1 00:00:00,040 --> 00:00:13,640 Speaker 1: Bloomberg Audio Studios, podcasts, radio news. Bloomberg Tech is alive 2 00:00:13,680 --> 00:00:17,479 Speaker 1: from coast to coast with Caroline Hide in New York 3 00:00:17,760 --> 00:00:20,560 Speaker 1: and Vow in San Francisco. 4 00:00:23,120 --> 00:00:26,119 Speaker 2: This is Bloomberg Tech coming up. Energy prices tumble and 5 00:00:26,160 --> 00:00:30,080 Speaker 2: global tech stocks rally after the US in Iran agreed 6 00:00:30,120 --> 00:00:31,280 Speaker 2: to a two week ceasefire. 7 00:00:31,440 --> 00:00:35,000 Speaker 3: Pas Anthropics new myth Us model is so powerful. It's 8 00:00:35,040 --> 00:00:38,040 Speaker 3: giving tech firms and competitors early access to get ahead 9 00:00:38,040 --> 00:00:40,519 Speaker 3: of possible cyber attacks that could reach havoc. 10 00:00:41,200 --> 00:00:44,080 Speaker 2: And Apple's first foldable phone it's still on track to 11 00:00:44,159 --> 00:00:48,040 Speaker 2: launch in September, despite reports of major manufacturing delays. 12 00:00:48,320 --> 00:00:50,960 Speaker 3: First check on those markets that you just pointed us 13 00:00:51,000 --> 00:00:55,600 Speaker 3: towards ed extraordinary moves as we have a tentative moves 14 00:00:55,640 --> 00:00:57,920 Speaker 3: towards a ceasefire. We of course have up to the 15 00:00:57,960 --> 00:01:01,960 Speaker 3: moment reporting of concerns in some ways that's being well 16 00:01:01,960 --> 00:01:05,119 Speaker 3: inflated upon at the moment, and Pakistan is speaking out. 17 00:01:05,120 --> 00:01:06,600 Speaker 3: But I'm looking at two and a half percent higher 18 00:01:06,600 --> 00:01:08,160 Speaker 3: on the nast that one hundred. We are off five 19 00:01:08,160 --> 00:01:09,880 Speaker 3: percentage point from the original highs. 20 00:01:10,080 --> 00:01:11,440 Speaker 4: As we worry about how this is. 21 00:01:11,400 --> 00:01:13,880 Speaker 3: Fire continues to unfold, but I'm looking at Brent crude 22 00:01:13,880 --> 00:01:16,440 Speaker 3: off like thirteen percent. This is a record move. At 23 00:01:16,480 --> 00:01:19,160 Speaker 3: one point it had fallen as much as seventeen percent 24 00:01:19,520 --> 00:01:22,000 Speaker 3: on the international contract. But let's get out to the 25 00:01:22,080 --> 00:01:25,440 Speaker 3: latest with Tyler. Kendall stands by in the White House 26 00:01:25,480 --> 00:01:28,760 Speaker 3: and Tyler, look, we are hopeful, but there are signs 27 00:01:28,760 --> 00:01:29,360 Speaker 3: of concerns. 28 00:01:31,680 --> 00:01:35,199 Speaker 5: Yeah, hey, Caroline, Well, at this point, it does appear 29 00:01:35,319 --> 00:01:38,240 Speaker 5: that the ceasefire is very fragile. This White House has 30 00:01:38,319 --> 00:01:42,320 Speaker 5: maintained that the US has achieved all of its military objectives, 31 00:01:42,400 --> 00:01:45,720 Speaker 5: touting tactical successes as what led to this truce in 32 00:01:45,760 --> 00:01:47,960 Speaker 5: the first place. But still we heard from Secretary Pete 33 00:01:47,960 --> 00:01:51,200 Speaker 5: Hegseth earlier today who said that the US military remains 34 00:01:51,240 --> 00:01:55,560 Speaker 5: ready to be redeployed and resume military operations if needed. 35 00:01:55,640 --> 00:01:59,240 Speaker 5: After the Vice President Jade Vance indicated earlier today that 36 00:01:59,280 --> 00:02:02,200 Speaker 5: the ceasefire or at this point, is something to watch 37 00:02:02,400 --> 00:02:05,560 Speaker 5: very closely because there's already been these reports that attacks 38 00:02:05,560 --> 00:02:08,000 Speaker 5: are continuing in the region, as a person familiar with 39 00:02:08,040 --> 00:02:11,119 Speaker 5: the matter confirms to Bloomberg News that Saudi Arabia's key 40 00:02:11,200 --> 00:02:15,040 Speaker 5: East West oil pipeline was hit by a drone earlier today, 41 00:02:15,120 --> 00:02:17,519 Speaker 5: though it appears that the damage is limited to pair 42 00:02:17,600 --> 00:02:21,000 Speaker 5: that with other reports about attacks against the UAE as 43 00:02:21,000 --> 00:02:24,040 Speaker 5: well as Kuwait now, Pakistan's Foreign minister, in a post 44 00:02:24,160 --> 00:02:27,000 Speaker 5: just within the last hour, said that Iran could be 45 00:02:27,040 --> 00:02:30,520 Speaker 5: at risk of pulling out of the truce if operations 46 00:02:30,639 --> 00:02:34,360 Speaker 5: against Lebanon continue. Keep in mind, Lebanon was not part 47 00:02:34,480 --> 00:02:37,760 Speaker 5: of this agreement. As Israel continues its military campaign against 48 00:02:37,760 --> 00:02:42,000 Speaker 5: the Iranian backed hesbela militant group ed in Caroline, Iranian 49 00:02:42,040 --> 00:02:44,640 Speaker 5: state media is now reporting that traffic in the Strait 50 00:02:44,639 --> 00:02:48,440 Speaker 5: of Remus has been halted amid Israeli attacks against Lebanon. 51 00:02:48,680 --> 00:02:50,720 Speaker 5: This is going to be something to watch very closely, 52 00:02:50,760 --> 00:02:53,520 Speaker 5: considering we know that this White House's position has been 53 00:02:53,800 --> 00:02:56,840 Speaker 5: that the strait needs to be reopened and there has 54 00:02:56,880 --> 00:02:59,760 Speaker 5: to be freedom of navigation in order for this truce 55 00:03:00,160 --> 00:03:01,200 Speaker 5: to be maintained. 56 00:03:01,440 --> 00:03:02,799 Speaker 4: President Trump is said to meet. 57 00:03:02,720 --> 00:03:05,720 Speaker 5: Later today with the NATO Secretary General Mark Rutta here 58 00:03:05,800 --> 00:03:08,160 Speaker 5: at the White House, and no doubt securing the straighter 59 00:03:08,240 --> 00:03:10,480 Speaker 5: removes will be the top topic of discussion. 60 00:03:10,760 --> 00:03:14,360 Speaker 3: Tyler Kennall, with the breakdown, we appreciate it. Look markets 61 00:03:14,760 --> 00:03:18,680 Speaker 3: are certainly higher after the Ceasphile has really boosted morale 62 00:03:18,720 --> 00:03:21,120 Speaker 3: on Wall Street. Now the analyst take Dan ives over 63 00:03:21,120 --> 00:03:23,919 Speaker 3: at web Bush. So the recent geopolitical tensions had pushed 64 00:03:23,919 --> 00:03:27,400 Speaker 3: tech into oversold territory, especially across the mag seven software 65 00:03:27,400 --> 00:03:30,200 Speaker 3: broader AI winners. He sees room for a remound on 66 00:03:30,240 --> 00:03:33,519 Speaker 3: the buy side. Let's get the take cook Dane gomman' 67 00:03:33,560 --> 00:03:38,200 Speaker 3: sacceass in management, who is leading on the bullishness, pointing 68 00:03:38,240 --> 00:03:42,040 Speaker 3: to AI Capex, saying stronger for longer. Brook, I'm so 69 00:03:42,120 --> 00:03:44,280 Speaker 3: happy to have you here because there is fragility that 70 00:03:44,520 --> 00:03:47,640 Speaker 3: is also potential to stop buying some beaten up stocks. 71 00:03:47,880 --> 00:03:49,000 Speaker 4: What do you make of today's news? 72 00:03:49,000 --> 00:03:50,360 Speaker 6: So, first, thanks for having me on. It's great to 73 00:03:50,400 --> 00:03:53,680 Speaker 6: be here today. So you know, our perspective is that 74 00:03:53,680 --> 00:03:56,480 Speaker 6: investors need to be focused on the size and the 75 00:03:56,520 --> 00:03:59,920 Speaker 6: magnitude of this shift that we're seeing and our baseline forecast. 76 00:04:00,040 --> 00:04:02,560 Speaker 6: So when we do our bottoms up work, continue to 77 00:04:02,600 --> 00:04:06,240 Speaker 6: point to acceleration and CAPEX spending and it being durable 78 00:04:06,240 --> 00:04:08,600 Speaker 6: for much longer. So we feel like investors need to 79 00:04:08,600 --> 00:04:11,000 Speaker 6: have exposure to the companies that are directly benefiting from that. 80 00:04:11,240 --> 00:04:13,480 Speaker 6: I don't think there's any question in anyone's mind right now, 81 00:04:13,520 --> 00:04:16,039 Speaker 6: as you're seeing these latest models come out about how 82 00:04:16,040 --> 00:04:17,920 Speaker 6: impactful they're going to be, how much they're going to change, 83 00:04:17,920 --> 00:04:19,800 Speaker 6: how much they're going to be able to do. And 84 00:04:19,880 --> 00:04:22,320 Speaker 6: so while there's uncertainty about how you know which sectors 85 00:04:22,320 --> 00:04:23,919 Speaker 6: get impacted in which way is and how it all 86 00:04:23,960 --> 00:04:27,440 Speaker 6: plays out, fundamentally, right now, we're in the start of 87 00:04:27,440 --> 00:04:29,440 Speaker 6: what's going to be a multi year CAPEX build out 88 00:04:29,720 --> 00:04:33,120 Speaker 6: and investors need exposure to that, and so our baseline is, 89 00:04:33,400 --> 00:04:35,400 Speaker 6: you know, you want to still be involved in the 90 00:04:35,440 --> 00:04:37,960 Speaker 6: semiconductor names, the networking names, things like that that are 91 00:04:38,000 --> 00:04:40,480 Speaker 6: the direct beneficiars of the spending. 92 00:04:40,880 --> 00:04:43,640 Speaker 2: Brook The corners of the technology market that are rallying 93 00:04:43,720 --> 00:04:48,400 Speaker 2: the hardest in this moment are memory, logic, storage. I'm 94 00:04:48,440 --> 00:04:52,080 Speaker 2: talking right now in response to the cease fire, but 95 00:04:52,200 --> 00:04:54,159 Speaker 2: that would indicate if you look at some of the 96 00:04:54,200 --> 00:04:56,640 Speaker 2: names that returned to the prices they were at prior 97 00:04:56,680 --> 00:04:59,159 Speaker 2: to the start of the war in Iran, that this 98 00:04:59,680 --> 00:05:05,040 Speaker 2: con has done nothing to divert the trajectory that you 99 00:05:05,080 --> 00:05:07,800 Speaker 2: believe capital expenditures are on when it comes to the 100 00:05:07,800 --> 00:05:08,760 Speaker 2: AI build out. 101 00:05:08,960 --> 00:05:11,680 Speaker 6: And our fundamental belief is is that there's been nothing 102 00:05:11,720 --> 00:05:15,760 Speaker 6: that's changed in that backdrop that the plans and expectations 103 00:05:15,760 --> 00:05:17,320 Speaker 6: that we had from these companies in terms of their 104 00:05:17,400 --> 00:05:20,760 Speaker 6: levels of spending, where they were putting the capacity, how 105 00:05:20,920 --> 00:05:24,200 Speaker 6: tight compute is right now. None of that has changed. 106 00:05:24,320 --> 00:05:26,840 Speaker 6: And yes, the world is more uncertain today than it 107 00:05:27,000 --> 00:05:30,600 Speaker 6: was four months ago, and risk premia have probably moved 108 00:05:30,640 --> 00:05:34,880 Speaker 6: higher structurally because of that. It doesn't change the direction 109 00:05:35,000 --> 00:05:37,400 Speaker 6: and the magnitude of what we're seeing. You remember that 110 00:05:37,440 --> 00:05:40,880 Speaker 6: in this cycle right now, Compute and all the associated 111 00:05:40,920 --> 00:05:43,880 Speaker 6: parts around compute are the most restricted and hardest to 112 00:05:43,920 --> 00:05:46,240 Speaker 6: get the thing in the marketplace. And if you want 113 00:05:46,240 --> 00:05:49,400 Speaker 6: these models to perform better and to keep you know, 114 00:05:49,440 --> 00:05:52,560 Speaker 6: doing more and more, the key ingredient is compute and 115 00:05:52,640 --> 00:05:54,480 Speaker 6: power and memory and all of the things that get 116 00:05:54,480 --> 00:05:56,760 Speaker 6: pulled on by that. So you know, there's nothing that 117 00:05:57,120 --> 00:06:00,480 Speaker 6: would change the outlook. We're very close to the big 118 00:06:00,480 --> 00:06:02,640 Speaker 6: spenders around this area. So when you think of the 119 00:06:02,640 --> 00:06:04,800 Speaker 6: mag seven companies, we spend a lot of time meeting 120 00:06:04,839 --> 00:06:07,359 Speaker 6: with those management teams and understanding what their plans look like. 121 00:06:08,120 --> 00:06:11,279 Speaker 6: None of them are wavering in their commitment and their 122 00:06:11,480 --> 00:06:13,440 Speaker 6: belief in the opportunity in front of them. 123 00:06:13,480 --> 00:06:18,039 Speaker 2: Behind this just very quickly did supply chains for chip 124 00:06:18,080 --> 00:06:21,560 Speaker 2: and chip manufacturing prove to be more resilient in this 125 00:06:21,680 --> 00:06:24,520 Speaker 2: moment than they were in the early days of the pandemic. 126 00:06:26,080 --> 00:06:28,960 Speaker 6: Yes, but I still think that those supply chains are 127 00:06:29,560 --> 00:06:32,000 Speaker 6: an area where there's a tremendous amount of focus in 128 00:06:32,040 --> 00:06:35,880 Speaker 6: building more resiliency, creating more chip capacity, doing it in 129 00:06:35,920 --> 00:06:39,240 Speaker 6: more geographies, and that build out is likely to persist. 130 00:06:39,600 --> 00:06:42,320 Speaker 6: But as of today and what we've seen through this 131 00:06:42,720 --> 00:06:46,039 Speaker 6: latest geopolitical crisis, there hasn't been any material impact on 132 00:06:46,040 --> 00:06:47,000 Speaker 6: supply chains there. 133 00:06:47,640 --> 00:06:49,800 Speaker 3: What's been really interesting is when we think about the 134 00:06:49,800 --> 00:06:53,599 Speaker 3: semis in video Marvel Broadcom names that you're in, you're like, 135 00:06:53,640 --> 00:06:58,520 Speaker 3: exposure to is there any hit longer term to the 136 00:06:58,600 --> 00:07:00,720 Speaker 3: margin or any of you point on how you get 137 00:07:00,760 --> 00:07:03,000 Speaker 3: into the picks and shovels still, because when we think 138 00:07:03,040 --> 00:07:04,000 Speaker 3: about the energy. 139 00:07:03,720 --> 00:07:06,040 Speaker 4: Prices, for example, it does inflate a little bit. 140 00:07:06,320 --> 00:07:09,040 Speaker 6: Yeah, So our perspective is just like you said, like 141 00:07:09,200 --> 00:07:12,040 Speaker 6: we need you investors' neat exposure to the picks and 142 00:07:12,040 --> 00:07:14,040 Speaker 6: shovels and this build up because it's still early and 143 00:07:14,080 --> 00:07:17,000 Speaker 6: still happening. You know, the margin structures right now are 144 00:07:17,200 --> 00:07:20,160 Speaker 6: high relative to history, but we think durable because of 145 00:07:20,200 --> 00:07:23,560 Speaker 6: the fact that you know, this relationship between your ability 146 00:07:23,600 --> 00:07:26,040 Speaker 6: to put compute in the ground and generate revenue for 147 00:07:26,160 --> 00:07:28,600 Speaker 6: the model companies and for the cloud vendors is still 148 00:07:28,680 --> 00:07:31,960 Speaker 6: very tight. So you know, there isn't a big discontinuity 149 00:07:31,960 --> 00:07:34,920 Speaker 6: event happening from either a supplier a demand side that 150 00:07:35,000 --> 00:07:37,400 Speaker 6: would upset margin structures. So you know, you mentioned a 151 00:07:37,400 --> 00:07:39,120 Speaker 6: couple of the names. One of them that we think 152 00:07:39,240 --> 00:07:42,360 Speaker 6: you know right now the market hasn't fully gotten its 153 00:07:42,440 --> 00:07:45,600 Speaker 6: arms around is what's happening with Marvel and last year 154 00:07:45,640 --> 00:07:47,960 Speaker 6: there's a lot of controversy around Marvel and their ability 155 00:07:48,000 --> 00:07:50,640 Speaker 6: to you know, hold share at Amazon and some of 156 00:07:50,640 --> 00:07:53,040 Speaker 6: the other big vendors. But we think the market's missing 157 00:07:53,160 --> 00:07:57,280 Speaker 6: is around all of the acic business. Marvel has this 158 00:07:57,320 --> 00:08:00,880 Speaker 6: great opportunity to you know, attach XPU attached products and 159 00:08:00,920 --> 00:08:03,960 Speaker 6: grow substantially. So we think that's a specific company where 160 00:08:03,960 --> 00:08:05,560 Speaker 6: there's a real opportunity around the upside. 161 00:08:05,600 --> 00:08:08,880 Speaker 3: I mean, I love the fundamental analysis and cunning through 162 00:08:08,920 --> 00:08:11,240 Speaker 3: the noise of the moment from a macro perspective, Go 163 00:08:11,240 --> 00:08:14,040 Speaker 3: go to the anthropic news. Then we started this conversation 164 00:08:14,080 --> 00:08:16,080 Speaker 3: while you're saying, like, look at the changes and the 165 00:08:16,160 --> 00:08:19,000 Speaker 3: in step changes are seeing in models you like the 166 00:08:19,160 --> 00:08:21,800 Speaker 3: likes of Well Pallel to Networks. They're one of the 167 00:08:21,880 --> 00:08:25,040 Speaker 3: key forty companies that's helping stress test and understand what 168 00:08:25,120 --> 00:08:25,840 Speaker 3: a cyber. 169 00:08:25,600 --> 00:08:26,240 Speaker 4: Risk is from that. 170 00:08:26,440 --> 00:08:28,520 Speaker 6: Yeah, so I describe myself a little bit as a 171 00:08:28,600 --> 00:08:29,880 Speaker 6: reformed software investor. 172 00:08:29,920 --> 00:08:30,600 Speaker 3: I grew up. 173 00:08:30,480 --> 00:08:33,640 Speaker 6: Following the entire software suite and IBOs. The software has 174 00:08:33,720 --> 00:08:35,840 Speaker 6: some headwinds right now that we can talk about, but 175 00:08:35,960 --> 00:08:39,240 Speaker 6: within that whole industry set, we think that companies that 176 00:08:39,240 --> 00:08:42,319 Speaker 6: are involved in cybersecurity have a massive opportunity in front 177 00:08:42,320 --> 00:08:43,920 Speaker 6: of them. Some of the companies that are helping with 178 00:08:44,000 --> 00:08:46,360 Speaker 6: data problems have really big opportunities in front of them. So, 179 00:08:46,640 --> 00:08:50,480 Speaker 6: you know, the news from Anthropic last night is, you know, 180 00:08:50,760 --> 00:08:53,440 Speaker 6: the power of those models is incredible and it is 181 00:08:53,520 --> 00:08:56,720 Speaker 6: exposing a whole bunch of vulnerabilities out there that people 182 00:08:56,760 --> 00:09:01,160 Speaker 6: didn't know existed. That is really good for the cyber companies. 183 00:09:01,840 --> 00:09:04,920 Speaker 6: You know, just from a business standpoint, there's going to 184 00:09:04,920 --> 00:09:07,160 Speaker 6: be more vulnerabilities exposed. Bad actors are going to try 185 00:09:07,200 --> 00:09:09,320 Speaker 6: and take advantage of that. Enterprises are going to need 186 00:09:09,360 --> 00:09:11,680 Speaker 6: to be defended in the right ways, and so owning 187 00:09:11,920 --> 00:09:15,679 Speaker 6: leading edge next gen security providers is a really good 188 00:09:15,720 --> 00:09:16,680 Speaker 6: place for INDUSTRISS to. 189 00:09:16,679 --> 00:09:18,199 Speaker 5: Be a brook. 190 00:09:18,440 --> 00:09:20,240 Speaker 2: I want to end this by going back to the 191 00:09:20,280 --> 00:09:25,320 Speaker 2: ceasefire and the war in Iran. You are an investor 192 00:09:25,360 --> 00:09:27,880 Speaker 2: in some of the biggest hyperscalers and the deployers of 193 00:09:27,880 --> 00:09:32,520 Speaker 2: capital expenditure, right and this president central policy was to 194 00:09:32,600 --> 00:09:37,760 Speaker 2: deregulate expedite permits and hold those names accountable to energy spending. 195 00:09:38,480 --> 00:09:41,400 Speaker 2: Then we had the war in Iran. And so just 196 00:09:41,679 --> 00:09:45,760 Speaker 2: from the technology investor's perspective, how your desk analysts and 197 00:09:45,800 --> 00:09:49,040 Speaker 2: you are modeling for the wild card that is a 198 00:09:49,080 --> 00:09:51,760 Speaker 2: president who on the one hand has an AI policy 199 00:09:52,160 --> 00:09:56,720 Speaker 2: grounded in deregulation and letting them move quickly and then 200 00:09:56,760 --> 00:09:59,400 Speaker 2: move to an escalated conflict in Iran. 201 00:09:59,679 --> 00:10:03,440 Speaker 6: Yes, So with every investment we're making, we're running lots 202 00:10:03,440 --> 00:10:06,240 Speaker 6: of scenarios on different parts of the business and what 203 00:10:06,280 --> 00:10:08,560 Speaker 6: can happen and what can change. So if you're talking 204 00:10:08,559 --> 00:10:11,960 Speaker 6: about like changes in the fundamental input costs of creating 205 00:10:12,040 --> 00:10:14,959 Speaker 6: compute to run these models and to do things as 206 00:10:15,000 --> 00:10:17,040 Speaker 6: we project out where we think, you know, p and 207 00:10:17,120 --> 00:10:19,240 Speaker 6: LS and cash flows can look like. That's one of 208 00:10:19,280 --> 00:10:21,400 Speaker 6: the factors that we're stress testing and we're saying, look, 209 00:10:21,600 --> 00:10:24,560 Speaker 6: if power you know, goes up by x factor, what 210 00:10:24,600 --> 00:10:26,960 Speaker 6: does it do to the ultimate return on a dollar 211 00:10:27,040 --> 00:10:29,720 Speaker 6: of GPU investment and how does that get priced in? 212 00:10:30,080 --> 00:10:32,920 Speaker 6: Similarly on other things, like you know, on memory supply, 213 00:10:33,080 --> 00:10:35,880 Speaker 6: and if we can't get enough memory to fund these things, 214 00:10:35,960 --> 00:10:37,960 Speaker 6: you know what happens there and how does that impact stuff. 215 00:10:38,080 --> 00:10:40,520 Speaker 6: So all of this comes into the fundamental frameworks and 216 00:10:40,559 --> 00:10:43,440 Speaker 6: the risk reward that we're managing, and we're looking through 217 00:10:43,480 --> 00:10:46,440 Speaker 6: with every investment we make. And what I would just 218 00:10:46,480 --> 00:10:48,160 Speaker 6: tell you, as a guy that's been investing in the 219 00:10:48,160 --> 00:10:51,480 Speaker 6: sector for a long time, I am skeptical of my 220 00:10:51,760 --> 00:10:55,120 Speaker 6: ability to be precise, but I'm really good about my 221 00:10:55,240 --> 00:10:58,800 Speaker 6: ability to understand ranges of outcome, and so we're always 222 00:10:58,880 --> 00:11:02,079 Speaker 6: looking for when those range is favor us and favor 223 00:11:02,160 --> 00:11:04,160 Speaker 6: us putting capital all the work. And so, you know, 224 00:11:04,240 --> 00:11:06,640 Speaker 6: to sort of end where I started with, we still 225 00:11:06,679 --> 00:11:09,880 Speaker 6: think this buildout of capex that were very early in it, 226 00:11:09,960 --> 00:11:13,000 Speaker 6: that were not late cycle, were early cycle to mid 227 00:11:13,000 --> 00:11:15,680 Speaker 6: cycle in this and that there's plenty of opportunity to 228 00:11:15,800 --> 00:11:18,760 Speaker 6: continue to compound wealth and growth in you know, the 229 00:11:18,800 --> 00:11:22,080 Speaker 6: fundamental chip pixel shovels, chip Ecosystems of AI. 230 00:11:23,280 --> 00:11:25,960 Speaker 2: Brooke Dayne of Goldman sax Asset Management doing the math 231 00:11:26,000 --> 00:11:28,400 Speaker 2: for us. I really appreciate that, Thank you very much. 232 00:11:28,600 --> 00:11:31,880 Speaker 2: Coming up, a massive Oracle data center could get roughly 233 00:11:31,920 --> 00:11:36,959 Speaker 2: fourteen billion dollars of debt financing from Pimco with the details. Next, 234 00:11:37,240 --> 00:11:38,360 Speaker 2: this has been big tech. 235 00:11:43,520 --> 00:11:47,480 Speaker 3: A flurry data center related news today. First, Fine Capital's 236 00:11:47,480 --> 00:11:52,880 Speaker 3: Bridge Data center has dropped neocloud supplier Magaspeed from Malaysia 237 00:11:52,920 --> 00:11:55,160 Speaker 3: Computing harm Now. That's according to the sources citing and 238 00:11:55,240 --> 00:11:58,319 Speaker 3: internal memo. The changes follow a US government probe into 239 00:11:58,360 --> 00:12:02,080 Speaker 3: Magaspeed over its ownerships, ructure and allegations it smuggled. 240 00:12:01,679 --> 00:12:03,360 Speaker 4: Advanced in video chips to China. 241 00:12:03,840 --> 00:12:06,760 Speaker 3: Meanwhile, here in the US, a city in Wisconsin has 242 00:12:06,840 --> 00:12:11,040 Speaker 3: passed the nation's first referendum restricting future data center construction, 243 00:12:11,200 --> 00:12:14,160 Speaker 3: according to the county's unofficial count now, Port Washington is 244 00:12:14,200 --> 00:12:16,440 Speaker 3: home to a one point three gigawatt project from opening 245 00:12:16,440 --> 00:12:19,200 Speaker 3: I on Oracle that won't be impacted by this vote. Still, 246 00:12:19,559 --> 00:12:21,840 Speaker 3: when the AI build out just getting underway, it's a 247 00:12:21,920 --> 00:12:25,559 Speaker 3: signal of the growing tension between local communities and developers. 248 00:12:25,559 --> 00:12:25,839 Speaker 4: Said. 249 00:12:26,800 --> 00:12:28,960 Speaker 2: The AI build out is still in early days and 250 00:12:29,000 --> 00:12:31,800 Speaker 2: over in Michigan, a massive Oracle data center could get 251 00:12:31,880 --> 00:12:35,880 Speaker 2: roughly fourteen billion dollars of debt financing from PIMCO. That's 252 00:12:35,920 --> 00:12:39,200 Speaker 2: according to sources. The deal, if finalized, would make the 253 00:12:39,200 --> 00:12:41,920 Speaker 2: bond giant a key backer of the project, to mark 254 00:12:42,160 --> 00:12:45,520 Speaker 2: its second jumbo financing for a data center within the 255 00:12:45,559 --> 00:12:48,440 Speaker 2: past year. Let's get out to bliding Boks Brady Ford Brady. 256 00:12:48,679 --> 00:12:50,760 Speaker 2: Let's go to the details here. And something that has 257 00:12:50,760 --> 00:12:53,120 Speaker 2: come up in conversation with you in the past is 258 00:12:53,440 --> 00:12:58,240 Speaker 2: the debt backs the project. It doesn't necessarily go straight 259 00:12:58,280 --> 00:13:01,400 Speaker 2: onto Oracle's debt load. Explain the mechanics of this and 260 00:13:01,440 --> 00:13:02,120 Speaker 2: what we need to know. 261 00:13:03,640 --> 00:13:07,800 Speaker 7: Yeah, zooming out. You know, these centers costs an incredible 262 00:13:07,840 --> 00:13:10,760 Speaker 7: amount of money that a company like Oracle, even with 263 00:13:10,800 --> 00:13:15,000 Speaker 7: its cash flowing database and SaaS applications, it needs to 264 00:13:15,000 --> 00:13:18,080 Speaker 7: bring a lot of partners in. So Pimco in this 265 00:13:18,160 --> 00:13:21,240 Speaker 7: case would be providing some debt essentially loans to get 266 00:13:21,240 --> 00:13:24,239 Speaker 7: this thing built. That would go on to the books 267 00:13:24,280 --> 00:13:28,040 Speaker 7: of related digital and intermediary who's actually building the site, 268 00:13:28,080 --> 00:13:30,680 Speaker 7: who then Oracle is going to rent from. If this 269 00:13:30,760 --> 00:13:32,959 Speaker 7: sounds like a lot of steps, it's because it is 270 00:13:33,040 --> 00:13:36,880 Speaker 7: because these projects are just so expensive, and so Oracle 271 00:13:36,960 --> 00:13:40,600 Speaker 7: has been kind of testing the market's appetite for financing 272 00:13:40,800 --> 00:13:43,760 Speaker 7: big data centers, and there have been moments where it 273 00:13:43,800 --> 00:13:47,640 Speaker 7: looked quite rocky. But you know, getting a cosign from 274 00:13:47,679 --> 00:13:50,680 Speaker 7: a company like Pimco which appears in the works, this 275 00:13:50,880 --> 00:13:52,400 Speaker 7: is a good sign for their build out. 276 00:13:53,040 --> 00:13:55,520 Speaker 3: And it's interesting as they name a new CFO that 277 00:13:55,559 --> 00:13:59,240 Speaker 3: they brought over from Schneider Electric, but also as they 278 00:13:59,440 --> 00:14:01,120 Speaker 3: tried to out lining to the market and put to 279 00:14:01,120 --> 00:14:03,000 Speaker 3: bed any anxiety that they were going to take on 280 00:14:03,040 --> 00:14:04,440 Speaker 3: more debt themselves. 281 00:14:05,120 --> 00:14:08,439 Speaker 7: Right, they've been doing some pretty incredible financial maneuvering, right, 282 00:14:08,440 --> 00:14:11,760 Speaker 7: I mean, they issued equity which nobody thought they would do, 283 00:14:11,880 --> 00:14:14,439 Speaker 7: they laid out the exact amount they're going to borrow, 284 00:14:14,920 --> 00:14:17,640 Speaker 7: and they named a CFO from the world of heavy 285 00:14:17,679 --> 00:14:21,160 Speaker 7: industry now from software. I mean that reflects that we're 286 00:14:21,160 --> 00:14:24,600 Speaker 7: in a new era where software companies are acting more 287 00:14:24,760 --> 00:14:28,800 Speaker 7: like ge or Boeing or industrial giants of the past 288 00:14:28,960 --> 00:14:31,640 Speaker 7: because of just these incredible capital outlays. 289 00:14:32,240 --> 00:14:35,200 Speaker 4: The most brady forward is always all across Oracle. We approciate, 290 00:14:35,280 --> 00:14:35,920 Speaker 4: appreciate it. 291 00:14:35,960 --> 00:14:40,200 Speaker 3: Now coming up Anthropping launches a new program aimed at 292 00:14:40,200 --> 00:14:43,160 Speaker 3: giving cyber defenders early access to its powerful new model 293 00:14:43,520 --> 00:14:45,800 Speaker 3: why to help them spot you threats were on that 294 00:14:45,920 --> 00:14:47,400 Speaker 3: Next is Bring Bad Tech. 295 00:14:52,400 --> 00:14:55,840 Speaker 2: Thropic is giving tech firms including Amazon, Apple, and Microsoft 296 00:14:55,880 --> 00:14:59,800 Speaker 2: access to a powerful but not widely released AI model 297 00:14:59,840 --> 00:15:02,040 Speaker 2: no as Mythos. It's part of an effort to prepare 298 00:15:02,280 --> 00:15:05,560 Speaker 2: for possible cyber attacks that could result from wider adoption 299 00:15:05,600 --> 00:15:08,520 Speaker 2: of the technology. Bloomberg cybersecurity report of Margie Murphy's with 300 00:15:08,640 --> 00:15:10,840 Speaker 2: us and MARGI yesterday just after this news here was 301 00:15:10,880 --> 00:15:13,600 Speaker 2: on stage with Mike Krieger, who co leads the labs Anthropic, 302 00:15:13,640 --> 00:15:17,120 Speaker 2: and he basically explained mythos at the model level incredibly 303 00:15:17,160 --> 00:15:20,120 Speaker 2: powerful relative to what's gone before it, and they don't 304 00:15:20,200 --> 00:15:22,600 Speaker 2: know if they'll release it generally. That was a big question. 305 00:15:22,920 --> 00:15:26,480 Speaker 2: So they have this set of limited technology companies and 306 00:15:26,560 --> 00:15:29,280 Speaker 2: one capability is cyber What do we need to know 307 00:15:29,320 --> 00:15:31,320 Speaker 2: about the kind of roll out and what on earth 308 00:15:31,360 --> 00:15:32,240 Speaker 2: these companies will. 309 00:15:32,120 --> 00:15:32,560 Speaker 8: Do with it? 310 00:15:32,680 --> 00:15:36,440 Speaker 9: Correct, So they created this model, realized it was incredibly 311 00:15:36,520 --> 00:15:40,440 Speaker 9: powerful and really really good at finding bugs and flaws 312 00:15:40,640 --> 00:15:43,280 Speaker 9: in a lot of open source software that's out there, 313 00:15:43,520 --> 00:15:47,040 Speaker 9: things that's been around for decades, and that companies. You know, 314 00:15:47,080 --> 00:15:49,320 Speaker 9: all the apps and services we use are dependent on 315 00:15:50,080 --> 00:15:53,840 Speaker 9: and they realized that not only could their model find 316 00:15:53,880 --> 00:15:57,240 Speaker 9: these bugs, it could automatically exploit them. And so their 317 00:15:57,280 --> 00:15:59,960 Speaker 9: concern was if we let that out into the general public, 318 00:16:00,240 --> 00:16:02,480 Speaker 9: that could be abused in some way. So we'll do 319 00:16:02,600 --> 00:16:06,080 Speaker 9: a kind of limited release, getting the people who might 320 00:16:06,120 --> 00:16:09,040 Speaker 9: be most affected by it, and will allow them to 321 00:16:09,080 --> 00:16:12,280 Speaker 9: play around, find some of those bugs, fix them, and 322 00:16:12,320 --> 00:16:14,720 Speaker 9: then feedback, and then they'll figure out from their what 323 00:16:14,720 --> 00:16:16,320 Speaker 9: they're going to do about a general release. 324 00:16:16,680 --> 00:16:17,520 Speaker 4: I mean, it's interesting. 325 00:16:17,560 --> 00:16:21,680 Speaker 3: There's cyber companies like Palata Networks. There's also potential competitors 326 00:16:22,080 --> 00:16:25,640 Speaker 3: to Anthropics own ai that are using it. What's interesting 327 00:16:25,680 --> 00:16:27,920 Speaker 3: is a competitor open Ai had already as well sort 328 00:16:27,960 --> 00:16:31,280 Speaker 3: of piloted the idea of getting its technology into the 329 00:16:31,320 --> 00:16:33,920 Speaker 3: hands and defenders before the main public. How much is 330 00:16:33,920 --> 00:16:35,960 Speaker 3: there something written large that everyone has to think about. 331 00:16:36,840 --> 00:16:37,760 Speaker 4: Yeah, you're right, it's. 332 00:16:37,680 --> 00:16:41,240 Speaker 9: A really unusual agreement for competitors to be coming together, 333 00:16:41,560 --> 00:16:45,160 Speaker 9: for someone to be sharing their IP in this way, 334 00:16:45,760 --> 00:16:47,800 Speaker 9: you know, and not you know, claiming not to be. 335 00:16:47,720 --> 00:16:48,520 Speaker 4: Profiting off it. 336 00:16:48,800 --> 00:16:53,400 Speaker 9: But it shows this wider, real existential threat not only 337 00:16:53,400 --> 00:16:56,760 Speaker 9: to the cybersecurity industry, but to software at large. And 338 00:16:56,960 --> 00:16:59,360 Speaker 9: I have just been hearing so much from my sources 339 00:16:59,400 --> 00:17:03,440 Speaker 9: about how can concerned they are about this incoming potential 340 00:17:03,560 --> 00:17:07,879 Speaker 9: vulnerability apocalypse. They're describing it as which could really impact 341 00:17:08,200 --> 00:17:10,320 Speaker 9: so many of the things that we know in our 342 00:17:10,320 --> 00:17:11,520 Speaker 9: normal day to day use. 343 00:17:12,720 --> 00:17:13,119 Speaker 4: And it. 344 00:17:14,600 --> 00:17:17,959 Speaker 9: Shows a kind of coming together of cyber defenders and 345 00:17:18,040 --> 00:17:21,280 Speaker 9: trying to do the right thing in this new age of. 346 00:17:21,240 --> 00:17:24,520 Speaker 3: AI and Probert continuing to work with the government, even 347 00:17:24,560 --> 00:17:27,280 Speaker 3: though of course they're in a legal battle with. 348 00:17:27,280 --> 00:17:28,080 Speaker 4: Part of it at the moment. 349 00:17:28,080 --> 00:17:31,080 Speaker 3: Bloomberg's Margaret Murphy, thanks so much for the reporting. Let's 350 00:17:31,119 --> 00:17:34,080 Speaker 3: stick with AI with cyber threats and bring in Teresa Payton. 351 00:17:34,160 --> 00:17:38,440 Speaker 3: She's the CEO of FORTELLSS Solutions, which brings strategic cybersecurity 352 00:17:38,480 --> 00:17:40,880 Speaker 3: services to leaders navigating high stakes environments. 353 00:17:41,040 --> 00:17:42,720 Speaker 4: You're also with the White House Chief. 354 00:17:42,520 --> 00:17:45,640 Speaker 3: Information Officer during the George Bush administration, and Teresa, I mean, 355 00:17:45,840 --> 00:17:48,760 Speaker 3: an apocalypse is what we just heard from, Margie. 356 00:17:48,800 --> 00:17:49,800 Speaker 4: Is that what you're thinking. 357 00:17:50,880 --> 00:17:54,880 Speaker 8: I mean, I think it's interesting. So it could be 358 00:17:54,920 --> 00:17:59,520 Speaker 8: from the standpoint of isn't it interesting that Anthropic self 359 00:17:59,560 --> 00:18:03,679 Speaker 8: report that they found thousands of vulnerabilities that could be 360 00:18:03,680 --> 00:18:06,680 Speaker 8: fixed very easily. In sort of the early release of 361 00:18:06,720 --> 00:18:10,120 Speaker 8: this tool, that means that it was outpacing the ability 362 00:18:10,200 --> 00:18:13,480 Speaker 8: of cybersecurity products all the big names on the market, 363 00:18:13,520 --> 00:18:16,560 Speaker 8: which happen to also be now in this cohort. And 364 00:18:16,640 --> 00:18:20,359 Speaker 8: so it is interesting from an apocalypse standpoint. It depends 365 00:18:20,400 --> 00:18:22,080 Speaker 8: on I think your point of view. Is it an 366 00:18:22,080 --> 00:18:25,119 Speaker 8: apocalypse or do we now finally have a tool that, 367 00:18:25,280 --> 00:18:29,720 Speaker 8: in the hands of cybersecurity practitioners can make sort of 368 00:18:29,760 --> 00:18:34,320 Speaker 8: this limited skill set even more powerful, to be able 369 00:18:34,320 --> 00:18:36,639 Speaker 8: to cover even more than they have in the past. 370 00:18:37,359 --> 00:18:40,720 Speaker 8: I see AI as being an incredible assistant here. The 371 00:18:40,800 --> 00:18:43,160 Speaker 8: one concern is if we don't have the right governance 372 00:18:43,200 --> 00:18:46,800 Speaker 8: and guardrails, we could find that this tool or a 373 00:18:46,840 --> 00:18:49,479 Speaker 8: tool like it could end up in the hands of 374 00:18:49,520 --> 00:18:53,159 Speaker 8: the cyber operatives who have negative and criminal intent. And 375 00:18:53,200 --> 00:18:55,359 Speaker 8: I think that's the concern here. 376 00:18:55,600 --> 00:18:58,199 Speaker 2: That last point is key, and it relates to what 377 00:18:58,200 --> 00:18:59,560 Speaker 2: we've been talking about in the show right the war 378 00:18:59,600 --> 00:19:02,280 Speaker 2: in ira On for many weeks. We've been saying that Iran, 379 00:19:02,359 --> 00:19:05,560 Speaker 2: as an example, has a competence in the domain where 380 00:19:05,840 --> 00:19:09,320 Speaker 2: if you're on the defense or you use a US 381 00:19:09,440 --> 00:19:13,400 Speaker 2: based technology for defense that there's every chance that those 382 00:19:13,440 --> 00:19:17,000 Speaker 2: going on the offense have access to similar technology, although 383 00:19:17,280 --> 00:19:20,000 Speaker 2: as Margie outlined, you know mythos is at the cutting 384 00:19:20,160 --> 00:19:23,399 Speaker 2: edge in terms of models. How do you see the 385 00:19:23,440 --> 00:19:26,320 Speaker 2: field in that conflict offense versus defense. 386 00:19:27,400 --> 00:19:30,080 Speaker 8: Yeah, well, first of all, even though there is thankfully 387 00:19:30,240 --> 00:19:34,080 Speaker 8: a ceasefire has been announced, that does not mean that 388 00:19:34,640 --> 00:19:38,040 Speaker 8: the Irani and cyber operatives have taken their fingers off keyboards. 389 00:19:38,080 --> 00:19:41,359 Speaker 8: So we need to continue to have a heightened sense 390 00:19:41,440 --> 00:19:45,800 Speaker 8: of vigilance and alertness around these issues. But what I 391 00:19:45,840 --> 00:19:48,840 Speaker 8: will say is, you know that they are obviously leveraging 392 00:19:48,960 --> 00:19:51,840 Speaker 8: AI and different types of tools to be able to 393 00:19:51,840 --> 00:19:54,919 Speaker 8: carry off attacks like they did against Striker and some 394 00:19:55,000 --> 00:19:58,000 Speaker 8: of the other attacks that are ongoing against critical infrastructure 395 00:19:58,720 --> 00:20:01,600 Speaker 8: in the United States. Bulletins have come out. If people 396 00:20:01,600 --> 00:20:03,600 Speaker 8: are listening to the show right now and you're thinking, 397 00:20:04,000 --> 00:20:06,040 Speaker 8: I don't feel like I'm in the know in this information, 398 00:20:06,280 --> 00:20:08,600 Speaker 8: I certainly can't give it to you publicly, but you 399 00:20:08,680 --> 00:20:11,440 Speaker 8: can contact your local FBI office and they will tell 400 00:20:11,440 --> 00:20:13,080 Speaker 8: you all the things that you need to look for 401 00:20:13,480 --> 00:20:15,840 Speaker 8: as it relates to Iran and some of the cyber 402 00:20:15,880 --> 00:20:20,000 Speaker 8: attacks on critical infrastructure. What I would say is you know, 403 00:20:20,080 --> 00:20:23,160 Speaker 8: I would love to hear more from Meanthropic once they 404 00:20:23,200 --> 00:20:26,119 Speaker 8: have this cohort, do sort of an alpha pilot, I 405 00:20:26,119 --> 00:20:28,199 Speaker 8: guess is the best way to put it. What is 406 00:20:28,240 --> 00:20:30,240 Speaker 8: the beta pilot and what is a you know, sort 407 00:20:30,240 --> 00:20:33,800 Speaker 8: of a rollout to trusted vetted you know, smaller maybe 408 00:20:33,800 --> 00:20:36,639 Speaker 8: privately held firms look like to make sure that all 409 00:20:36,760 --> 00:20:40,520 Speaker 8: offensive and defensive teams have these capabilities and are able 410 00:20:40,560 --> 00:20:44,240 Speaker 8: to use them against the attackers. Time is of the essence. 411 00:20:44,280 --> 00:20:45,480 Speaker 4: We're in a race, and we. 412 00:20:45,440 --> 00:20:47,600 Speaker 8: Want to make sure that everybody who is on the 413 00:20:47,640 --> 00:20:52,800 Speaker 8: ethical side of cybersecurity is actually empowered, engaged and has 414 00:20:52,840 --> 00:20:56,840 Speaker 8: all the best state of the art tooling possible to 415 00:20:56,880 --> 00:20:58,119 Speaker 8: fight back these attacks. 416 00:20:58,440 --> 00:21:00,600 Speaker 3: And Teresa, you make a good point, and it's actually 417 00:21:00,640 --> 00:21:04,520 Speaker 3: we've been hearing Fromanthropic as they've outlined what glass wing 418 00:21:04,760 --> 00:21:07,400 Speaker 3: really looks like. And you can see what certain Newton 419 00:21:07,520 --> 00:21:09,879 Speaker 3: Cheng who's the Anthropics Frontier Red Teams cyber lead, has 420 00:21:09,920 --> 00:21:12,320 Speaker 3: been talking about how this is an industry wide problem, 421 00:21:12,760 --> 00:21:15,480 Speaker 3: you know, for the smaller companies that would like to 422 00:21:15,520 --> 00:21:17,439 Speaker 3: be able to help. There is obviously a chunk of 423 00:21:17,480 --> 00:21:20,400 Speaker 3: change being offered Bianthropic as well. In terms of computing tokens. 424 00:21:20,760 --> 00:21:22,760 Speaker 3: How much have you been working perhaps with open ai 425 00:21:22,920 --> 00:21:26,520 Speaker 3: with its well prototype, it's been letting and working with 426 00:21:26,800 --> 00:21:29,679 Speaker 3: others in the security space to try and ensure that 427 00:21:29,720 --> 00:21:32,080 Speaker 3: defense has its tools ahead of others as well. 428 00:21:33,000 --> 00:21:36,200 Speaker 8: Yeah, I think there's a lot going on in this industry, 429 00:21:36,359 --> 00:21:38,639 Speaker 8: and so I have been talking to all of the 430 00:21:38,680 --> 00:21:41,919 Speaker 8: different sort of leading AI providers to find out what 431 00:21:42,000 --> 00:21:45,520 Speaker 8: are the best offensive defensive tools, what are some things 432 00:21:45,520 --> 00:21:47,760 Speaker 8: that are yet to be built, what is on the roadmap, 433 00:21:47,800 --> 00:21:50,719 Speaker 8: and then what are really the best solutions depending on 434 00:21:50,760 --> 00:21:53,679 Speaker 8: the size company you are, how much money you have. 435 00:21:53,760 --> 00:21:57,760 Speaker 8: Because you're right, everything is token based, how can companies 436 00:21:57,880 --> 00:22:01,000 Speaker 8: learn if they don't have a big technology team, how 437 00:22:01,000 --> 00:22:03,360 Speaker 8: can they learn to optimize the use of the tools 438 00:22:03,800 --> 00:22:07,080 Speaker 8: so that they're spending their token sparingly and get the 439 00:22:07,160 --> 00:22:09,280 Speaker 8: most value out of the money that they're spent. And 440 00:22:09,320 --> 00:22:12,119 Speaker 8: so there's definitely an arms race, you know, a healthy 441 00:22:12,160 --> 00:22:16,479 Speaker 8: competition across the different AI companies, and some of them 442 00:22:16,480 --> 00:22:19,960 Speaker 8: are now cooperating as part of this anthropic rollout. The 443 00:22:20,000 --> 00:22:22,920 Speaker 8: other thing I do want to mention is that answers 444 00:22:23,160 --> 00:22:27,040 Speaker 8: did actually self report that Mythos went rogue and so 445 00:22:27,119 --> 00:22:30,639 Speaker 8: I think that's something also that we need to be discussing, 446 00:22:30,760 --> 00:22:33,600 Speaker 8: is that although it's going to help, it also has 447 00:22:33,640 --> 00:22:35,280 Speaker 8: the fine tunity that needs to happen. 448 00:22:36,080 --> 00:22:38,119 Speaker 2: Teresa, Paide and CEO for the list Solutions. We'll have 449 00:22:38,119 --> 00:22:39,800 Speaker 2: you back on the program and we'll discuss it more 450 00:22:39,840 --> 00:22:42,640 Speaker 2: detail coming up. Apple is on track to launch its 451 00:22:42,640 --> 00:22:46,639 Speaker 2: foldable phone later this year. Right now, markets off session 452 00:22:46,680 --> 00:22:57,360 Speaker 2: highs but rallying. This Bloomberg Tech Welcome back to Bloomberg Tech. 453 00:22:57,400 --> 00:23:01,240 Speaker 2: Technology socks are rallying. There is outside performance, particularly in 454 00:23:01,320 --> 00:23:04,560 Speaker 2: chip names in memory and the chip name's extent to 455 00:23:04,640 --> 00:23:05,520 Speaker 2: logic as well. 456 00:23:05,800 --> 00:23:06,439 Speaker 4: The other part of. 457 00:23:06,440 --> 00:23:09,800 Speaker 2: The story is that oil is down Brent the global 458 00:23:09,840 --> 00:23:13,600 Speaker 2: benchmark below one hundred dollars per barrel. The story is 459 00:23:13,600 --> 00:23:15,560 Speaker 2: that we are in a two week ceasefire in the 460 00:23:15,560 --> 00:23:19,359 Speaker 2: war in Iran, broken by Pakistan. There is still ongoing 461 00:23:19,400 --> 00:23:23,960 Speaker 2: headlines about activity in the region and supply chain disruption, 462 00:23:24,359 --> 00:23:26,560 Speaker 2: but that is what the main driver and main story 463 00:23:26,640 --> 00:23:29,080 Speaker 2: is at this moment. We're off session highs when it 464 00:23:29,119 --> 00:23:32,879 Speaker 2: comes to the Nazak one hundred, but actually the Philadelphia 465 00:23:32,880 --> 00:23:36,479 Speaker 2: Semiconductor in edental socks, those chip stocks continue to rally 466 00:23:36,640 --> 00:23:38,399 Speaker 2: pretty hardcater I certainly do. 467 00:23:38,880 --> 00:23:41,200 Speaker 3: And let's focus in on kind of area of supply 468 00:23:41,359 --> 00:23:44,480 Speaker 3: chain right now, because Apple's first foldable phone, it's still 469 00:23:44,520 --> 00:23:46,919 Speaker 3: on track to launched in September now, despite some reports 470 00:23:46,920 --> 00:23:48,080 Speaker 3: of major manufacturing delays. 471 00:23:48,119 --> 00:23:50,920 Speaker 4: Bluemogs Mark German is here with the latest reporting. Yesterday 472 00:23:50,960 --> 00:23:51,800 Speaker 4: Apple was lower. 473 00:23:52,119 --> 00:23:54,080 Speaker 3: Today of course we rise with the rest of the market, 474 00:23:54,200 --> 00:23:56,760 Speaker 3: but there is an anticipation here Mark that you'll make 475 00:23:56,800 --> 00:23:59,480 Speaker 3: it clear they're still going to have that foldable phone 476 00:23:59,520 --> 00:24:00,000 Speaker 3: on time. 477 00:24:01,240 --> 00:24:01,480 Speaker 1: Yeah. 478 00:24:01,480 --> 00:24:03,280 Speaker 10: So a couple of days ago, the Nickay out of 479 00:24:03,320 --> 00:24:05,760 Speaker 10: Asia put out a report saying that there are major 480 00:24:05,920 --> 00:24:10,960 Speaker 10: engineering and manufacturing snags related to the company's first foldable iPhone, 481 00:24:11,000 --> 00:24:15,280 Speaker 10: which could mean a several month delay. What I'm told 482 00:24:15,480 --> 00:24:18,560 Speaker 10: is that the foldable iPhone will be introduced in September 483 00:24:19,160 --> 00:24:22,280 Speaker 10: during the normal cadence, alongside the iPhone eighteen Pro and 484 00:24:22,359 --> 00:24:26,880 Speaker 10: eighteen promacs, And as of this week, there's a very 485 00:24:26,880 --> 00:24:29,879 Speaker 10: strong likelihood that the foldable phone will actually even go 486 00:24:29,960 --> 00:24:34,080 Speaker 10: on sale at the same time as the new eighteen 487 00:24:34,119 --> 00:24:38,360 Speaker 10: Pro models, which means no shipment delayed. Now things are fluid. 488 00:24:38,520 --> 00:24:41,840 Speaker 10: Mass production has not begun yet, but everything that I 489 00:24:41,880 --> 00:24:45,080 Speaker 10: have been told by several people close to the situation 490 00:24:45,280 --> 00:24:48,080 Speaker 10: is that there are no major engineering snags at this point. 491 00:24:48,520 --> 00:24:51,879 Speaker 10: Things are moving as planned. Again, these things can change, 492 00:24:52,200 --> 00:24:55,320 Speaker 10: but as of now, I do anticipate the foldable phone 493 00:24:55,960 --> 00:24:58,680 Speaker 10: to go on sale either at the same time as 494 00:24:58,720 --> 00:25:02,240 Speaker 10: the non new foldable madel or very soon thereafter. 495 00:25:02,960 --> 00:25:04,760 Speaker 2: The other way I'm looking at it, Mark is that 496 00:25:05,040 --> 00:25:07,520 Speaker 2: come September, we're going to get a foldable phone from Apple. 497 00:25:07,760 --> 00:25:10,200 Speaker 2: You know, what do we need to know about? Nice speeches? 498 00:25:11,200 --> 00:25:14,320 Speaker 2: If it's super exciting, if it's on track, how big 499 00:25:14,400 --> 00:25:16,840 Speaker 2: is product? Is this going to be in Apple's evolution? 500 00:25:18,040 --> 00:25:20,320 Speaker 10: Well, I think initially it's going to be quite niche 501 00:25:20,640 --> 00:25:23,560 Speaker 10: I anticipate this to be, in the words of one person, 502 00:25:24,640 --> 00:25:29,240 Speaker 10: very very expensive. Apple believes it has solved some of 503 00:25:29,280 --> 00:25:32,520 Speaker 10: the quirks of the current foldable models. A few things 504 00:25:32,520 --> 00:25:35,159 Speaker 10: I'll point out. One, some of the current foldables from 505 00:25:35,200 --> 00:25:38,640 Speaker 10: other players, they have lots of durability issues. You get 506 00:25:38,720 --> 00:25:41,840 Speaker 10: dust or stand in it, you could break the screen. Obviously, 507 00:25:41,880 --> 00:25:46,320 Speaker 10: Apple has very high specifications for durability, and so the 508 00:25:46,359 --> 00:25:50,920 Speaker 10: foldable phone meets that, I'm told. The other quirk related 509 00:25:50,920 --> 00:25:53,800 Speaker 10: to displays, there are creases in many of the foldable 510 00:25:53,800 --> 00:25:56,880 Speaker 10: phones today when you open and shut it. Apple has 511 00:25:56,960 --> 00:26:02,440 Speaker 10: reduced that crease working with Samsung some new display technology Samsung, 512 00:26:02,440 --> 00:26:05,159 Speaker 10: of course, is going to bring to its devices. The 513 00:26:05,280 --> 00:26:07,640 Speaker 10: third thing, if you look at the phone, it looks 514 00:26:07,680 --> 00:26:11,080 Speaker 10: like a small little reporter notebook. You open it up 515 00:26:11,160 --> 00:26:14,320 Speaker 10: and it goes wider than the narrower foldable phones you 516 00:26:14,359 --> 00:26:17,000 Speaker 10: see on the market today, which means it's an excellent 517 00:26:17,040 --> 00:26:21,280 Speaker 10: device for watching video and for gaming, but overall it's 518 00:26:21,600 --> 00:26:24,159 Speaker 10: a foldable phone. Samsung has had these on the market 519 00:26:24,280 --> 00:26:27,440 Speaker 10: since around twenty nineteen twenty twenty. All the Chinese phone 520 00:26:27,480 --> 00:26:30,199 Speaker 10: makers have them, Google has it, But now that Apple's 521 00:26:30,320 --> 00:26:32,640 Speaker 10: entering the market, it's going to get a lot more popular. 522 00:26:33,960 --> 00:26:37,720 Speaker 2: Bloomberg, Smart German, thank you very much. What was driving 523 00:26:37,720 --> 00:26:41,840 Speaker 2: Apples stock in yesterday's session was that concern manufacturing engineering 524 00:26:41,880 --> 00:26:44,800 Speaker 2: delays and how it impacted a foldable phone. The stock 525 00:26:44,840 --> 00:26:47,679 Speaker 2: today's rallying for a very different reason, right, and that 526 00:26:47,840 --> 00:26:51,960 Speaker 2: is the cease far with Iran. As geopolitical tensions rise. 527 00:26:52,160 --> 00:26:54,800 Speaker 2: In the context of that ongoing war in Iran, global 528 00:26:54,840 --> 00:26:57,800 Speaker 2: tech supply chains and as well as US China relations 529 00:26:57,840 --> 00:27:01,160 Speaker 2: are being tested. As bringing Reva Goose on, director at 530 00:27:01,160 --> 00:27:05,080 Speaker 2: the policy and economic research firm Rhodium Group. You know, 531 00:27:05,440 --> 00:27:07,639 Speaker 2: I'm looking again. I'm going to go back to what's rallying. 532 00:27:07,640 --> 00:27:10,359 Speaker 2: We're showing it on the screen. It's technology stocks that 533 00:27:10,440 --> 00:27:14,399 Speaker 2: are pushing higher right now. And one of the reasons is, Okay, 534 00:27:15,119 --> 00:27:20,760 Speaker 2: helium comes from the Gulf, critical in stabilization, cooling in 535 00:27:21,160 --> 00:27:25,280 Speaker 2: the etching and deposition process and chip manufacturing. Energy prices 536 00:27:25,320 --> 00:27:28,040 Speaker 2: have been impacted by this war in Iran. Why is 537 00:27:28,080 --> 00:27:30,400 Speaker 2: it to your mind that the tech industry is seeing 538 00:27:30,440 --> 00:27:32,879 Speaker 2: some relief from the idea of a ceasefire. 539 00:27:34,240 --> 00:27:37,399 Speaker 11: Well, because we're not going to all out war is 540 00:27:37,480 --> 00:27:41,040 Speaker 11: the short answer. And so I think everyone is taking 541 00:27:41,080 --> 00:27:43,240 Speaker 11: this bit of good news that there's a two week 542 00:27:43,320 --> 00:27:46,760 Speaker 11: seeds fire and running with it. But the reality is 543 00:27:46,800 --> 00:27:51,240 Speaker 11: that the Pandora's box and the Persian Gulf has been opened. 544 00:27:51,520 --> 00:27:54,439 Speaker 11: And by that I mean, you know, even if we 545 00:27:54,560 --> 00:27:58,600 Speaker 11: have an extended seasfire and cessation of hostilities that last 546 00:27:58,840 --> 00:28:04,119 Speaker 11: longer term. Here, we have now seen Iran demonstrate capability 547 00:28:04,160 --> 00:28:09,000 Speaker 11: and political will to strike critical infrastructure across this region, 548 00:28:09,320 --> 00:28:13,959 Speaker 11: to include very high value targets like data centers multi 549 00:28:13,960 --> 00:28:18,880 Speaker 11: billion data centers, including explicit threats against Opening Eyes Stargate 550 00:28:18,920 --> 00:28:21,800 Speaker 11: project in the UAE, not to mention the number of 551 00:28:21,840 --> 00:28:26,960 Speaker 11: AWS sites that have been struck already in Bahrain and UAE. 552 00:28:27,119 --> 00:28:31,320 Speaker 11: So that's just a completely different investor climate overall for 553 00:28:31,440 --> 00:28:34,520 Speaker 11: this region. But coming to supply chains though, as you said, 554 00:28:34,760 --> 00:28:37,000 Speaker 11: when we talk about helium supply, bro mind just energy 555 00:28:37,000 --> 00:28:42,080 Speaker 11: supply overall, Iran also has an interest in allowing transit 556 00:28:42,120 --> 00:28:44,240 Speaker 11: through the strait. But that's where we come back to 557 00:28:44,920 --> 00:28:48,120 Speaker 11: political conditions that will be attached to the ceasefire. 558 00:28:49,040 --> 00:28:53,120 Speaker 2: Caroline and I discussed in detail the vulnerability of data 559 00:28:53,120 --> 00:28:56,560 Speaker 2: center assets in the Gulf as it relates to China. 560 00:28:56,680 --> 00:29:01,000 Speaker 2: For example, there is a long term threat to Taiwan 561 00:29:01,520 --> 00:29:06,560 Speaker 2: and the concentration of semiconductor manufacturing. With that taken all 562 00:29:06,880 --> 00:29:11,720 Speaker 2: into account, how conscious are companies like Apple on now 563 00:29:11,840 --> 00:29:15,640 Speaker 2: changing where they plan to build future products and have 564 00:29:15,720 --> 00:29:18,800 Speaker 2: a footprint in the future for their manufacturing operations. 565 00:29:19,760 --> 00:29:22,360 Speaker 11: Well, there are multiple drivers in play here, not just 566 00:29:22,600 --> 00:29:27,080 Speaker 11: geopolitical risk obviously, but also the evolution of US trade strategy, 567 00:29:27,560 --> 00:29:30,120 Speaker 11: which is still in play as we know. There's a 568 00:29:30,280 --> 00:29:34,120 Speaker 11: pending Section two thirty two still on semiconductors that could 569 00:29:34,200 --> 00:29:38,880 Speaker 11: contain component tariffs right that implicate the devices that contain 570 00:29:39,160 --> 00:29:44,040 Speaker 11: those semiconductors, not to mention looking beyond tariffs content requirements 571 00:29:44,160 --> 00:29:48,040 Speaker 11: right so far, for example, we have regional value content 572 00:29:48,200 --> 00:29:53,600 Speaker 11: requirements for automotive parts. Imagine that applied to electronics more broadly. 573 00:29:53,960 --> 00:29:57,120 Speaker 11: So I think you know companies like Apple are looking 574 00:29:57,200 --> 00:30:01,440 Speaker 11: at these these issues in parallel to say, yes, there 575 00:30:01,520 --> 00:30:07,080 Speaker 11: is a lot concentrated still in Taiwan device manufacturers also 576 00:30:07,320 --> 00:30:12,840 Speaker 11: in China, that geopolitical risk is is only amplified. But 577 00:30:13,200 --> 00:30:15,360 Speaker 11: at the same time there are going to be a 578 00:30:15,400 --> 00:30:17,800 Speaker 11: lot of disruptors coming from the US who tried to 579 00:30:18,360 --> 00:30:22,200 Speaker 11: drive more not just onshoing, but more regionalized trade and 580 00:30:22,280 --> 00:30:26,080 Speaker 11: bring more of that manufacturing based to the Western hemisphere 581 00:30:26,080 --> 00:30:26,440 Speaker 11: as well. 582 00:30:27,280 --> 00:30:29,040 Speaker 3: I want to just go a little bit further with 583 00:30:29,080 --> 00:30:32,360 Speaker 3: where it took us in terms of China, because there's 584 00:30:32,400 --> 00:30:34,960 Speaker 3: a lot of really well read stories on the Bloomberg 585 00:30:34,960 --> 00:30:38,640 Speaker 3: today thinking about how maybe the US has reduced its 586 00:30:38,680 --> 00:30:43,520 Speaker 3: overall well power among allies and adversaries at the moment 587 00:30:43,520 --> 00:30:45,640 Speaker 3: because of what's happened in Iran in particular when it 588 00:30:45,680 --> 00:30:48,160 Speaker 3: comes to China and the Iran campaign has been quote 589 00:30:48,200 --> 00:30:51,560 Speaker 3: a serious setback for Trump is what one particular Chinese 590 00:30:51,560 --> 00:30:53,720 Speaker 3: foreign ministry advisor. 591 00:30:53,560 --> 00:30:55,800 Speaker 4: Had told us. What does that mean for. 592 00:30:55,840 --> 00:30:58,720 Speaker 3: US China relations What does that mean about AI access, 593 00:30:58,840 --> 00:31:00,720 Speaker 3: cheer access, the relationship that we have. 594 00:31:01,760 --> 00:31:02,080 Speaker 4: Yeah. 595 00:31:02,120 --> 00:31:06,800 Speaker 11: Absolutely So if you contrast where we are today in 596 00:31:06,840 --> 00:31:12,040 Speaker 11: the wake of this Iran miscalculation on part of the US, 597 00:31:12,280 --> 00:31:15,400 Speaker 11: and then compare it to where we were late last 598 00:31:15,440 --> 00:31:18,760 Speaker 11: year when the Busan truth was struck. Back then, the 599 00:31:18,880 --> 00:31:22,320 Speaker 11: US was caught flat footed by Chinese expert controls on 600 00:31:22,360 --> 00:31:24,600 Speaker 11: critical raw materials, and so the US was in a 601 00:31:24,680 --> 00:31:27,560 Speaker 11: very reactive mode when it went to Busan. Those truce 602 00:31:27,640 --> 00:31:30,920 Speaker 11: terms were set, and I think the whole intent going 603 00:31:30,960 --> 00:31:34,400 Speaker 11: into the next summit in Beijing was to come with 604 00:31:34,440 --> 00:31:38,240 Speaker 11: a more assertive position with China, to say, look, we're 605 00:31:38,240 --> 00:31:42,680 Speaker 11: doing things to reindustrialize the US, protect critical infrastructure. We're 606 00:31:42,720 --> 00:31:46,240 Speaker 11: not explicitly calling out China in these measures. So you 607 00:31:46,280 --> 00:31:48,560 Speaker 11: know what, you don't get a vote on these areas, 608 00:31:48,680 --> 00:31:51,760 Speaker 11: and the US is going to be forging ahead now 609 00:31:51,800 --> 00:31:56,719 Speaker 11: because of the Iran dynamic. That certainly tamps down the confidence, 610 00:31:57,280 --> 00:32:00,360 Speaker 11: right that the ability of the US to assert itself 611 00:32:00,400 --> 00:32:04,160 Speaker 11: in the same way, and Beijing sees that. Beijing certainly 612 00:32:04,720 --> 00:32:08,080 Speaker 11: is in a more confident position going into the summit. 613 00:32:08,240 --> 00:32:13,440 Speaker 11: It has very explicit demands. For example, not just to reduce, 614 00:32:13,640 --> 00:32:16,560 Speaker 11: you know, maintain tariff levels at the Busan Truce level, 615 00:32:16,560 --> 00:32:19,360 Speaker 11: but to reduce them, not just to restrain the US 616 00:32:19,400 --> 00:32:23,080 Speaker 11: technology controls, but to actually roll them back. And this 617 00:32:23,200 --> 00:32:25,440 Speaker 11: is where the US has been doing more country agnostic 618 00:32:25,520 --> 00:32:28,520 Speaker 11: measures as well, trying to say, look, China, we're not 619 00:32:28,600 --> 00:32:31,640 Speaker 11: out to get you, but Beijing's messages I see what 620 00:32:31,680 --> 00:32:34,120 Speaker 11: you're doing here and it's got to stop. And so 621 00:32:34,200 --> 00:32:37,760 Speaker 11: their interest is to make everything a red line, which 622 00:32:37,800 --> 00:32:40,200 Speaker 11: means that we have a number of friction points heading 623 00:32:40,240 --> 00:32:44,480 Speaker 11: into the summit. Not to mention there where the status 624 00:32:44,480 --> 00:32:46,720 Speaker 11: of the Iran issue is is also going to consume 625 00:32:46,720 --> 00:32:48,600 Speaker 11: a lot of oxygen going into that meeting. 626 00:32:48,880 --> 00:32:52,000 Speaker 3: This has such broad implications and you help break down 627 00:32:52,000 --> 00:32:54,320 Speaker 3: many of them today. Riva Boujon, thank you, director of 628 00:32:54,320 --> 00:32:57,560 Speaker 3: the Rodium Group. Now coming up, let's return to all 629 00:32:57,640 --> 00:32:59,720 Speaker 3: things AI because that's old. That's talking about it the 630 00:32:59,760 --> 00:33:01,960 Speaker 3: Human Next conference in San Francisco. We speak with the 631 00:33:02,000 --> 00:33:04,800 Speaker 3: Snowflake CEO Rabaswami as. 632 00:33:04,640 --> 00:33:15,240 Speaker 4: A blom Meg Tech. It's time now for talking tech 633 00:33:15,280 --> 00:33:15,760 Speaker 4: and first. 634 00:33:15,600 --> 00:33:18,760 Speaker 3: S up Resolve AI it's going straight to shareholders with 635 00:33:18,760 --> 00:33:21,720 Speaker 3: a hostile bid for Commerce dot Com. This after negotiations 636 00:33:21,720 --> 00:33:24,040 Speaker 3: with the board our management brote down. According to a 637 00:33:24,120 --> 00:33:27,600 Speaker 3: letter to investors seen by Bloomberg, Resolve is now offering 638 00:33:27,760 --> 00:33:30,040 Speaker 3: one of its own shares for every two Commerce dot 639 00:33:30,080 --> 00:33:32,640 Speaker 3: Com shares and says it's earlier one for one proposal 640 00:33:32,680 --> 00:33:34,920 Speaker 3: that gave to the board is no longer valid. Plus, 641 00:33:34,920 --> 00:33:38,040 Speaker 3: Morgan Stanley has become the first Wall Street bank to 642 00:33:38,120 --> 00:33:40,600 Speaker 3: launch its own bitcoin tracking ETF. Now, the Morgan Stanley 643 00:33:40,640 --> 00:33:44,920 Speaker 3: Bitcoin Trust is debuting under the ticker MSBT Bitcoin, touching 644 00:33:44,960 --> 00:33:46,640 Speaker 3: three week highs today. Of course, because the US and 645 00:33:46,680 --> 00:33:49,280 Speaker 3: Iran tentively agreeing that sees fire and prompting a surge 646 00:33:49,280 --> 00:33:52,760 Speaker 3: and risk assets and super Micro says a committee of 647 00:33:52,760 --> 00:33:55,560 Speaker 3: independent directors hired an outside law firm. 648 00:33:55,400 --> 00:33:57,200 Speaker 4: To investigate the circumstances tied. 649 00:33:57,200 --> 00:34:00,000 Speaker 3: To last month's indictment of two employees and a contract 650 00:34:00,360 --> 00:34:03,280 Speaker 3: a religed server sales to China. In a statement, giper 651 00:34:03,360 --> 00:34:05,640 Speaker 3: Micro said it will share an update what's the inquiry 652 00:34:05,720 --> 00:34:06,280 Speaker 3: is complete. 653 00:34:06,520 --> 00:34:07,959 Speaker 4: It won't comment further until then. 654 00:34:08,239 --> 00:34:12,320 Speaker 2: Aad okay human X is underway here in San Francisco, 655 00:34:12,360 --> 00:34:16,479 Speaker 2: bringing together industry leaders focused on turning AI into real 656 00:34:16,520 --> 00:34:20,040 Speaker 2: world results. Snowflake CEO Tridah Ramaswami is set to speak 657 00:34:20,040 --> 00:34:22,800 Speaker 2: at the conference and says it's not about big bets 658 00:34:23,480 --> 00:34:26,439 Speaker 2: but execution. He joins us, now there's something I want 659 00:34:26,440 --> 00:34:28,680 Speaker 2: to take up with you, Shreda. And you and I 660 00:34:28,719 --> 00:34:32,040 Speaker 2: have discussed this before. You have projects snow work, and 661 00:34:32,080 --> 00:34:34,040 Speaker 2: the idea is that you have this big customer base 662 00:34:34,120 --> 00:34:38,840 Speaker 2: that goes from running agents that give you answers to 663 00:34:39,000 --> 00:34:42,840 Speaker 2: running workloads, And I go back to is an agent 664 00:34:42,880 --> 00:34:45,799 Speaker 2: that only gives answers? Actually an agent? Is it a 665 00:34:45,840 --> 00:34:49,120 Speaker 2: real piece of agentic AI that acts with autonomy? It 666 00:34:49,160 --> 00:34:51,000 Speaker 2: seems like now we're getting to that part. 667 00:34:51,880 --> 00:34:53,479 Speaker 4: Yeah, I mean a lu storm. 668 00:34:54,040 --> 00:34:58,000 Speaker 12: It is a looser. It's all about it's fundamentally ABODA 669 00:34:58,120 --> 00:35:02,120 Speaker 12: models using tools and learning from how they are using it. 670 00:35:02,719 --> 00:35:06,840 Speaker 12: And for example, or first generation of agents with Snowflake intelligence, 671 00:35:06,840 --> 00:35:10,320 Speaker 12: which gave analytic insights into data, and then Cortex code 672 00:35:10,680 --> 00:35:14,200 Speaker 12: made acting on Snowflake a lot easier. Snowork is a 673 00:35:14,239 --> 00:35:17,239 Speaker 12: natural culmination. What it's able to do is take the 674 00:35:17,239 --> 00:35:19,480 Speaker 12: one hundred tabs that you have open on your computer 675 00:35:19,880 --> 00:35:23,879 Speaker 12: and create a single environment that brings together things like that. 676 00:35:24,160 --> 00:35:27,279 Speaker 12: So not only can you get information about something you want, 677 00:35:27,360 --> 00:35:29,560 Speaker 12: how is the customer doing, she can also take an action, 678 00:35:29,640 --> 00:35:32,640 Speaker 12: send them an email, or make a recommendation for them. 679 00:35:33,000 --> 00:35:36,000 Speaker 12: It's that power that's increasing by the day as AI 680 00:35:36,000 --> 00:35:37,520 Speaker 12: models get stronger and stronger. 681 00:35:37,600 --> 00:35:39,799 Speaker 2: The idea, you know, every time you go to an 682 00:35:39,800 --> 00:35:41,480 Speaker 2: event like human X, you're trying to say, what is 683 00:35:41,520 --> 00:35:44,040 Speaker 2: the thing that people are excited about what they're talking about. 684 00:35:44,040 --> 00:35:46,560 Speaker 2: And one of the concepts is you go to bed 685 00:35:46,640 --> 00:35:49,160 Speaker 2: and you leave your computer to do work overnight and 686 00:35:49,200 --> 00:35:52,680 Speaker 2: you wake up to tangible results. Is that the idea 687 00:35:52,760 --> 00:35:54,040 Speaker 2: with this that. 688 00:35:53,960 --> 00:35:56,520 Speaker 12: Can be It's one of the consequences we've always said 689 00:35:56,520 --> 00:35:59,040 Speaker 12: scheduled things like scheduled reports. I'm sure you get something 690 00:35:59,040 --> 00:36:01,239 Speaker 12: in the morning that's a summary of all of the 691 00:36:01,280 --> 00:36:02,239 Speaker 12: things that you should know. 692 00:36:02,960 --> 00:36:04,600 Speaker 4: You can think of agents as very summary that. 693 00:36:04,680 --> 00:36:08,160 Speaker 2: Has taken an action. It hasn't performed a piece of 694 00:36:08,200 --> 00:36:09,719 Speaker 2: work on the microphaf. 695 00:36:09,239 --> 00:36:10,040 Speaker 4: That's right, that's right. 696 00:36:10,080 --> 00:36:13,160 Speaker 12: But with these agents, absolutely you can. I have people 697 00:36:13,160 --> 00:36:15,239 Speaker 12: that are writing code twenty four to seven. They just 698 00:36:15,400 --> 00:36:18,440 Speaker 12: like set the agent on the task of writing new 699 00:36:18,480 --> 00:36:21,400 Speaker 12: code to get some feature done, and they give you 700 00:36:21,480 --> 00:36:23,719 Speaker 12: some instructions, come back check in the morning. I think 701 00:36:23,760 --> 00:36:27,680 Speaker 12: absolutely these agents are getting better and better at taking 702 00:36:27,840 --> 00:36:30,640 Speaker 12: actions on your behalf. But part of the magic is 703 00:36:30,880 --> 00:36:32,480 Speaker 12: how do you do that in a government bay. How 704 00:36:32,520 --> 00:36:35,239 Speaker 12: do you have a control plane that says these kinds 705 00:36:35,239 --> 00:36:39,040 Speaker 12: of actions are fine, but those have large outside world consequence. 706 00:36:39,120 --> 00:36:42,040 Speaker 12: They really need to be subject to human scrutiny. Increasingly, 707 00:36:42,120 --> 00:36:45,880 Speaker 12: that's going to be what is needed to make AI succeed. 708 00:36:45,520 --> 00:36:50,319 Speaker 3: At scale outside world consequences. You are the man to 709 00:36:50,320 --> 00:36:53,000 Speaker 3: speak to you today about well, the few or the 710 00:36:53,040 --> 00:36:56,880 Speaker 3: concern the anxiety of the latest Anthropic model that is 711 00:36:56,920 --> 00:37:00,120 Speaker 3: being stress tested by forty tech companies because it's so 712 00:37:00,160 --> 00:37:02,320 Speaker 3: powerfully I can't get into the hands of US made models. 713 00:37:03,280 --> 00:37:05,239 Speaker 3: What do you make about the step changes that we're 714 00:37:05,239 --> 00:37:07,279 Speaker 3: seeing models right now and how powerful they were coming. 715 00:37:08,640 --> 00:37:12,600 Speaker 12: I think the impact that they're having on software is profound. 716 00:37:12,760 --> 00:37:17,640 Speaker 12: Software is getting easier to create, and this also means 717 00:37:17,640 --> 00:37:21,440 Speaker 12: that people can create value faster. I think Anthropic is 718 00:37:21,440 --> 00:37:24,880 Speaker 12: doing an incredibly responsible thing by taking a model like 719 00:37:24,920 --> 00:37:30,080 Speaker 12: that which has breakthrough capabilities, especially in code generation, and 720 00:37:30,200 --> 00:37:33,000 Speaker 12: giving it to a set of their closest partners to 721 00:37:33,200 --> 00:37:35,719 Speaker 12: ensure that when it comes to security, for example, we 722 00:37:35,719 --> 00:37:38,279 Speaker 12: are able to plug gaps. Security has always been a 723 00:37:38,320 --> 00:37:42,719 Speaker 12: difficult hand done, human led kind of motion. And actually 724 00:37:42,760 --> 00:37:44,879 Speaker 12: think of this as a big advance where we can 725 00:37:44,920 --> 00:37:48,880 Speaker 12: systematically make all of our systems they're so critical for 726 00:37:49,000 --> 00:37:52,000 Speaker 12: our nation, for our own survival a lot more stronger. 727 00:37:52,040 --> 00:37:53,960 Speaker 12: I actually see this as a positive outcome. 728 00:37:54,560 --> 00:37:55,720 Speaker 4: You're at the cutting edge of models. 729 00:37:55,760 --> 00:37:57,440 Speaker 3: Of course, you're the co founder Niva, you came on 730 00:37:57,480 --> 00:37:59,680 Speaker 3: to Snowflake and you work for a long time at Google. 731 00:38:00,080 --> 00:38:01,560 Speaker 4: I'm interested actually in how. 732 00:38:01,440 --> 00:38:04,160 Speaker 3: You're finding the narrative around Snowflake right now. We actually 733 00:38:04,160 --> 00:38:06,440 Speaker 3: had the top of the show Brooke Dane on from 734 00:38:06,480 --> 00:38:09,279 Speaker 3: Golm and Sachs as the management. They love Snowflake and 735 00:38:09,280 --> 00:38:12,040 Speaker 3: in particular they're liking and wanting to see from application 736 00:38:12,160 --> 00:38:16,439 Speaker 3: names and from software companies well, more cost optimization, more 737 00:38:16,480 --> 00:38:18,799 Speaker 3: capital return. Is that something you're having to think about 738 00:38:18,840 --> 00:38:21,359 Speaker 3: when you've got some of your coders working twenty four 739 00:38:21,360 --> 00:38:23,279 Speaker 3: hours because they're able to just plug and play while 740 00:38:23,320 --> 00:38:23,840 Speaker 3: they sleep. 741 00:38:24,360 --> 00:38:27,880 Speaker 4: How is that affecting your costs? Well? 742 00:38:28,480 --> 00:38:30,960 Speaker 12: I think the return that we get from investing in 743 00:38:31,000 --> 00:38:34,680 Speaker 12: AI comes back to us many many manyfold. Used to 744 00:38:34,680 --> 00:38:37,000 Speaker 12: be Last year we had maybe a dozen demos that 745 00:38:37,040 --> 00:38:39,000 Speaker 12: we would show our customers. They were all kind of 746 00:38:39,040 --> 00:38:43,000 Speaker 12: cookie cutter. But today I and my sales team is 747 00:38:43,000 --> 00:38:45,560 Speaker 12: a lot more capable can generate a demo for you 748 00:38:45,680 --> 00:38:47,880 Speaker 12: in ten minutes if I want. But and that is 749 00:38:47,920 --> 00:38:50,520 Speaker 12: a lot more personal, It's much more relevant to your business. 750 00:38:50,719 --> 00:38:53,600 Speaker 12: AI is making a real difference in how we show 751 00:38:53,640 --> 00:38:57,520 Speaker 12: people what is possible with Snowflake, and we are getting 752 00:38:57,560 --> 00:39:00,799 Speaker 12: huge returns in areas like support or even now what 753 00:39:00,840 --> 00:39:02,839 Speaker 12: we call our SR team. There's a team that keeps 754 00:39:02,840 --> 00:39:05,720 Speaker 12: Snowflake the system up. They're using the power of Agent 755 00:39:05,760 --> 00:39:10,000 Speaker 12: DKI to be a lot more effective. SAI has cost, 756 00:39:10,080 --> 00:39:12,680 Speaker 12: but the returns we get from it. Problems that used 757 00:39:12,719 --> 00:39:15,000 Speaker 12: to take multiple days to debug and get to the 758 00:39:15,040 --> 00:39:17,719 Speaker 12: bottom er we can get answers now in ten to 759 00:39:17,760 --> 00:39:20,840 Speaker 12: fifteen minutes. There is cost, but I think the returns 760 00:39:20,920 --> 00:39:23,920 Speaker 12: are more than worth it, definitely for us and for 761 00:39:24,000 --> 00:39:25,920 Speaker 12: a lot of our customers that are able to do 762 00:39:26,000 --> 00:39:29,680 Speaker 12: things that honestly were just inconceivable. Just like a year 763 00:39:29,719 --> 00:39:33,960 Speaker 12: ago where companies like United Drenttals basically just released an 764 00:39:34,000 --> 00:39:36,960 Speaker 12: app to their sixteen hundred branches, and what are questions 765 00:39:37,000 --> 00:39:39,680 Speaker 12: these branches have about the current state of the business. 766 00:39:39,960 --> 00:39:42,520 Speaker 12: They can just get it interactively. That is game changing 767 00:39:42,520 --> 00:39:42,799 Speaker 12: for that. 768 00:39:42,800 --> 00:39:46,080 Speaker 2: That's Bloomberg Intelligence are in house research arms thesis that 769 00:39:46,239 --> 00:39:49,200 Speaker 2: as company use of agents grows, you will benefit just 770 00:39:49,320 --> 00:39:54,200 Speaker 2: very quickly. Where you seeing that written true, The most. 771 00:39:54,040 --> 00:39:57,280 Speaker 12: Beautiful part right now is across the data life cycle. 772 00:39:57,840 --> 00:40:01,360 Speaker 12: It's increasing migrations because getting data into a place like 773 00:40:01,400 --> 00:40:04,600 Speaker 12: Snowflake where the data is AI ready has an added 774 00:40:04,640 --> 00:40:08,280 Speaker 12: impetus Because of AI, it's also automating the process of migration. 775 00:40:08,520 --> 00:40:10,800 Speaker 12: So people are bringing in more data sets, they're building 776 00:40:10,800 --> 00:40:13,799 Speaker 12: more with products like Snowflake Intelligence because they can get 777 00:40:13,880 --> 00:40:18,520 Speaker 12: value faster, distributed more broadly. Just here was I wanted 778 00:40:18,520 --> 00:40:20,839 Speaker 12: to look up what exactly is United Rentals doing in 779 00:40:20,960 --> 00:40:23,960 Speaker 12: other areas. That was just a single question into my 780 00:40:24,000 --> 00:40:27,760 Speaker 12: Snowflake Intelligence app. That's right on my phone. That's value 781 00:40:27,840 --> 00:40:30,560 Speaker 12: when I want it, where I want it. That's the 782 00:40:30,680 --> 00:40:33,120 Speaker 12: kind of thing that every CEO that I show this 783 00:40:33,239 --> 00:40:36,480 Speaker 12: too wants because they want those insights to be available 784 00:40:36,560 --> 00:40:40,640 Speaker 12: at their fingertips. So it's actually a broad acceleration across 785 00:40:40,840 --> 00:40:43,759 Speaker 12: many many things that we do. And as models like 786 00:40:43,840 --> 00:40:46,800 Speaker 12: mythos come along, we are able to take advantage of 787 00:40:46,840 --> 00:40:49,360 Speaker 12: them in our products and get our customers to do 788 00:40:49,520 --> 00:40:50,799 Speaker 12: even more with Snowflake. 789 00:40:51,120 --> 00:40:54,560 Speaker 3: Snowflake CEO should I Ramaswami. Ahead of your talkover at 790 00:40:54,600 --> 00:40:57,080 Speaker 3: human X, we so appreciate you joining the show. Now 791 00:40:57,080 --> 00:40:59,520 Speaker 3: coming up, social media companies. 792 00:40:59,320 --> 00:41:02,719 Speaker 4: Trying to hand and a rebrand and a work. There's 793 00:41:02,760 --> 00:41:03,640 Speaker 4: a bluebog tech. 794 00:41:07,719 --> 00:41:10,520 Speaker 2: TikTok plans to invest one point two billion dollars in 795 00:41:10,520 --> 00:41:13,000 Speaker 2: a second data center in Finland. It's part of the 796 00:41:13,040 --> 00:41:16,760 Speaker 2: company's Project Clover effort to move data storage for European 797 00:41:16,840 --> 00:41:19,680 Speaker 2: users to the continent, the company says. Earlier this month, 798 00:41:19,880 --> 00:41:24,200 Speaker 2: TikTok abandoned plans for a second data center in Ireland Cara. 799 00:41:24,920 --> 00:41:29,800 Speaker 3: Meanwhile, there's more on big tech's new spin in social media, 800 00:41:30,000 --> 00:41:33,400 Speaker 3: the new Big Tobacco. Tech giants like Meta, TikTok, YouTube 801 00:41:33,480 --> 00:41:36,600 Speaker 3: are trying to redefine themselves to escape the social stigma. 802 00:41:36,880 --> 00:41:39,160 Speaker 3: Merg's Alex Lavine is here to talk about your story 803 00:41:39,200 --> 00:41:42,160 Speaker 3: your tech in depth today. That really builds upon a 804 00:41:42,200 --> 00:41:45,080 Speaker 3: narrative that maybe was started by Snapchat almost a decade ago, 805 00:41:45,120 --> 00:41:46,960 Speaker 3: but basically a lot of companies trying to tell us 806 00:41:47,000 --> 00:41:49,200 Speaker 3: that they're not actually social media companies at all. 807 00:41:49,920 --> 00:41:52,759 Speaker 13: It's striking because fifteen or twenty years ago social media 808 00:41:52,880 --> 00:41:54,760 Speaker 13: was a sexy thing to be the way that today 809 00:41:54,800 --> 00:41:57,520 Speaker 13: now everyone is an AI need avis or an AI 810 00:41:57,640 --> 00:42:01,680 Speaker 13: power that as sentiment has and we've entered this moment 811 00:42:01,760 --> 00:42:05,719 Speaker 13: where it's common vernacular we talk about social media as 812 00:42:05,760 --> 00:42:08,520 Speaker 13: brain rod and doom scrolling, and we say going analog 813 00:42:08,680 --> 00:42:11,480 Speaker 13: is very in vogue. The sentiment has soured and now 814 00:42:11,480 --> 00:42:13,720 Speaker 13: the social media companies want to be known for something 815 00:42:13,760 --> 00:42:16,480 Speaker 13: other than either what they started as or in some 816 00:42:16,520 --> 00:42:18,480 Speaker 13: cases we could argue what they actually are. 817 00:42:19,480 --> 00:42:22,920 Speaker 2: Alex reinventions possible are remember twenty twenty one being at 818 00:42:23,000 --> 00:42:25,279 Speaker 2: number one, hack away and the top is pulled off 819 00:42:25,320 --> 00:42:29,640 Speaker 2: and Facebook is then meta. There's some case studies you've 820 00:42:29,640 --> 00:42:31,000 Speaker 2: looked at and your techond depth. 821 00:42:31,880 --> 00:42:34,480 Speaker 13: Yeah, so, I mean we looked at how meta, as 822 00:42:34,520 --> 00:42:37,440 Speaker 13: you said, was trying to rebrand to a metaverse company, 823 00:42:37,440 --> 00:42:39,280 Speaker 13: and since then we've seen them get into chip making, 824 00:42:39,680 --> 00:42:43,200 Speaker 13: into AI, into consumer tech. YouTube very much wants to 825 00:42:43,200 --> 00:42:45,800 Speaker 13: be known just as TV or as really an entertainment 826 00:42:45,800 --> 00:42:48,560 Speaker 13: and streaming company. TikTok similarly also wants to be known 827 00:42:48,600 --> 00:42:51,040 Speaker 13: as an entertainment company. Snap has said in the pass 828 00:42:51,080 --> 00:42:53,919 Speaker 13: it it's a camera company. I think all these are. 829 00:42:54,280 --> 00:42:56,800 Speaker 13: It's understandable why the companies want to move in this direction, 830 00:42:56,840 --> 00:42:59,800 Speaker 13: but the corporate spin is just not seemingly working. 831 00:43:00,840 --> 00:43:03,239 Speaker 2: Bloomberg's Alex Lavine with a must red Tech in debt, 832 00:43:03,280 --> 00:43:05,879 Speaker 2: thank you very much. That does it for this edition 833 00:43:05,920 --> 00:43:07,799 Speaker 2: of Bloomberg Tech. But this is the story. We have 834 00:43:07,800 --> 00:43:10,280 Speaker 2: a two week cease fire with the war in Iran 835 00:43:10,480 --> 00:43:17,560 Speaker 2: broken by Pakistan. Tech stocks pushing higher, particularly chip stocks, AI, infrastructure, hardware, memory, etc. 836 00:43:18,360 --> 00:43:22,160 Speaker 2: Bitcoin also pushing to the upside. Carro, but we still 837 00:43:22,160 --> 00:43:23,880 Speaker 2: see continued activity in the region. 838 00:43:24,280 --> 00:43:25,680 Speaker 4: Yeah, we do kuwait sair defenses. 839 00:43:25,719 --> 00:43:28,120 Speaker 3: For example, I've been dealing with intense attacks from Iran 840 00:43:28,160 --> 00:43:31,720 Speaker 3: since eight am Wednesday and it's still ongoing, the country's 841 00:43:31,760 --> 00:43:33,759 Speaker 3: army says in a post on x So we keep 842 00:43:33,760 --> 00:43:36,480 Speaker 3: abreast on the latest. When it comes to geopolitics and 843 00:43:36,560 --> 00:43:38,759 Speaker 3: the effect on tech and supply chains, you have to 844 00:43:38,760 --> 00:43:40,080 Speaker 3: bring it back to that to a certain extent. I 845 00:43:40,120 --> 00:43:42,400 Speaker 3: forget to check out our podcast finding on the terminal 846 00:43:42,400 --> 00:43:44,799 Speaker 3: as well as online on Apple, Spotify, an iHeart from 847 00:43:44,840 --> 00:43:45,840 Speaker 3: New York and San Francisco. 848 00:43:46,200 --> 00:43:47,239 Speaker 4: This is Bloomberg Tech,