1 00:00:00,040 --> 00:00:13,480 Speaker 1: Bloomberg Audio Studios, podcasts, radio news. Bloomberg Tech is alive 2 00:00:13,520 --> 00:00:17,319 Speaker 1: from coast to coast with Caroline Hide in New York 3 00:00:17,600 --> 00:00:20,080 Speaker 1: and Eva Low in sentences. 4 00:00:19,520 --> 00:00:25,720 Speaker 2: Go this is Bloomberg Tech coming up. AMD signs a 5 00:00:25,840 --> 00:00:28,720 Speaker 2: historic deal with open Ai to roll out AI infrastructure, 6 00:00:28,760 --> 00:00:31,880 Speaker 2: a packed the chipmakers said could generate tens of billions 7 00:00:31,880 --> 00:00:33,320 Speaker 2: in dollars of new revenue. 8 00:00:33,440 --> 00:00:33,720 Speaker 3: Plus. 9 00:00:33,760 --> 00:00:35,680 Speaker 4: We speak with the White House AI and Crypto Zar 10 00:00:35,800 --> 00:00:38,280 Speaker 4: David Sachs about relations between the US and China in 11 00:00:38,320 --> 00:00:40,000 Speaker 4: the context of the AI race. 12 00:00:40,680 --> 00:00:43,760 Speaker 2: And metas cmo joins us to discuss how AI is 13 00:00:43,840 --> 00:00:48,000 Speaker 2: changing the advertising landscape what that means for small businesses. 14 00:00:48,080 --> 00:00:50,960 Speaker 4: Here and now we turn our attention to the broader markets. 15 00:00:50,960 --> 00:00:52,760 Speaker 3: Bitcoin at a record high. Then, now's that one. 16 00:00:52,720 --> 00:00:54,760 Speaker 4: Hundred at a record high. We're up six tenths of 17 00:00:54,800 --> 00:00:58,160 Speaker 4: a percent, the moon music around AIAI infrastructure, and the 18 00:00:58,200 --> 00:01:01,920 Speaker 4: frenzy therein ed it continues. I'm looking at the socks 19 00:01:02,040 --> 00:01:05,000 Speaker 4: up three point five percent, a new record high for 20 00:01:05,080 --> 00:01:07,640 Speaker 4: the index that tracks the semiconductor industry. 21 00:01:07,920 --> 00:01:10,240 Speaker 3: But what is leading a game? Of course, you know 22 00:01:10,280 --> 00:01:11,080 Speaker 3: it's one key name. 23 00:01:11,240 --> 00:01:13,240 Speaker 4: We look at AMD at one point hitting a record 24 00:01:13,280 --> 00:01:15,520 Speaker 4: high We are up the most on this stock in 25 00:01:15,640 --> 00:01:18,319 Speaker 4: nine years, twenty eight percent surge. At one point it 26 00:01:18,360 --> 00:01:21,000 Speaker 4: was thirty seven percent higher. All as we hear that 27 00:01:21,120 --> 00:01:22,920 Speaker 4: tens of billions are going to be added in terms 28 00:01:22,959 --> 00:01:27,080 Speaker 4: of revenue because the open ai infrastructure pat a sign 29 00:01:27,520 --> 00:01:31,480 Speaker 4: of real confidence in Lisa SU's AI accelerator and build 30 00:01:31,480 --> 00:01:34,440 Speaker 4: out ed. We've got such an interesting conversation coming out 31 00:01:34,640 --> 00:01:35,840 Speaker 4: about this new deal. 32 00:01:39,040 --> 00:01:42,360 Speaker 2: Welcome to Bloomberg radio and TV audiences around the world. 33 00:01:42,400 --> 00:01:45,520 Speaker 2: AMD has signed a definitive agreement with open Ai to 34 00:01:45,560 --> 00:01:49,760 Speaker 2: deploy six gigawatts of amdgpus. AMD says it will equate 35 00:01:49,800 --> 00:01:52,720 Speaker 2: to tens of billions of dollars in revenue. Open Ai 36 00:01:52,960 --> 00:01:55,200 Speaker 2: will get up to one hundred and sixty million AMD 37 00:01:55,320 --> 00:02:00,000 Speaker 2: shares in tranches and set against both operational and financial milestone. 38 00:02:00,320 --> 00:02:03,520 Speaker 2: The focus is inference. Let's bring in a m D 39 00:02:03,640 --> 00:02:06,840 Speaker 2: CEO Lisa Too and open Ai president Greg Brockman, both 40 00:02:06,840 --> 00:02:09,000 Speaker 2: of whom join us on set here at Bloomberg Tech 41 00:02:09,120 --> 00:02:09,880 Speaker 2: in San Francisco. 42 00:02:09,960 --> 00:02:12,680 Speaker 5: Good morning, Good morning, It's great to see you here. 43 00:02:12,800 --> 00:02:15,760 Speaker 2: Let's frame the opportunity. Lisa, you know that the market 44 00:02:15,800 --> 00:02:18,680 Speaker 2: reaction is very clear, but for a m D and 45 00:02:18,760 --> 00:02:21,639 Speaker 2: the AI industry at large. What do you think this represents. 46 00:02:22,080 --> 00:02:24,679 Speaker 6: Well, look, this is a huge milestone for a m D. 47 00:02:24,919 --> 00:02:27,200 Speaker 6: You know, we are so thrilled with the partnership with 48 00:02:27,520 --> 00:02:29,560 Speaker 6: the open Ai team, and it's also you know a 49 00:02:29,639 --> 00:02:32,440 Speaker 6: huge moment for the AI industry because you know, when 50 00:02:32,480 --> 00:02:34,400 Speaker 6: you get to the right down to it, you need 51 00:02:34,440 --> 00:02:36,799 Speaker 6: more AI compute. I mean, that's where we are today. 52 00:02:36,800 --> 00:02:39,240 Speaker 6: Compute is a foundation for all of the intelligence we 53 00:02:39,280 --> 00:02:41,800 Speaker 6: can get from AI. And you know, we are a 54 00:02:41,880 --> 00:02:45,079 Speaker 6: compute provider. We have spent years on our roadmap, We've 55 00:02:45,120 --> 00:02:48,680 Speaker 6: spent years working with open Ai and the team, and 56 00:02:48,840 --> 00:02:51,359 Speaker 6: you know, together now we're embarking on you know, a 57 00:02:51,440 --> 00:02:55,320 Speaker 6: massive build out of six gigawatts of AI compute, and 58 00:02:55,680 --> 00:02:58,320 Speaker 6: it's it's a big deal for us, for our shareholders, 59 00:02:58,400 --> 00:03:01,239 Speaker 6: for our teams, and for you know, the partnership and 60 00:03:01,560 --> 00:03:03,000 Speaker 6: the overall AI ecosystem. 61 00:03:03,240 --> 00:03:05,720 Speaker 2: Greg I say that the top the focus is inference. 62 00:03:05,840 --> 00:03:09,200 Speaker 2: I think that's really important to be specific about what 63 00:03:09,240 --> 00:03:11,840 Speaker 2: you will do with this capacity. So so literally explain 64 00:03:11,919 --> 00:03:14,000 Speaker 2: that part. And I'm conscious that you know, in the 65 00:03:14,000 --> 00:03:17,080 Speaker 2: first instance, the first target is one gigawatt and then 66 00:03:17,120 --> 00:03:19,880 Speaker 2: eventually six gigawatts, but what will you use it for? 67 00:03:20,600 --> 00:03:23,960 Speaker 7: Well, I think that the world continues to underestimate the 68 00:03:24,000 --> 00:03:27,880 Speaker 7: amount of demand for AI compute. Right that we've seen 69 00:03:27,960 --> 00:03:30,920 Speaker 7: this explosion of demand with things like chat GBT. You know, 70 00:03:30,919 --> 00:03:33,080 Speaker 7: we're at eight hundred million weekly active users. Now this 71 00:03:33,080 --> 00:03:35,680 Speaker 7: product didn't even exist three years ago, and we're in 72 00:03:35,720 --> 00:03:38,840 Speaker 7: a position where we cannot launch futures, we cannot launch 73 00:03:38,920 --> 00:03:42,120 Speaker 7: new products simply because of lack of computational power. And 74 00:03:42,200 --> 00:03:45,480 Speaker 7: we see these models continuing to get exponentially better, and 75 00:03:45,560 --> 00:03:46,960 Speaker 7: I think we're just heading to a world where so 76 00:03:47,040 --> 00:03:48,680 Speaker 7: much of the economy is going to be lifted up 77 00:03:48,680 --> 00:03:51,040 Speaker 7: and driven by progress and AI. And so we're very 78 00:03:51,120 --> 00:03:52,800 Speaker 7: much heading to a world by default that I think 79 00:03:52,840 --> 00:03:54,960 Speaker 7: looks like a compute desert, right that there's just not 80 00:03:55,080 --> 00:03:56,920 Speaker 7: enough compute to go around, and so we're trying to 81 00:03:56,920 --> 00:03:59,440 Speaker 7: build as much as possible, as quickly as possible. So 82 00:03:59,480 --> 00:04:01,600 Speaker 7: we're starting with one gigawatt simply because you got to 83 00:04:01,600 --> 00:04:04,520 Speaker 7: start somewhere. But honestly, we're building as fast as we 84 00:04:04,560 --> 00:04:07,440 Speaker 7: possibly can and trying to bring as much computational power 85 00:04:07,880 --> 00:04:09,720 Speaker 7: to bear for the economy and for the world. 86 00:04:10,480 --> 00:04:14,040 Speaker 4: Lisa, this is such a big commitment to Instinct in 87 00:04:14,080 --> 00:04:17,240 Speaker 4: particular as a customer. Does it make open ai the 88 00:04:17,320 --> 00:04:19,320 Speaker 4: largest for that particular product. 89 00:04:20,160 --> 00:04:23,120 Speaker 6: Well, this is certainly the largest deployment that we have 90 00:04:23,160 --> 00:04:26,400 Speaker 6: announced by far. I mean, you know, six gigawatts of compute. 91 00:04:26,800 --> 00:04:28,400 Speaker 6: As Greg said, we're going to start with the first 92 00:04:28,400 --> 00:04:31,080 Speaker 6: gigawatt in the second half of twenty twenty six on 93 00:04:31,160 --> 00:04:34,920 Speaker 6: our new next generation m I four to fifty chip. 94 00:04:35,400 --> 00:04:37,400 Speaker 6: I think the thing to understand is, you know, these 95 00:04:37,440 --> 00:04:40,920 Speaker 6: types of partnerships actually take you know, years to really 96 00:04:40,960 --> 00:04:43,359 Speaker 6: get comfortable with the idea that we're going to you know, 97 00:04:43,400 --> 00:04:46,360 Speaker 6: go all in together. And this isn't all in partnership 98 00:04:46,839 --> 00:04:50,000 Speaker 6: in terms of building out you know, the AI compute 99 00:04:50,480 --> 00:04:53,520 Speaker 6: that open ai needs for everything that they're offering to 100 00:04:53,520 --> 00:04:55,640 Speaker 6: the world. So yes, it's a huge deal, and it 101 00:04:56,040 --> 00:04:58,640 Speaker 6: also says a lot about you know, how much needs 102 00:04:58,640 --> 00:05:01,839 Speaker 6: to come together for you know, this entire ecosystem to operate. 103 00:05:01,920 --> 00:05:03,800 Speaker 6: So you know, we are setting up you know, certainly 104 00:05:03,839 --> 00:05:06,400 Speaker 6: there's a lot of engineering work, but our teams are 105 00:05:06,400 --> 00:05:10,000 Speaker 6: working together on hardware, software, We're ensuring the supply chain, 106 00:05:10,120 --> 00:05:12,800 Speaker 6: all of those elements are set up and ready to 107 00:05:13,120 --> 00:05:15,480 Speaker 6: deliver on this massive commitment. 108 00:05:16,040 --> 00:05:18,960 Speaker 4: Greg, talk us through a little bit about the players 109 00:05:19,000 --> 00:05:20,600 Speaker 4: that you need to also lean on. This has been 110 00:05:20,720 --> 00:05:22,880 Speaker 4: years in the making, as you say, with AMD, but 111 00:05:23,320 --> 00:05:25,799 Speaker 4: what other cloud providers were involved. How are you thinking 112 00:05:25,839 --> 00:05:28,800 Speaker 4: about this working with an Oracle all others out there. 113 00:05:29,600 --> 00:05:32,200 Speaker 7: Yeah, we really think of this as an industry wide effort, 114 00:05:32,480 --> 00:05:34,720 Speaker 7: and in general, we think that compute is something that 115 00:05:35,080 --> 00:05:37,600 Speaker 7: does require the entire supply chain to really wake up 116 00:05:37,600 --> 00:05:40,280 Speaker 7: and to really to start building much more than people 117 00:05:40,279 --> 00:05:43,200 Speaker 7: we're planning on. I think this starts from energy to 118 00:05:43,240 --> 00:05:46,200 Speaker 7: try to get far more power to be built on. 119 00:05:46,400 --> 00:05:48,040 Speaker 7: Things like nuclear I think are going to be very 120 00:05:48,040 --> 00:05:50,480 Speaker 7: important to come online. The cloud providers are an important 121 00:05:50,480 --> 00:05:51,760 Speaker 7: part of this as well. So we're going to be 122 00:05:51,800 --> 00:05:54,279 Speaker 7: deploying AMD in our own data centers. We'll be deploying 123 00:05:54,320 --> 00:05:56,840 Speaker 7: them together with cloud providers. You know, we have a 124 00:05:56,920 --> 00:05:59,960 Speaker 7: deal with Oracle, lots of other cloud providers out there. 125 00:06:00,360 --> 00:06:02,359 Speaker 7: You can really see that we're very much in the 126 00:06:02,680 --> 00:06:05,080 Speaker 7: We just want compute as much compute as possible. We 127 00:06:05,080 --> 00:06:07,039 Speaker 7: think this is important for the economy, We think this 128 00:06:07,080 --> 00:06:08,960 Speaker 7: is important for the nation, we think this is important 129 00:06:09,000 --> 00:06:11,800 Speaker 7: for humanity, and so really we're working with everyone in 130 00:06:11,839 --> 00:06:14,440 Speaker 7: this whole industry in order to get as much compute 131 00:06:14,480 --> 00:06:16,160 Speaker 7: power online as quickly as we can. 132 00:06:16,800 --> 00:06:20,960 Speaker 2: Lisa, I'm sorry specifics where is this data center going 133 00:06:21,000 --> 00:06:25,440 Speaker 2: to be? Is it one single site? Is it orical 134 00:06:26,320 --> 00:06:27,560 Speaker 2: that we'll partner with you on this? 135 00:06:28,040 --> 00:06:31,200 Speaker 6: Well, actually, what this really is is an announcement of 136 00:06:31,240 --> 00:06:32,720 Speaker 6: what you know, A m D and open a are 137 00:06:32,720 --> 00:06:35,080 Speaker 6: going to do together. You know, open ai has a 138 00:06:35,120 --> 00:06:37,800 Speaker 6: lot of partners in terms of you know, where they deploy. 139 00:06:38,160 --> 00:06:40,719 Speaker 6: I imagine a lot of it will be in cloud 140 00:06:40,720 --> 00:06:43,320 Speaker 6: service providers. It's really up to you know open Ai 141 00:06:43,440 --> 00:06:46,640 Speaker 6: and Greg and Sam and the team. But the way 142 00:06:46,680 --> 00:06:49,799 Speaker 6: to think about it is, for this you know, amount 143 00:06:49,800 --> 00:06:51,359 Speaker 6: of compute, it's going to have to be in a 144 00:06:51,360 --> 00:06:52,839 Speaker 6: lot of different places. 145 00:06:52,880 --> 00:06:55,000 Speaker 5: It's a massive amount, multiple locations. 146 00:06:54,640 --> 00:06:58,800 Speaker 6: Multiple locations, I would imagine, you know, multiple providers to 147 00:06:58,880 --> 00:07:00,720 Speaker 6: really get this online as fast as possible. 148 00:07:01,040 --> 00:07:05,200 Speaker 2: Greg, there is a lot of focus on where open 149 00:07:05,240 --> 00:07:07,440 Speaker 2: ai is going to get the money from to fund 150 00:07:07,440 --> 00:07:12,480 Speaker 2: all of this. Sam Altman's big picture commitment is well documented, 151 00:07:12,560 --> 00:07:15,520 Speaker 2: right and the numbers to his mind are in the trillions. 152 00:07:16,040 --> 00:07:19,520 Speaker 2: But have you specifically thought about debt financing for this 153 00:07:19,600 --> 00:07:24,160 Speaker 2: relationship with AMD? Have you thought about doing a specific 154 00:07:24,240 --> 00:07:28,160 Speaker 2: equity raise. You are very committed across multiple projects. 155 00:07:28,400 --> 00:07:28,600 Speaker 8: Yeah. 156 00:07:28,640 --> 00:07:30,040 Speaker 7: Look, the way that I would the way that I 157 00:07:30,040 --> 00:07:33,160 Speaker 7: would look at this is that AI revenue is growing 158 00:07:33,200 --> 00:07:36,880 Speaker 7: faster than I think almost any product in history, and 159 00:07:37,120 --> 00:07:39,239 Speaker 7: that ultimately, at the end of the day, the reason 160 00:07:39,240 --> 00:07:42,160 Speaker 7: this compute power is so important, it's so worthwhile for 161 00:07:42,400 --> 00:07:45,880 Speaker 7: everyone to build, is because the revenue ultimately will be there. 162 00:07:45,960 --> 00:07:46,200 Speaker 5: Now. 163 00:07:46,480 --> 00:07:48,880 Speaker 7: As a company that is trying to move as fast 164 00:07:48,880 --> 00:07:50,880 Speaker 7: as we can, we look at everything, right, we look 165 00:07:50,880 --> 00:07:54,760 Speaker 7: at equity debt, we look at trying to find creative 166 00:07:54,840 --> 00:07:58,000 Speaker 7: ways of financing all of this. That's been actually a 167 00:07:58,040 --> 00:07:59,880 Speaker 7: huge focus of us for the past couple of years, 168 00:08:00,000 --> 00:08:02,920 Speaker 7: thinking about how can we possibly build the amount of 169 00:08:02,920 --> 00:08:05,680 Speaker 7: compute that is required in order to really transform this 170 00:08:05,720 --> 00:08:08,640 Speaker 7: whole economy into an aipowered economy. And so I think 171 00:08:08,680 --> 00:08:12,120 Speaker 7: you'll see lots of creative ideas, but fundamentally, I think 172 00:08:12,120 --> 00:08:13,920 Speaker 7: at the end of the day, it is because we believe. 173 00:08:14,320 --> 00:08:16,440 Speaker 2: Sorry to jump in an interrupt and carriage, just forgive 174 00:08:16,480 --> 00:08:21,360 Speaker 2: me on this one. The condition of AMD issuing the 175 00:08:21,440 --> 00:08:25,080 Speaker 2: stock to open AI requires you to spend money basically 176 00:08:25,120 --> 00:08:29,720 Speaker 2: because you have to deliver that gig awad of capacity. First, Lisa, 177 00:08:29,760 --> 00:08:32,560 Speaker 2: I have to ask you if you have assurances that 178 00:08:32,640 --> 00:08:33,319 Speaker 2: open ai. 179 00:08:33,240 --> 00:08:33,840 Speaker 5: Is good for it. 180 00:08:34,160 --> 00:08:36,720 Speaker 6: Well, let me be clear. I mean, this deal is 181 00:08:36,760 --> 00:08:39,680 Speaker 6: a win for AMD, it's a win for open Ai, 182 00:08:39,880 --> 00:08:42,120 Speaker 6: and it's a win for our shareholders. And that's kind 183 00:08:42,120 --> 00:08:44,680 Speaker 6: of the way we put this together. I have full 184 00:08:44,720 --> 00:08:48,800 Speaker 6: confidence in open Ai, Sam, Greg Sarah. I mean, this 185 00:08:48,880 --> 00:08:51,960 Speaker 6: is a massive opportunity for us right now, right here. 186 00:08:52,120 --> 00:08:55,320 Speaker 6: It's about who has the most compute and how fast 187 00:08:55,360 --> 00:08:57,720 Speaker 6: can we get it online? And we're committing to doing 188 00:08:57,760 --> 00:09:00,920 Speaker 6: this together. And the fact is as open ai ad 189 00:09:00,920 --> 00:09:03,760 Speaker 6: buys chips, that's great for AMD. Our revenue goes up, 190 00:09:03,840 --> 00:09:06,600 Speaker 6: our earnings go up. You know, we expect that it 191 00:09:06,640 --> 00:09:11,160 Speaker 6: will also be very very accretive to our shareholders from 192 00:09:11,280 --> 00:09:13,760 Speaker 6: day one. And as we do that, you know, we're 193 00:09:13,840 --> 00:09:15,960 Speaker 6: very happy to have open Ai as a deep partner 194 00:09:16,000 --> 00:09:19,040 Speaker 6: and we win together. So it's like a virtuous positive 195 00:09:19,080 --> 00:09:21,360 Speaker 6: cycle in how we build out. You know, this big 196 00:09:21,440 --> 00:09:23,880 Speaker 6: vision for having all this compute out there, right. 197 00:09:24,280 --> 00:09:27,319 Speaker 4: And yet we still question, as you were just talking about, Greg, 198 00:09:27,600 --> 00:09:29,040 Speaker 4: some of the other supply chain elements. 199 00:09:29,040 --> 00:09:30,760 Speaker 3: You're talking about the need for nuclear for power. 200 00:09:31,200 --> 00:09:35,080 Speaker 4: What's really interesting is we are you feeling confident enough 201 00:09:35,320 --> 00:09:37,839 Speaker 4: about the rest of the compute the supply chain is there, 202 00:09:37,960 --> 00:09:41,080 Speaker 4: is this going to be US manufactured? From your perspective, 203 00:09:41,120 --> 00:09:43,840 Speaker 4: were you looking and also building out internationally with AMD. 204 00:09:44,840 --> 00:09:48,440 Speaker 7: Yeah, we've been looking at really all options our preference 205 00:09:48,440 --> 00:09:50,400 Speaker 7: and really the core thing that we try to do 206 00:09:50,440 --> 00:09:52,080 Speaker 7: is build as much as possible in the US. And 207 00:09:52,120 --> 00:09:53,800 Speaker 7: you can see the commitments that we've made over the 208 00:09:53,800 --> 00:09:56,480 Speaker 7: past year, you know, five hundred billion dollars of investment 209 00:09:56,520 --> 00:09:59,520 Speaker 7: in the US, and that's not stopping. We're continuing to build. 210 00:10:00,080 --> 00:10:03,240 Speaker 7: Think the international that there it is also going to 211 00:10:03,280 --> 00:10:05,600 Speaker 7: be important for the world to have compute. I think 212 00:10:05,600 --> 00:10:08,120 Speaker 7: that computer is going to become this like national security 213 00:10:08,280 --> 00:10:11,600 Speaker 7: strategic resource, and every country is going to need computational 214 00:10:11,679 --> 00:10:15,400 Speaker 7: power and so that we are really not limiting our 215 00:10:15,559 --> 00:10:17,640 Speaker 7: sort of sites in terms of where to build, but 216 00:10:17,720 --> 00:10:19,480 Speaker 7: we do think it is important that the US leads 217 00:10:19,480 --> 00:10:22,360 Speaker 7: in this technology, leads in computational power, and we're expanding 218 00:10:22,360 --> 00:10:24,240 Speaker 7: the supply chain. But you can see that we've really 219 00:10:24,280 --> 00:10:26,520 Speaker 7: been working with partners across the globe in order to 220 00:10:26,920 --> 00:10:28,960 Speaker 7: actually meet the demand that we expect to becoming and 221 00:10:29,040 --> 00:10:29,720 Speaker 7: upcoming years. 222 00:10:30,240 --> 00:10:33,360 Speaker 4: LISTA the manufacturing of these chips. Will do you look 223 00:10:33,360 --> 00:10:34,520 Speaker 4: to Intel at all for it? 224 00:10:34,520 --> 00:10:37,880 Speaker 6: Do you think in the future, Well, as you know, 225 00:10:38,000 --> 00:10:40,880 Speaker 6: the supply chain is something that we work on, you know, 226 00:10:41,080 --> 00:10:44,400 Speaker 6: very very meticulously. I think we have a very strong 227 00:10:44,400 --> 00:10:48,199 Speaker 6: supply chain. We're certainly deeply partnered with you know, TSMC 228 00:10:48,400 --> 00:10:51,760 Speaker 6: across the supply chain. You know, just to that earlier question, 229 00:10:51,960 --> 00:10:55,559 Speaker 6: we're absolutely prioritizing building in the United States because I 230 00:10:55,640 --> 00:10:59,720 Speaker 6: think that's super important. This is the US AI stack. 231 00:11:00,080 --> 00:11:01,360 Speaker 6: We want to have as much of it in the 232 00:11:01,440 --> 00:11:04,160 Speaker 6: US as possible. And you know, we continue to really 233 00:11:04,160 --> 00:11:06,040 Speaker 6: look at, you know, how do we ensure that there 234 00:11:06,080 --> 00:11:07,600 Speaker 6: will be a strong supply chain, you know. 235 00:11:07,600 --> 00:11:08,160 Speaker 9: Going forward. 236 00:11:08,440 --> 00:11:11,920 Speaker 2: Greg Sam posted on x that this deal with AMD 237 00:11:12,080 --> 00:11:15,320 Speaker 2: is incremental to what's already being done with Nvidia, but 238 00:11:15,679 --> 00:11:17,200 Speaker 2: as least enough. So I spent quite a lot of 239 00:11:17,200 --> 00:11:19,959 Speaker 2: time looking at the MII family and the newer generations 240 00:11:19,960 --> 00:11:23,199 Speaker 2: of products to come. Is there a very clear specific 241 00:11:23,280 --> 00:11:27,960 Speaker 2: benefit to using a MD technology for inference relative to 242 00:11:28,200 --> 00:11:31,000 Speaker 2: the capabilities of Nvidia, or do you just see it 243 00:11:31,080 --> 00:11:34,360 Speaker 2: broadly as some sort of diversifying factor. 244 00:11:34,760 --> 00:11:36,440 Speaker 7: Well, I would look at it this way, that there's 245 00:11:36,480 --> 00:11:39,600 Speaker 7: a huge fixed cost to getting AI models running on 246 00:11:39,640 --> 00:11:42,920 Speaker 7: any platform, and so that when we look at what's 247 00:11:42,960 --> 00:11:45,760 Speaker 7: out there, that actually getting AI training to work is 248 00:11:46,160 --> 00:11:48,680 Speaker 7: a huge, huge amount of lift. That's something we've really 249 00:11:48,720 --> 00:11:51,200 Speaker 7: only done the work for in VideA, but for inference 250 00:11:51,240 --> 00:11:55,199 Speaker 7: that that's something that's much more that there's an easier 251 00:11:55,360 --> 00:11:57,600 Speaker 7: barrier to entry there. And one thing we found is 252 00:11:57,640 --> 00:12:00,320 Speaker 7: that I think that the work that Lisai team have 253 00:12:00,360 --> 00:12:02,640 Speaker 7: been doing on the m I four fifty series, it's 254 00:12:02,679 --> 00:12:04,480 Speaker 7: looking like it's going to be a really incredible chip. 255 00:12:04,720 --> 00:12:07,040 Speaker 7: I think that that there's the way that these things 256 00:12:07,040 --> 00:12:10,360 Speaker 7: work is there's niches for different balances of memory and 257 00:12:10,960 --> 00:12:14,480 Speaker 7: computational power, and so as we have a diversity of workload, 258 00:12:14,520 --> 00:12:17,280 Speaker 7: we're finding that having a diversity of chips also really 259 00:12:17,320 --> 00:12:18,600 Speaker 7: accelerates what we're able to do. 260 00:12:18,920 --> 00:12:20,800 Speaker 2: Lisa, at the beginning of this conversation, I said they 261 00:12:20,800 --> 00:12:24,160 Speaker 2: were both operational and financial milestones to be met, and 262 00:12:24,400 --> 00:12:26,520 Speaker 2: Greg explained, you've got to start somewhere. So in the 263 00:12:26,520 --> 00:12:29,600 Speaker 2: first instance, one giga what, but would you just sort 264 00:12:29,600 --> 00:12:31,800 Speaker 2: of draw out the pathway to that first giga what? 265 00:12:32,720 --> 00:12:35,560 Speaker 2: You know, it seems like you're prepared to move quickly here. 266 00:12:35,720 --> 00:12:38,320 Speaker 6: Yeah, absolutely, and maybe if I can just build on 267 00:12:38,440 --> 00:12:41,439 Speaker 6: some things that Greg said. I think he's absolutely right. 268 00:12:41,520 --> 00:12:44,080 Speaker 6: You know, we're a believer in there's a diversity of 269 00:12:44,080 --> 00:12:46,880 Speaker 6: workloads and there will be a diversity of workloads across 270 00:12:47,240 --> 00:12:51,559 Speaker 6: you know, customers, models, use cases, and from that standpoint, 271 00:12:52,000 --> 00:12:54,240 Speaker 6: you know, we feel really good about how we're positioned. 272 00:12:54,280 --> 00:12:57,319 Speaker 6: You know, we we love the work here because you know, Frankly, 273 00:12:57,840 --> 00:13:00,400 Speaker 6: you know, open Ai is the ultimate power user of 274 00:13:00,400 --> 00:13:03,440 Speaker 6: our chips and and test us in very good ways. 275 00:13:03,880 --> 00:13:06,800 Speaker 6: So I think that's that's what gives us confidence that 276 00:13:06,880 --> 00:13:08,800 Speaker 6: you know, the technology is there. And then to your 277 00:13:08,880 --> 00:13:12,600 Speaker 6: point about milestones, yes, I mean this is clearly a 278 00:13:12,679 --> 00:13:16,480 Speaker 6: case where we are tied to each other. The first 279 00:13:16,520 --> 00:13:19,560 Speaker 6: gigawatt of deployment is super important. We're going to start that, 280 00:13:20,160 --> 00:13:22,080 Speaker 6: you know, second half of next year, and we're going 281 00:13:22,160 --> 00:13:25,360 Speaker 6: to build on from there. And it really is not 282 00:13:25,600 --> 00:13:29,480 Speaker 6: just the technology, but your commercial milestones, adoption milestones, and 283 00:13:29,880 --> 00:13:32,600 Speaker 6: just how we proliferate the capability going forward. But I'm 284 00:13:32,640 --> 00:13:34,840 Speaker 6: looking forward to building this as fast as possible. You know, 285 00:13:34,840 --> 00:13:38,479 Speaker 6: we're already working with a number of cloud service providers 286 00:13:38,679 --> 00:13:41,400 Speaker 6: who are also very active on our technology, and I 287 00:13:41,400 --> 00:13:44,280 Speaker 6: think this is a great catalyst to get the industry 288 00:13:44,360 --> 00:13:45,200 Speaker 6: to build faster. 289 00:13:46,160 --> 00:13:49,160 Speaker 4: Tied to each other is such an interesting ton of phrase, 290 00:13:49,240 --> 00:13:54,000 Speaker 4: and Greg, you are seeing more AI users and chip 291 00:13:54,400 --> 00:13:57,400 Speaker 4: makers and designers becoming more financially tied to each other. 292 00:13:57,880 --> 00:13:59,160 Speaker 3: Is this going to continue? 293 00:13:59,360 --> 00:14:01,320 Speaker 4: Is this the staf forward for how you see this 294 00:14:01,360 --> 00:14:02,400 Speaker 4: financing going forward? 295 00:14:03,360 --> 00:14:06,920 Speaker 7: Well, I really see the world transitioning to this AI 296 00:14:07,000 --> 00:14:09,719 Speaker 7: powered economy. And the interesting thing is within open AI 297 00:14:10,040 --> 00:14:13,719 Speaker 7: that we've really seen what it's like when your progress 298 00:14:14,040 --> 00:14:17,000 Speaker 7: is limited and accelerated as two sides of the coin 299 00:14:17,880 --> 00:14:20,720 Speaker 7: by computational power. Like teams within open Ai, that their 300 00:14:20,800 --> 00:14:23,560 Speaker 7: ability to deliver really is tied to the amount of 301 00:14:23,560 --> 00:14:25,520 Speaker 7: compute that they get. And I think we're heading to 302 00:14:25,560 --> 00:14:28,240 Speaker 7: a world where that is how the whole economy will function. 303 00:14:28,480 --> 00:14:31,040 Speaker 7: And we're starting to see it right that people having 304 00:14:31,080 --> 00:14:33,600 Speaker 7: access to better AI tools. If you're a coder, you're 305 00:14:33,640 --> 00:14:36,040 Speaker 7: able to do far more if you have access to 306 00:14:37,120 --> 00:14:39,080 Speaker 7: better AI models. And we're heading to a world where 307 00:14:39,080 --> 00:14:41,560 Speaker 7: if you can have ten times as much AI power 308 00:14:41,600 --> 00:14:44,600 Speaker 7: behind you, you will probably be ten times more productive. 309 00:14:44,880 --> 00:14:46,360 Speaker 7: And so I think that we're moving to a world 310 00:14:46,360 --> 00:14:48,680 Speaker 7: where the whole industry is waking up to the fact 311 00:14:48,720 --> 00:14:51,560 Speaker 7: that we have just not planned. We have not planned 312 00:14:51,560 --> 00:14:55,080 Speaker 7: for this moment where this explosion in AI demand is happening. 313 00:14:55,240 --> 00:14:57,440 Speaker 7: So it's happening all the way from the power to 314 00:14:57,520 --> 00:14:59,920 Speaker 7: the silicon, and I think this whole industry has to 315 00:15:00,040 --> 00:15:02,920 Speaker 7: to find a way to actually rise to meet the occasion. 316 00:15:03,480 --> 00:15:06,280 Speaker 2: Lisa, you have given us a look into the future 317 00:15:06,320 --> 00:15:09,320 Speaker 2: before about how you see the total addressable market the 318 00:15:09,400 --> 00:15:13,320 Speaker 2: industry now that the ink is dry with open AI 319 00:15:13,440 --> 00:15:17,800 Speaker 2: and Greg, are you rethinking either your bigger picture analysis 320 00:15:18,280 --> 00:15:21,240 Speaker 2: of the market for AI accelerators and GPUs or do 321 00:15:21,320 --> 00:15:24,400 Speaker 2: you see AMD now having an improved position in that 322 00:15:24,440 --> 00:15:27,280 Speaker 2: market relative of course to your friends at Nvidia. 323 00:15:27,800 --> 00:15:30,880 Speaker 6: Well, again, I think, and I've told you before, I 324 00:15:30,920 --> 00:15:33,960 Speaker 6: believe that this is a huge market. You know, we 325 00:15:34,080 --> 00:15:38,640 Speaker 6: have size just the AI accelerator TAM being you know, 326 00:15:38,720 --> 00:15:41,360 Speaker 6: over five hundred billion dollars in TAM over the next 327 00:15:41,640 --> 00:15:44,040 Speaker 6: few years. I think some might say, you know, maybe 328 00:15:44,080 --> 00:15:47,240 Speaker 6: I was a little conservative in that TAM analysis, but 329 00:15:47,320 --> 00:15:49,840 Speaker 6: the way to think about it is there's so much 330 00:15:49,880 --> 00:15:52,080 Speaker 6: need from compute. I mean, you just heard it from Greg, 331 00:15:52,240 --> 00:15:55,120 Speaker 6: so you know this is a huge pie and you're 332 00:15:55,120 --> 00:15:57,520 Speaker 6: going to see the need for you know, more players 333 00:15:57,560 --> 00:16:00,920 Speaker 6: coming into it, and you know, from my standpoint, this 334 00:16:01,000 --> 00:16:04,320 Speaker 6: is a big validation of our technology and our capability. 335 00:16:04,800 --> 00:16:07,000 Speaker 6: You know, as much as we love the work with OpenAI, 336 00:16:07,080 --> 00:16:09,400 Speaker 6: we're working with a lot of other customers as well. 337 00:16:09,440 --> 00:16:11,440 Speaker 6: There's a lot of excitement in the industry around m 338 00:16:11,480 --> 00:16:13,280 Speaker 6: I four fifty, so we're ready for it. 339 00:16:14,040 --> 00:16:17,880 Speaker 4: MDCO Lisa Sou Open Ai President Greg Brockman, it's been 340 00:16:17,920 --> 00:16:19,000 Speaker 4: a joy having you on the show. 341 00:16:19,160 --> 00:16:20,280 Speaker 3: Thank you both very much. 342 00:16:21,280 --> 00:16:24,080 Speaker 4: Thank you, and coming up more on AMD and its 343 00:16:24,200 --> 00:16:26,200 Speaker 4: impact on the broader tech markets. We've got a key 344 00:16:26,200 --> 00:16:28,480 Speaker 4: investor for you, Tony one of t row Price. 345 00:16:28,800 --> 00:16:29,680 Speaker 3: This is really metech. 346 00:16:36,480 --> 00:16:39,800 Speaker 4: Let's stay on that AMD open Ai story, it's impact 347 00:16:39,800 --> 00:16:42,320 Speaker 4: of the broader tech markets. Tony Huang t row Price, 348 00:16:42,400 --> 00:16:44,240 Speaker 4: Science and Technology from Portfolio Manager. 349 00:16:44,400 --> 00:16:46,360 Speaker 3: You have exposure to AMD. 350 00:16:46,600 --> 00:16:48,880 Speaker 4: It's added more than seventeen billion dollars in market cap 351 00:16:48,920 --> 00:16:49,760 Speaker 4: on one day alone. 352 00:16:49,760 --> 00:16:50,800 Speaker 3: What does this deal signal? 353 00:16:52,200 --> 00:16:55,080 Speaker 10: Yeah, Well, I think it's a really exciting deal here 354 00:16:55,120 --> 00:16:59,960 Speaker 10: for AMD and validates their roadmap and their silica and software's. 355 00:17:00,560 --> 00:17:03,240 Speaker 10: I think it gives them a really big lead customer 356 00:17:03,520 --> 00:17:06,639 Speaker 10: to really scale and I think that this will probably 357 00:17:06,680 --> 00:17:10,960 Speaker 10: attract other customers to their roadmap and continue to build 358 00:17:10,960 --> 00:17:11,400 Speaker 10: out ROCK. 359 00:17:11,800 --> 00:17:13,639 Speaker 8: And I've always had this like thesis. 360 00:17:13,280 --> 00:17:17,280 Speaker 10: That the TAM is really big and if companies can 361 00:17:17,359 --> 00:17:21,159 Speaker 10: deliver on that compute, that there's there's a lot of 362 00:17:21,400 --> 00:17:24,639 Speaker 10: companies that can win in the overall compute. TAM so 363 00:17:24,880 --> 00:17:28,160 Speaker 10: overall very exciting, and you know, I think it's got 364 00:17:28,200 --> 00:17:30,879 Speaker 10: a good setup for the next few years for the 365 00:17:30,880 --> 00:17:31,720 Speaker 10: company's runway. 366 00:17:31,840 --> 00:17:33,720 Speaker 4: I mean, Lisa Sue just telling us that maybe she's 367 00:17:33,760 --> 00:17:36,000 Speaker 4: being conservative when she puts the AI accelerate to TAM 368 00:17:36,040 --> 00:17:37,160 Speaker 4: at five hundred billion dollars. 369 00:17:37,240 --> 00:17:38,400 Speaker 3: Tony, I'm interested. 370 00:17:38,080 --> 00:17:40,920 Speaker 4: Though, at opening eye telling us Greg saying that we're 371 00:17:40,920 --> 00:17:42,720 Speaker 4: looking at all kinds of way of financing this. 372 00:17:43,800 --> 00:17:45,280 Speaker 3: How are you feeling. 373 00:17:45,160 --> 00:17:50,840 Speaker 4: About this knitting together of purchases and creators, designers of chips. 374 00:17:52,440 --> 00:17:54,440 Speaker 10: Yeah, well, I think that taking your step back over 375 00:17:54,480 --> 00:17:58,200 Speaker 10: the last kind of ten years, everybody's underestimated this TAM. 376 00:17:58,560 --> 00:17:59,480 Speaker 8: And I think what's. 377 00:17:59,280 --> 00:18:02,360 Speaker 10: Exciting about AI is that there are a lot of 378 00:18:02,440 --> 00:18:05,800 Speaker 10: productivity use cases here and we're starting to see AI 379 00:18:05,920 --> 00:18:08,800 Speaker 10: agents really form. And then I think physical AI, so 380 00:18:09,040 --> 00:18:11,280 Speaker 10: I think as long as like the n use cases 381 00:18:11,359 --> 00:18:13,880 Speaker 10: are delivering a lot of value, and there's promise there. 382 00:18:13,920 --> 00:18:16,280 Speaker 10: I think that's what matters in terms of the ROI, 383 00:18:16,320 --> 00:18:18,400 Speaker 10: and I think a lot of these kind of contracts, 384 00:18:18,400 --> 00:18:21,240 Speaker 10: I mean they are backstop by really blue chip companies 385 00:18:21,280 --> 00:18:22,680 Speaker 10: like Mirosov or Oracle. 386 00:18:23,600 --> 00:18:25,760 Speaker 8: So in terms of you know, how I'm feeling. 387 00:18:25,520 --> 00:18:28,320 Speaker 10: About the financing, I think it's it continues to be 388 00:18:28,359 --> 00:18:30,480 Speaker 10: probably a good attractive way to do as long as 389 00:18:30,520 --> 00:18:32,080 Speaker 10: the use cases are intact. 390 00:18:32,760 --> 00:18:36,120 Speaker 2: Tony, the mechanics of the deal are the open AI 391 00:18:36,359 --> 00:18:38,600 Speaker 2: has the right to buy up to one hundred and 392 00:18:38,680 --> 00:18:42,399 Speaker 2: sixty million shares of AMD at a penny apiece, but 393 00:18:42,480 --> 00:18:46,840 Speaker 2: that right contingent on the operational and financial milestones that 394 00:18:46,840 --> 00:18:51,080 Speaker 2: we discussed. It is not circular financing, it's something different. 395 00:18:51,560 --> 00:18:55,200 Speaker 2: But how comfortable are you with that mechanism. 396 00:18:55,320 --> 00:18:57,320 Speaker 8: Well, I think it could make a lot of sense, right. 397 00:18:57,840 --> 00:19:00,639 Speaker 10: I mean it puts the two companies of in co 398 00:19:00,840 --> 00:19:05,440 Speaker 10: development and partnership, and I think there's shared economics and 399 00:19:05,440 --> 00:19:07,719 Speaker 10: it's a win win for both companies. I think some 400 00:19:07,800 --> 00:19:10,040 Speaker 10: of the prices the warrants are stricted at a pretty 401 00:19:11,600 --> 00:19:14,159 Speaker 10: you know high price on the AMB stock price, So 402 00:19:14,200 --> 00:19:17,080 Speaker 10: I think it's good for both sides and if they win, 403 00:19:17,119 --> 00:19:17,879 Speaker 10: they win together. 404 00:19:18,040 --> 00:19:20,200 Speaker 8: So to me, I think it can make sense. 405 00:19:21,720 --> 00:19:26,840 Speaker 2: Does this alter AMD's position in the AI accelerator market 406 00:19:26,840 --> 00:19:30,320 Speaker 2: in other words, they're closing the gap a little within video? 407 00:19:31,880 --> 00:19:32,440 Speaker 8: Yeah, well, I. 408 00:19:32,359 --> 00:19:35,920 Speaker 10: Think that both companies are doing really well and they're 409 00:19:35,920 --> 00:19:37,840 Speaker 10: doing things a little bit differently. 410 00:19:38,400 --> 00:19:39,800 Speaker 8: But I think that you're. 411 00:19:39,640 --> 00:19:42,560 Speaker 10: Alluding to kind of the market narrative of AMB was 412 00:19:42,640 --> 00:19:45,679 Speaker 10: kind of in between kind of in video on the 413 00:19:45,720 --> 00:19:49,200 Speaker 10: GPU side and then Broadcom on the custom side, and 414 00:19:49,280 --> 00:19:51,439 Speaker 10: I think this can help lift the narrative to go 415 00:19:51,520 --> 00:19:55,639 Speaker 10: from like a second place GPU player to a co developed. 416 00:19:55,320 --> 00:19:58,960 Speaker 8: Partner of open Ai on the AI compute side. 417 00:19:59,000 --> 00:20:02,440 Speaker 10: So I think it positions them well, gives them credibility scale, 418 00:20:03,400 --> 00:20:06,879 Speaker 10: an additional kind of kind of co development R and 419 00:20:06,960 --> 00:20:10,280 Speaker 10: D to improve their roadmaps. So to me, it definitely 420 00:20:10,359 --> 00:20:13,199 Speaker 10: is a big positive. And I don't think it's a 421 00:20:13,320 --> 00:20:15,800 Speaker 10: zero sum game. I mean, I think it just further 422 00:20:15,960 --> 00:20:18,679 Speaker 10: signifies the demand for AI computing. It's good for the 423 00:20:18,680 --> 00:20:22,040 Speaker 10: country to build this much capacity because I do think 424 00:20:22,200 --> 00:20:24,360 Speaker 10: the and payoffs are immense. 425 00:20:24,800 --> 00:20:27,240 Speaker 4: I mean, Tony, we're a new record high for the 426 00:20:27,320 --> 00:20:29,320 Speaker 4: NASA that one hundred. At one point, MD was at 427 00:20:29,320 --> 00:20:31,680 Speaker 4: a record high. The valuations kind of a higher. When 428 00:20:31,680 --> 00:20:34,040 Speaker 4: I'm looking at your holdings top holding, Apple, Alphabet and 429 00:20:34,080 --> 00:20:37,480 Speaker 4: Video Meta, do feel confident that these are going to 430 00:20:37,480 --> 00:20:42,000 Speaker 4: continue to gain in value with this AI story. 431 00:20:42,359 --> 00:20:45,440 Speaker 10: I think there's a good long term story behind all 432 00:20:45,480 --> 00:20:48,080 Speaker 10: our holdings, and I think a lot of them are 433 00:20:48,160 --> 00:20:52,320 Speaker 10: beneficiars of AI. They're strong companies that are are great 434 00:20:52,359 --> 00:20:55,879 Speaker 10: platforms and as a result, I think that you know, 435 00:20:55,960 --> 00:20:59,040 Speaker 10: we are entering this inflection in AI driving a ton 436 00:20:59,080 --> 00:21:03,119 Speaker 10: of productivity. You think about you know, output being productivity 437 00:21:03,160 --> 00:21:05,480 Speaker 10: and labor. I think pun game is going up and 438 00:21:05,680 --> 00:21:09,200 Speaker 10: labor is going to be relatively uncapped, I think as 439 00:21:09,440 --> 00:21:12,800 Speaker 10: a limiting factors. So to me, I think it's exciting. 440 00:21:12,880 --> 00:21:15,080 Speaker 10: It's never been a better time to be a tech investor. 441 00:21:15,119 --> 00:21:18,520 Speaker 10: I think there's just so much change in dynamism, So 442 00:21:18,520 --> 00:21:20,240 Speaker 10: I'm excited for the multi. 443 00:21:20,040 --> 00:21:23,920 Speaker 2: Year here, Tony, if your humor me please, My colin 444 00:21:24,000 --> 00:21:26,360 Speaker 2: today was about all of the debt deals that are 445 00:21:26,440 --> 00:21:31,080 Speaker 2: underpinning the AI infrastructure build out that we're seeing. You know, 446 00:21:31,359 --> 00:21:33,800 Speaker 2: everyone seems to have a different comfort level with that 447 00:21:33,880 --> 00:21:36,679 Speaker 2: as well. How sanguine are you about the role of 448 00:21:36,720 --> 00:21:38,520 Speaker 2: debt in what we're seeing? 449 00:21:39,960 --> 00:21:42,080 Speaker 10: Yeah, I think that you know, the key signal here 450 00:21:42,200 --> 00:21:46,119 Speaker 10: is you know, what what is this AI capacity, this 451 00:21:46,200 --> 00:21:47,880 Speaker 10: AI infrastructed capacity. 452 00:21:47,520 --> 00:21:49,560 Speaker 8: Worth, and what's the useful life? 453 00:21:49,640 --> 00:21:52,560 Speaker 10: I think a few years ago we're debating is the 454 00:21:52,640 --> 00:21:55,679 Speaker 10: useful life for a GPU data center three or four years? 455 00:21:55,880 --> 00:21:58,600 Speaker 10: And I think that was just too short. You look 456 00:21:58,640 --> 00:22:00,960 Speaker 10: at a lot of the GP use you know from 457 00:22:01,000 --> 00:22:04,000 Speaker 10: seven eight years ago. They're still being fully utilized because 458 00:22:04,040 --> 00:22:06,919 Speaker 10: the AI workloads continue to evolve. And then when you 459 00:22:06,960 --> 00:22:09,840 Speaker 10: fully appreciate these assets, like they're at a price. 460 00:22:09,640 --> 00:22:13,280 Speaker 8: Point that could make sense for other lms be running 461 00:22:13,280 --> 00:22:13,600 Speaker 8: on them. 462 00:22:13,600 --> 00:22:16,000 Speaker 10: And so I think that you know, the useful life 463 00:22:16,000 --> 00:22:19,200 Speaker 10: of these gps are definitely more than four or five years. 464 00:22:19,680 --> 00:22:23,720 Speaker 10: And you know, there's just more and more use cases. 465 00:22:23,840 --> 00:22:30,040 Speaker 10: You think about robotics, like you know, diagnostics, simulation, there's 466 00:22:30,080 --> 00:22:32,280 Speaker 10: just a ton more than we thought there were three 467 00:22:32,359 --> 00:22:34,080 Speaker 10: years ago. And I think that's what We'll keep the 468 00:22:34,200 --> 00:22:37,040 Speaker 10: utilization high of these data sounds as long as that 469 00:22:37,240 --> 00:22:39,560 Speaker 10: you know continues, I think it makes a lot of sense. 470 00:22:39,760 --> 00:22:42,040 Speaker 3: Totally a good pivot another part of your portfolio. 471 00:22:42,200 --> 00:22:44,560 Speaker 4: Verizon change at the top, hands verse bag, handing over 472 00:22:44,600 --> 00:22:46,480 Speaker 4: the rainsidantialman, how'd you feel about it. 473 00:22:48,840 --> 00:22:50,880 Speaker 8: You can you come again now that question. 474 00:22:51,119 --> 00:22:54,120 Speaker 4: With Verizon, we are seeing executive changes at the top 475 00:22:54,160 --> 00:22:55,200 Speaker 4: on certain companies. 476 00:22:55,600 --> 00:22:57,080 Speaker 3: How do you feel, for example of. 477 00:22:57,160 --> 00:23:00,600 Speaker 4: Verizon seeing hands verse bag changing over the rain today. 478 00:23:01,760 --> 00:23:04,480 Speaker 10: Yeah, actually we don't own verses, so I probably that's 479 00:23:04,480 --> 00:23:05,840 Speaker 10: a little bit outside of my wheelhouse. 480 00:23:05,840 --> 00:23:06,680 Speaker 8: To carment. 481 00:23:09,640 --> 00:23:11,919 Speaker 2: Tony Wong with tro Price, we're trying to get you 482 00:23:11,960 --> 00:23:13,760 Speaker 2: on all the news of the day, but we really 483 00:23:13,800 --> 00:23:16,320 Speaker 2: appreciate the insight into the AMD open aideal. 484 00:23:16,400 --> 00:23:17,560 Speaker 5: Thank you very much, Carrot. 485 00:23:17,840 --> 00:23:19,080 Speaker 3: Time now for talking tech. 486 00:23:19,160 --> 00:23:22,439 Speaker 4: First up, Apple ed is facing an investigation in France 487 00:23:22,640 --> 00:23:25,800 Speaker 4: over the storing of voice recordings made by its Sirie assistant. 488 00:23:26,080 --> 00:23:29,360 Speaker 4: Probe is related to claims that subcontractors and had access 489 00:23:29,400 --> 00:23:33,440 Speaker 4: to sensitive recordings. The iphonemaker says it uses Syrian interactions 490 00:23:33,440 --> 00:23:37,040 Speaker 4: to improve services and only stores them in users opt in. 491 00:23:37,359 --> 00:23:39,040 Speaker 3: Apple declined to comment on the investigation. 492 00:23:39,160 --> 00:23:41,399 Speaker 4: Plus Fox con reported sales growth of eleven percent in 493 00:23:41,440 --> 00:23:44,840 Speaker 4: the third quarter and projective further sales growth this quarter. 494 00:23:45,240 --> 00:23:47,840 Speaker 3: The gains by and video partner officially named Honhai. 495 00:23:47,680 --> 00:23:50,639 Speaker 4: Signals that demand for AI chips and servers remains strong, 496 00:23:51,240 --> 00:23:55,160 Speaker 4: and Tesla shares jumped after the company posted cryptic videos 497 00:23:55,240 --> 00:23:58,640 Speaker 4: teasing a possible product. Unveil Saysa's last addition to its 498 00:23:58,640 --> 00:24:01,000 Speaker 4: product lineup was a cyber truck two years ago. 499 00:24:01,359 --> 00:24:05,440 Speaker 3: Executives have said, an affordable model. Why it's on the works. 500 00:24:10,560 --> 00:24:12,359 Speaker 3: Welcome back to Blue med Tech. We check in on 501 00:24:12,440 --> 00:24:13,040 Speaker 3: these markets. 502 00:24:13,320 --> 00:24:17,159 Speaker 4: New record highs October apparently has been coined, certainly for 503 00:24:17,240 --> 00:24:19,000 Speaker 4: looking at the crypto world, Bitcoin one hundred and twenty 504 00:24:19,400 --> 00:24:21,600 Speaker 4: two hundred and thirty up two percent. We had a 505 00:24:21,680 --> 00:24:24,120 Speaker 4: risk on rally throughout the weekend. That risk on rally 506 00:24:24,119 --> 00:24:25,959 Speaker 4: continues into money with the NAZ that one hundred and 507 00:24:25,960 --> 00:24:26,960 Speaker 4: seven ten seven percent. 508 00:24:27,200 --> 00:24:28,080 Speaker 3: And you know why. 509 00:24:28,400 --> 00:24:30,159 Speaker 4: There is a key player at the moment that is 510 00:24:30,200 --> 00:24:31,960 Speaker 4: helping us push up and to the right. It is 511 00:24:32,040 --> 00:24:34,119 Speaker 4: AMD up twenty six percent at one point, hitting a 512 00:24:34,119 --> 00:24:37,800 Speaker 4: record high, up the most in nine years. Extraordinary new 513 00:24:37,840 --> 00:24:41,080 Speaker 4: deal soaring after of course, it's doing this deal with 514 00:24:41,119 --> 00:24:44,399 Speaker 4: open Ai for AI infrastructure that can generate tens of 515 00:24:44,480 --> 00:24:46,920 Speaker 4: billions of dollars in new revenue. 516 00:24:47,160 --> 00:24:47,200 Speaker 5: Ed. 517 00:24:47,280 --> 00:24:49,640 Speaker 4: We had a great conversation with Lisa Su and Greg 518 00:24:49,680 --> 00:24:51,920 Speaker 4: Brockman a moment ago. Another great conversation coming up. 519 00:24:52,680 --> 00:24:54,359 Speaker 5: Yeah, really looking forward to this one. 520 00:24:54,680 --> 00:24:57,200 Speaker 2: Joining us now as the White House, AI and cryptos 521 00:24:57,240 --> 00:24:59,880 Speaker 2: are David Sachs. He's also the co founder and partner 522 00:25:00,119 --> 00:25:03,080 Speaker 2: of Craft Ventures all of course as well. David, there 523 00:25:03,119 --> 00:25:04,640 Speaker 2: is something that I wanted to talk to you about 524 00:25:04,680 --> 00:25:06,159 Speaker 2: for a long time, and we will get to it, 525 00:25:06,240 --> 00:25:09,439 Speaker 2: the debate around should the United States or should the 526 00:25:09,520 --> 00:25:12,960 Speaker 2: United States not export some form of AI chip to China, 527 00:25:13,160 --> 00:25:15,200 Speaker 2: And we will get to it. But I'm sure you 528 00:25:15,280 --> 00:25:18,600 Speaker 2: can appreciate, you know, the AMD and Open AI agreement 529 00:25:18,760 --> 00:25:21,480 Speaker 2: is very interesting, and so as a place to start, 530 00:25:21,600 --> 00:25:24,960 Speaker 2: could I just ask your sort of interpretation of what 531 00:25:25,040 --> 00:25:28,080 Speaker 2: it signals when you have a US chip maker like 532 00:25:28,119 --> 00:25:31,880 Speaker 2: AMD after Nvidia, and one of the frontier model shops 533 00:25:32,160 --> 00:25:35,600 Speaker 2: like open Ai working so closely together on infrastructure. 534 00:25:37,119 --> 00:25:38,680 Speaker 11: Well, and I think the main thing here is this, 535 00:25:38,720 --> 00:25:41,000 Speaker 11: there's an AI boom going on and the market is 536 00:25:41,040 --> 00:25:44,000 Speaker 11: highly competitive. We have a number of leading model companies, 537 00:25:44,000 --> 00:25:45,840 Speaker 11: we have a number of leading chip companies, and they're 538 00:25:45,840 --> 00:25:48,359 Speaker 11: all competing with each other and they're all booming. And 539 00:25:48,680 --> 00:25:52,320 Speaker 11: what we see here is that this AI boom just 540 00:25:52,400 --> 00:25:56,240 Speaker 11: keeps going and going and keeps driving businesses to new highs. 541 00:25:56,640 --> 00:25:59,640 Speaker 11: And this is all a result of President Trump's pro innovation, 542 00:26:00,359 --> 00:26:04,240 Speaker 11: pro export, pro AI policy. We saw this help drive 543 00:26:04,760 --> 00:26:07,119 Speaker 11: us GDP to three point eight percent growth rate in 544 00:26:07,240 --> 00:26:10,200 Speaker 11: the second quarter. So I think this seems this boom 545 00:26:10,200 --> 00:26:11,680 Speaker 11: just seems like you going to keep going and going, 546 00:26:11,960 --> 00:26:13,040 Speaker 11: and we just heard I think it's. 547 00:26:12,920 --> 00:26:15,560 Speaker 4: Just justson One was the latest on a podcast shouting 548 00:26:15,560 --> 00:26:17,520 Speaker 4: you out, shouting out those that are in the team 549 00:26:17,520 --> 00:26:19,439 Speaker 4: at the White House at the moment guiding when it 550 00:26:19,440 --> 00:26:22,080 Speaker 4: comes to technology, David. But I'm interested as to whether you, 551 00:26:22,080 --> 00:26:26,280 Speaker 4: as an investor, have great comfort in these relationships being built, 552 00:26:26,359 --> 00:26:30,280 Speaker 4: financial relationships between chip designers and the chip users. 553 00:26:32,400 --> 00:26:33,720 Speaker 11: Well, I think it's up to them. You know, I 554 00:26:33,720 --> 00:26:37,360 Speaker 11: don't really take sides in these deals. We want all 555 00:26:37,400 --> 00:26:40,240 Speaker 11: of our American air companies to be successful, and they're 556 00:26:40,280 --> 00:26:42,600 Speaker 11: all competing with each other and cooperating with each other. 557 00:26:42,680 --> 00:26:45,680 Speaker 11: I guess it's called coopetition, and that's a great thing 558 00:26:45,680 --> 00:26:48,840 Speaker 11: to see. We just want, you know, putting my AI 559 00:26:48,880 --> 00:26:52,359 Speaker 11: hat on, we just want these markets to be competitive, 560 00:26:52,359 --> 00:26:54,280 Speaker 11: and we want American companies to be successful, and that's 561 00:26:54,280 --> 00:26:55,160 Speaker 11: what's happening right now. 562 00:26:56,440 --> 00:27:00,359 Speaker 4: I'm looking therefore at the future rollout, the buildouting that 563 00:27:00,400 --> 00:27:03,400 Speaker 4: gives you pause in terms of actually the infrastructure needs 564 00:27:03,400 --> 00:27:03,960 Speaker 4: here at the moment. 565 00:27:04,040 --> 00:27:08,040 Speaker 11: David, Well, anytime you're working in the world of atoms, 566 00:27:08,240 --> 00:27:10,800 Speaker 11: it's going to be more complicated than when you're working 567 00:27:11,000 --> 00:27:13,359 Speaker 11: in the world of bits, and so it's very important 568 00:27:13,400 --> 00:27:15,720 Speaker 11: for us to scale up the amount of power that's 569 00:27:15,760 --> 00:27:18,919 Speaker 11: available to AI companies. And you're seeing that thanks to 570 00:27:19,000 --> 00:27:24,240 Speaker 11: President Trump's policies that are pro energy. The President back 571 00:27:24,280 --> 00:27:26,320 Speaker 11: the idea of drill baby drill going back many years. 572 00:27:26,359 --> 00:27:28,359 Speaker 11: I think he was very far sighted in this regard. 573 00:27:28,400 --> 00:27:31,120 Speaker 11: He understood that energy is a basis for everything. It's 574 00:27:31,119 --> 00:27:34,679 Speaker 11: certainly the basis for this AI boom. And we're in 575 00:27:34,720 --> 00:27:37,800 Speaker 11: the process of allowing a lot more power generation in 576 00:27:37,840 --> 00:27:41,199 Speaker 11: the US. President Trump is allowing so called behind the 577 00:27:41,200 --> 00:27:44,040 Speaker 11: meter power generation. In other words, these AI companies can 578 00:27:44,240 --> 00:27:46,720 Speaker 11: stand up their own power generation. We need to squeeze 579 00:27:46,760 --> 00:27:50,399 Speaker 11: more out of the grid, and we're enabling new oil 580 00:27:50,400 --> 00:27:53,720 Speaker 11: and gas and nuclear so all the above. So I 581 00:27:53,760 --> 00:27:57,080 Speaker 11: think that energy is the main thing, and President Trump 582 00:27:57,080 --> 00:27:57,639 Speaker 11: is supporting that. 583 00:27:58,760 --> 00:27:59,080 Speaker 5: David. 584 00:27:59,119 --> 00:28:02,080 Speaker 2: Final question on on AMD and open AI. Did the 585 00:28:02,119 --> 00:28:05,439 Speaker 2: parties consult the White House about moving forward with this 586 00:28:05,560 --> 00:28:07,760 Speaker 2: arrangement and their plan to work together? 587 00:28:09,240 --> 00:28:09,600 Speaker 8: No? 588 00:28:09,600 --> 00:28:11,800 Speaker 11: No, I mean not with me. And I wouldn't expect 589 00:28:11,840 --> 00:28:14,400 Speaker 11: them to. We just don't get involved in the deal 590 00:28:14,400 --> 00:28:18,000 Speaker 11: making between private companies. Again, We're here to create a 591 00:28:18,040 --> 00:28:21,280 Speaker 11: policy environment that is supportive for all of RII companies. 592 00:28:22,200 --> 00:28:22,840 Speaker 5: I appreciate that. 593 00:28:22,880 --> 00:28:24,720 Speaker 2: Okay, So, David, I asked you to come on the 594 00:28:24,720 --> 00:28:27,760 Speaker 2: program because we spoke at the end of last week 595 00:28:27,840 --> 00:28:30,359 Speaker 2: for a story that was done out of DC on 596 00:28:30,680 --> 00:28:33,960 Speaker 2: the idea of well, what is a China Hawk? But 597 00:28:34,160 --> 00:28:39,880 Speaker 2: also within that broad description, should the United States or 598 00:28:39,920 --> 00:28:45,200 Speaker 2: should the United States not export deprecated AI chips to China. 599 00:28:45,600 --> 00:28:48,120 Speaker 2: You've been very generous with your time, but I just 600 00:28:48,200 --> 00:28:51,480 Speaker 2: invite you again for our Bloomberg Tech audience. Explain how 601 00:28:51,520 --> 00:28:54,440 Speaker 2: you view yourself in this debate and what you think 602 00:28:54,520 --> 00:28:56,760 Speaker 2: is strategically most important for this country. 603 00:28:58,000 --> 00:29:00,320 Speaker 11: Well, first of all, I would say that I consider 604 00:29:00,400 --> 00:29:02,600 Speaker 11: myself to be a China Hawk. I want the US 605 00:29:02,680 --> 00:29:06,800 Speaker 11: to win this AI race. We understand that China is 606 00:29:06,840 --> 00:29:10,360 Speaker 11: our main competition globally in this AI race, and we 607 00:29:10,400 --> 00:29:13,280 Speaker 11: want to do everything we can to win. And President 608 00:29:13,280 --> 00:29:16,600 Speaker 11: Trump and his AI speech on July twenty third laid 609 00:29:16,600 --> 00:29:18,360 Speaker 11: out some of the tempoles of that strategy. We need 610 00:29:18,360 --> 00:29:20,560 Speaker 11: to be pro innovation, We need to be pro infrastructure 611 00:29:20,560 --> 00:29:23,040 Speaker 11: and pro energy, and we need to be pro exports 612 00:29:23,040 --> 00:29:26,440 Speaker 11: so that the American technology stack dominates the world. So 613 00:29:26,560 --> 00:29:29,120 Speaker 11: this is all in the service of the United States 614 00:29:29,160 --> 00:29:32,440 Speaker 11: winning this AI race. We understand that it's going to 615 00:29:32,480 --> 00:29:36,560 Speaker 11: have major economic and national security ratifications, and we're in 616 00:29:36,600 --> 00:29:37,120 Speaker 11: it to win it. 617 00:29:38,360 --> 00:29:40,120 Speaker 2: There is a lot of focus on the direct to 618 00:29:40,240 --> 00:29:42,600 Speaker 2: China part. But the other way that some look at 619 00:29:42,600 --> 00:29:45,600 Speaker 2: it is there is a marketplace outside of America and 620 00:29:45,640 --> 00:29:49,240 Speaker 2: outside of China, the Middle East being an example. What's 621 00:29:49,320 --> 00:29:52,200 Speaker 2: your position on that and whether the United States wants 622 00:29:52,240 --> 00:29:55,560 Speaker 2: to kind of leave the world open to China to 623 00:29:55,600 --> 00:29:56,800 Speaker 2: sell its own technology. 624 00:29:58,400 --> 00:30:01,960 Speaker 11: Well, there was of you in the previous administration that 625 00:30:02,000 --> 00:30:05,480 Speaker 11: we shouldn't sell chips to many countries, including the resource 626 00:30:05,640 --> 00:30:07,880 Speaker 11: rich Gulf States, and I think that was a major 627 00:30:07,920 --> 00:30:10,800 Speaker 11: mistake because every time you tell a country that they 628 00:30:10,840 --> 00:30:14,240 Speaker 11: can't buy the American tech stack, what's their action going 629 00:30:14,280 --> 00:30:16,440 Speaker 11: to be? They're going to turn to China and adopt 630 00:30:16,480 --> 00:30:20,040 Speaker 11: the Chinese tech stack. I have a very simple metric 631 00:30:20,200 --> 00:30:24,800 Speaker 11: for measuring weatherwoiting the AI race, which is global market share. 632 00:30:25,120 --> 00:30:27,120 Speaker 11: If we look around the world and say five years 633 00:30:27,120 --> 00:30:30,280 Speaker 11: and we see that the American technology stacks has say 634 00:30:30,320 --> 00:30:33,080 Speaker 11: eighty percent market share, that means that we won. But 635 00:30:33,160 --> 00:30:34,920 Speaker 11: if we look around the world in five years and 636 00:30:34,960 --> 00:30:37,320 Speaker 11: we see that the Chinese technology stack and I'm talking 637 00:30:37,320 --> 00:30:40,720 Speaker 11: about Huawei chips and Deepseak models for example, has a 638 00:30:40,720 --> 00:30:43,880 Speaker 11: eighty percent market share, then obviously we lost. By the way, 639 00:30:43,880 --> 00:30:45,680 Speaker 11: that's what happened in five g We don't want to 640 00:30:45,720 --> 00:30:49,040 Speaker 11: repeat of that. So again, the strategy here should be 641 00:30:49,040 --> 00:30:51,160 Speaker 11: for the US to dominate the world and have the 642 00:30:51,200 --> 00:30:54,400 Speaker 11: greatest market share. And I think this is pretty obvious 643 00:30:54,440 --> 00:30:56,720 Speaker 11: to everyone in Silicon Valley because we understand that the 644 00:30:56,720 --> 00:30:59,720 Speaker 11: way to win technology races is to have the biggest 645 00:30:59,720 --> 00:31:02,640 Speaker 11: eco system. If you're a technology platform, you want to 646 00:31:02,640 --> 00:31:04,880 Speaker 11: have the most developers using your API. If you're an 647 00:31:04,920 --> 00:31:07,320 Speaker 11: app store, you want to have the most apps in 648 00:31:07,360 --> 00:31:09,680 Speaker 11: your app store. In a similar way, we want as 649 00:31:09,680 --> 00:31:14,760 Speaker 11: many users on the American technology stack as possible. And 650 00:31:14,040 --> 00:31:18,880 Speaker 11: I find it hard to understand why the previous administration 651 00:31:19,520 --> 00:31:23,760 Speaker 11: would exclude these rich countries from participating on our tech stack. 652 00:31:24,160 --> 00:31:27,240 Speaker 11: It certainly didn't help us in the race with China. 653 00:31:27,360 --> 00:31:31,000 Speaker 11: If anything, President Trump's policy boxes out China from the 654 00:31:31,040 --> 00:31:35,440 Speaker 11: Middle East, whereas the previous administration's policy forced these countries 655 00:31:35,480 --> 00:31:36,520 Speaker 11: into China's arms. 656 00:31:37,200 --> 00:31:42,360 Speaker 4: You've reworked, therefore the diffusion ideals set by the previous administration. 657 00:31:42,920 --> 00:31:46,080 Speaker 3: But when you're allowing only less. 658 00:31:45,840 --> 00:31:48,360 Speaker 4: Powerful chips, I mean, I think the President even called 659 00:31:48,360 --> 00:31:51,880 Speaker 4: them obsolete versions of in Video's chips into China. Does 660 00:31:51,880 --> 00:31:54,560 Speaker 4: that mean that ship of innovation in China's already sailed 661 00:31:54,600 --> 00:31:56,720 Speaker 4: because they need to have the most sophisticated and they're 662 00:31:56,720 --> 00:31:57,760 Speaker 4: going to have to build it themselves. 663 00:31:59,400 --> 00:32:02,480 Speaker 11: Well, but when you're talking about what we export to China, 664 00:32:02,520 --> 00:32:05,640 Speaker 11: that's obviously going to be a very complicated question, and 665 00:32:05,200 --> 00:32:08,880 Speaker 11: there's arguments on both sides. I think there's a pretty 666 00:32:08,880 --> 00:32:12,000 Speaker 11: strong argument for not selling China our latest and greatest 667 00:32:12,040 --> 00:32:15,600 Speaker 11: chips because that would be too beneficial for them. However, 668 00:32:15,680 --> 00:32:17,840 Speaker 11: if you don't sell them anything, then, like you're saying, 669 00:32:18,080 --> 00:32:23,080 Speaker 11: that will accelerate their desire to be independent of the 670 00:32:23,120 --> 00:32:25,400 Speaker 11: American stack. And so I do think there is a 671 00:32:25,440 --> 00:32:29,640 Speaker 11: compelling argument for selling them a let's call it deprecated 672 00:32:29,680 --> 00:32:33,440 Speaker 11: American chip or a less great American chip. And by 673 00:32:33,480 --> 00:32:35,760 Speaker 11: the way, this is why the Biden administration approved the 674 00:32:35,920 --> 00:32:41,000 Speaker 11: H twenty in unlimited quantities for export to China. Now, 675 00:32:41,000 --> 00:32:43,040 Speaker 11: when President Trump came in, he said that those H 676 00:32:43,080 --> 00:32:46,000 Speaker 11: twenty sales have to be licensed and they're subject to 677 00:32:46,040 --> 00:32:49,880 Speaker 11: a fifteen percent surcharge, And nonetheless, all the people who 678 00:32:49,880 --> 00:32:53,160 Speaker 11: approved the Biden policy started attacking President Trump. I think 679 00:32:53,160 --> 00:32:55,440 Speaker 11: this is a classic case of no one had a 680 00:32:55,440 --> 00:32:57,880 Speaker 11: problem with it until President Trump agreed to do it. 681 00:32:59,400 --> 00:33:02,920 Speaker 2: David Jensen Woe went on Brad Gersner's podcast, and there 682 00:33:02,960 --> 00:33:05,240 Speaker 2: was a reaction to what he said. You would point 683 00:33:05,240 --> 00:33:08,040 Speaker 2: out you are an official of the government. Jensen is 684 00:33:08,040 --> 00:33:10,880 Speaker 2: a private citizen and CEO. But at the root of 685 00:33:10,880 --> 00:33:12,760 Speaker 2: what he was saying was also the idea of talent. 686 00:33:13,160 --> 00:33:17,360 Speaker 2: Many talented computer scientists and engineers come from China. I 687 00:33:17,400 --> 00:33:19,640 Speaker 2: think that was the point he was making. But the 688 00:33:19,680 --> 00:33:23,880 Speaker 2: personnel part of this story, where do you stand on that? 689 00:33:25,920 --> 00:33:28,080 Speaker 11: Well, it's true that something like half the world's AI 690 00:33:28,120 --> 00:33:30,960 Speaker 11: researchers are from China, and so I do think that 691 00:33:31,000 --> 00:33:34,520 Speaker 11: we have to somehow be open to working with this 692 00:33:34,600 --> 00:33:37,280 Speaker 11: talent or allowing this talent to use the American chip 693 00:33:37,360 --> 00:33:40,280 Speaker 11: stack to some degree. So it's a complicated question always 694 00:33:40,280 --> 00:33:42,760 Speaker 11: of what you allow China to do. But I think 695 00:33:42,840 --> 00:33:46,360 Speaker 11: Jensen was making a point there about again the talent pipeline, 696 00:33:46,360 --> 00:33:49,440 Speaker 11: and we have to be open to it to some degree. Again, 697 00:33:49,440 --> 00:33:51,200 Speaker 11: I could serder myself to be a China Hawk. But 698 00:33:51,240 --> 00:33:54,600 Speaker 11: I wasn't triggered by what Jensen said because I watched 699 00:33:54,640 --> 00:33:57,800 Speaker 11: the entire two hour podcast, not just the thirty second clip, 700 00:33:57,800 --> 00:34:00,880 Speaker 11: and I understood what he was trying to say in context, 701 00:34:00,960 --> 00:34:03,880 Speaker 11: and it is a nuanced question. And look, let me 702 00:34:03,920 --> 00:34:06,320 Speaker 11: just say to all these people who are criticizing Jensen, 703 00:34:07,040 --> 00:34:08,919 Speaker 11: what have you done to help us win the AI race? 704 00:34:09,000 --> 00:34:11,480 Speaker 11: I can't think of a more strategic asset in the 705 00:34:11,520 --> 00:34:15,040 Speaker 11: I race than in Nvidia and Jensen himself, who for 706 00:34:15,080 --> 00:34:18,760 Speaker 11: thirty years have been working on these GPUs and Jensen 707 00:34:19,040 --> 00:34:22,480 Speaker 11: is a source of huge American advantage in this AI race. 708 00:34:22,480 --> 00:34:24,000 Speaker 11: So let's trow them a little bit of grace here. 709 00:34:24,640 --> 00:34:27,800 Speaker 4: Context is everything. We appreciate you talking about the complicated 710 00:34:27,800 --> 00:34:30,800 Speaker 4: as well. White House, AI and Cryptois are David Sachs great. 711 00:34:30,640 --> 00:34:31,480 Speaker 3: To have you with us. 712 00:34:32,000 --> 00:34:34,880 Speaker 4: Now coming up met his chief marketing officer, Alex Schultz, 713 00:34:34,880 --> 00:34:38,120 Speaker 4: discusses how AI is changing the advertising landscape. Joins us 714 00:34:38,160 --> 00:34:47,640 Speaker 4: next as a Bluebertet. A new book is out click here, 715 00:34:48,000 --> 00:34:51,040 Speaker 4: The Art and Science of digital marketing and advertising. It's 716 00:34:51,040 --> 00:34:53,960 Speaker 4: by Alex Schultz, and he says that AI is going 717 00:34:53,960 --> 00:34:57,480 Speaker 4: to be a game changing for advertisers if used correctly. 718 00:34:57,600 --> 00:34:59,600 Speaker 4: Schultz the chief marketing officer, of course a meta there's 719 00:34:59,600 --> 00:35:02,719 Speaker 4: a thing all too about direct advertising and directing it. 720 00:35:02,840 --> 00:35:06,000 Speaker 4: You also understand entirely the analytics and the raw data 721 00:35:06,040 --> 00:35:06,560 Speaker 4: pine all of this. 722 00:35:06,600 --> 00:35:07,640 Speaker 3: Alex, welcome to the show. 723 00:35:07,960 --> 00:35:10,920 Speaker 4: I'm so interested in how you have seen the unfolding 724 00:35:11,040 --> 00:35:14,080 Speaker 4: over the last decade or two. Incremental growth is something 725 00:35:14,120 --> 00:35:17,080 Speaker 4: that you talk about so much, having direct impact with 726 00:35:17,480 --> 00:35:19,879 Speaker 4: direct targeted ads in many ways, but then you've got 727 00:35:19,880 --> 00:35:22,400 Speaker 4: to think of how a company has got long term vision, 728 00:35:22,440 --> 00:35:24,600 Speaker 4: a north star, as you put it. How are company 729 00:35:24,680 --> 00:35:26,239 Speaker 4: small and large managing to balance. 730 00:35:26,000 --> 00:35:26,560 Speaker 7: This right now? 731 00:35:26,760 --> 00:35:28,200 Speaker 12: Yeah, thank you for having me. This is awesome to 732 00:35:28,239 --> 00:35:30,560 Speaker 12: be here. I mean, look, incremental growth is about being 733 00:35:30,600 --> 00:35:33,440 Speaker 12: incremental to what would have happened without you doing your actions. 734 00:35:33,480 --> 00:35:35,640 Speaker 12: You need to be incremental to what would have happened. 735 00:35:35,719 --> 00:35:37,280 Speaker 12: You need to see what the lift is of your work. 736 00:35:37,520 --> 00:35:40,319 Speaker 12: That fits totally with thinking long term and what is 737 00:35:40,320 --> 00:35:42,640 Speaker 12: your north star? What is the goal you want to achieve? 738 00:35:43,200 --> 00:35:45,759 Speaker 12: That is absolutely a question of like I have a goal. 739 00:35:45,920 --> 00:35:47,799 Speaker 12: How do I incrementally lift that goal? 740 00:35:48,719 --> 00:35:51,480 Speaker 4: Your north star was writing this book. I laid in 741 00:35:51,520 --> 00:35:53,200 Speaker 4: many ways as well as the day to day in 742 00:35:53,239 --> 00:35:56,400 Speaker 4: which you're understanding the analytics underneath what's happening in terms 743 00:35:56,400 --> 00:35:58,200 Speaker 4: of marketing for meta and. 744 00:35:58,040 --> 00:36:01,000 Speaker 3: The growth story. Why how did you put this out here? 745 00:36:01,280 --> 00:36:01,440 Speaker 8: Yeah? 746 00:36:01,719 --> 00:36:03,200 Speaker 12: Well, I mean my north staff of the book is 747 00:36:03,239 --> 00:36:05,560 Speaker 12: to be useful, Like I want small businesses and large 748 00:36:05,560 --> 00:36:07,719 Speaker 12: businesses to be able to actually be better at using the 749 00:36:07,760 --> 00:36:10,800 Speaker 12: tools like you go back Ogilvie on advertising nineteen eighty 750 00:36:10,800 --> 00:36:14,440 Speaker 12: five Bible, Absolutely incredible book, but there are new channels 751 00:36:14,480 --> 00:36:17,239 Speaker 12: that have evolved since then. Digital marketing is seventy five 752 00:36:17,239 --> 00:36:20,040 Speaker 12: percent of the global advertising industry. There's no book for that. 753 00:36:20,120 --> 00:36:21,640 Speaker 12: I kept being asked for a book for that, so 754 00:36:21,680 --> 00:36:23,920 Speaker 12: I wanted to produce a useful book to help people 755 00:36:23,920 --> 00:36:24,480 Speaker 12: be good at that. 756 00:36:26,200 --> 00:36:26,600 Speaker 5: Alex. 757 00:36:26,680 --> 00:36:29,279 Speaker 2: In the back part of the book on AI and 758 00:36:29,320 --> 00:36:31,959 Speaker 2: its impact on marketing, you talk about an audience of one, 759 00:36:32,640 --> 00:36:35,359 Speaker 2: the idea that the AI can make the ad so 760 00:36:35,560 --> 00:36:37,120 Speaker 2: targeted to the individual. 761 00:36:37,680 --> 00:36:38,759 Speaker 5: What are the risks with that? 762 00:36:39,000 --> 00:36:41,800 Speaker 2: You're clearly needing a lot of data about that person 763 00:36:42,320 --> 00:36:44,760 Speaker 2: to make that audience of one relevant ad. 764 00:36:45,400 --> 00:36:48,040 Speaker 12: Yeah, I mean, from my perspective, there are clear Now 765 00:36:48,120 --> 00:36:50,160 Speaker 12: we've reached the point where there are clear laws out there. 766 00:36:50,320 --> 00:36:53,160 Speaker 12: There are clear platform policies out there about purpose use, 767 00:36:53,239 --> 00:36:55,239 Speaker 12: limitation of data, and so it's very clear what you 768 00:36:55,239 --> 00:36:58,360 Speaker 12: can use the data for. And people like personalized ads, 769 00:36:58,520 --> 00:37:01,960 Speaker 12: so people have control. You can get young people really 770 00:37:02,000 --> 00:37:05,040 Speaker 12: understand how to manage their algorithm, and they love the 771 00:37:05,080 --> 00:37:07,480 Speaker 12: personalized ads that they get where there's something relevant that's 772 00:37:07,480 --> 00:37:09,800 Speaker 12: actually useful for them. So I think the balance is 773 00:37:09,840 --> 00:37:11,560 Speaker 12: actually in a way better place than it was, say 774 00:37:11,600 --> 00:37:14,080 Speaker 12: ten years ago, in terms of people's understanding of these things. 775 00:37:14,239 --> 00:37:16,239 Speaker 12: So I think the risks are much much lower, and 776 00:37:16,280 --> 00:37:18,680 Speaker 12: I think the biggest risk is not taking advantage of it. 777 00:37:18,760 --> 00:37:20,440 Speaker 12: You look where Europe is, You look at how they 778 00:37:20,440 --> 00:37:23,400 Speaker 12: regret the current regulations, You look at the draggy report. 779 00:37:23,640 --> 00:37:26,840 Speaker 12: They are regretting the lack of growth they have versus 780 00:37:26,840 --> 00:37:28,400 Speaker 12: the growth the US has, and I think one big 781 00:37:28,480 --> 00:37:30,600 Speaker 12: part of that is they've hurt advertising. 782 00:37:32,200 --> 00:37:35,120 Speaker 2: The big picture with meta has been that all of 783 00:37:35,120 --> 00:37:37,360 Speaker 2: the work in AI has paid off for like the 784 00:37:37,360 --> 00:37:39,879 Speaker 2: bread and butter core business. Can you give us any 785 00:37:39,880 --> 00:37:43,640 Speaker 2: sort of insight into how much more valuable an AI 786 00:37:43,840 --> 00:37:46,960 Speaker 2: powered ad is what the conversion on it is relative 787 00:37:47,000 --> 00:37:50,920 Speaker 2: to traditional advertising, how it drives growth on the top line. 788 00:37:51,120 --> 00:37:53,320 Speaker 12: Yeah, I mean you can step back completely. Our business 789 00:37:53,320 --> 00:37:57,200 Speaker 12: has been completely transformed by AI. When TikTok came along, 790 00:37:57,440 --> 00:38:00,279 Speaker 12: like five years ago, everything that you were looking out 791 00:38:00,320 --> 00:38:03,399 Speaker 12: on Facebook and Instagram was connected content. You joined, you'd liked, 792 00:38:03,400 --> 00:38:06,880 Speaker 12: you'd friended, you'd followed. Today, the majority of Instagram and 793 00:38:07,000 --> 00:38:10,360 Speaker 12: Facebook are unconnected content. And you can see in our 794 00:38:10,440 --> 00:38:14,000 Speaker 12: earnings results what the impact of AI allowing you to 795 00:38:14,080 --> 00:38:17,840 Speaker 12: rank content based on semantic understanding of content and semantic 796 00:38:17,920 --> 00:38:21,640 Speaker 12: understanding of the individual does for engagement. And then you 797 00:38:21,680 --> 00:38:23,520 Speaker 12: see it too on the ads in terms of the 798 00:38:23,520 --> 00:38:26,520 Speaker 12: revenue that's coming through and the results were providing for advertisers. 799 00:38:26,520 --> 00:38:29,840 Speaker 12: So it's very high double digit percentage uplifts if you 800 00:38:29,920 --> 00:38:33,680 Speaker 12: adopt things like advantage plus shopping campaigns, make sure you 801 00:38:33,680 --> 00:38:36,279 Speaker 12: feed the data through with cappy and do the basics right, 802 00:38:36,320 --> 00:38:39,279 Speaker 12: which I describe in this book, both with meta but 803 00:38:39,320 --> 00:38:41,120 Speaker 12: also using anyone else's tools. 804 00:38:41,760 --> 00:38:45,680 Speaker 4: What about human creativity at this moment, Yeah, one of 805 00:38:45,680 --> 00:38:48,600 Speaker 4: my struggles is creativity is always about pixels and look, 806 00:38:48,640 --> 00:38:50,280 Speaker 4: I love my creative seemed they're incredible. 807 00:38:50,280 --> 00:38:52,560 Speaker 12: We closed Fifth Avenue. We had Lewis Hamilton do donuts 808 00:38:52,560 --> 00:38:55,160 Speaker 12: on Fifth Avenue. It was amazing. Things were great and 809 00:38:55,239 --> 00:38:57,960 Speaker 12: it really really worked. But there's a ton of creativity 810 00:38:58,000 --> 00:39:00,279 Speaker 12: with data. There's a ton of creativity with target. There's 811 00:39:00,280 --> 00:39:03,719 Speaker 12: a ton of creativity with conversion rate optimization and flow optimization. 812 00:39:04,000 --> 00:39:06,160 Speaker 12: And what I get at in this book is creativity 813 00:39:06,239 --> 00:39:09,200 Speaker 12: in the factors where it isn't David Ogilvy saying stick 814 00:39:09,239 --> 00:39:11,600 Speaker 12: a car to a billboard with super glue to sell superglue. 815 00:39:11,760 --> 00:39:14,160 Speaker 12: It's creativity and how you use data to give people 816 00:39:14,239 --> 00:39:18,320 Speaker 12: amazing personalized experiences and high conversion rate flows. 817 00:39:19,719 --> 00:39:22,960 Speaker 2: Alex Schultz, chief marketing officer for Meta, thank you very 818 00:39:23,080 --> 00:39:25,920 Speaker 2: much and tune in to Bloomberg's screen Time on Thursday 819 00:39:26,239 --> 00:39:28,680 Speaker 2: for a deep dive on the creator economy with the 820 00:39:28,719 --> 00:39:31,560 Speaker 2: head of Meta's Instagram Adam Missi. 821 00:39:31,640 --> 00:39:32,960 Speaker 5: That's what we're really looking forward to. 822 00:39:33,120 --> 00:39:34,640 Speaker 2: Okay, coming up, we're going to get back to the 823 00:39:34,680 --> 00:39:37,440 Speaker 2: deal between AMD and open Ai and what that means 824 00:39:37,680 --> 00:39:41,399 Speaker 2: for concerns around circular financing, the build out of AI infrastructure, 825 00:39:41,719 --> 00:39:42,479 Speaker 2: and everything else. 826 00:39:42,640 --> 00:39:43,560 Speaker 5: This is Bloomberg Tech. 827 00:39:54,480 --> 00:39:56,040 Speaker 4: We'm going to get back to open AI and am 828 00:39:56,360 --> 00:39:57,880 Speaker 4: the deal pushing broader. 829 00:39:57,640 --> 00:40:00,000 Speaker 3: Markets high, certainly pushing it AMD high. 830 00:40:00,120 --> 00:40:04,440 Speaker 4: Joining Usskaya is with US Swiss quote senior market analyst IBEC. 831 00:40:05,000 --> 00:40:09,920 Speaker 4: These ongoing deals between chip designers and ultimately those that 832 00:40:09,960 --> 00:40:11,160 Speaker 4: are perching the GPUs. 833 00:40:11,320 --> 00:40:12,040 Speaker 3: What do you make of it? 834 00:40:13,480 --> 00:40:16,879 Speaker 9: Well, actually, it's very amazing, especially for those who are 835 00:40:17,000 --> 00:40:20,160 Speaker 9: questioning the circular nature of the business and the deals 836 00:40:20,160 --> 00:40:22,759 Speaker 9: that are being announced right now but infinite when. And 837 00:40:22,760 --> 00:40:25,320 Speaker 9: we'll look at the MD and open Ai deal today. 838 00:40:25,400 --> 00:40:29,080 Speaker 9: While open Ai is outstriking deals with data centers and 839 00:40:29,200 --> 00:40:31,880 Speaker 9: chip makers from inside the US and outside of the 840 00:40:31,960 --> 00:40:34,239 Speaker 9: US in order to stay ahead of the game and 841 00:40:34,360 --> 00:40:37,920 Speaker 9: make sure that they are not constrained by capacity constraints. 842 00:40:37,920 --> 00:40:40,839 Speaker 9: So that basically means that this company today is more 843 00:40:40,840 --> 00:40:44,560 Speaker 9: worried about not having enough supply for the huge demand 844 00:40:44,560 --> 00:40:47,880 Speaker 9: than the contrary. And that's outright positive for the aiballs and. 845 00:40:47,920 --> 00:40:49,680 Speaker 13: The sentiment here and for AMD. 846 00:40:49,800 --> 00:40:51,799 Speaker 9: There is nothing else to say that the jackpot their 847 00:40:51,880 --> 00:40:53,320 Speaker 9: moment looks like it has come. 848 00:40:53,640 --> 00:40:56,440 Speaker 4: Yeah, certainly biggest move in nine years Epech. What's interesting 849 00:40:56,480 --> 00:40:58,319 Speaker 4: is in video did sync pulled back a little bit 850 00:40:58,960 --> 00:41:01,000 Speaker 4: after it had risen higher on the hopes that on 851 00:41:01,040 --> 00:41:02,680 Speaker 4: hire and the server demand was clearly there. 852 00:41:02,680 --> 00:41:04,319 Speaker 3: But then we question market share. 853 00:41:04,360 --> 00:41:06,120 Speaker 4: I mean, is there anything that gives you any anxiety 854 00:41:06,160 --> 00:41:07,240 Speaker 4: about videos dominance? 855 00:41:08,080 --> 00:41:11,239 Speaker 9: No, absolutely not, and that's due to the context. As 856 00:41:11,239 --> 00:41:14,279 Speaker 9: you always say, context is everything. The sequence on which 857 00:41:14,280 --> 00:41:18,000 Speaker 9: we are receiving the information has been very insightful in 858 00:41:18,080 --> 00:41:21,120 Speaker 9: what's coming. Actually, NVDA announced last week that they would 859 00:41:21,120 --> 00:41:23,920 Speaker 9: be investing up to one hundred billion US dollars in 860 00:41:24,040 --> 00:41:26,640 Speaker 9: open Ai. That is announcing today that they will be 861 00:41:26,719 --> 00:41:29,319 Speaker 9: taking a ten percent's sake in AMD. So there is 862 00:41:29,320 --> 00:41:32,920 Speaker 9: a circularity there that suggests that m viderships are not 863 00:41:32,960 --> 00:41:35,799 Speaker 9: going to replace in the context of this open Ai deal, 864 00:41:35,880 --> 00:41:39,560 Speaker 9: but rather complemented now beyond the still this could give 865 00:41:39,640 --> 00:41:42,880 Speaker 9: some leverage to AMD, but it looks like the companies 866 00:41:42,920 --> 00:41:45,680 Speaker 9: are considering today that the AIPI is big enough to 867 00:41:45,680 --> 00:41:47,280 Speaker 9: feed everyone grandly. 868 00:41:48,800 --> 00:41:51,480 Speaker 5: Epek for public market investors. 869 00:41:52,600 --> 00:41:56,520 Speaker 2: Is open Ai the private company becoming some kind of 870 00:41:56,600 --> 00:41:59,560 Speaker 2: macro level factor that they have to model in. 871 00:42:01,200 --> 00:42:03,799 Speaker 9: Well, absolutely, I mean they are so huge now, they 872 00:42:03,800 --> 00:42:06,080 Speaker 9: are the biggest startup in the world. 873 00:42:06,120 --> 00:42:07,640 Speaker 13: They are worth five hundred. 874 00:42:07,360 --> 00:42:09,920 Speaker 9: Billion US dollars and d do you have these huge 875 00:42:09,960 --> 00:42:14,080 Speaker 9: deals with publicly traded companies, and those are the market 876 00:42:14,360 --> 00:42:19,000 Speaker 9: moving deals of market moving stocks. We are talking about NVIDIAs, 877 00:42:19,080 --> 00:42:22,719 Speaker 9: we are talking about AMD So definitely, open ai is 878 00:42:23,160 --> 00:42:26,759 Speaker 9: pretty much the center of this AI revolution. It has 879 00:42:26,880 --> 00:42:30,000 Speaker 9: been the well the starting point, and it is gaining 880 00:42:30,040 --> 00:42:34,600 Speaker 9: importance every single day, and with each deal that they announced, 881 00:42:34,640 --> 00:42:37,920 Speaker 9: they are actually securing their position at the center of 882 00:42:37,960 --> 00:42:38,439 Speaker 9: this game. 883 00:42:39,360 --> 00:42:41,759 Speaker 2: Should we be asking more questions about how open ai 884 00:42:41,840 --> 00:42:43,200 Speaker 2: is going to pay for all of this, then. 885 00:42:44,719 --> 00:42:47,839 Speaker 9: Well yes, I mean some automol recently say that they 886 00:42:47,880 --> 00:42:51,400 Speaker 9: will be looking at some funding possibilities that he didn't 887 00:42:51,400 --> 00:42:52,520 Speaker 9: give details about. 888 00:42:52,840 --> 00:42:55,680 Speaker 13: But one of the questions here is. 889 00:42:55,600 --> 00:43:00,160 Speaker 9: That ai is actually a very capital intensive place, so 890 00:43:00,200 --> 00:43:04,480 Speaker 9: this company needs funding now. Nvidia coming to the rescue 891 00:43:04,719 --> 00:43:08,360 Speaker 9: could open the way for other companies also looking to 892 00:43:08,400 --> 00:43:11,000 Speaker 9: help open ai and take a sake in this company 893 00:43:11,000 --> 00:43:13,719 Speaker 9: that has not yet gone public. So I think that 894 00:43:13,840 --> 00:43:16,560 Speaker 9: open Ai, if anything, is not going to really having 895 00:43:16,640 --> 00:43:19,080 Speaker 9: any funding problems at the stage of the game because 896 00:43:19,120 --> 00:43:23,080 Speaker 9: they are dominating the AI business right now, and even 897 00:43:23,120 --> 00:43:26,560 Speaker 9: though it was it's it's a private company, it does 898 00:43:26,680 --> 00:43:28,879 Speaker 9: have all the fundings that it needs. I think from 899 00:43:28,920 --> 00:43:33,319 Speaker 9: private and public investors are just eager to take take 900 00:43:33,440 --> 00:43:36,080 Speaker 9: part of this company actually because. 901 00:43:35,840 --> 00:43:38,560 Speaker 3: The growth story is so clear in terms of revenue. 902 00:43:38,600 --> 00:43:41,680 Speaker 4: But when you compare thirteen billion dollars in revenue per 903 00:43:41,760 --> 00:43:43,799 Speaker 4: year and then you're thinking the trillions and dollars it 904 00:43:43,840 --> 00:43:46,680 Speaker 4: has to spend. What gives you comfort that that revenue 905 00:43:46,760 --> 00:43:50,560 Speaker 4: will match the amount that ultimately needs to be financed on. 906 00:43:50,520 --> 00:43:54,960 Speaker 9: It, Well, it's it's definitely the reach that they have. 907 00:43:55,560 --> 00:43:58,840 Speaker 9: We have seen over the past three years other companies 908 00:43:58,920 --> 00:44:02,920 Speaker 9: like Meta or a Chinese company's aside, but many companies 909 00:44:03,000 --> 00:44:07,720 Speaker 9: right now perplexity met her. We have x for example, 910 00:44:07,760 --> 00:44:10,480 Speaker 9: wet Grog and other models that are trying to compete 911 00:44:10,560 --> 00:44:13,960 Speaker 9: open Ai, and so far they have not been successful 912 00:44:14,040 --> 00:44:17,240 Speaker 9: in meeting the level of enthusiasm. 913 00:44:16,640 --> 00:44:19,759 Speaker 13: That open Ai had so far. So open Ai is 914 00:44:19,800 --> 00:44:20,680 Speaker 13: actually surfing. 915 00:44:20,880 --> 00:44:23,040 Speaker 9: It has been to first to come in and it 916 00:44:23,160 --> 00:44:25,920 Speaker 9: still has this popularity and this leverage of being the 917 00:44:25,960 --> 00:44:28,920 Speaker 9: first comer. So I think that they do have a 918 00:44:29,120 --> 00:44:32,520 Speaker 9: very good leverage in expanding their business and making more 919 00:44:32,600 --> 00:44:33,680 Speaker 9: revenue in the future. 920 00:44:33,719 --> 00:44:35,399 Speaker 13: They just need to play the game right. 921 00:44:35,640 --> 00:44:37,960 Speaker 9: They just need to find the right partners and that's 922 00:44:38,040 --> 00:44:39,760 Speaker 9: exactly what they are doing right now. 923 00:44:41,000 --> 00:44:43,520 Speaker 2: Eupe Oscar Deskaya Swiss. Quite great to have you back 924 00:44:43,560 --> 00:44:45,960 Speaker 2: on the program. Appreciate it a lot more a program 925 00:44:45,960 --> 00:44:49,680 Speaker 2: it was that does it for this edition of Bloomberg Tech, Caro, 926 00:44:50,320 --> 00:44:53,080 Speaker 2: tune in later today we have more news due to 927 00:44:53,080 --> 00:44:54,640 Speaker 2: come from Open AI. We're going to speak with the 928 00:44:54,680 --> 00:44:58,080 Speaker 2: CEO Brad Lightcap on the sidelines of Open AIS. 929 00:44:58,160 --> 00:44:58,919 Speaker 5: Develop a day. 930 00:44:59,480 --> 00:45:02,440 Speaker 4: Gosh a lot going, can't stop, won't stop, ed you 931 00:45:02,480 --> 00:45:04,719 Speaker 4: know it for us, we so appreciate it. Meanwhile, check 932 00:45:04,719 --> 00:45:06,439 Speaker 4: out our podcast why don't you There are so many 933 00:45:06,440 --> 00:45:09,000 Speaker 4: conversations you've got to dial back in. Get into Lisa Sue, 934 00:45:09,080 --> 00:45:12,000 Speaker 4: get into Greg Brotman, get into David Sachs and of 935 00:45:12,000 --> 00:45:13,400 Speaker 4: course Alex Schultz. 936 00:45:13,040 --> 00:45:16,360 Speaker 3: Of Meta as well all online. This is a Bloomberg 937 00:45:16,400 --> 00:45:16,600 Speaker 3: Tech