5 00:00:36,400 --> 00:00:39,440 Speaker 3: This is Bloomberg Technology coming up in Nvidio CEO Jensen 6 00:00:39,479 --> 00:00:43,680 Speaker 3: Wang slams US restrictions on AI chip sales to China 7 00:00:44,040 --> 00:00:47,479 Speaker 3: and calls them a failure. Plus, President Trump vows to 8 00:00:47,479 --> 00:00:51,440 Speaker 3: make a Golden Dome missile defense shield fully operational by 9 00:00:51,479 --> 00:00:54,200 Speaker 3: the end of his term. And President Trump invites Elon 10 00:00:54,320 --> 00:00:57,200 Speaker 3: Musk to meet with South African leader Cyril Ramafosa with 11 00:00:57,440 --> 00:01:01,600 Speaker 3: starlink a point of discussion, Let's get straight to today's 12 00:01:01,600 --> 00:01:05,200 Speaker 3: top story and video. CEO Jensen Wong slamming US restrictions 13 00:01:05,440 --> 00:01:08,640 Speaker 3: on AI chip sales to China, calling them quote a 14 00:01:08,720 --> 00:01:12,640 Speaker 3: failure and urging the White House to lower barriers, saying, quote, 15 00:01:12,680 --> 00:01:16,360 Speaker 3: the US should maximize the speed of AI diffusion because 16 00:01:16,400 --> 00:01:20,200 Speaker 3: if we don't, the competition will come. Bloomberg's Ian King, 17 00:01:20,240 --> 00:01:23,720 Speaker 3: who leads our coverage as Semiconductors, is here in San Francisco. 18 00:01:24,360 --> 00:01:27,640 Speaker 3: What is Jensen Wong's grievance and how does it relate 19 00:01:27,680 --> 00:01:28,840 Speaker 3: to in Nvidia's business. 20 00:01:29,160 --> 00:01:32,200 Speaker 4: Yeah, that's the key point here is that China is 21 00:01:32,240 --> 00:01:36,880 Speaker 4: the largest market for semiconductors in video obviously has faced 22 00:01:36,959 --> 00:01:39,680 Speaker 4: increasing regulations from Washington, which. 23 00:01:39,520 --> 00:01:41,720 Speaker 2: Is restricting its ability to do business there. 24 00:01:41,920 --> 00:01:44,679 Speaker 4: The latest of which we saw was the H twenty chip, 25 00:01:44,720 --> 00:01:46,760 Speaker 4: which he can no longer ship there. And this was 26 00:01:46,800 --> 00:01:48,800 Speaker 4: a chip remember that he scaled down so that it 27 00:01:48,800 --> 00:01:51,880 Speaker 4: wouldn't trigger those restrictions. So life is getting harder and 28 00:01:51,920 --> 00:01:54,279 Speaker 4: harder for Jensen and for Invidia in China. 29 00:01:54,320 --> 00:01:57,120 Speaker 3: Scale down in the sense that he lowered the performance 30 00:01:57,160 --> 00:01:59,400 Speaker 3: of that chip so that it met parameises outlined by 31 00:01:59,440 --> 00:02:02,000 Speaker 3: the US government. And it's kind of an interesting change 32 00:02:02,000 --> 00:02:05,160 Speaker 3: of tone because it was just literally seven days ago 33 00:02:05,480 --> 00:02:08,120 Speaker 3: that Jensen one was with the President in the Middle East, 34 00:02:08,600 --> 00:02:11,720 Speaker 3: and at that time, when the Commerce Department rescinded the 35 00:02:11,720 --> 00:02:15,080 Speaker 3: diffusion rule, Jensen was quite happy about it. But there 36 00:02:15,120 --> 00:02:19,440 Speaker 3: is a distinction between in Vidia's ability to sell semiconductors 37 00:02:19,480 --> 00:02:22,440 Speaker 3: to the Gulf States and its hope to do business 38 00:02:22,520 --> 00:02:23,400 Speaker 3: directly with China. 39 00:02:23,560 --> 00:02:23,960 Speaker 5: That's right. 40 00:02:24,000 --> 00:02:26,480 Speaker 4: I mean, he's trying to parlay that sort of momentum 41 00:02:26,919 --> 00:02:29,720 Speaker 4: into a new a fresh look at what's going on 42 00:02:29,800 --> 00:02:33,800 Speaker 4: in China. The diffusion rule was basically going to restrict 43 00:02:33,800 --> 00:02:35,560 Speaker 4: an area of the world where there's a lot of 44 00:02:35,600 --> 00:02:39,040 Speaker 4: money going to be devoted to AI infrastructure, and you know, 45 00:02:39,360 --> 00:02:41,320 Speaker 4: there was a victory that the US companies are going 46 00:02:41,360 --> 00:02:43,240 Speaker 4: to be allowed to be part of that. His argument 47 00:02:43,360 --> 00:02:46,400 Speaker 4: is that should be the case everywhere, because then that'll 48 00:02:46,400 --> 00:02:50,519 Speaker 4: give you US technology the position that it needs and 49 00:02:50,560 --> 00:02:52,640 Speaker 4: the position it should occupy, which is that they're a 50 00:02:52,680 --> 00:02:53,880 Speaker 4: center of this revolution. 51 00:02:53,960 --> 00:02:56,400 Speaker 3: And a reminder that in Nvidia has its quarterly earnings 52 00:02:56,440 --> 00:02:57,200 Speaker 3: report next week. 53 00:02:57,240 --> 00:02:59,240 Speaker 2: Bloombogzine King, thank you very much. 54 00:02:59,280 --> 00:03:03,160 Speaker 3: Now, last week were multiple deals between US technology companies, 55 00:03:03,160 --> 00:03:05,920 Speaker 3: including in Nvidia and the Golf States as part of 56 00:03:05,960 --> 00:03:07,640 Speaker 3: President Trump's visit to the region. 57 00:03:08,120 --> 00:03:09,440 Speaker 2: In the week that's followed. 58 00:03:09,600 --> 00:03:13,359 Speaker 3: China's access to US technology via third countries or channels 59 00:03:13,680 --> 00:03:16,320 Speaker 3: like the Golf has since become a concern for the 60 00:03:16,320 --> 00:03:20,280 Speaker 3: technology industry to discuss. We're joined by White House Senior 61 00:03:20,320 --> 00:03:23,920 Speaker 3: Policy Advisor for AI, Sri Ram Krishnan, who was with 62 00:03:24,000 --> 00:03:27,160 Speaker 3: the President in the Middle East and participated in the 63 00:03:27,160 --> 00:03:30,200 Speaker 3: formation of some of those deals. Mister Christian, welcome to 64 00:03:30,200 --> 00:03:33,119 Speaker 3: Bloomberg Technology. Thank you for your time and for joining us. 65 00:03:33,639 --> 00:03:37,960 Speaker 3: You heard there the reporting overnight about Nvidia CEO Jensen 66 00:03:38,000 --> 00:03:43,400 Speaker 3: Wung's complaint about the state of American business and being 67 00:03:43,440 --> 00:03:46,680 Speaker 3: able to do business with China. You're familiar with that story. 68 00:03:47,200 --> 00:03:50,400 Speaker 3: What is your reaction to start to mister Hung's comments 69 00:03:50,640 --> 00:03:52,400 Speaker 3: and his line of argument. 70 00:03:55,160 --> 00:03:56,800 Speaker 6: Thank you for having me on the show. I'm a 71 00:03:56,800 --> 00:04:00,840 Speaker 6: big fan excited to finally be here. It may be 72 00:04:00,960 --> 00:04:04,600 Speaker 6: useful to set some context about how we got here. 73 00:04:05,360 --> 00:04:08,600 Speaker 6: In the first week of President Trump taking office, he 74 00:04:09,240 --> 00:04:10,880 Speaker 6: issued an executive. 75 00:04:10,480 --> 00:04:12,760 Speaker 2: Order which did away with the old. 76 00:04:12,640 --> 00:04:16,080 Speaker 6: Biden executive order on AI, and President Trump's executive order 77 00:04:16,120 --> 00:04:20,320 Speaker 6: had very clear marching orders. It said, we need dominance 78 00:04:20,600 --> 00:04:24,280 Speaker 6: for American AI. And since that first week, that's what 79 00:04:24,880 --> 00:04:28,880 Speaker 6: me the AIS are David Sachs, Michael Katsios, the head 80 00:04:28,880 --> 00:04:31,880 Speaker 6: of OSTB, and many many others have been working towards 81 00:04:31,920 --> 00:04:34,600 Speaker 6: and what you saw last week in the Gulf as 82 00:04:34,640 --> 00:04:37,480 Speaker 6: part of this historic trip was a milestone as part 83 00:04:37,480 --> 00:04:40,200 Speaker 6: of the dominance of American AI. And if I could 84 00:04:40,200 --> 00:04:43,960 Speaker 6: maybe quickly recap what we announced last week, so under 85 00:04:44,000 --> 00:04:47,960 Speaker 6: the Biden regime, it was pretty much impossible for our 86 00:04:48,080 --> 00:04:52,680 Speaker 6: allies to get access to our best and brightest advance 87 00:04:52,680 --> 00:04:56,080 Speaker 6: semi conductor technology, whether it be from Nvidia or AMD 88 00:04:56,520 --> 00:05:00,240 Speaker 6: or a whole set of other companies. We basically wide 89 00:05:00,360 --> 00:05:03,200 Speaker 6: the world into a set of GPU haves and a 90 00:05:03,200 --> 00:05:06,640 Speaker 6: bunch of GPU have nots. And essentially you really told 91 00:05:06,680 --> 00:05:10,160 Speaker 6: our allies to go pound sand and last week's trip, 92 00:05:10,200 --> 00:05:15,239 Speaker 6: which was historic, changed all that. And this deal, these 93 00:05:15,279 --> 00:05:18,279 Speaker 6: deals that we struck in the Middle East, has three 94 00:05:18,440 --> 00:05:22,359 Speaker 6: essential components which are very very important to understand. The 95 00:05:22,440 --> 00:05:26,160 Speaker 6: first is they are America first in the sense that 96 00:05:26,360 --> 00:05:29,440 Speaker 6: every one of these investments is going to have a 97 00:05:29,680 --> 00:05:34,800 Speaker 6: matching investment in American data centers and American AI infrastructure 98 00:05:35,120 --> 00:05:39,840 Speaker 6: right here in the United States. So if there's a 99 00:05:39,960 --> 00:05:43,120 Speaker 6: UAE or any of the countries are building data center 100 00:05:43,200 --> 00:05:46,039 Speaker 6: capacity over there, as a part of this framework, we 101 00:05:46,120 --> 00:05:49,400 Speaker 6: are going to get equal to build out here in. 102 00:05:49,279 --> 00:05:50,000 Speaker 2: The United States. 103 00:05:50,040 --> 00:05:52,000 Speaker 6: And I think that is a very very critical piece 104 00:05:52,040 --> 00:05:55,479 Speaker 6: of this in terms of how this deal is America first. 105 00:05:56,160 --> 00:06:01,520 Speaker 6: Number two, these deals and these gbus are predominantly going 106 00:06:01,560 --> 00:06:05,359 Speaker 6: to be run by American hyperscalers and American cloud service 107 00:06:05,360 --> 00:06:09,240 Speaker 6: providers and American companies. So for our businesses, a lot 108 00:06:09,240 --> 00:06:12,600 Speaker 6: of whom I'm probably watching the show right now. This 109 00:06:12,720 --> 00:06:17,440 Speaker 6: represents a amazing expansion opportunity, a new channel, if you will, 110 00:06:17,640 --> 00:06:21,520 Speaker 6: for a region which has obviously resource rich. 111 00:06:21,640 --> 00:06:22,520 Speaker 2: So that's number two. 112 00:06:22,560 --> 00:06:25,240 Speaker 6: A lot of these GPUs, most of these GPUs are 113 00:06:25,279 --> 00:06:29,800 Speaker 6: going to be run, hosted, controlled by American companies. And 114 00:06:29,839 --> 00:06:31,839 Speaker 6: the third part, and this is maybe the most important part, 115 00:06:31,839 --> 00:06:34,479 Speaker 6: because I think Ian and you touchdown, is on security. 116 00:06:34,720 --> 00:06:38,960 Speaker 6: So we took security incredibly seriously, and this agreement have 117 00:06:39,240 --> 00:06:42,480 Speaker 6: two key security productions, which is about how do we 118 00:06:42,560 --> 00:06:46,680 Speaker 6: stop the physical diversion of these GPUs to countries of concern? 119 00:06:47,000 --> 00:06:49,000 Speaker 6: And two, how do we make sure there is no 120 00:06:49,200 --> 00:06:52,600 Speaker 6: illegal or illicit remote access. So that's kind of the deal, 121 00:06:52,640 --> 00:06:54,880 Speaker 6: and that's why it's historic and the reason why this 122 00:06:54,920 --> 00:06:56,880 Speaker 6: is important, and again this is we're in a business show. 123 00:06:56,920 --> 00:06:59,640 Speaker 6: This will make sense is what is the history of 124 00:06:59,680 --> 00:07:02,839 Speaker 6: silicon value teach us? It teaches us that the companies 125 00:07:02,880 --> 00:07:06,320 Speaker 6: which have the biggest ecosystem, the biggest network effect, they 126 00:07:06,360 --> 00:07:10,679 Speaker 6: wind up winning and America AI Inc. Is no different. 127 00:07:10,960 --> 00:07:13,200 Speaker 6: So would this piece of AI diplomacy. What we are 128 00:07:13,240 --> 00:07:18,280 Speaker 6: doing is essentially locking in the American AI stack in 129 00:07:18,360 --> 00:07:19,920 Speaker 6: the Middle East. And I think that's going to set 130 00:07:19,960 --> 00:07:21,800 Speaker 6: us up for years and decades to come. 131 00:07:22,920 --> 00:07:26,800 Speaker 3: Mister Krishtnan, what Jensen Huang is essentially calling for is 132 00:07:26,840 --> 00:07:32,360 Speaker 3: to maximize speed up diffusion, and his argument is predicated 133 00:07:32,400 --> 00:07:34,960 Speaker 3: on the idea that there's a lot of activity in China. 134 00:07:35,360 --> 00:07:37,880 Speaker 3: He has consistently said, like more than fifty percent of 135 00:07:37,880 --> 00:07:41,280 Speaker 3: the research or AI development is done there. What is 136 00:07:41,320 --> 00:07:44,680 Speaker 3: your current thinking and what will you advise the president 137 00:07:44,840 --> 00:07:48,760 Speaker 3: on either ending limits to China's ability to buy US 138 00:07:48,880 --> 00:07:50,880 Speaker 3: chips or adjusting the current rules. 139 00:07:52,720 --> 00:07:55,640 Speaker 6: Look, we obviously a huge aspect for Jensen. I was 140 00:07:55,680 --> 00:07:58,720 Speaker 6: with him and several other founders and CEOs in the 141 00:07:58,760 --> 00:08:02,720 Speaker 6: Middle East of last week. When it comes to I 142 00:08:02,720 --> 00:08:06,040 Speaker 6: think it's important to make a distinction between advanced of 143 00:08:06,120 --> 00:08:09,800 Speaker 6: my conductors being used inside China and everywhere else. When 144 00:08:09,800 --> 00:08:12,160 Speaker 6: it comes to inside China, I do think there is 145 00:08:12,240 --> 00:08:18,120 Speaker 6: still bipartisan and broad concern about what can happen to 146 00:08:18,160 --> 00:08:20,520 Speaker 6: these GPUs when they are physically inside Because a lot 147 00:08:20,520 --> 00:08:22,760 Speaker 6: of these are obviously dual use. They can be used 148 00:08:22,760 --> 00:08:25,720 Speaker 6: for a broad range of applications, and we've tried to 149 00:08:25,800 --> 00:08:28,680 Speaker 6: kind of thread the needle when it comes to export regulation. 150 00:08:29,240 --> 00:08:31,520 Speaker 6: But one of things to point out is I think 151 00:08:31,640 --> 00:08:33,800 Speaker 6: Jensen is right on a couple of these items. The 152 00:08:33,840 --> 00:08:37,760 Speaker 6: first is our previous export controls have failed from the 153 00:08:37,760 --> 00:08:40,920 Speaker 6: Biden regime, So if those export controls at work, we 154 00:08:40,920 --> 00:08:43,400 Speaker 6: would not be looking at deep c coming out the 155 00:08:43,400 --> 00:08:47,160 Speaker 6: first week of President Trump's being in office, which basically 156 00:08:47,240 --> 00:08:49,200 Speaker 6: kind of skirted around a lot of our export controls. 157 00:08:49,240 --> 00:08:52,400 Speaker 6: So the Biden regime did not enforces export controls correctly, 158 00:08:52,440 --> 00:08:54,920 Speaker 6: and we are taking measures to stop that. But again 159 00:08:55,000 --> 00:08:58,640 Speaker 6: it is important to distinguish between GPUs inside China to 160 00:08:58,720 --> 00:09:00,600 Speaker 6: the rest of the world. When it comes to the 161 00:09:00,640 --> 00:09:04,200 Speaker 6: rest of the world, we want American AI stack, starting 162 00:09:04,240 --> 00:09:07,680 Speaker 6: from the GPUs to the models to everything on top. 163 00:09:08,040 --> 00:09:10,680 Speaker 6: We want our companies a lot of them, obviously, you know, 164 00:09:11,000 --> 00:09:13,280 Speaker 6: sitting very very close to where are now in Silicon Valley, 165 00:09:13,520 --> 00:09:15,800 Speaker 6: to good diffuse around the world. And I think on 166 00:09:15,880 --> 00:09:19,680 Speaker 6: that Jensen and I and US are in violent agreement. 167 00:09:20,880 --> 00:09:24,320 Speaker 3: You raise an interesting point that many technologists saw, just 168 00:09:25,040 --> 00:09:28,520 Speaker 3: technology company founders in the Bayer particularly make to me, 169 00:09:28,600 --> 00:09:31,679 Speaker 3: which is why even bother with the infrastructure build out 170 00:09:31,679 --> 00:09:34,440 Speaker 3: in the Middle East, why not retain the full stack 171 00:09:34,559 --> 00:09:38,280 Speaker 3: and capacity here in America, both not just from a 172 00:09:38,320 --> 00:09:41,880 Speaker 3: national security point of view, but just the market have 173 00:09:42,040 --> 00:09:45,480 Speaker 3: all of the workload availability here on domestic soil. 174 00:09:47,280 --> 00:09:49,199 Speaker 6: I'm happy you asked THATID because I saw this come 175 00:09:49,280 --> 00:09:50,880 Speaker 6: up on X a couple of times saying here's what 176 00:09:50,880 --> 00:09:52,880 Speaker 6: I would say. The first is going back to the 177 00:09:52,920 --> 00:09:55,520 Speaker 6: first thing I said, as a part of these deals, 178 00:09:55,800 --> 00:09:58,920 Speaker 6: every bit of investment that goes into the Middle East 179 00:09:58,960 --> 00:10:00,480 Speaker 6: into infrastructure is. 180 00:10:00,360 --> 00:10:02,760 Speaker 2: Going to be matched in the United States. 181 00:10:02,800 --> 00:10:05,120 Speaker 6: So if we have a lead, we're going to continue 182 00:10:05,120 --> 00:10:07,440 Speaker 6: having that lead, and we are going full throttle on 183 00:10:07,559 --> 00:10:10,680 Speaker 6: build baby build, as what maybe the President would say 184 00:10:10,840 --> 00:10:14,319 Speaker 6: in term making sure we have enough infrastructure and enough capacity. 185 00:10:14,640 --> 00:10:16,320 Speaker 6: But the other hand, you have to think about the 186 00:10:16,360 --> 00:10:18,839 Speaker 6: fact that these and other countries in the world are 187 00:10:18,840 --> 00:10:22,360 Speaker 6: our allies. They are resource which they have ambitions, they 188 00:10:22,480 --> 00:10:25,400 Speaker 6: want to use American tech. What do you think winds 189 00:10:25,440 --> 00:10:27,880 Speaker 6: up happening if we basically say hey, we can't do 190 00:10:27,960 --> 00:10:29,880 Speaker 6: business with you, go look elsewhere. 191 00:10:30,040 --> 00:10:31,400 Speaker 2: They're going to go look elsewhere. 192 00:10:31,640 --> 00:10:34,480 Speaker 6: We would much rather have them locked in to the 193 00:10:34,520 --> 00:10:37,920 Speaker 6: American stack for years and dCas and what that would 194 00:10:37,960 --> 00:10:40,920 Speaker 6: represent to the founders since us watching this is an opportunity, 195 00:10:40,920 --> 00:10:44,360 Speaker 6: an opportunity for partnerships, an opportunity market expansion rather than 196 00:10:44,559 --> 00:10:47,520 Speaker 6: a market which is locked to an adversary. 197 00:10:48,760 --> 00:10:51,480 Speaker 3: An AI action plan is due in mid July. There 198 00:10:51,559 --> 00:10:54,640 Speaker 3: is a new AI policy document. Could you give any 199 00:10:54,720 --> 00:10:57,360 Speaker 3: preview for our audience of what you miss the sas 200 00:10:57,480 --> 00:11:01,040 Speaker 3: the President will ask Congress to leg slate on in 201 00:11:01,080 --> 00:11:03,400 Speaker 3: the field of AI and what policy is coming down 202 00:11:03,640 --> 00:11:04,120 Speaker 3: the pipe. 203 00:11:04,360 --> 00:11:09,080 Speaker 6: M Well, I know you had a White House Officer 204 00:11:09,160 --> 00:11:12,000 Speaker 6: Science and Technology Policy Director, Michael Tratzios on the show 205 00:11:12,040 --> 00:11:15,520 Speaker 6: who talked about this and him, you know, David Saxey, 206 00:11:15,559 --> 00:11:18,079 Speaker 6: Izar and me and many others are working on this now. 207 00:11:18,160 --> 00:11:20,559 Speaker 6: I don't want to sort of giveaway what is going 208 00:11:20,600 --> 00:11:23,800 Speaker 6: to be announced in a month and a half, but it's. 209 00:11:25,200 --> 00:11:28,719 Speaker 2: Please Ram. I would say a. 210 00:11:28,640 --> 00:11:31,160 Speaker 6: Few things right, and I think Michael talked about promote 211 00:11:31,160 --> 00:11:33,640 Speaker 6: and protect. I think here's what we would say. 212 00:11:33,920 --> 00:11:34,319 Speaker 2: First. 213 00:11:34,600 --> 00:11:38,840 Speaker 6: It's it's critical to have innovation and domination when it 214 00:11:38,880 --> 00:11:41,000 Speaker 6: comes to this technology. And what we saw in the 215 00:11:41,000 --> 00:11:44,800 Speaker 6: Biden regime was an incredibly complicated executive order, an incredibly 216 00:11:44,840 --> 00:11:49,400 Speaker 6: complicated diffusion rule, a patchwork of state laws. We we 217 00:11:49,440 --> 00:11:52,320 Speaker 6: need less of that. We need more of entrepreneurs like 218 00:11:52,360 --> 00:11:54,400 Speaker 6: the ones who are on your show are watching now. 219 00:11:54,559 --> 00:11:56,720 Speaker 6: We need we need them to go build right and 220 00:11:56,720 --> 00:11:58,560 Speaker 6: that's what we need, and I think that's probably going 221 00:11:58,600 --> 00:12:01,880 Speaker 6: to be a big theme. Second, build Baby built. We 222 00:12:01,920 --> 00:12:05,240 Speaker 6: need infrastructure. We need to make sure we get a 223 00:12:05,440 --> 00:12:07,800 Speaker 6: bureaucracy and red tape out of the way in making 224 00:12:07,880 --> 00:12:11,360 Speaker 6: sure this infrastructure is built out as quickly as we can. 225 00:12:11,600 --> 00:12:14,640 Speaker 6: And the third thing I would say is AI diplomacy. 226 00:12:15,240 --> 00:12:18,199 Speaker 6: We need our allies with the resources who want to 227 00:12:18,240 --> 00:12:20,640 Speaker 6: work with us. We want them to pick the American 228 00:12:20,720 --> 00:12:24,120 Speaker 6: stack and not be frustrated and wait for somebody else 229 00:12:24,160 --> 00:12:26,800 Speaker 6: to go build a competing solution. So I think those 230 00:12:26,840 --> 00:12:31,040 Speaker 6: would be three kind of mini previews of what's going 231 00:12:31,040 --> 00:12:32,480 Speaker 6: to be in the AI Action Plan. 232 00:12:33,080 --> 00:12:33,520 Speaker 2: So realm. 233 00:12:33,559 --> 00:12:35,400 Speaker 3: I think there's a lot of interest in your day 234 00:12:35,440 --> 00:12:38,079 Speaker 3: to day work with the administration. There are a number 235 00:12:38,080 --> 00:12:41,040 Speaker 3: of names from Silicon Valley, so to speak, you David Sachs, 236 00:12:41,440 --> 00:12:45,400 Speaker 3: but also Elon Musk. Has Elon been engaged with you 237 00:12:45,600 --> 00:12:49,040 Speaker 3: and had a seat at the table specifically with the 238 00:12:49,080 --> 00:12:53,079 Speaker 3: President's considerations around AI. I asked, because yesterday I was 239 00:12:53,120 --> 00:12:56,400 Speaker 3: at Dell Technologies World talking to them about the rapid 240 00:12:56,440 --> 00:13:00,360 Speaker 3: buildout of Colossus in the XAI Data Center. I wonder 241 00:13:00,360 --> 00:13:02,880 Speaker 3: how much influence he has on your thinking and on 242 00:13:02,920 --> 00:13:03,920 Speaker 3: the President's thinking. 243 00:13:04,400 --> 00:13:07,839 Speaker 6: Well, look, I mean Elon is a long term friend 244 00:13:07,960 --> 00:13:12,280 Speaker 6: of mine, of David Sachs, of the President, and he's 245 00:13:12,320 --> 00:13:14,760 Speaker 6: obviously one of the greatest entrepreneurs of our era, and 246 00:13:14,800 --> 00:13:17,040 Speaker 6: if he's hanging around the White House, we would be 247 00:13:17,120 --> 00:13:20,080 Speaker 6: stupid did not seek out his advice. So he you know, 248 00:13:20,200 --> 00:13:23,120 Speaker 6: he's definitely had useful input. But at the end of 249 00:13:23,120 --> 00:13:26,240 Speaker 6: the day, I think this comes from the President himself, 250 00:13:26,320 --> 00:13:28,880 Speaker 6: you know, the first or the second day in office, 251 00:13:29,200 --> 00:13:32,240 Speaker 6: his executive order said we need American AI dominance. And 252 00:13:32,280 --> 00:13:36,120 Speaker 6: there's a lot of people from Michael Katia's, David Sachs, obviously, 253 00:13:36,400 --> 00:13:39,480 Speaker 6: the Secretary of Commerce, Howard Lutnik, who was instrumental in 254 00:13:39,520 --> 00:13:42,239 Speaker 6: seeing these deals happen, many many, many others. 255 00:13:42,000 --> 00:13:44,599 Speaker 3: Who are all working towards this gold Syriam Krishna and 256 00:13:44,640 --> 00:13:47,040 Speaker 3: the White House Senior Policy Advisor Forray, I thank you 257 00:13:47,160 --> 00:13:50,240 Speaker 3: very much for joining us humbling big technology. Now coming out, 258 00:13:50,280 --> 00:13:53,760 Speaker 3: President Trump fast tracks his Golden Dome defense plan. We're 259 00:13:53,800 --> 00:13:56,880 Speaker 3: going to discuss that more and some breaking news. Bitcoin 260 00:13:57,000 --> 00:14:01,160 Speaker 3: has hit a fresh record high, largely driven by optimism 261 00:14:01,520 --> 00:14:03,800 Speaker 3: around coming US regulations. 262 00:14:03,960 --> 00:14:05,240 Speaker 2: This is Bloomberg technology. 263 00:14:08,960 --> 00:14:12,000 Speaker 3: President Trump has vowed that a new Golden Dome missile 264 00:14:12,000 --> 00:14:15,240 Speaker 3: defense would be fully operational in three years time, saying 265 00:14:15,240 --> 00:14:17,680 Speaker 3: it would be able to protect the US from ballistic, 266 00:14:17,960 --> 00:14:20,720 Speaker 3: hypersonic and advanced cruise missiles. I want to bring in 267 00:14:20,720 --> 00:14:24,160 Speaker 3: bloombergs Michael Shepherd in DC. Michael, what are the details 268 00:14:24,200 --> 00:14:25,400 Speaker 3: of the president's pledge? 269 00:14:26,600 --> 00:14:29,520 Speaker 7: Well, you know, actually the details were a little bit short. 270 00:14:29,600 --> 00:14:32,520 Speaker 7: He did outline what he expected would be the cost 271 00:14:32,600 --> 00:14:35,960 Speaker 7: for this system, which would be space based and designed 272 00:14:36,000 --> 00:14:39,360 Speaker 7: to protect a huge area of the United States territory 273 00:14:39,360 --> 00:14:43,960 Speaker 7: from Alaska to Florida to Hawaii from ballistic and hypersonic 274 00:14:44,040 --> 00:14:47,400 Speaker 7: missile attacks. He gave that three year timeline that you outlined. 275 00:14:47,720 --> 00:14:50,880 Speaker 7: Many believed that to be optimistic. His price tag for 276 00:14:50,960 --> 00:14:53,880 Speaker 7: it is one hundred and seventy five billion dollars over 277 00:14:53,920 --> 00:14:56,920 Speaker 7: the next several years. But the Congressional Budget Office has 278 00:14:56,960 --> 00:14:59,720 Speaker 7: a different view on the cost. It sees it going 279 00:14:59,800 --> 00:15:02,840 Speaker 7: up as much as five hundred and forty two billion 280 00:15:02,880 --> 00:15:06,120 Speaker 7: dollars in taking as many as twenty years to deploy. 281 00:15:06,400 --> 00:15:08,760 Speaker 7: And there are a lot of uncertainties about this technology 282 00:15:08,920 --> 00:15:12,280 Speaker 7: ed including how it would get deployed, and who would 283 00:15:12,320 --> 00:15:15,160 Speaker 7: be in the lead to do it. We see some 284 00:15:15,240 --> 00:15:18,040 Speaker 7: of the biggest names that we know from American defense, 285 00:15:18,080 --> 00:15:21,880 Speaker 7: including Raytheon and Lockheed Martin and EL three Harris getting 286 00:15:21,960 --> 00:15:24,200 Speaker 7: name checked at the event yesterday at the White House 287 00:15:24,480 --> 00:15:27,680 Speaker 7: where parts of this were unveiled together with Defense Secretary 288 00:15:27,680 --> 00:15:30,040 Speaker 7: Pete Hegseth. But we also know that there are other 289 00:15:30,160 --> 00:15:32,800 Speaker 7: startups that are trying to get involved in this as well, 290 00:15:32,960 --> 00:15:35,960 Speaker 7: and of course there is Elon. Musk SpaceX is one 291 00:15:36,040 --> 00:15:39,040 Speaker 7: of the top Pentagon contractors when it comes to defense 292 00:15:39,800 --> 00:15:43,680 Speaker 7: satellites and launches of satellites, and they've also done a 293 00:15:43,960 --> 00:15:47,640 Speaker 7: significant amount of work for the intelligence sector here in 294 00:15:47,840 --> 00:15:48,880 Speaker 7: Washington as well. 295 00:15:49,280 --> 00:15:51,000 Speaker 5: They would be a likely contender. 296 00:15:51,240 --> 00:15:53,720 Speaker 7: But there's a lot that we don't know about this system, 297 00:15:53,800 --> 00:15:56,720 Speaker 7: and the three year timeline, according to many experts, is 298 00:15:56,840 --> 00:16:00,640 Speaker 7: really optimistic given the task that the President is outlining 299 00:16:00,720 --> 00:16:03,800 Speaker 7: and the expectations he's putting on it. Remember, this is 300 00:16:03,920 --> 00:16:06,840 Speaker 7: a campaign pledge that he made much in the same 301 00:16:06,880 --> 00:16:09,880 Speaker 7: way that he is promising this big, beautiful tax cut 302 00:16:10,000 --> 00:16:14,280 Speaker 7: for all those tariffs, and so this missile defense program 303 00:16:14,360 --> 00:16:16,520 Speaker 7: is another pillar of what he had promised the. 304 00:16:16,560 --> 00:16:20,640 Speaker 3: Voters Bloomberg's Michael Shepard, thank you very much. Staying in Washington, 305 00:16:20,800 --> 00:16:24,920 Speaker 3: President Trump is meeting with South African leader similar Cyril Ramaposa, 306 00:16:25,160 --> 00:16:28,480 Speaker 3: with Elon Musk also set to join. Bloomberg's reported that 307 00:16:28,560 --> 00:16:31,360 Speaker 3: the South African government is set to offer Musk a 308 00:16:31,400 --> 00:16:35,440 Speaker 3: workaround of black ownership laws, allowing his Starlink service to 309 00:16:35,480 --> 00:16:38,440 Speaker 3: operate in the country. Musk has been a staunch critic 310 00:16:38,640 --> 00:16:42,280 Speaker 3: of Ramaphosa's government and what he calls racist laws. Here's 311 00:16:42,280 --> 00:16:45,920 Speaker 3: what must hold Bloomberg yesterday at the Katar Economic Forum. 312 00:16:46,360 --> 00:16:48,960 Speaker 8: It's improper for the racist laws in South Africa. All 313 00:16:49,040 --> 00:16:52,320 Speaker 8: races retweet it equally and there should be no preference 314 00:16:52,320 --> 00:16:55,120 Speaker 8: given to one or the other. Whereas there are now 315 00:16:55,120 --> 00:16:57,320 Speaker 8: one hundred and forty laws in South Africa that to give. 316 00:16:58,800 --> 00:16:59,760 Speaker 2: That basically. 317 00:17:02,120 --> 00:17:05,040 Speaker 8: Give strong preface to to if you're black South African 318 00:17:05,400 --> 00:17:08,480 Speaker 8: and not otherwise. And so now I'm in the sobsid 319 00:17:08,560 --> 00:17:10,440 Speaker 8: situation where I was born in South Africa but can't 320 00:17:10,440 --> 00:17:13,440 Speaker 8: get a license to our greats in Starlink because I'm 321 00:17:13,440 --> 00:17:14,040 Speaker 8: not black. 322 00:17:15,240 --> 00:17:17,680 Speaker 3: From all, we joined by Bloomberg's Max Chafkin, a member 323 00:17:17,720 --> 00:17:18,520 Speaker 3: of our Elon Inc. 324 00:17:18,560 --> 00:17:19,240 Speaker 2: Team as well. 325 00:17:19,320 --> 00:17:22,520 Speaker 3: And the crux of that that sound bite was a 326 00:17:22,600 --> 00:17:28,000 Speaker 3: question about whether Trump's relationship with Musk opens doors for 327 00:17:28,080 --> 00:17:31,960 Speaker 3: Starlink in certain jurisdictions because those countries want to get 328 00:17:31,960 --> 00:17:34,040 Speaker 3: close to President Trump in this administration. 329 00:17:34,520 --> 00:17:34,800 Speaker 2: Yeah. 330 00:17:34,840 --> 00:17:38,880 Speaker 9: Absolutely, And look, there are reasons why South Africa might 331 00:17:39,080 --> 00:17:42,199 Speaker 9: be willing to do this outside of the influence of 332 00:17:42,240 --> 00:17:44,760 Speaker 9: Donald Trump and and perhaps an effort to you know, 333 00:17:44,920 --> 00:17:46,280 Speaker 9: make make the White House happy. 334 00:17:46,960 --> 00:17:47,119 Speaker 5: You know. 335 00:17:47,200 --> 00:17:50,800 Speaker 9: Starlink of course offers a compelling service. You know, We've 336 00:17:50,800 --> 00:17:54,280 Speaker 9: seen a lot of countries and people, especially if they 337 00:17:54,280 --> 00:17:55,760 Speaker 9: can afford it, you know, embrace this. 338 00:17:56,200 --> 00:17:56,359 Speaker 10: You know. 339 00:17:56,400 --> 00:17:59,560 Speaker 9: On the other hand, this is clearly a case where 340 00:17:59,600 --> 00:18:03,280 Speaker 9: Elon Musk is benefiting from his connections with the White House. 341 00:18:03,280 --> 00:18:06,200 Speaker 9: We're seeing this not just in South Africa, but lots 342 00:18:06,200 --> 00:18:09,679 Speaker 9: of countries over the last couple of months changing their 343 00:18:09,760 --> 00:18:13,879 Speaker 9: rules essentially allowing this company to operate. You know, telecom 344 00:18:13,960 --> 00:18:17,280 Speaker 9: satellite telecom in particular is a heavily regulated field, not 345 00:18:17,640 --> 00:18:19,679 Speaker 9: just in South Africa but around the world. 346 00:18:19,840 --> 00:18:22,159 Speaker 2: So this is part of a larger pattern. 347 00:18:22,200 --> 00:18:25,960 Speaker 3: I would say, Musk borne in pretorious South Africa. This 348 00:18:26,119 --> 00:18:28,960 Speaker 3: is just one case study, right, Elon Musk has met 349 00:18:29,000 --> 00:18:32,560 Speaker 3: with several heads of state, either directly because he's with 350 00:18:32,640 --> 00:18:36,400 Speaker 3: Trump in person or all on the president's behalf since 351 00:18:36,440 --> 00:18:37,440 Speaker 3: Trump took office. 352 00:18:37,640 --> 00:18:40,600 Speaker 9: Yeah, and what's interesting about this meeting today, I mean, 353 00:18:40,640 --> 00:18:43,280 Speaker 9: you know, besides the dynamic with Musk as a critic, 354 00:18:43,680 --> 00:18:46,440 Speaker 9: is just that you know, Musk has said in various ways, 355 00:18:46,480 --> 00:18:48,720 Speaker 9: I'm backing away from the White House. You know, he 356 00:18:48,760 --> 00:18:51,119 Speaker 9: said he's spending less time there, he's maybe going to 357 00:18:51,200 --> 00:18:53,760 Speaker 9: donate less money, at least accordinator his comments yesterday. On 358 00:18:53,800 --> 00:18:56,359 Speaker 9: the other hand, here he is at the White House 359 00:18:56,520 --> 00:18:58,760 Speaker 9: with Donald Trump. Also yesterday he said he was going 360 00:18:58,800 --> 00:19:00,760 Speaker 9: to spend two days this week I think mentioned a 361 00:19:00,840 --> 00:19:03,680 Speaker 9: cabinet meeting as well. So you know, on one hand, 362 00:19:03,720 --> 00:19:06,720 Speaker 9: he may be making noises about backing away, but it 363 00:19:06,840 --> 00:19:09,080 Speaker 9: sure doesn't seem like he's doing a ton of. 364 00:19:09,040 --> 00:19:12,640 Speaker 2: That right now. Bloomberg's Max Chafkin, thank you very much. 365 00:19:21,160 --> 00:19:25,200 Speaker 3: Google is rolling out its experimental AI agent project Mariner 366 00:19:25,359 --> 00:19:28,760 Speaker 3: two subscribers in the United States. The research prototype can 367 00:19:28,800 --> 00:19:30,440 Speaker 3: browse the Internet and use websites. 368 00:19:30,440 --> 00:19:32,440 Speaker 2: Bloomberg' Jackie Davilas heard. 369 00:19:32,240 --> 00:19:35,399 Speaker 3: More about it from Jacqueline Conslman, the director of product 370 00:19:35,440 --> 00:19:36,440 Speaker 3: management at Google. 371 00:19:36,520 --> 00:19:36,760 Speaker 5: Lab. 372 00:19:37,400 --> 00:19:41,320 Speaker 11: Project Mariner is a research prototype that's exploring the future 373 00:19:41,359 --> 00:19:44,800 Speaker 11: of human agent interactions starting with browsers. And what we 374 00:19:44,880 --> 00:19:47,679 Speaker 11: launched today was a web app that allows anybody to 375 00:19:47,760 --> 00:19:50,160 Speaker 11: give Project Mariner a task and it will be able 376 00:19:50,200 --> 00:19:52,800 Speaker 11: to go and navigate across the Internet to complete it. 377 00:19:53,119 --> 00:19:55,800 Speaker 11: And the specific version that we launch today can also 378 00:19:55,840 --> 00:19:57,760 Speaker 11: handle up to ten tasks that it can do at 379 00:19:57,800 --> 00:20:00,719 Speaker 11: the same time, allowing people to really try to multitask. 380 00:20:01,640 --> 00:20:05,119 Speaker 12: Agents are really kind of the next frontier of AI. 381 00:20:05,320 --> 00:20:06,720 Speaker 5: We talk a lot about. 382 00:20:06,400 --> 00:20:10,560 Speaker 12: How artificial intelligence can boost productivity, and agents seem to 383 00:20:10,600 --> 00:20:13,359 Speaker 12: be really at the forefront of that. But I'm curious 384 00:20:13,400 --> 00:20:17,720 Speaker 12: about how this moves from the innovation lab to prime time, 385 00:20:17,880 --> 00:20:20,879 Speaker 12: especially because if you're asking in AI to do something, 386 00:20:20,920 --> 00:20:22,440 Speaker 12: you really want to make sure it gets it right. 387 00:20:23,160 --> 00:20:25,320 Speaker 11: Absolutely, So I think that that's part of the power 388 00:20:25,359 --> 00:20:29,680 Speaker 11: of launching Project Mariner as this you know, research prototype, 389 00:20:29,720 --> 00:20:32,760 Speaker 11: as an experiment. Right now, we are putting users at 390 00:20:32,760 --> 00:20:35,600 Speaker 11: the forefront of being able to try this out firsthand. 391 00:20:36,240 --> 00:20:40,719 Speaker 11: We are pairing nicely the ability for Project Mariner to multitask, 392 00:20:40,920 --> 00:20:43,399 Speaker 11: but also a user at any time can watch everything 393 00:20:43,440 --> 00:20:44,040 Speaker 11: that it's doing. 394 00:20:44,080 --> 00:20:45,680 Speaker 2: It'll get a live view of. 395 00:20:45,600 --> 00:20:47,879 Speaker 11: The task as it's happening, and a user can take 396 00:20:47,920 --> 00:20:50,360 Speaker 11: over the task at any time. They can also pause 397 00:20:50,480 --> 00:20:52,679 Speaker 11: the agent from doing the task and resume it at 398 00:20:52,720 --> 00:20:55,159 Speaker 11: a later point in time, or they could choose to 399 00:20:55,320 --> 00:20:58,080 Speaker 11: trust Project Mariner if they're giving it something that's you know, 400 00:20:58,119 --> 00:21:00,280 Speaker 11: a simpler task, and go back to whatever it is 401 00:21:00,280 --> 00:21:02,960 Speaker 11: that they're doing. So we've really tried to blend user 402 00:21:03,080 --> 00:21:06,479 Speaker 11: control with pushing forward what these agents are capable of 403 00:21:06,720 --> 00:21:09,360 Speaker 11: and launching it as that early experiment kind of let's 404 00:21:09,400 --> 00:21:11,159 Speaker 11: us take it to that next phase of bringing it 405 00:21:11,200 --> 00:21:11,720 Speaker 11: to people. 406 00:21:13,240 --> 00:21:16,720 Speaker 3: That was Jacqueline Conzelman, Director Product management at Google Lab, 407 00:21:16,760 --> 00:21:20,320 Speaker 3: speaking to Bloomberg's Jackie Devalo's look at shares of Google. 408 00:21:20,359 --> 00:21:23,240 Speaker 3: By the way, this is easily the best performer on 409 00:21:23,240 --> 00:21:26,080 Speaker 3: the Nazek one hundred, outperforming the other mag seven names, 410 00:21:26,240 --> 00:21:29,000 Speaker 3: and it's all reaction to what was happening at Google 411 00:21:29,040 --> 00:21:31,520 Speaker 3: Io on track for its biggest jump since early April, 412 00:21:31,720 --> 00:21:34,760 Speaker 3: the stock trading at its highest level since early March. 413 00:21:35,160 --> 00:21:38,760 Speaker 3: Analysts are very positive about what they heard, and they 414 00:21:38,800 --> 00:21:42,800 Speaker 3: note that the company's updates have basically eased concerns about 415 00:21:42,880 --> 00:21:46,240 Speaker 3: its ability to win in this crowded AI space. We 416 00:21:46,280 --> 00:21:50,200 Speaker 3: will continue to talk about Google throughout the air and what. 417 00:21:50,160 --> 00:21:51,280 Speaker 2: We learned out at Google Io. 418 00:21:51,400 --> 00:21:54,000 Speaker 3: Now, coming up, we speak with CRUSO CEO Chase lock 419 00:21:54,080 --> 00:21:57,399 Speaker 3: Miller as the company sees the second phase beginning for 420 00:21:57,480 --> 00:22:01,800 Speaker 3: its Texas AI data centers. Really interesting and timely conversation 421 00:22:01,880 --> 00:22:02,879 Speaker 3: coming up here. 422 00:22:02,760 --> 00:22:05,480 Speaker 2: From San Francisco. This is Bloomberg Technology. 423 00:22:16,400 --> 00:22:18,520 Speaker 3: Welcome back to Bloomberg Technology, Ed love loow here in 424 00:22:18,520 --> 00:22:20,680 Speaker 3: San Francisco. Let's talk a little bit more about what's 425 00:22:20,720 --> 00:22:23,240 Speaker 3: going on in financial markets center tech sector. Martin Norton, 426 00:22:23,320 --> 00:22:26,880 Speaker 3: chief investment strategist at Empire, joins us. Now I'm trying 427 00:22:26,920 --> 00:22:28,800 Speaker 3: to find a common thread of what's going on because 428 00:22:28,800 --> 00:22:31,680 Speaker 3: when I woke up this morning, a lot of the 429 00:22:31,720 --> 00:22:35,320 Speaker 3: headlines on the terminal were about economics and concerns about growth. 430 00:22:35,359 --> 00:22:37,760 Speaker 3: There's been a lot of FED speak and then Tech 431 00:22:37,800 --> 00:22:41,920 Speaker 3: brings us back baby Google Io developers conference headlines, alphabets 432 00:22:41,960 --> 00:22:44,639 Speaker 3: up five percent. What is it that you're paying attention 433 00:22:44,720 --> 00:22:45,399 Speaker 3: to right now? 434 00:22:45,560 --> 00:22:48,200 Speaker 10: Well, I think you're making an interesting point because really 435 00:22:48,280 --> 00:22:51,320 Speaker 10: not just this morning, but the past few weeks, the 436 00:22:51,320 --> 00:22:54,200 Speaker 10: past few months, maybe you could say the entire twenty 437 00:22:54,240 --> 00:22:56,640 Speaker 10: twenty five years, so far has been focused on news 438 00:22:56,640 --> 00:23:01,040 Speaker 10: out of Washington and kind of that macroeconomic and fiscal dynamic. 439 00:23:01,160 --> 00:23:04,199 Speaker 10: And what we're seeing, at least for the moment, is 440 00:23:04,280 --> 00:23:08,000 Speaker 10: a return back to the fundamentals, the individual story and 441 00:23:08,040 --> 00:23:10,960 Speaker 10: the AI narrative, which has been lost for a while 442 00:23:11,640 --> 00:23:14,840 Speaker 10: as we dealt with deep seek and as we've turned macroeconomics. 443 00:23:14,880 --> 00:23:17,120 Speaker 10: So I think what we're seeing here is and move 444 00:23:17,200 --> 00:23:18,800 Speaker 10: back to the individual companies a. 445 00:23:18,720 --> 00:23:21,960 Speaker 3: Bit, right, And then it reminds me of all of 446 00:23:22,000 --> 00:23:24,439 Speaker 3: the focus we had earlier in the year on mag seven. 447 00:23:25,119 --> 00:23:28,080 Speaker 3: Perhaps actually more to the downside, but you know, if 448 00:23:28,119 --> 00:23:29,960 Speaker 3: you have a stock like Alphabet and it is up 449 00:23:29,960 --> 00:23:33,080 Speaker 3: that much, just purely from a points perspective, it's giving 450 00:23:33,119 --> 00:23:37,359 Speaker 3: us a little lift this morning. That's the moment in time. Yeah, sorry, 451 00:23:37,400 --> 00:23:39,119 Speaker 3: I don't mean to interrupt my All I was going 452 00:23:39,160 --> 00:23:40,879 Speaker 3: to go on to say is like it was a 453 00:23:40,880 --> 00:23:43,240 Speaker 3: moment in time where investors wanted to say, what is 454 00:23:43,240 --> 00:23:44,720 Speaker 3: it you're going to do in AI? And they gave 455 00:23:44,840 --> 00:23:48,800 Speaker 3: enough evidence I think over in Mountain View yesterday, Well, 456 00:23:48,880 --> 00:23:49,400 Speaker 3: I think. 457 00:23:49,200 --> 00:23:51,760 Speaker 10: That's absolutely the case. And I think what's also interesting 458 00:23:51,800 --> 00:23:54,320 Speaker 10: is when we look back to earlier in the year 459 00:23:55,080 --> 00:23:57,359 Speaker 10: and really Up until the present moment, a lot of 460 00:23:57,359 --> 00:24:00,480 Speaker 10: the focus has been on the spend for AI and 461 00:24:00,640 --> 00:24:03,639 Speaker 10: questions on whether that's reasonable, and at least with the 462 00:24:03,680 --> 00:24:06,080 Speaker 10: most recent news, and I think this has been kind 463 00:24:06,080 --> 00:24:08,920 Speaker 10: of filtering out over the course of the year. We're 464 00:24:08,920 --> 00:24:11,400 Speaker 10: seeing a bit more of that use case and that 465 00:24:11,600 --> 00:24:15,560 Speaker 10: gives kind of a shinier tinge to what we can 466 00:24:15,560 --> 00:24:18,440 Speaker 10: think about when it comes to AI. So it's shifting 467 00:24:18,480 --> 00:24:20,399 Speaker 10: the narrative a bit. Now, you know, whether this is 468 00:24:20,440 --> 00:24:23,160 Speaker 10: long lasting. I think there's a lot more to see 469 00:24:23,200 --> 00:24:25,639 Speaker 10: on that front, but it certainly is shifting the narrative 470 00:24:25,680 --> 00:24:28,520 Speaker 10: to the use case rather than just the spend on AI. 471 00:24:28,680 --> 00:24:33,800 Speaker 3: On the Capex front, Liberation Day is kind of it 472 00:24:33,840 --> 00:24:35,680 Speaker 3: feels like it's in the rear view mirror a bit. 473 00:24:35,800 --> 00:24:39,000 Speaker 3: But if there's one thing that's consistent about this administration 474 00:24:39,240 --> 00:24:42,639 Speaker 3: and trade, it's that anything can happen. Is the market 475 00:24:42,840 --> 00:24:45,840 Speaker 3: still braced for that? Is it checking over its shoulder 476 00:24:45,880 --> 00:24:47,639 Speaker 3: for some uncertainty in the tech sector. 477 00:24:48,080 --> 00:24:52,200 Speaker 10: Well, this is what I'm wrestling with. I think Liberation Day, 478 00:24:52,440 --> 00:24:55,520 Speaker 10: you know, the seismic change and trade where we don't 479 00:24:55,560 --> 00:24:58,359 Speaker 10: really even fully know where it lands quite yet. 480 00:24:58,960 --> 00:24:59,840 Speaker 5: It feels like. 481 00:25:00,080 --> 00:25:03,159 Speaker 10: Your point distant history, a blip on the radar, and 482 00:25:03,400 --> 00:25:07,359 Speaker 10: markets have rallied really powerfully. That's at the broad market level, 483 00:25:07,400 --> 00:25:10,119 Speaker 10: and that's also within technology. If we're looking at technology 484 00:25:10,160 --> 00:25:13,840 Speaker 10: as a sector on a valuation basis relative to its history, 485 00:25:13,880 --> 00:25:17,000 Speaker 10: it's back in those upper echelon of death stiles where 486 00:25:17,040 --> 00:25:20,480 Speaker 10: the valuations are really stretched. So that gives me some 487 00:25:20,800 --> 00:25:23,960 Speaker 10: concern that there is a bit of complacency or as 488 00:25:23,960 --> 00:25:26,720 Speaker 10: we're thinking about the future the range of outcomes, people 489 00:25:26,760 --> 00:25:29,879 Speaker 10: are really focused on the positive and less focused on 490 00:25:30,200 --> 00:25:32,679 Speaker 10: what could go wrong. That is the concern. But ed, 491 00:25:32,720 --> 00:25:35,080 Speaker 10: I think a really important point here is that the 492 00:25:35,160 --> 00:25:38,880 Speaker 10: mag seven actually have looked cheaper. They've been harder hit 493 00:25:39,200 --> 00:25:41,160 Speaker 10: over the course of twenty twenty five, and that's brought 494 00:25:41,200 --> 00:25:44,040 Speaker 10: their valuations down. So maybe there's a bit more margin 495 00:25:44,040 --> 00:25:46,359 Speaker 10: of safety in those megacap tech names. 496 00:25:47,160 --> 00:25:47,320 Speaker 2: Right. 497 00:25:47,359 --> 00:25:50,000 Speaker 3: That takes us to a broader conversation about valuations, right 498 00:25:50,080 --> 00:25:52,439 Speaker 3: that on any given day you can forget that the 499 00:25:52,480 --> 00:25:55,560 Speaker 3: technology sector is more than the MAG seven. Is there 500 00:25:55,600 --> 00:25:58,440 Speaker 3: anywhere that you see opportunity or that you're super cautious 501 00:25:58,480 --> 00:26:01,119 Speaker 3: on outside of that buck of megacap names? 502 00:26:01,680 --> 00:26:05,000 Speaker 10: I think I had more characterize my view as cautious. 503 00:26:05,040 --> 00:26:07,159 Speaker 10: Outside of the MAG seven. If you're breaking down the 504 00:26:07,160 --> 00:26:11,399 Speaker 10: technology sector going industry by industry, most of the industries 505 00:26:11,480 --> 00:26:14,240 Speaker 10: look a bit expensive. And it's really that, you know, 506 00:26:14,400 --> 00:26:16,879 Speaker 10: it's such a strange world to be talking about the 507 00:26:16,920 --> 00:26:21,080 Speaker 10: MAG seven as the valuation opportunity within tech within the market, 508 00:26:21,119 --> 00:26:23,760 Speaker 10: but that's kind of where our analysis is pointing us today. 509 00:26:25,400 --> 00:26:28,560 Speaker 3: Marta, I know that this is an entirely fair question, 510 00:26:28,760 --> 00:26:31,680 Speaker 3: but i'd ask you to have a crystal ball what 511 00:26:31,760 --> 00:26:33,800 Speaker 3: happens in the second half of this year for the 512 00:26:33,840 --> 00:26:34,760 Speaker 3: technology sector. 513 00:26:35,240 --> 00:26:37,800 Speaker 10: Well, you know, I think a big part of what 514 00:26:37,840 --> 00:26:39,840 Speaker 10: we're going to see over the second half of the 515 00:26:39,920 --> 00:26:44,920 Speaker 10: year is maybe not resolution, but additional clarity on what's 516 00:26:44,960 --> 00:26:48,639 Speaker 10: happening on the TARA front for technology, what happens on 517 00:26:48,680 --> 00:26:51,720 Speaker 10: the regulation front, what happens in terms of as you 518 00:26:51,800 --> 00:26:55,280 Speaker 10: were recently chatting about the administration's focus on building out 519 00:26:55,359 --> 00:26:58,160 Speaker 10: kind of the US as an AI superpower. I think 520 00:26:58,240 --> 00:27:00,840 Speaker 10: all of those different dynamics are going to have a 521 00:27:00,960 --> 00:27:04,240 Speaker 10: role to play in terms of how technology performs over 522 00:27:04,240 --> 00:27:06,439 Speaker 10: the course of twenty twenty five. And what we have 523 00:27:06,520 --> 00:27:10,080 Speaker 10: to ask ourselves as investors is what of those outcomes 524 00:27:10,080 --> 00:27:12,159 Speaker 10: are embedded in the price and where is there a 525 00:27:12,200 --> 00:27:15,600 Speaker 10: margin of safety? And so that's the kind of chemistry 526 00:27:15,640 --> 00:27:17,199 Speaker 10: that we have to do as we look at the 527 00:27:17,240 --> 00:27:19,400 Speaker 10: different movers coming over the course of. 528 00:27:19,320 --> 00:27:22,600 Speaker 3: Twenty five MARTA real quick a big moment in the 529 00:27:22,680 --> 00:27:26,600 Speaker 3: show earlier, Bitcoin at almost one hundred and nine thousand 530 00:27:26,720 --> 00:27:27,800 Speaker 3: US dollars for token. 531 00:27:28,240 --> 00:27:32,040 Speaker 2: You interested there. It's tough with this, It's real. 532 00:27:32,080 --> 00:27:34,439 Speaker 10: I mean what I really struggle with is where is 533 00:27:34,520 --> 00:27:37,240 Speaker 10: the fair value here? And I think that's really hard 534 00:27:37,240 --> 00:27:39,520 Speaker 10: to determine. So I think, yes, we're getting a lot 535 00:27:39,560 --> 00:27:43,400 Speaker 10: of news on the regulation front that's going to charge bitcoin, 536 00:27:43,440 --> 00:27:45,639 Speaker 10: that's going to charge crypto. But I think what we 537 00:27:45,680 --> 00:27:48,400 Speaker 10: have to ask ourselves is can we predict what's coming next? 538 00:27:48,400 --> 00:27:50,000 Speaker 10: And I think that's a bit harder question. 539 00:27:50,960 --> 00:27:53,399 Speaker 3: Martin Alton, Empower, Chief Investment Strategy. It's great to have 540 00:27:53,400 --> 00:27:55,640 Speaker 3: you back here on Bloomberg Technology. Thank you very much. 541 00:27:55,720 --> 00:27:57,440 Speaker 2: Now, Cruso blue out. 542 00:27:57,280 --> 00:28:02,120 Speaker 3: Capital and primary digital infrastructure funding for the second phase 543 00:28:02,359 --> 00:28:05,800 Speaker 3: of it's AI data center joint venture, raising now fifteen 544 00:28:05,840 --> 00:28:06,680 Speaker 3: billion dollars in. 545 00:28:06,640 --> 00:28:07,600 Speaker 2: Total for the project. 546 00:28:07,640 --> 00:28:10,879 Speaker 3: For more, Chase Lock Miller, Cruso CEO joins us here 547 00:28:10,880 --> 00:28:13,560 Speaker 3: in San Francisco. Actually, you featured in the show twenty 548 00:28:13,600 --> 00:28:17,400 Speaker 3: four hours ago because Emily Chang went to the Texas 549 00:28:17,400 --> 00:28:19,800 Speaker 3: Stargate site and brought us that kind of deep look 550 00:28:19,840 --> 00:28:25,480 Speaker 3: at its construction and your involvement there. Very simply explain 551 00:28:25,880 --> 00:28:29,879 Speaker 3: your role and this joint initiative that has raised this 552 00:28:30,000 --> 00:28:31,480 Speaker 3: funding and where the funding goes. 553 00:28:32,200 --> 00:28:35,480 Speaker 13: Sure, the announcement that we had this morning was fifteen 554 00:28:35,520 --> 00:28:40,240 Speaker 13: billion dollars in funding for developing a very large, purpose 555 00:28:40,280 --> 00:28:43,080 Speaker 13: built AI data center. You know, we're calling them AI factories, 556 00:28:43,160 --> 00:28:49,600 Speaker 13: these large scale data centers with chips that can manufacture intelligence. 557 00:28:50,680 --> 00:28:53,520 Speaker 13: So our role in all of this is, you know, 558 00:28:53,560 --> 00:28:58,320 Speaker 13: we are basically we developed the data center itself and 559 00:28:59,280 --> 00:29:02,040 Speaker 13: we're in charge of holding it and we will own 560 00:29:02,080 --> 00:29:05,080 Speaker 13: it in a joint venture partnership with Blue. 561 00:29:05,120 --> 00:29:11,080 Speaker 3: All data centers, to some are large buildings with concrete 562 00:29:11,120 --> 00:29:14,800 Speaker 3: foundations and metal around the outside, and no one really 563 00:29:14,840 --> 00:29:17,200 Speaker 3: knows what's inside them. The other way of looking at 564 00:29:17,200 --> 00:29:20,920 Speaker 3: it is that these are like massive infrastructure projects where 565 00:29:20,960 --> 00:29:23,920 Speaker 3: it's not just the compute inside all of the associated 566 00:29:23,920 --> 00:29:27,520 Speaker 3: infrastructure around Which part are you more responsible for? 567 00:29:29,000 --> 00:29:32,080 Speaker 13: So Cruso is a business, we'll handle the entire stack 568 00:29:32,720 --> 00:29:37,320 Speaker 13: for this specific project. We're focused on the design, build, 569 00:29:37,400 --> 00:29:43,880 Speaker 13: construction of this large scale data center. We you know, 570 00:29:44,200 --> 00:29:47,520 Speaker 13: I think it's worth appreciating that data centers as a 571 00:29:47,560 --> 00:29:52,600 Speaker 13: product have changed very dramatically in line with the changes 572 00:29:52,640 --> 00:29:55,840 Speaker 13: in compute architecture computer architectures over the last twenty years. 573 00:29:55,840 --> 00:29:58,400 Speaker 13: So if you look twenty years ago, a standard data 574 00:29:58,400 --> 00:30:00,560 Speaker 13: center rack might have been two kilo, it's maybe four 575 00:30:00,640 --> 00:30:04,080 Speaker 13: killowats today for the GB two hundred m VL seventy 576 00:30:04,080 --> 00:30:06,800 Speaker 13: two configuration, you know, we're budgeting one hundred and thirty 577 00:30:06,880 --> 00:30:09,760 Speaker 13: kilowatts per rack all what direct to chip liquid cooled, 578 00:30:11,440 --> 00:30:14,960 Speaker 13: water cooling, And if you look forward to the next generation, 579 00:30:15,040 --> 00:30:18,600 Speaker 13: what what Jensen announced recently at GtC with the Verra 580 00:30:18,720 --> 00:30:22,160 Speaker 13: Rubin Ultra Those are six hundred kilwatts per rack, right, 581 00:30:22,200 --> 00:30:25,680 Speaker 13: so you're looking at more than one hundred x increase 582 00:30:25,720 --> 00:30:29,520 Speaker 13: in overall power density. That fundamentally shifts the what you're 583 00:30:29,560 --> 00:30:31,480 Speaker 13: building in terms of when you when you're when you're 584 00:30:31,480 --> 00:30:33,320 Speaker 13: thinking about building and constructing a data center. 585 00:30:33,720 --> 00:30:36,360 Speaker 2: The other piece I'll highlight is that these. 586 00:30:36,320 --> 00:30:41,080 Speaker 13: Data centers are no longer a bunch of individual computers 587 00:30:41,560 --> 00:30:45,760 Speaker 13: all plugged in operating independent work. You know, workflows and 588 00:30:45,800 --> 00:30:49,120 Speaker 13: tasks you really should be thinking about. The data center 589 00:30:49,280 --> 00:30:52,280 Speaker 13: is the computer, right, All of these, all of these 590 00:30:52,360 --> 00:30:57,040 Speaker 13: scale link exactly are interconnected on the same high performance 591 00:30:57,120 --> 00:30:59,080 Speaker 13: r em A fabric so that they can really be 592 00:30:59,160 --> 00:31:01,000 Speaker 13: thinking together and sharing informations. 593 00:31:01,400 --> 00:31:03,840 Speaker 3: The thing is like in aggregate, right, we're talking about 594 00:31:03,880 --> 00:31:07,680 Speaker 3: ten thousand megawatt facilities plus or minus up to you know, 595 00:31:07,720 --> 00:31:10,600 Speaker 3: the future generation maybe twenty thousand megawat. Something that came 596 00:31:10,640 --> 00:31:13,400 Speaker 3: out at earnings of the hyperscalers is that they raise 597 00:31:13,440 --> 00:31:16,560 Speaker 3: capital expenditures in some cases like Meta, but there's real 598 00:31:16,600 --> 00:31:19,640 Speaker 3: evidence that that wasn't just like a bigger commitment. It's 599 00:31:19,640 --> 00:31:22,720 Speaker 3: because the cost of building a data center is higher. 600 00:31:23,280 --> 00:31:25,040 Speaker 2: Do you see that in your activities? 601 00:31:25,680 --> 00:31:27,840 Speaker 13: The cost of building it is higher. 602 00:31:27,600 --> 00:31:29,520 Speaker 2: Because materials, I mean labor. 603 00:31:29,560 --> 00:31:31,960 Speaker 13: Materials in labor are kind of the labor is honestly 604 00:31:32,000 --> 00:31:34,400 Speaker 13: the biggest expense here. And what we're seeing because there's 605 00:31:34,560 --> 00:31:38,479 Speaker 13: such a big boom, there's a shortage of labor, right, 606 00:31:38,520 --> 00:31:41,680 Speaker 13: There's a shortage of construction workers, there's a shortage of 607 00:31:41,720 --> 00:31:46,480 Speaker 13: electricians that can really make these facilities happen at scale. 608 00:31:47,320 --> 00:31:49,479 Speaker 3: Crystal Ball for you as well, you know, you've kind 609 00:31:49,480 --> 00:31:50,800 Speaker 3: of put Cruso on the map here. 610 00:31:50,840 --> 00:31:52,640 Speaker 2: But this is one project. 611 00:31:53,760 --> 00:31:55,800 Speaker 3: Are we seeing year on year on year all the 612 00:31:55,840 --> 00:31:59,240 Speaker 3: way through the twenty thirty just capital expenditure, increase more 613 00:31:59,320 --> 00:32:03,640 Speaker 3: data centers. There's no sort of plateauing of the activity. 614 00:32:03,680 --> 00:32:06,000 Speaker 13: You know, I can speak for what we're seeing, which 615 00:32:06,080 --> 00:32:10,720 Speaker 13: is no slowdown. We are seeing incredible demand from across 616 00:32:11,160 --> 00:32:14,360 Speaker 13: a diverse set of customers that want to build scaled 617 00:32:14,400 --> 00:32:17,000 Speaker 13: infrastructure to support their own. 618 00:32:16,840 --> 00:32:18,000 Speaker 5: AI computing needs. 619 00:32:18,760 --> 00:32:21,840 Speaker 13: And we see that scaling, you know, by orders of magnitude. 620 00:32:22,200 --> 00:32:25,200 Speaker 13: So from our perspective, you know, this is the largest 621 00:32:25,200 --> 00:32:28,080 Speaker 13: infrastructure investment in human history and it's not slowing down. 622 00:32:28,080 --> 00:32:31,320 Speaker 3: How big a challenge is securing energy and also the 623 00:32:31,440 --> 00:32:33,600 Speaker 3: utility infrastructure to support your sites. 624 00:32:33,960 --> 00:32:37,600 Speaker 13: So energy is the core of this in terms of 625 00:32:37,640 --> 00:32:39,880 Speaker 13: you know, we've never seen data center infrastructure built at 626 00:32:39,880 --> 00:32:40,680 Speaker 13: this scale before. 627 00:32:40,960 --> 00:32:41,720 Speaker 5: If you look at. 628 00:32:41,560 --> 00:32:44,440 Speaker 13: Northern Virginia as of the end of last year, a 629 00:32:44,520 --> 00:32:47,360 Speaker 13: report from JLLL, you know, shared that it was four 630 00:32:47,360 --> 00:32:50,040 Speaker 13: and a half gigawats of total capacity in Northern Virginia. 631 00:32:50,360 --> 00:32:53,520 Speaker 13: What we're building in Applene, Texas is one point two gigawatts, right, 632 00:32:53,560 --> 00:32:55,960 Speaker 13: so you're talking about one facility that's a quarter of 633 00:32:56,040 --> 00:32:58,600 Speaker 13: the size of all of the data center capacity that's 634 00:32:58,640 --> 00:33:01,400 Speaker 13: been built in the center of Data Center MECA in 635 00:33:01,440 --> 00:33:05,560 Speaker 13: Northern Virginia, So you know, and now we're looking at 636 00:33:05,600 --> 00:33:10,120 Speaker 13: facilities that are five times bigger. Right. So, energy and 637 00:33:10,160 --> 00:33:13,080 Speaker 13: the amount of energy required to support AI infrastructure at 638 00:33:13,080 --> 00:33:16,320 Speaker 13: scale is truly mind boggling. So it is a critical 639 00:33:16,360 --> 00:33:20,320 Speaker 13: component to solving and you know, I know we're focused 640 00:33:20,320 --> 00:33:24,200 Speaker 13: on doing it here in America and expanding you know, 641 00:33:24,200 --> 00:33:25,120 Speaker 13: also internationally. 642 00:33:25,720 --> 00:33:27,680 Speaker 3: Greuse so CEO Chase lock Miller. Great to have you 643 00:33:27,720 --> 00:33:30,200 Speaker 3: on the show. Thank you very much. Real quick call. 644 00:33:30,280 --> 00:33:32,600 Speaker 3: We've shared soaring today, by the way, on pace for 645 00:33:32,640 --> 00:33:35,920 Speaker 3: a four straight day of gains after City raised its 646 00:33:36,000 --> 00:33:39,160 Speaker 3: price target on the data center builder to ninety four 647 00:33:39,240 --> 00:33:43,000 Speaker 3: dollars from forty three dollars. Keeping in neutral rating, City 648 00:33:43,040 --> 00:33:46,120 Speaker 3: says the company's first quarter was quote out of the gate, 649 00:33:46,400 --> 00:33:49,440 Speaker 3: though the rest of results were mixed, and let's just 650 00:33:49,440 --> 00:33:53,480 Speaker 3: saying the earnings report reinforces call we've's high growth status, 651 00:33:53,600 --> 00:33:57,360 Speaker 3: especially with that recent four billion dollar Open AI expansion deal, 652 00:33:57,600 --> 00:34:01,080 Speaker 3: but more progress on profitability is to be seen, something 653 00:34:01,120 --> 00:34:02,480 Speaker 3: we discussed with the company this week. 654 00:34:02,480 --> 00:34:03,280 Speaker 2: All right, coming up. 655 00:34:03,920 --> 00:34:09,440 Speaker 3: These smart glasses are x Reel and Google's answer to 656 00:34:09,640 --> 00:34:12,640 Speaker 3: Meta's Ray Bands smart glasses. We're going to hear from 657 00:34:12,680 --> 00:34:16,080 Speaker 3: the CEO of the startup next. This is Bloomberg Technology. 658 00:34:27,880 --> 00:34:32,200 Speaker 3: Google's host of AI announcements yesterday included plans for augmented 659 00:34:32,320 --> 00:34:37,480 Speaker 3: reality glasses that could rival Meta's upcoming smart glasses releases. 660 00:34:37,600 --> 00:34:40,759 Speaker 3: To do this, the search giant has partnered with x Reel, 661 00:34:41,080 --> 00:34:45,719 Speaker 3: Chinese startup which makes popular AI sunglasses, and this new partnership, 662 00:34:45,760 --> 00:34:49,960 Speaker 3: called Project Aura, would be the first extended reality glasses 663 00:34:50,360 --> 00:34:54,799 Speaker 3: to run Google's Android x A operating system. 664 00:34:54,920 --> 00:34:56,040 Speaker 2: Let's bring into you the. 665 00:34:56,000 --> 00:34:59,640 Speaker 3: CEO and founder of x Well, welcome back to the program. 666 00:34:58,800 --> 00:34:59,960 Speaker 5: And thank you for having me. 667 00:35:00,400 --> 00:35:02,279 Speaker 3: I don't know that I saw this one coming. This is 668 00:35:02,320 --> 00:35:05,680 Speaker 3: a very big deal for x Rael with a big partner. 669 00:35:06,680 --> 00:35:08,480 Speaker 3: What was the logic behind doing it? 670 00:35:08,960 --> 00:35:12,360 Speaker 14: Okay, so first of all is I don't think it's 671 00:35:12,400 --> 00:35:14,279 Speaker 14: not just a big deal for us. I think it's 672 00:35:14,320 --> 00:35:17,920 Speaker 14: a big deal for the future of ARM Android. Finally, 673 00:35:17,960 --> 00:35:20,360 Speaker 14: speaking of the spatial computing, I feel. 674 00:35:20,080 --> 00:35:21,600 Speaker 5: Like all the pieces are ready. 675 00:35:21,800 --> 00:35:25,520 Speaker 14: The hardware is ready, phone factor is right, and you know, 676 00:35:25,680 --> 00:35:26,760 Speaker 14: the platform is ready. 677 00:35:26,960 --> 00:35:30,720 Speaker 5: ECO system is there, and most importantly Gemni AI. 678 00:35:30,880 --> 00:35:32,960 Speaker 14: We believe it's going to make the whole XR experience 679 00:35:33,080 --> 00:35:35,480 Speaker 14: so much better to the next level. So that's why 680 00:35:35,600 --> 00:35:37,320 Speaker 14: we're so excited to see this coming. 681 00:35:37,760 --> 00:35:38,480 Speaker 2: When's it coming? 682 00:35:38,560 --> 00:35:40,719 Speaker 3: Tell me about the product, like how real is this 683 00:35:41,200 --> 00:35:43,920 Speaker 3: in terms of when a consumer or a developer can 684 00:35:43,960 --> 00:35:45,160 Speaker 3: get their hands on the hardware. 685 00:35:45,280 --> 00:35:47,839 Speaker 14: I think developer will get them, you know, as soon 686 00:35:47,920 --> 00:35:49,560 Speaker 14: as you know the end of this year, is that 687 00:35:49,640 --> 00:35:52,640 Speaker 14: it No, this is the current generation or actual one 688 00:35:52,640 --> 00:35:54,520 Speaker 14: pro the one you try the last time. 689 00:35:54,480 --> 00:35:56,840 Speaker 3: For our audience, there's a set of glasses on the 690 00:35:56,880 --> 00:35:58,480 Speaker 3: desk in front of us, but that's not it. 691 00:35:58,480 --> 00:35:59,000 Speaker 2: I keep going. 692 00:35:59,440 --> 00:36:01,840 Speaker 14: So we believe, you know, for the consumer, it's going 693 00:36:01,920 --> 00:36:02,959 Speaker 14: to be early twenty six. 694 00:36:03,840 --> 00:36:07,160 Speaker 3: Has the relationship with Google allowed you to accelerate your 695 00:36:07,160 --> 00:36:08,760 Speaker 3: go to market on a new product? 696 00:36:08,840 --> 00:36:11,359 Speaker 14: Absolutely absolutely, And you know this is the first time. 697 00:36:11,400 --> 00:36:13,400 Speaker 14: I believe you're going to have this. You know, for 698 00:36:13,960 --> 00:36:15,799 Speaker 14: special computing, you're going to have this and to any 699 00:36:15,840 --> 00:36:18,719 Speaker 14: curius and you've got a chance to try all the 700 00:36:18,760 --> 00:36:21,719 Speaker 14: first party Google applications of XRI as well. 701 00:36:23,280 --> 00:36:23,600 Speaker 2: Meta. 702 00:36:24,239 --> 00:36:28,759 Speaker 3: I am a ray band metas user regularly that is 703 00:36:28,800 --> 00:36:30,560 Speaker 3: not a VRAR set of glasses. 704 00:36:30,600 --> 00:36:31,520 Speaker 2: They're smart glasses. 705 00:36:31,560 --> 00:36:33,719 Speaker 3: But as you know, I spent a bit of time 706 00:36:33,719 --> 00:36:39,799 Speaker 3: with a projector Ryan right, augmented reality glasses. Why or 707 00:36:39,840 --> 00:36:42,839 Speaker 3: why not? Is your offering going to compete with that? 708 00:36:43,160 --> 00:36:44,960 Speaker 3: Is it similar? Why is it dissimilar? 709 00:36:45,320 --> 00:36:46,360 Speaker 5: Okay, I will. 710 00:36:46,239 --> 00:36:49,520 Speaker 14: Say it's a competition, you know, but I really enjoy 711 00:36:49,680 --> 00:36:53,200 Speaker 14: working with Google because I believe the last you know, 712 00:36:53,360 --> 00:36:57,080 Speaker 14: meeting pieces will be ecosystem right, and Google always offer 713 00:36:57,200 --> 00:37:00,799 Speaker 14: a really open developer friendly kind of equals great. And 714 00:37:01,200 --> 00:37:04,200 Speaker 14: you know, the theory line I like about Android QSR 715 00:37:04,400 --> 00:37:07,919 Speaker 14: is if you already develop on Android, you develop on Android, Well. 716 00:37:07,800 --> 00:37:10,879 Speaker 3: How simple should we make this for the Bloomberg technology audience? Right, 717 00:37:11,040 --> 00:37:15,720 Speaker 3: Android from a smart phone perspective or you know in cars, 718 00:37:16,200 --> 00:37:20,280 Speaker 3: the os is everywhere around the world. So you're basically 719 00:37:20,440 --> 00:37:26,240 Speaker 3: arguing that I'm going to make the Androids smart glass simple. 720 00:37:26,520 --> 00:37:27,399 Speaker 5: That's right, I think. 721 00:37:27,440 --> 00:37:29,600 Speaker 14: You know, if you look at the success of Android, 722 00:37:30,080 --> 00:37:33,680 Speaker 14: Google just want to repeat that for XR. And you know, 723 00:37:33,960 --> 00:37:36,600 Speaker 14: this time they recognize they need to work with a 724 00:37:36,600 --> 00:37:39,680 Speaker 14: lot of other kind of OM partners and starting with 725 00:37:39,920 --> 00:37:41,680 Speaker 14: some song the Project Mohang. 726 00:37:41,480 --> 00:37:44,440 Speaker 5: Will be the first XR devices. And you know, we 727 00:37:44,520 --> 00:37:45,239 Speaker 5: have x rayal. 728 00:37:45,360 --> 00:37:48,759 Speaker 14: We have even gentle mounseer Bbie Parker announcing yesterday as well. 729 00:37:49,040 --> 00:37:50,319 Speaker 5: I see we're putting a lie. 730 00:37:50,320 --> 00:37:53,600 Speaker 14: It's working together to deliver a variety of different kinds 731 00:37:53,600 --> 00:37:55,080 Speaker 14: of devices for Android XR. 732 00:37:55,160 --> 00:37:57,319 Speaker 3: I've got a question for you from our audience when 733 00:37:57,360 --> 00:37:59,200 Speaker 3: I said on social media coming on the show, this 734 00:37:59,239 --> 00:38:02,200 Speaker 3: is from at VR on the score, Tonio on x 735 00:38:02,719 --> 00:38:05,040 Speaker 3: can you ask them if the new X one pro 736 00:38:05,320 --> 00:38:09,080 Speaker 3: needs the beam puck for a spatial experience? It is 737 00:38:09,200 --> 00:38:10,560 Speaker 3: vague on your website. 738 00:38:11,239 --> 00:38:14,680 Speaker 14: Well, okay, So for x real, we always have two 739 00:38:14,719 --> 00:38:15,720 Speaker 14: different product visions. 740 00:38:15,760 --> 00:38:17,960 Speaker 5: One is what we call spatial display. 741 00:38:18,080 --> 00:38:21,960 Speaker 14: It's basically easy to connect accessory that can plug into 742 00:38:22,040 --> 00:38:23,719 Speaker 14: virtually any devices, right. 743 00:38:23,800 --> 00:38:25,600 Speaker 5: And then there's another one we're. 744 00:38:25,520 --> 00:38:29,320 Speaker 14: Called spatial computing, which is onto an experience just similar 745 00:38:29,360 --> 00:38:31,520 Speaker 14: to a RYE or maybe even Apple Vision. 746 00:38:31,520 --> 00:38:35,800 Speaker 3: Problem real quick financials? How much money you're making right now? 747 00:38:35,880 --> 00:38:37,600 Speaker 3: Do you need to raise money? How's that looking? 748 00:38:37,880 --> 00:38:40,520 Speaker 5: Okay? So we do have a plan to go public 749 00:38:41,040 --> 00:38:44,919 Speaker 5: in twenty six. Okay, so I'm definitely in the US. 750 00:38:45,239 --> 00:38:45,600 Speaker 2: All right? 751 00:38:45,760 --> 00:38:48,680 Speaker 3: That was you do CEO of x reel with up 752 00:38:48,719 --> 00:38:51,319 Speaker 3: to the minute updates, not just on the latest technology. 753 00:38:51,560 --> 00:39:01,560 Speaker 3: But what's happening for the company too. I want to 754 00:39:01,560 --> 00:39:03,480 Speaker 3: go back to our top story and Video shares are 755 00:39:03,480 --> 00:39:07,080 Speaker 3: actually now markedly higher having opened under pressure. CEO Jensen 756 00:39:07,160 --> 00:39:11,320 Speaker 3: Wang overnight slam the US over chip sale restrictions to China. 757 00:39:11,840 --> 00:39:13,600 Speaker 3: For more, I want to bring in Mandy Singh of 758 00:39:13,640 --> 00:39:18,560 Speaker 3: Bloomberg Intelligence and Mandy Helper Bloomberg Technology audience understand the 759 00:39:18,600 --> 00:39:22,239 Speaker 3: state of the market in China. Basically, in Video the 760 00:39:22,280 --> 00:39:27,080 Speaker 3: world's leader in AI accelerators, but others like AMD right they. 761 00:39:26,960 --> 00:39:28,560 Speaker 2: Can't access that market. 762 00:39:28,760 --> 00:39:31,400 Speaker 3: But it's in that market where a lot of the 763 00:39:31,440 --> 00:39:35,359 Speaker 3: activity of AI development is taking place, and there are 764 00:39:35,440 --> 00:39:37,080 Speaker 3: data centers being built as well. 765 00:39:37,920 --> 00:39:38,160 Speaker 2: Yeah. 766 00:39:38,239 --> 00:39:41,640 Speaker 15: Look, I think we all know about Deepseeker and what 767 00:39:41,719 --> 00:39:44,000 Speaker 15: it did at the start of the year and all 768 00:39:44,080 --> 00:39:46,880 Speaker 15: these models. When you think about the training of deep 769 00:39:46,920 --> 00:39:51,560 Speaker 15: Seek or Ali Baba's models, it's happening within China and 770 00:39:52,400 --> 00:39:55,280 Speaker 15: a lot of the in video GPUs are still used 771 00:39:55,280 --> 00:39:58,480 Speaker 15: for training off LLM. So you know the fact that 772 00:39:59,120 --> 00:40:02,000 Speaker 15: in video on be able to sell to China, which 773 00:40:02,040 --> 00:40:05,520 Speaker 15: is about twenty percent of the overall GPU spend off 774 00:40:05,520 --> 00:40:08,799 Speaker 15: around two hundred and fifty billion, It tells you that 775 00:40:09,000 --> 00:40:13,879 Speaker 15: you know, they don't have that sovereign AI opportunity which 776 00:40:13,920 --> 00:40:17,480 Speaker 15: everyone is very excited about with Nvidia, and Nvidia did 777 00:40:17,600 --> 00:40:19,920 Speaker 15: sign up a few deals in the Middle East, But 778 00:40:19,960 --> 00:40:22,920 Speaker 15: when I look at the large language model landscape, I 779 00:40:22,920 --> 00:40:25,440 Speaker 15: mean the four or five of the best models are 780 00:40:25,440 --> 00:40:28,640 Speaker 15: here in the US, and then you have the remainder 781 00:40:28,719 --> 00:40:32,440 Speaker 15: of the other most prominent models in China. All the 782 00:40:32,600 --> 00:40:36,480 Speaker 15: other countries are still kind of investing building the infrastructure, 783 00:40:36,520 --> 00:40:39,719 Speaker 15: but they don't have the llms like these two countries have. 784 00:40:40,920 --> 00:40:44,080 Speaker 3: Mandy, when you publish research on the Bloomberg terminal, I 785 00:40:44,200 --> 00:40:45,840 Speaker 3: print it and I put it on my desk, and 786 00:40:45,880 --> 00:40:48,520 Speaker 3: I have this piece on my desk from March seventeenth. 787 00:40:48,840 --> 00:40:50,600 Speaker 2: Deep Seek to recast. 788 00:40:50,120 --> 00:40:53,360 Speaker 3: Five hundred and twenty five billion dollar CAPEX for hyperscalers 789 00:40:53,360 --> 00:40:57,239 Speaker 3: by twenty thirty two. In the context of the Nvidia story. 790 00:40:57,320 --> 00:41:00,000 Speaker 3: Just explain that thesis at the time and what's happened 791 00:41:00,360 --> 00:41:02,839 Speaker 3: as a result of the Deep Seek volatility. 792 00:41:03,239 --> 00:41:06,400 Speaker 15: Yeah, I mean, what Deep Seak did was really forced 793 00:41:06,719 --> 00:41:09,640 Speaker 15: everyone who was on the training side to look at, 794 00:41:09,680 --> 00:41:12,239 Speaker 15: you know, how much they wanted to spend upfront on 795 00:41:12,360 --> 00:41:16,480 Speaker 15: pre training and really pivot more towards reasoning and inferencing, 796 00:41:16,520 --> 00:41:18,640 Speaker 15: and that's what we have seen with Open Ai you 797 00:41:18,680 --> 00:41:21,640 Speaker 15: know one and three models and now with Google two 798 00:41:21,719 --> 00:41:24,920 Speaker 15: point five pro a lot of emphasis on reasoning. And 799 00:41:25,080 --> 00:41:28,600 Speaker 15: our thesis at that time was there is this shift 800 00:41:28,640 --> 00:41:32,399 Speaker 15: from training to inferencing that's happening now, and what will 801 00:41:32,440 --> 00:41:35,400 Speaker 15: happen as a result of that is that inferencing opportunity 802 00:41:35,440 --> 00:41:38,160 Speaker 15: will be a lot bigger, that part will grow a 803 00:41:38,200 --> 00:41:41,880 Speaker 15: lot faster, and the net result is that overall AI 804 00:41:42,000 --> 00:41:44,600 Speaker 15: market will continue to grow, but there will be that 805 00:41:44,680 --> 00:41:47,920 Speaker 15: shift from training to inferencing, and I think we are seeing. 806 00:41:47,640 --> 00:41:50,920 Speaker 3: That pan out and in China a part of it 807 00:41:51,000 --> 00:41:55,040 Speaker 3: maybe domestic names, using domestic champion Huawei, some of their 808 00:41:55,080 --> 00:41:57,279 Speaker 3: training chips men keep seeing Bloomberg Intelligence. 809 00:41:57,480 --> 00:41:58,160 Speaker 2: Thank you so much. 810 00:41:58,239 --> 00:42:01,520 Speaker 3: That does it for this edition of Bloomberg Technology. Don't 811 00:42:01,520 --> 00:42:03,880 Speaker 3: forget check out our podcasts. So many of you listen 812 00:42:03,920 --> 00:42:06,120 Speaker 3: to the show in podcast format. You can find it 813 00:42:06,160 --> 00:42:09,000 Speaker 3: on the terminal and online on Apple, Spotify and iHeart 814 00:42:09,160 --> 00:42:12,360 Speaker 3: from San Francisco. This is Bloomberg Technology.