1 00:00:01,880 --> 00:00:07,160 Speaker 1: From Bahard where Innovation, Money and power Collie in Silicon Valley, NBN. 2 00:00:07,480 --> 00:00:11,520 Speaker 1: This is Bloomberg Technology with Caroline Hyde and Ed Luod Love. 3 00:00:25,400 --> 00:00:27,880 Speaker 2: I'm Caroline Heide of Bloomberg's world headquarters in New York, 4 00:00:28,360 --> 00:00:30,000 Speaker 2: and I'm Met Ludlow in San Francisco. 5 00:00:30,160 --> 00:00:31,800 Speaker 3: This is Bloomberg Technology. 6 00:00:32,000 --> 00:00:35,080 Speaker 2: Coming up. We'll give you the key takeaways from Apple's 7 00:00:35,120 --> 00:00:39,160 Speaker 2: scary fast unveil, which featured new chips, laptops, so much more. 8 00:00:40,080 --> 00:00:42,600 Speaker 4: Plus we'll speak to an AI advisor to the White 9 00:00:42,640 --> 00:00:45,440 Speaker 4: House as President Biden takes the most significant step to 10 00:00:45,560 --> 00:00:48,960 Speaker 4: date in regulating artificial intelligence technology. 11 00:00:49,120 --> 00:00:52,320 Speaker 2: Meanwhile, in Vidia, it dips below one trillion dollars today 12 00:00:52,400 --> 00:00:55,640 Speaker 2: as concerns linger over US curbs on China and how 13 00:00:55,720 --> 00:00:58,960 Speaker 2: that may impact billions of dollars of existing orders. We'll 14 00:00:58,960 --> 00:01:01,440 Speaker 2: have that and so much more, but first let's just 15 00:01:01,480 --> 00:01:03,360 Speaker 2: check on on these marks. It is the end of 16 00:01:03,360 --> 00:01:05,959 Speaker 2: the month. We all know there's plenty of candy being 17 00:01:06,000 --> 00:01:07,360 Speaker 2: eyed for the day, but it is the end of 18 00:01:07,400 --> 00:01:09,679 Speaker 2: October and it was down, down, down, if you see 19 00:01:09,760 --> 00:01:12,120 Speaker 2: for the Really the Nasdaq currently off by some three 20 00:01:12,120 --> 00:01:14,000 Speaker 2: point three percent end the month, but it's the third 21 00:01:14,040 --> 00:01:16,160 Speaker 2: straight months of losses as we see for some of 22 00:01:16,160 --> 00:01:18,520 Speaker 2: the key big benchmarks across the US at the moment 23 00:01:18,680 --> 00:01:21,720 Speaker 2: ed and that as we worry about earnings, we worry 24 00:01:21,760 --> 00:01:24,640 Speaker 2: about of course geopolitics as well. This is the worst 25 00:01:24,680 --> 00:01:27,280 Speaker 2: therefore three month loss set of three months or so 26 00:01:27,360 --> 00:01:30,959 Speaker 2: that we've seen since back in what June twenty twenty two. Overall, 27 00:01:31,000 --> 00:01:33,480 Speaker 2: though this isn't as bad as September was. That's moving 28 00:01:33,520 --> 00:01:35,240 Speaker 2: on to some of the other benchmarks that we're keeping 29 00:01:35,240 --> 00:01:36,760 Speaker 2: an eye on. I'm looking the Nasdaq one hundred on 30 00:01:36,800 --> 00:01:38,520 Speaker 2: the day basically flat, which is pushing off of those 31 00:01:38,520 --> 00:01:40,760 Speaker 2: loads for the day, but the tenure yield is coming 32 00:01:40,760 --> 00:01:42,240 Speaker 2: down a little bit, so we've got a little bit 33 00:01:42,240 --> 00:01:44,280 Speaker 2: of buying head of the all important FED meeting that 34 00:01:44,280 --> 00:01:46,919 Speaker 2: comes tomorrow, and of course how much the Treasury Department 35 00:01:46,920 --> 00:01:48,800 Speaker 2: will actually be issuing a new debt. We've got a 36 00:01:48,800 --> 00:01:50,360 Speaker 2: little bit of a bid, but we're seeing the dollar 37 00:01:50,760 --> 00:01:53,280 Speaker 2: having a big bid against the Japanese end. This is 38 00:01:53,280 --> 00:01:55,280 Speaker 2: the key macro story of the day, really, the fact 39 00:01:55,280 --> 00:01:57,360 Speaker 2: that the back of Japan with Japan was not going 40 00:01:57,400 --> 00:01:59,560 Speaker 2: to be changing any of those yield curve controls, but 41 00:02:00,080 --> 00:02:03,320 Speaker 2: into the micro particularly the tech focus micro ed Yeah. 42 00:02:03,120 --> 00:02:04,360 Speaker 3: And earnings is still a story. 43 00:02:04,440 --> 00:02:07,200 Speaker 4: Pinterest a real surprise up more than eighteen percent, on 44 00:02:07,280 --> 00:02:10,359 Speaker 4: track for its biggest jump since August of twenty twenty two. 45 00:02:10,680 --> 00:02:13,240 Speaker 4: Top and bottom line beat, but a record monthly active 46 00:02:13,639 --> 00:02:16,200 Speaker 4: user base. We will get a key analyst conversation later 47 00:02:16,280 --> 00:02:19,239 Speaker 4: in the show on that name. Tesla's moving to the 48 00:02:19,320 --> 00:02:21,720 Speaker 4: upside modestly four tens and one percent, but it's kind 49 00:02:21,720 --> 00:02:25,400 Speaker 4: of rebounding from this slump since October eighteenth, and earnings 50 00:02:25,400 --> 00:02:27,440 Speaker 4: where it's down more than twenty percent. There's a lot 51 00:02:27,440 --> 00:02:31,359 Speaker 4: of concern about interest rates and how that's impacting EV demand. 52 00:02:31,360 --> 00:02:33,560 Speaker 4: We will have that conversation later in the program. And 53 00:02:33,600 --> 00:02:35,600 Speaker 4: on Semi we had the CEO on the show twenty 54 00:02:35,600 --> 00:02:39,360 Speaker 4: four hours ago, a tepid forecast for the fourth quarter, 55 00:02:39,760 --> 00:02:43,160 Speaker 4: a mystery OEM customer they wouldn't name that dragged down 56 00:02:43,200 --> 00:02:45,880 Speaker 4: results and the stock a second day of pressure. 57 00:02:45,919 --> 00:02:46,160 Speaker 1: There. 58 00:02:46,520 --> 00:02:48,720 Speaker 4: The key name that we're looking keeping it real this 59 00:02:48,800 --> 00:02:50,240 Speaker 4: Halloween is Apple. 60 00:02:50,360 --> 00:02:51,720 Speaker 3: They're scary fast event. 61 00:02:51,760 --> 00:02:54,320 Speaker 4: We got what we expected because of Bloomberg's Mark German 62 00:02:54,360 --> 00:02:57,239 Speaker 4: in his reporting. And I'm at a trio of Matt 63 00:02:57,240 --> 00:03:01,360 Speaker 4: book pros and the first volume may curve PCs with 64 00:03:01,480 --> 00:03:05,000 Speaker 4: a three nanometer processor, the M three, M three Pro 65 00:03:05,080 --> 00:03:06,359 Speaker 4: and M three Max chip. 66 00:03:06,680 --> 00:03:07,600 Speaker 3: What does this mean? 67 00:03:07,840 --> 00:03:10,440 Speaker 4: Does it put the mac back on track? We're joining 68 00:03:10,480 --> 00:03:14,320 Speaker 4: us now is Carolina Milianacy Creative Strategies, President and principle, 69 00:03:14,360 --> 00:03:17,440 Speaker 4: And listen, you just heard me Carolina. Does this put 70 00:03:17,480 --> 00:03:19,040 Speaker 4: the MacBook back on track? 71 00:03:20,320 --> 00:03:24,240 Speaker 5: I think it certainly puts Mac and ARM back on track. 72 00:03:24,440 --> 00:03:28,440 Speaker 5: I think coming out of the snap Dragon qualcom event 73 00:03:28,800 --> 00:03:31,800 Speaker 5: last week and seeing what they're able to do with 74 00:03:31,880 --> 00:03:35,520 Speaker 5: their platform and now having Apple with the N three 75 00:03:35,720 --> 00:03:39,600 Speaker 5: clearly puts the ARM architecture as a whole in a 76 00:03:39,680 --> 00:03:43,280 Speaker 5: much better place compared to Intel an X eighty six. 77 00:03:43,480 --> 00:03:47,280 Speaker 5: So I think that's what we're looking at, the excitement 78 00:03:47,400 --> 00:03:50,560 Speaker 5: around what this opportunity brings to the PC market. 79 00:03:50,800 --> 00:03:52,240 Speaker 2: What did you make at the price points? 80 00:03:52,320 --> 00:03:56,160 Speaker 5: Krolena, It was very interesting to see that the entry 81 00:03:56,280 --> 00:04:00,200 Speaker 5: level MacBook Pro was dropped in price visits. The one that, 82 00:04:00,720 --> 00:04:04,280 Speaker 5: together with the Mac Bugere as being the volume maker 83 00:04:04,680 --> 00:04:07,360 Speaker 5: and one that now at a lower price of one 84 00:04:07,440 --> 00:04:11,120 Speaker 5: five ninety nine starting point is going to really worry 85 00:04:11,200 --> 00:04:15,600 Speaker 5: the pcoems. You know the power that Apple has to 86 00:04:15,680 --> 00:04:19,520 Speaker 5: convince people that now that they own every from the 87 00:04:19,560 --> 00:04:22,120 Speaker 5: silicon to the harder to the software, and Tim Cook 88 00:04:22,200 --> 00:04:27,040 Speaker 5: pointed to that in the event yesterday, is really reassuring 89 00:04:27,279 --> 00:04:30,800 Speaker 5: customers that basically they're buying something that is in full 90 00:04:30,839 --> 00:04:35,000 Speaker 5: control of Apple and not like before at the mercy 91 00:04:35,040 --> 00:04:37,800 Speaker 5: of the Intel silicon, it. 92 00:04:37,839 --> 00:04:41,920 Speaker 2: Really was Apple silicon that was a selling point here, Carolina, 93 00:04:42,080 --> 00:04:45,279 Speaker 2: is that enough to drive significant well desire to be 94 00:04:45,320 --> 00:04:48,040 Speaker 2: buying in particularly as we had towards the holidays, You think. 95 00:04:48,880 --> 00:04:51,000 Speaker 5: Well, I think if the silicon doesn't convince you, the 96 00:04:51,040 --> 00:04:54,960 Speaker 5: black collar will was a lot of assignment on the 97 00:04:55,480 --> 00:04:59,400 Speaker 5: black color for the higher end s Q of the 98 00:05:00,200 --> 00:05:04,000 Speaker 5: book pro and I think that where Apple is pointing 99 00:05:04,240 --> 00:05:09,880 Speaker 5: is really upgrading current Intel based Mac owners and then 100 00:05:10,200 --> 00:05:15,520 Speaker 5: m one owners to really get into the newest and 101 00:05:15,760 --> 00:05:22,039 Speaker 5: fastest of their silicon product based. But you know, from 102 00:05:23,120 --> 00:05:26,680 Speaker 5: a holiday perspective, we are still in a tough economic 103 00:05:26,760 --> 00:05:31,799 Speaker 5: environment and that needs to be reconciled with We're looking 104 00:05:31,839 --> 00:05:34,960 Speaker 5: at also enterprise, not only consumers for Apple though, and 105 00:05:35,040 --> 00:05:38,920 Speaker 5: so there's definitely more opportunity there as we get into 106 00:05:38,960 --> 00:05:39,520 Speaker 5: the new year. 107 00:05:40,880 --> 00:05:45,440 Speaker 4: By the way, Caroline and Carolina, the entire event, according 108 00:05:45,480 --> 00:05:48,520 Speaker 4: to a blog post that Apple posted six minutes ago, 109 00:05:49,040 --> 00:05:52,720 Speaker 4: was shot on an i phone fifteen Promax the entire thing, 110 00:05:53,200 --> 00:05:55,080 Speaker 4: and it was edited on Max, So they're trying to 111 00:05:55,080 --> 00:05:59,000 Speaker 4: flex their product line across the board. What I think 112 00:05:59,080 --> 00:06:02,640 Speaker 4: is interesting, Carolina, is how they compare the performance of 113 00:06:02,760 --> 00:06:04,840 Speaker 4: M three and there's kind of an arc, right, M three, 114 00:06:04,960 --> 00:06:08,479 Speaker 4: M three Pro, M three Max relative to M one, 115 00:06:08,640 --> 00:06:11,720 Speaker 4: but also Intel, right, the top tier MacBook Pro they 116 00:06:11,720 --> 00:06:15,680 Speaker 4: said was eleven times faster than the most powerful Intel 117 00:06:15,800 --> 00:06:16,400 Speaker 4: Power PC. 118 00:06:17,160 --> 00:06:18,719 Speaker 3: So who are they going after here? 119 00:06:19,400 --> 00:06:19,560 Speaker 6: Right? 120 00:06:19,600 --> 00:06:23,160 Speaker 4: Are they going after just the high end PC market Intel, 121 00:06:23,839 --> 00:06:26,000 Speaker 4: you know, users that want an upgrade, or are they 122 00:06:26,040 --> 00:06:28,560 Speaker 4: just trying to keep refreshing those existing install base of 123 00:06:28,600 --> 00:06:29,400 Speaker 4: Apple devices. 124 00:06:30,520 --> 00:06:33,760 Speaker 5: Both is really trying to capture market share in the 125 00:06:33,839 --> 00:06:37,520 Speaker 5: higher end of a PC market, getting as many people 126 00:06:37,520 --> 00:06:41,360 Speaker 5: as possible off of Intel so they can experience the 127 00:06:41,440 --> 00:06:47,760 Speaker 5: higher seamless workflows that comes from owning all the pieces 128 00:06:47,760 --> 00:06:52,360 Speaker 5: of the equation and continue to vent upgrade from there. 129 00:06:53,480 --> 00:06:55,640 Speaker 5: I do think that, as you said, there's a lot 130 00:06:55,680 --> 00:06:59,560 Speaker 5: of flex there on. They had actually footnotes at the 131 00:06:59,720 --> 00:07:03,320 Speaker 5: end of the event about being shot on iPhone and 132 00:07:04,160 --> 00:07:08,240 Speaker 5: editor on Max, so they are really talking to creators. 133 00:07:08,560 --> 00:07:13,400 Speaker 5: By the same time, without ever mentioning jen Ai. They 134 00:07:13,480 --> 00:07:19,440 Speaker 5: were talking to AI developers two very high demanding workflows, 135 00:07:19,480 --> 00:07:22,960 Speaker 5: so there's definitely an attention there to where the conversation 136 00:07:23,160 --> 00:07:23,680 Speaker 5: is today. 137 00:07:25,120 --> 00:07:25,600 Speaker 3: Carolyn E. 138 00:07:25,640 --> 00:07:28,680 Speaker 4: MILNECI of Creative Strategies. Great to catch up here on 139 00:07:28,720 --> 00:07:29,800 Speaker 4: Bloomberg Technology. 140 00:07:29,800 --> 00:07:30,120 Speaker 3: Thank you. 141 00:07:30,160 --> 00:07:33,160 Speaker 4: Another Apple story that we've been tracking, the company could 142 00:07:33,160 --> 00:07:36,000 Speaker 4: be forced to scale back its app store fees for 143 00:07:36,240 --> 00:07:39,720 Speaker 4: developers after one of the EU's antitrust watchdogs said its 144 00:07:39,800 --> 00:07:43,720 Speaker 4: commissions violate the block's rules. The Dutch Authority for Consumers 145 00:07:43,720 --> 00:07:47,320 Speaker 4: and Market's ruled that Apple's commission on certain app subscriptions 146 00:07:47,520 --> 00:07:50,200 Speaker 4: are an abuse of the company's market power. 147 00:07:50,280 --> 00:07:54,120 Speaker 2: Caroc ongoing with the Dutch agency. Meanwhile, let's talk about 148 00:07:54,120 --> 00:07:58,080 Speaker 2: pinterest shares absolutely rallying today after reporting sales that men 149 00:07:58,120 --> 00:08:00,680 Speaker 2: expectations and mid of course, a push by the company 150 00:08:00,680 --> 00:08:03,840 Speaker 2: to make the platform basically a more shoppable. Tom Forte 151 00:08:03,960 --> 00:08:07,240 Speaker 2: da Davison, senior research analyst is with us and I mean, boy, 152 00:08:07,680 --> 00:08:10,560 Speaker 2: you had a by rating you reierated. This is the 153 00:08:10,560 --> 00:08:12,440 Speaker 2: best day for the stock since three years ago. 154 00:08:12,480 --> 00:08:16,920 Speaker 7: Tom, Yeah, they'd certainly delivered the treats that Pinterest Halloween. 155 00:08:17,320 --> 00:08:20,320 Speaker 7: So what you're seeing is the company's making it easier 156 00:08:20,360 --> 00:08:23,520 Speaker 7: for advertisers to prove that they're able to get conversion 157 00:08:23,800 --> 00:08:27,000 Speaker 7: or sales off of their platform. This is having a 158 00:08:27,080 --> 00:08:30,760 Speaker 7: huge positive impact to their revenue. You're seeing accelerating revenue 159 00:08:30,840 --> 00:08:33,960 Speaker 7: top line growth, and at the same time, like other 160 00:08:34,040 --> 00:08:37,200 Speaker 7: big tech companies, they have very much you know, belt 161 00:08:37,240 --> 00:08:41,880 Speaker 7: tightening going on managing expenses. So the combination of accelerating 162 00:08:41,920 --> 00:08:46,840 Speaker 7: revenue growth with belt tightening is tremendous revenue and tremendous earnings. 163 00:08:48,240 --> 00:08:51,480 Speaker 4: What was interesting Tom was monthly active uses hit a record, 164 00:08:51,720 --> 00:08:53,680 Speaker 4: or another way of looking at it is they eclipse 165 00:08:53,760 --> 00:08:57,600 Speaker 4: their pandemic peage. What was the more important metric for 166 00:08:57,679 --> 00:09:00,679 Speaker 4: you that monthly active use number or the fineinantals where 167 00:09:00,720 --> 00:09:02,600 Speaker 4: they beat top and bottom line. 168 00:09:03,280 --> 00:09:05,800 Speaker 7: What's very important on the monthly active users is they're 169 00:09:05,920 --> 00:09:10,480 Speaker 7: generating gen Z interest. So when you think about having 170 00:09:10,679 --> 00:09:13,360 Speaker 7: more engagement than they had in the early stages of 171 00:09:13,360 --> 00:09:16,760 Speaker 7: the pandemic, that's wonderful. How are they able to do 172 00:09:16,840 --> 00:09:20,320 Speaker 7: it so they stop trying to be TikTok. You've seen 173 00:09:20,400 --> 00:09:23,960 Speaker 7: meta platforms and others basically come up with metso products 174 00:09:24,160 --> 00:09:28,040 Speaker 7: to stay on parity with TikTok and they leaned into 175 00:09:28,080 --> 00:09:31,800 Speaker 7: what made pinterest unique and what made pinterest special, and 176 00:09:31,840 --> 00:09:34,720 Speaker 7: that's driven use in gen Z and it's had a 177 00:09:34,760 --> 00:09:37,600 Speaker 7: halo effect to the extent that the gen z's posting 178 00:09:37,640 --> 00:09:41,439 Speaker 7: content that other generations are liking as well. So very 179 00:09:41,440 --> 00:09:43,720 Speaker 7: impressive engagement growth for Pinterest. 180 00:09:44,400 --> 00:09:45,800 Speaker 3: Tom help us out with this one. 181 00:09:46,200 --> 00:09:48,480 Speaker 4: The CFO is saying that the investments they've made are 182 00:09:48,520 --> 00:09:50,080 Speaker 4: increasing shopability. 183 00:09:50,960 --> 00:09:52,280 Speaker 3: What is that? What does that mean? 184 00:09:53,120 --> 00:09:54,920 Speaker 7: The simple way to think about it is they have 185 00:09:54,960 --> 00:09:59,199 Speaker 7: a third party advertising initiative with Amazon. So let's say 186 00:09:59,200 --> 00:10:01,640 Speaker 7: you went to pinterest in the search for a green dress. 187 00:10:02,240 --> 00:10:06,280 Speaker 7: Among the many pictures you would see would be items 188 00:10:06,320 --> 00:10:09,200 Speaker 7: from Amazon. You could then click on those items and 189 00:10:09,320 --> 00:10:10,320 Speaker 7: buy those items. 190 00:10:10,559 --> 00:10:11,199 Speaker 6: So think about it. 191 00:10:11,200 --> 00:10:13,720 Speaker 7: In the past, you would see wonderful pictures on Pinterest, 192 00:10:14,080 --> 00:10:17,200 Speaker 7: but it wasn't always easy. Not easy to buy the sofa, 193 00:10:17,280 --> 00:10:19,840 Speaker 7: not easy to buy the apparel. They made it a 194 00:10:19,840 --> 00:10:23,880 Speaker 7: lot easier. That's improved the shoppability of Pinterest. 195 00:10:24,160 --> 00:10:26,000 Speaker 2: I mean, that's really what Bill Ready has been saying 196 00:10:26,040 --> 00:10:27,760 Speaker 2: since it's come aboard a little over a year ago. 197 00:10:27,800 --> 00:10:31,880 Speaker 2: It's about people come to Pinterest with intent. So clearly 198 00:10:31,960 --> 00:10:34,200 Speaker 2: advertiser seeing that for now, Tom Forte. Great to have 199 00:10:34,280 --> 00:10:36,040 Speaker 2: some time with you, Thank you, Da Davison. 200 00:10:36,080 --> 00:10:47,240 Speaker 8: Then one thing is clear. To realize the promise of 201 00:10:47,280 --> 00:10:50,239 Speaker 8: AI and avoid the risk, we need to govern this technology. 202 00:10:50,360 --> 00:10:54,120 Speaker 8: We face a genuine inflection point in history, one of 203 00:10:54,120 --> 00:10:56,360 Speaker 8: those moments where the decisions we make in the very 204 00:10:56,400 --> 00:10:59,280 Speaker 8: near term are going to set the course for the 205 00:10:59,280 --> 00:11:00,000 Speaker 8: next decade. 206 00:11:01,000 --> 00:11:04,839 Speaker 2: President Biden there after signing an Executive Order on Artificial Intelligence, 207 00:11:04,920 --> 00:11:08,079 Speaker 2: the most significant step to date in regulating the technology. 208 00:11:08,280 --> 00:11:09,880 Speaker 2: Joining us for more on all of this is doctor 209 00:11:09,920 --> 00:11:12,680 Speaker 2: Joey Bolowini, founder of the Algorithmic Justice League and author 210 00:11:12,720 --> 00:11:15,720 Speaker 2: of Unmasking AI. My Mission to Protect What is Human 211 00:11:16,040 --> 00:11:19,240 Speaker 2: in a World of Machines is published today. Congratulations spending 212 00:11:19,240 --> 00:11:21,080 Speaker 2: some time with us as well. It's great to have 213 00:11:21,160 --> 00:11:25,000 Speaker 2: you here. Equitable and Accountable AI, that's all about what 214 00:11:25,280 --> 00:11:28,480 Speaker 2: your Algorithmic Justice League is about. Does this EO get 215 00:11:28,520 --> 00:11:30,559 Speaker 2: us there in some way? Absolutely? 216 00:11:30,760 --> 00:11:35,400 Speaker 1: I will actually say this is a commendable step forward. 217 00:11:35,640 --> 00:11:39,560 Speaker 1: And part of this is because the EO that's been 218 00:11:39,720 --> 00:11:45,760 Speaker 1: issued includes teeth when it comes to government agencies using AI. 219 00:11:45,880 --> 00:11:50,160 Speaker 1: So once you're tying federal dollars too, who can procure 220 00:11:50,360 --> 00:11:54,199 Speaker 1: particular systems? Then we're actually starting to see something that's 221 00:11:54,240 --> 00:11:58,400 Speaker 1: going to compel change. So I am excited to see 222 00:11:58,440 --> 00:12:01,480 Speaker 1: the civil rights peace within this EO. 223 00:12:02,120 --> 00:12:04,240 Speaker 2: It's interesting you talk about teeth and there's sort of 224 00:12:04,280 --> 00:12:06,800 Speaker 2: the carrot teeth situation the US is using. There's the 225 00:12:06,880 --> 00:12:10,040 Speaker 2: teeth of finds. The EUAI Act has been talking about. 226 00:12:10,480 --> 00:12:13,720 Speaker 2: Everyone wants in on this pile of regulation or talking 227 00:12:13,720 --> 00:12:16,360 Speaker 2: about it. We've got the UKI Summit about to happen. 228 00:12:16,600 --> 00:12:18,720 Speaker 2: One of the critiques of that AI Summit is there's 229 00:12:18,760 --> 00:12:23,360 Speaker 2: not enough actual people from civil society at that summit. 230 00:12:23,600 --> 00:12:26,480 Speaker 2: How much do you think the US and lawmakers are 231 00:12:26,480 --> 00:12:27,960 Speaker 2: engaging with civil society on this. 232 00:12:28,120 --> 00:12:30,400 Speaker 1: I think that is a fair critique. I am glad 233 00:12:30,440 --> 00:12:33,480 Speaker 1: to share that a member of the Algorithmic Justice League 234 00:12:33,480 --> 00:12:36,440 Speaker 1: will be there, But when we look at the overall roster, 235 00:12:36,960 --> 00:12:40,920 Speaker 1: we don't want the tech companies writing the regulations, and 236 00:12:41,000 --> 00:12:43,800 Speaker 1: so I certainly think there's more work to be done, 237 00:12:43,920 --> 00:12:46,280 Speaker 1: and I also do think it's a good step forward 238 00:12:46,320 --> 00:12:49,920 Speaker 1: to have the AI Safety Summit at this time. 239 00:12:51,120 --> 00:12:51,800 Speaker 3: Doctor Joy. 240 00:12:51,920 --> 00:12:56,440 Speaker 4: The main mechanism or tool that the Executive Order puts 241 00:12:56,480 --> 00:13:00,920 Speaker 4: in place is big technology companies having to hand over 242 00:13:01,400 --> 00:13:05,880 Speaker 4: an LLM or foundation model for safety review before it 243 00:13:05,960 --> 00:13:09,000 Speaker 4: is released to the wider public. Just explain how you 244 00:13:09,080 --> 00:13:13,080 Speaker 4: think that will work as a safeguard as we develop 245 00:13:13,160 --> 00:13:15,320 Speaker 4: next generations of large language models. 246 00:13:15,960 --> 00:13:18,160 Speaker 1: I think it could go one of two ways. If 247 00:13:18,200 --> 00:13:23,320 Speaker 1: you actually have the companies themselves setting what the safety 248 00:13:23,360 --> 00:13:26,040 Speaker 1: measures are and what the safety test will be, will 249 00:13:26,080 --> 00:13:29,360 Speaker 1: have a situation of grading your own homework, which I 250 00:13:29,360 --> 00:13:32,200 Speaker 1: don't necessarily think is going to get us where we 251 00:13:32,280 --> 00:13:35,480 Speaker 1: want to go. If it is a situation where you 252 00:13:35,559 --> 00:13:39,599 Speaker 1: actually have third parties that are establishing what the standards 253 00:13:39,600 --> 00:13:43,199 Speaker 1: are and the guidelines are, then I think you actually 254 00:13:43,240 --> 00:13:47,200 Speaker 1: have a situation where the companies will have to meet 255 00:13:47,280 --> 00:13:51,120 Speaker 1: specific regulations that they themselves are not setting. So it 256 00:13:51,200 --> 00:13:55,680 Speaker 1: remains to be seen. I am cautiously optimistic, but companies 257 00:13:55,720 --> 00:13:58,400 Speaker 1: cannot be grading their own homework and then. 258 00:13:58,280 --> 00:14:00,880 Speaker 3: Turning in, Sorry to interrupt you. 259 00:14:00,920 --> 00:14:03,880 Speaker 4: The executive order kind of gives power to the agencies. 260 00:14:03,960 --> 00:14:07,360 Speaker 4: Twenty four hours ago, we had doctor faith A Lee 261 00:14:07,360 --> 00:14:09,360 Speaker 4: of Stanford on the show, and she basically said, I 262 00:14:09,400 --> 00:14:13,040 Speaker 4: want to see a more even distribution of funding and 263 00:14:13,120 --> 00:14:17,960 Speaker 4: control across academia, industry, and the public sector. Where do 264 00:14:18,000 --> 00:14:20,680 Speaker 4: you stand on that balance of who is working on 265 00:14:20,720 --> 00:14:23,400 Speaker 4: AI and who has the power to regulate it. 266 00:14:24,360 --> 00:14:28,640 Speaker 1: I think I would give more power to the government 267 00:14:28,840 --> 00:14:32,000 Speaker 1: and then to companies. But I do think it's commendable 268 00:14:32,360 --> 00:14:36,000 Speaker 1: to see increased AI capacity for academics. As we know, 269 00:14:36,400 --> 00:14:38,520 Speaker 1: a lot of the compute power and a lot of 270 00:14:38,560 --> 00:14:42,440 Speaker 1: the data access does go mainly to the large tech company, 271 00:14:42,520 --> 00:14:46,000 Speaker 1: so it does make it difficult for researchers and outsiders 272 00:14:46,040 --> 00:14:48,960 Speaker 1: to even have a sense of how to assess the 273 00:14:49,040 --> 00:14:51,440 Speaker 1: risk if you don't have that access. So that's a 274 00:14:51,440 --> 00:14:52,280 Speaker 1: good step forward. 275 00:14:52,600 --> 00:14:56,080 Speaker 2: Your whole Mighty thesis methodology basically showed the level of 276 00:14:56,120 --> 00:15:00,280 Speaker 2: bias whether it comes racial agenda within sort of AI 277 00:15:00,360 --> 00:15:03,520 Speaker 2: services from certain companies in the US. You are going 278 00:15:03,520 --> 00:15:05,160 Speaker 2: out on your book tour and talking to some of 279 00:15:05,200 --> 00:15:07,680 Speaker 2: the most notable names in technology, some Lpment for example 280 00:15:07,720 --> 00:15:09,840 Speaker 2: being one of them. How much they wanted to engage 281 00:15:10,280 --> 00:15:12,760 Speaker 2: with you. How much have they been trying to well 282 00:15:12,880 --> 00:15:15,560 Speaker 2: improve their homework before turning it in and having someone 283 00:15:15,560 --> 00:15:18,400 Speaker 2: else market I think it's a mixed bag. 284 00:15:18,480 --> 00:15:21,200 Speaker 1: I will say for Open AI, they have reached out 285 00:15:21,240 --> 00:15:24,280 Speaker 1: to the Algorithmic Justice League multiple times and when they 286 00:15:24,320 --> 00:15:28,200 Speaker 1: had their bug bounty program to try to look at 287 00:15:28,240 --> 00:15:32,560 Speaker 1: AI vulnerabilities, they mention one of our white papers and 288 00:15:32,600 --> 00:15:35,360 Speaker 1: so forth. So I think there is a good faith 289 00:15:35,800 --> 00:15:38,840 Speaker 1: effort to reach out, but I certainly don't think it's enough, 290 00:15:38,960 --> 00:15:42,680 Speaker 1: and we certainly have different views. For example, how do 291 00:15:42,840 --> 00:15:47,479 Speaker 1: you credit and compensate artist? I do think many companies 292 00:15:47,520 --> 00:15:51,400 Speaker 1: have gotten a free pass on building llms with data 293 00:15:51,480 --> 00:15:55,800 Speaker 1: that's been collected without consent and without compensation yourself. 294 00:15:55,840 --> 00:15:58,160 Speaker 2: Being an artist, a poet, not only just researcher and 295 00:15:58,240 --> 00:16:02,920 Speaker 2: policy guidance is important. Rate the media for a minute, 296 00:16:03,240 --> 00:16:05,880 Speaker 2: because I think there is this desire sometimes to go 297 00:16:05,960 --> 00:16:09,480 Speaker 2: for the biggest headline, which often is you know, terminator style, 298 00:16:09,560 --> 00:16:11,600 Speaker 2: AI is going to end us all. But much of 299 00:16:11,640 --> 00:16:14,160 Speaker 2: the criticism has been in the here and the now. 300 00:16:14,200 --> 00:16:16,360 Speaker 2: We need to tackle the bias in the real data 301 00:16:16,400 --> 00:16:19,040 Speaker 2: the application. Where we are today, are we managing to 302 00:16:19,080 --> 00:16:21,880 Speaker 2: move away from hyperboley and get to actually what really 303 00:16:21,920 --> 00:16:23,200 Speaker 2: is the risk when it comes to AI? 304 00:16:23,920 --> 00:16:28,200 Speaker 1: It depends on the outlet, right, and so I commend 305 00:16:29,320 --> 00:16:32,400 Speaker 1: news outlets that actually bring in voices like the Algorithmic 306 00:16:32,600 --> 00:16:37,120 Speaker 1: Justice League CDT and others. So we're getting a different perspective. 307 00:16:37,200 --> 00:16:39,520 Speaker 1: But I still think there is a lot of fear 308 00:16:39,680 --> 00:16:43,680 Speaker 1: mongering and numerism, and that's one of the concerns I 309 00:16:43,720 --> 00:16:47,840 Speaker 1: have even with the framing of the AI Safety Summit, 310 00:16:48,000 --> 00:16:50,480 Speaker 1: and so I certainly want to make sure we are 311 00:16:50,520 --> 00:16:54,400 Speaker 1: focused on immediate and emerging harms. It's not to say 312 00:16:54,400 --> 00:16:56,640 Speaker 1: we don't want to be forward looking, but we don't 313 00:16:56,640 --> 00:16:59,560 Speaker 1: want to be so distracted that we don't attend to 314 00:16:59,640 --> 00:17:02,160 Speaker 1: the thing things we already know how to address. 315 00:17:03,600 --> 00:17:07,240 Speaker 4: Doctor Joy bolh and Weeni, founder of the Algorithmic Justice League. 316 00:17:07,240 --> 00:17:09,439 Speaker 4: Great to have you here on Bloomberg Technology. Thank you 317 00:17:09,480 --> 00:17:11,760 Speaker 4: so much. Now coming up on the program in video. 318 00:17:11,800 --> 00:17:14,439 Speaker 4: Shares taking a hit today on reports they may have 319 00:17:14,520 --> 00:17:17,680 Speaker 4: to cancel billions of dollars worth of advanced chip sales 320 00:17:18,119 --> 00:17:21,080 Speaker 4: to China. More on how those new US export rules 321 00:17:21,080 --> 00:17:23,320 Speaker 4: are putting the company in a state of limbo. The 322 00:17:23,359 --> 00:17:27,159 Speaker 4: stockdown about one percent, more than two percent in the session. 323 00:17:27,240 --> 00:17:45,440 Speaker 4: This is Bloomberg Technology. Okay, it's time for talking tech 324 00:17:45,480 --> 00:17:48,160 Speaker 4: and first up, Vodaphone is agreeing to sell its Spanish 325 00:17:48,240 --> 00:17:51,359 Speaker 4: unit to Zagonda Communications. The deal is said to be valued 326 00:17:51,359 --> 00:17:54,760 Speaker 4: at roughly five point three billion dollars including debt. The 327 00:17:54,760 --> 00:17:58,399 Speaker 4: European Commission is currently examining that deal, which would shift 328 00:17:58,400 --> 00:18:03,320 Speaker 4: the telecommunications and open the door to more consolidation across Europe. 329 00:18:03,520 --> 00:18:06,400 Speaker 4: And Bunge, the Sony owned game studio behind the popular 330 00:18:06,400 --> 00:18:10,000 Speaker 4: Destiny Too franchise, has let go of an undisclosed number 331 00:18:10,000 --> 00:18:13,560 Speaker 4: of its workforce. This comes after Bungee delayed the upcoming 332 00:18:13,600 --> 00:18:16,520 Speaker 4: expansion of Destiny to, pushing it out of Sony's current 333 00:18:16,520 --> 00:18:19,280 Speaker 4: fiscal year. The move is part of a wider restructuring 334 00:18:19,280 --> 00:18:22,119 Speaker 4: at the company and reflects a recent bout of layoffs 335 00:18:22,320 --> 00:18:26,320 Speaker 4: across the gaming industry. Plus, Samsung is seeing encouraging signs 336 00:18:26,320 --> 00:18:29,760 Speaker 4: of recovery in the chips market after reporting profit for 337 00:18:29,840 --> 00:18:32,320 Speaker 4: the third quarter, which was well ahead of estimates. The 338 00:18:32,400 --> 00:18:34,920 Speaker 4: South Korean giant raked in more than four billion dollars 339 00:18:35,080 --> 00:18:38,120 Speaker 4: of net income in the September quarter, more than double 340 00:18:38,520 --> 00:18:42,520 Speaker 4: what was expected. Executives say that artificial intelligence is driving 341 00:18:42,560 --> 00:18:46,800 Speaker 4: demand and it plans increased spending on advanced chip making. 342 00:18:46,960 --> 00:18:50,520 Speaker 2: Caroline, Meanwhile, let's look at the fresh US curbs on 343 00:18:50,840 --> 00:18:53,880 Speaker 2: the advanced chip making that's going on to China. They're 344 00:18:53,880 --> 00:18:56,120 Speaker 2: presenting a new risk to in video. We understand now 345 00:18:56,160 --> 00:18:58,280 Speaker 2: a deep and dim in the shares that you've currently 346 00:18:58,320 --> 00:19:01,280 Speaker 2: see to push the market value just below that one 347 00:19:01,320 --> 00:19:03,200 Speaker 2: trillion dollar mark. So all off of the Wall Street 348 00:19:03,240 --> 00:19:07,000 Speaker 2: Journal reported, well, well that potentially in Video might have 349 00:19:07,040 --> 00:19:09,439 Speaker 2: to cancel five billion dollars worth of sales to China. 350 00:19:09,880 --> 00:19:12,640 Speaker 2: But please to welcome Conjan Zivani. He's been bag intelligence 351 00:19:12,640 --> 00:19:16,359 Speaker 2: analyst covering all things semiconductors. And was this in some 352 00:19:16,440 --> 00:19:18,840 Speaker 2: way obvious that they were going to have to pull 353 00:19:18,880 --> 00:19:21,760 Speaker 2: back from selling China sooner than some had anticipated. 354 00:19:23,560 --> 00:19:26,240 Speaker 9: Yeah, I mean the cancelation of the grace periods was 355 00:19:26,280 --> 00:19:29,560 Speaker 9: definitely a surprise for them, and we think now it 356 00:19:29,640 --> 00:19:32,280 Speaker 9: is the restrictions have reached a point where this does 357 00:19:32,320 --> 00:19:34,719 Speaker 9: start having some long term impacts. So when we look 358 00:19:34,720 --> 00:19:38,680 Speaker 9: at long term impacts next year, which is fiscal twenty five, 359 00:19:38,680 --> 00:19:41,680 Speaker 9: which is equal on to Calnar twenty four for them, 360 00:19:42,000 --> 00:19:44,520 Speaker 9: you know, we estimate about five to eight billion that 361 00:19:44,680 --> 00:19:47,120 Speaker 9: they would have shipped to China, so you can put 362 00:19:47,119 --> 00:19:50,600 Speaker 9: a five billion threshold baseline impact that it will have 363 00:19:50,720 --> 00:19:53,720 Speaker 9: to the top line. However, there are some avenues we 364 00:19:53,800 --> 00:19:58,120 Speaker 9: think where Nvidia can offset that loss, one being shipping 365 00:19:58,160 --> 00:20:01,560 Speaker 9: to other customers outside of because they're still as of 366 00:20:01,680 --> 00:20:04,760 Speaker 9: right now and things can change. Given the uncertain macro 367 00:20:04,960 --> 00:20:07,640 Speaker 9: there are still but there are still demand for Nvidia chips. 368 00:20:07,760 --> 00:20:10,920 Speaker 9: And second, there might be ways that n Media can 369 00:20:10,960 --> 00:20:14,680 Speaker 9: still get exposure to the Chinese market through alternate means. 370 00:20:15,119 --> 00:20:18,000 Speaker 4: What we're talking about in the AI chick context is 371 00:20:18,040 --> 00:20:21,960 Speaker 4: GPUs that are basically assemble components that go into data centers, 372 00:20:22,400 --> 00:20:24,800 Speaker 4: and for China that's always been twenty to twenty five 373 00:20:24,800 --> 00:20:28,480 Speaker 4: percent of its data center business right but their supply constrained, 374 00:20:28,480 --> 00:20:31,080 Speaker 4: and they never talk about it. Is it just simply 375 00:20:31,119 --> 00:20:34,560 Speaker 4: that they can sell to other nations, other startups around 376 00:20:34,560 --> 00:20:36,560 Speaker 4: the world the same product that they would have otherwise 377 00:20:36,560 --> 00:20:37,240 Speaker 4: sent to China. 378 00:20:37,720 --> 00:20:39,800 Speaker 9: Well, not exactly the same product, but a version of 379 00:20:39,840 --> 00:20:40,439 Speaker 9: the same product. 380 00:20:40,520 --> 00:20:40,720 Speaker 1: Yes. 381 00:20:41,000 --> 00:20:43,240 Speaker 9: And then the second key point is when we look 382 00:20:43,280 --> 00:20:46,600 Speaker 9: at these cloud providers, their CAPEX is not only the 383 00:20:46,680 --> 00:20:49,400 Speaker 9: only avenue that n media can get exposure. For example, 384 00:20:49,680 --> 00:20:53,280 Speaker 9: Microsoft has a compute offload through open AI, so when 385 00:20:53,320 --> 00:20:55,960 Speaker 9: we look at Microsoft Capex, that's not the only piece 386 00:20:56,000 --> 00:20:58,399 Speaker 9: that n media gets, but also through open AI. You 387 00:20:58,440 --> 00:21:00,919 Speaker 9: have the same thing going on with Alibab announcing that 388 00:21:00,960 --> 00:21:03,719 Speaker 9: they are going to offer Meda's large language models. So 389 00:21:03,760 --> 00:21:06,679 Speaker 9: there are still different avenues that sort of Nvidia can 390 00:21:06,720 --> 00:21:08,919 Speaker 9: get an indirect exposure to the Chinese market. 391 00:21:09,359 --> 00:21:12,879 Speaker 4: Counjen Savannia, Bloomberg Intelligence, the other mister chip out here 392 00:21:13,200 --> 00:21:26,360 Speaker 4: in San Francisco, Caroline. 393 00:21:22,240 --> 00:21:24,840 Speaker 2: Welcome back to Bloomberg Technology. I'm Caroline had in New York. 394 00:21:25,160 --> 00:21:27,040 Speaker 3: And I'm ed Louvelow in San Francisco. 395 00:21:27,119 --> 00:21:29,320 Speaker 4: Let's go to check on European markets which have just 396 00:21:29,400 --> 00:21:31,919 Speaker 4: closed all this week. Will bring you the European market 397 00:21:31,960 --> 00:21:34,280 Speaker 4: closed and the stock six hundred Europe kind of this 398 00:21:34,800 --> 00:21:38,320 Speaker 4: continent wide gauge of equities rising for a second straight day. 399 00:21:38,440 --> 00:21:40,880 Speaker 4: But we're ending the month of October, and this has 400 00:21:40,960 --> 00:21:43,880 Speaker 4: been the worst October going back to two thousand and four. 401 00:21:43,960 --> 00:21:46,960 Speaker 4: European stocks. A bright spot with Stilantis and their earnings. 402 00:21:47,200 --> 00:21:50,520 Speaker 4: That stock in its European trading up three point six percent. 403 00:21:50,840 --> 00:21:52,320 Speaker 3: Oil rebounding modesty. 404 00:21:52,400 --> 00:21:54,520 Speaker 4: Brent, which is kind of more of the global benchmark 405 00:21:54,520 --> 00:21:58,320 Speaker 4: as opposed to WTI, have been moving lower Monday over 406 00:21:58,359 --> 00:22:01,400 Speaker 4: the weekend due to the activity in the Israel hamass war, 407 00:22:01,680 --> 00:22:04,760 Speaker 4: rebounding modestly four tenths of one percent in euro dollar 408 00:22:04,800 --> 00:22:08,120 Speaker 4: one oh five six two softer the euro by half 409 00:22:08,160 --> 00:22:10,080 Speaker 4: a percentage point against the dollar. 410 00:22:10,119 --> 00:22:13,080 Speaker 2: Caroline tell you also, what's a little bit softer today 411 00:22:13,200 --> 00:22:15,480 Speaker 2: has been the Nasdaq and the NaSTA one hundred and 412 00:22:15,520 --> 00:22:17,280 Speaker 2: we're down by another tenth of a percent, but actually 413 00:22:17,320 --> 00:22:19,760 Speaker 2: all told, on the month of October we're down more 414 00:22:19,760 --> 00:22:21,960 Speaker 2: than three percent, not as bad as it was the 415 00:22:21,960 --> 00:22:23,760 Speaker 2: five percent sell off back in September, but it's three 416 00:22:23,760 --> 00:22:25,640 Speaker 2: straight months of losses for the NASDAK at the moment, 417 00:22:25,880 --> 00:22:28,280 Speaker 2: worse set of losing streets that we've seen all the 418 00:22:28,320 --> 00:22:30,800 Speaker 2: way back to, of course about June of last year. 419 00:22:30,880 --> 00:22:32,880 Speaker 2: So clearly pressure is on some of these big tech 420 00:22:32,920 --> 00:22:36,040 Speaker 2: naus A mid earnings amid these geopolitical stories that you 421 00:22:36,040 --> 00:22:38,080 Speaker 2: talk us through ed looking in in video, of course, 422 00:22:38,119 --> 00:22:40,040 Speaker 2: we're just talking about the JYMP politics there, the sales 423 00:22:40,080 --> 00:22:42,600 Speaker 2: into China and the issue with well maybe not being 424 00:22:42,640 --> 00:22:45,439 Speaker 2: able to put some of those orders through to that country. 425 00:22:45,600 --> 00:22:47,879 Speaker 2: So stock under pressure there. It just only tests that 426 00:22:47,960 --> 00:22:50,120 Speaker 2: up what about tenth of percent, two tens of percent, 427 00:22:50,200 --> 00:22:52,800 Speaker 2: let's call it. What a painful two weeks that's been 428 00:22:52,840 --> 00:22:54,199 Speaker 2: the wed dof You see the story at what one 429 00:22:54,280 --> 00:22:56,560 Speaker 2: hundred and forty five billion dollars wiped off in terms 430 00:22:56,600 --> 00:22:59,399 Speaker 2: of its market capitalization over the course of those two weeks. 431 00:23:00,560 --> 00:23:01,920 Speaker 2: That does seem to be in some way to do 432 00:23:02,040 --> 00:23:02,800 Speaker 2: with its earnings. 433 00:23:03,640 --> 00:23:05,800 Speaker 4: Yeah, I think if we go back to October eighteenth 434 00:23:05,800 --> 00:23:08,919 Speaker 4: when they reported earnings, until this point, we're down twenty 435 00:23:08,960 --> 00:23:11,479 Speaker 4: percent on Tesla's shares. The other way looking at it, though, 436 00:23:11,600 --> 00:23:14,120 Speaker 4: is that Tesla's up more than sixty percent year to date, 437 00:23:14,160 --> 00:23:16,400 Speaker 4: while the S and P five hundred is up much 438 00:23:16,440 --> 00:23:17,200 Speaker 4: more modestly. 439 00:23:17,280 --> 00:23:19,440 Speaker 3: But the concern is interest rates. 440 00:23:19,480 --> 00:23:21,760 Speaker 4: You know Elon Musk talking about the impact of interest 441 00:23:21,880 --> 00:23:24,719 Speaker 4: rates and they've used price cuts earlier in the year 442 00:23:24,760 --> 00:23:27,639 Speaker 4: to unlock demand. How much do we put emphasis on 443 00:23:27,840 --> 00:23:31,840 Speaker 4: high rate environment hurting consumers hitting that ev demand? But yeah, 444 00:23:31,880 --> 00:23:34,239 Speaker 4: that's a pretty telling chart, isn't it. The story at 445 00:23:34,280 --> 00:23:36,400 Speaker 4: least in the last two weeks has changed that we're 446 00:23:36,400 --> 00:23:37,800 Speaker 4: a little modestly higher. 447 00:23:38,840 --> 00:23:41,840 Speaker 2: It's Tuesday, and it has been interesting, hasn't it? Ultimately 448 00:23:41,880 --> 00:23:44,800 Speaker 2: that they've had pressure from China as well byd But 449 00:23:45,040 --> 00:23:47,320 Speaker 2: I don't know if you noticed overall that there was, 450 00:23:47,359 --> 00:23:49,440 Speaker 2: of course, well, the idea that we're going to be 451 00:23:49,480 --> 00:23:52,159 Speaker 2: seeing Elon Musk flying over to the UK. I think 452 00:23:52,200 --> 00:23:54,480 Speaker 2: he's actually just touched down. This is all about not 453 00:23:54,600 --> 00:23:57,119 Speaker 2: Tesla but art Intelligence, which of course, is a long 454 00:23:57,200 --> 00:23:58,119 Speaker 2: term theme for Tesla. 455 00:23:59,080 --> 00:24:01,320 Speaker 4: Yeah, it is a long term theme because the ultimate 456 00:24:01,560 --> 00:24:04,520 Speaker 4: business model is to have a fleet of robotaxis, which 457 00:24:04,520 --> 00:24:06,960 Speaker 4: are Tesla vehicles centrally controlled by Tesla. 458 00:24:07,320 --> 00:24:08,240 Speaker 3: The training of that is. 459 00:24:08,200 --> 00:24:11,400 Speaker 4: Done on Dojo, their supercomputer, and the neural network they're 460 00:24:11,440 --> 00:24:14,120 Speaker 4: building is their version of AI, which they have high 461 00:24:14,160 --> 00:24:16,480 Speaker 4: degree of confidence on. And AI is really top of 462 00:24:16,480 --> 00:24:17,639 Speaker 4: mind for mister Mask right. 463 00:24:17,520 --> 00:24:21,280 Speaker 2: Now, instantly is as he touches down for the UKAI Summit, 464 00:24:21,400 --> 00:24:23,560 Speaker 2: and of course all things AI have just been happening 465 00:24:23,600 --> 00:24:26,000 Speaker 2: right here in the United States with that Executive Order 466 00:24:26,040 --> 00:24:28,879 Speaker 2: as well, I speak to someone who's been there helping 467 00:24:28,960 --> 00:24:31,360 Speaker 2: craft what a NEO might look like with the President. 468 00:24:31,640 --> 00:24:34,000 Speaker 2: You are at this signing of the Executive Order, credo 469 00:24:34,080 --> 00:24:37,960 Speaker 2: a CEO Laverena singh Laverna. Really important to have some 470 00:24:38,080 --> 00:24:41,040 Speaker 2: time with you today because you're in the business of governance, 471 00:24:41,080 --> 00:24:45,560 Speaker 2: governance of AI systems. Ultimately, this Executive Order, I feel, 472 00:24:45,640 --> 00:24:48,360 Speaker 2: leaves a lot still yet to be ironed out when 473 00:24:48,359 --> 00:24:53,760 Speaker 2: it comes to actual guardrails, actual benchmarks, actual well really auditing. 474 00:24:54,160 --> 00:24:55,639 Speaker 2: How do you see this EO doing. 475 00:24:57,600 --> 00:25:00,320 Speaker 10: Thank you so much for having me here today. It 476 00:25:00,440 --> 00:25:03,560 Speaker 10: truly is a historic moment, and I am reporting here 477 00:25:03,680 --> 00:25:08,280 Speaker 10: life from DC. You know, this Executive Order was one 478 00:25:08,320 --> 00:25:13,160 Speaker 10: of the most comprehensive and expansive commitments to responsible AI 479 00:25:13,240 --> 00:25:16,439 Speaker 10: that we've seen from the federal government. And something that 480 00:25:16,480 --> 00:25:19,239 Speaker 10: I've stated in the past is we need to be 481 00:25:19,520 --> 00:25:22,919 Speaker 10: okay with adaptive policy making, and that's not exactly what 482 00:25:22,960 --> 00:25:25,560 Speaker 10: we are seeing at this moment in time. This year 483 00:25:25,800 --> 00:25:29,199 Speaker 10: is so comprehensive. I've barely made it through half of 484 00:25:29,240 --> 00:25:32,440 Speaker 10: it right now, but this is a strong signal from 485 00:25:32,520 --> 00:25:35,960 Speaker 10: the government that now it is not about just talking 486 00:25:36,000 --> 00:25:40,000 Speaker 10: about responsible AI, but making sure that not only the 487 00:25:40,119 --> 00:25:46,160 Speaker 10: US agencies but also private sector is showing success against 488 00:25:46,160 --> 00:25:49,320 Speaker 10: what those guardrails are. So Carolyn, absolutely, this is a 489 00:25:49,359 --> 00:25:50,119 Speaker 10: starting point. 490 00:25:50,359 --> 00:25:51,840 Speaker 2: There's a lot to unpack. 491 00:25:51,520 --> 00:25:53,919 Speaker 10: In this executive Order, but I think they're going to 492 00:25:53,920 --> 00:25:58,560 Speaker 10: see ripple effects across the world with this historic executive 493 00:25:58,640 --> 00:26:02,440 Speaker 10: order from the President the Vice President Navarna. 494 00:26:02,480 --> 00:26:05,679 Speaker 4: Of the fifty percent of the extensive EO that you 495 00:26:05,760 --> 00:26:09,520 Speaker 4: have read, what's your biggest criticism of the final form? 496 00:26:09,800 --> 00:26:11,960 Speaker 4: Where would you have liked to have seen the administration 497 00:26:12,119 --> 00:26:12,560 Speaker 4: do more? 498 00:26:14,600 --> 00:26:14,679 Speaker 11: So? 499 00:26:14,960 --> 00:26:17,520 Speaker 10: Ed, I think this goes back to we always want 500 00:26:17,600 --> 00:26:20,600 Speaker 10: more and especially as technologists, we want it to be precise. 501 00:26:20,880 --> 00:26:23,200 Speaker 10: But I think let's focus on what is so great 502 00:26:23,200 --> 00:26:26,400 Speaker 10: about the CEO. I think this EO think about it 503 00:26:26,480 --> 00:26:28,600 Speaker 10: as not a rule of the law, but this is 504 00:26:28,640 --> 00:26:32,480 Speaker 10: a work plan that is providing a very compressed timeline 505 00:26:32,520 --> 00:26:36,080 Speaker 10: for US agencies as well as the private sector to 506 00:26:36,200 --> 00:26:39,359 Speaker 10: start not only defining what does good look like in 507 00:26:39,440 --> 00:26:44,119 Speaker 10: terms of benchmarking and evaluation of these powerfolised systems, but 508 00:26:44,240 --> 00:26:48,000 Speaker 10: also holding them accountable in a very short timeframe to 509 00:26:48,040 --> 00:26:50,679 Speaker 10: come up with a game plan as to how we 510 00:26:50,720 --> 00:26:53,080 Speaker 10: are going to make sure that that whatever is defined 511 00:26:53,119 --> 00:26:56,439 Speaker 10: as good is actually adhere to. So two things that 512 00:26:56,440 --> 00:26:58,200 Speaker 10: I would like to call out in the EO that 513 00:26:58,240 --> 00:27:02,679 Speaker 10: we are very excited about is first and foremost, you know, 514 00:27:02,720 --> 00:27:08,520 Speaker 10: creation of standards led by NEST, especially for benchmarking and 515 00:27:08,600 --> 00:27:12,840 Speaker 10: auditing of artificial intelligence systems. We have about you know, 516 00:27:12,840 --> 00:27:16,200 Speaker 10: two hundred and seventy days for NEST to actually start 517 00:27:16,240 --> 00:27:18,879 Speaker 10: building a game plan for that. And then secondly, the 518 00:27:19,000 --> 00:27:22,919 Speaker 10: United States government agencies, you know, we're looking for the 519 00:27:22,960 --> 00:27:25,440 Speaker 10: OMB guidance which is going to come out any day, 520 00:27:25,880 --> 00:27:28,959 Speaker 10: which is going to put further focus on how procurement 521 00:27:29,400 --> 00:27:33,240 Speaker 10: of AI systems by US government and the agencies happen. 522 00:27:33,680 --> 00:27:35,600 Speaker 10: So I would say that right now we should be 523 00:27:35,640 --> 00:27:37,639 Speaker 10: focusing on all the great things that are in this 524 00:27:37,760 --> 00:27:41,000 Speaker 10: ear and then yes, there's lots more work ahead of us, 525 00:27:41,240 --> 00:27:44,200 Speaker 10: and we're going to continuously see over the coming months 526 00:27:44,600 --> 00:27:47,119 Speaker 10: more and more getting defined in terms of what is 527 00:27:47,160 --> 00:27:48,400 Speaker 10: really possible. 528 00:27:48,480 --> 00:27:51,240 Speaker 2: MIST being the national standard strategy for critical and Emerging 529 00:27:51,240 --> 00:27:56,159 Speaker 2: Technology in AVRENA. I'm interested that they're setting basically benchmark guidelines. 530 00:27:56,520 --> 00:28:00,760 Speaker 2: Your company already is kind of a private sector addition 531 00:28:01,040 --> 00:28:02,880 Speaker 2: that could be taken if you suddenly got a lot 532 00:28:02,880 --> 00:28:06,760 Speaker 2: of inbound from this, because when you're helping companies basically 533 00:28:07,200 --> 00:28:09,160 Speaker 2: build AI accountablity. 534 00:28:10,119 --> 00:28:14,399 Speaker 10: Yes, Carolyn so crew DOOAI is an AI governance platform 535 00:28:14,440 --> 00:28:18,880 Speaker 10: that provides continuous oversight and accountability of artificial intelligence systems. 536 00:28:19,480 --> 00:28:22,440 Speaker 10: I would say, not only because of this executive action, 537 00:28:22,720 --> 00:28:26,240 Speaker 10: but the need for trust, which is so critical as 538 00:28:26,280 --> 00:28:32,439 Speaker 10: companies are adopting building using artificial intelligence at scale, that 539 00:28:32,480 --> 00:28:36,080 Speaker 10: has really propelled and increased the need for CREDOI in 540 00:28:36,119 --> 00:28:38,920 Speaker 10: the market as well. And I think just to underscore 541 00:28:39,040 --> 00:28:42,719 Speaker 10: in CREDOAI, we are operationalizing what does good look like? 542 00:28:42,800 --> 00:28:47,560 Speaker 10: These might come from standards like nisst AI rmf ISO standards. 543 00:28:47,760 --> 00:28:50,560 Speaker 10: It could come from regulations like New York City Low 544 00:28:50,640 --> 00:28:53,240 Speaker 10: number one forty four. It can come from company policies 545 00:28:53,560 --> 00:28:57,520 Speaker 10: or industry best practices, and CREDOAI already has been at 546 00:28:57,560 --> 00:29:00,080 Speaker 10: the front of it. So now I would say the 547 00:29:00,120 --> 00:29:04,160 Speaker 10: companies as well as US government agencies are really requiring 548 00:29:04,240 --> 00:29:06,640 Speaker 10: creed OI as a core platform that's going to enable 549 00:29:06,720 --> 00:29:08,640 Speaker 10: trust with their consumers. 550 00:29:09,800 --> 00:29:10,240 Speaker 3: Nevrina. 551 00:29:10,320 --> 00:29:13,600 Speaker 4: In that light and in pursuit of a range of views, 552 00:29:13,600 --> 00:29:15,160 Speaker 4: I want to ask you the same question I've asked 553 00:29:15,200 --> 00:29:17,680 Speaker 4: all our guests in the last twenty four hours. The 554 00:29:17,720 --> 00:29:21,360 Speaker 4: main mechanism that the EO spells out and the Defense 555 00:29:21,640 --> 00:29:25,440 Speaker 4: Procurement Act empowers them to do, is to require makers 556 00:29:25,440 --> 00:29:28,840 Speaker 4: of large language models to offer them up for government 557 00:29:28,920 --> 00:29:33,000 Speaker 4: safety review before they are released to the wider public. 558 00:29:33,760 --> 00:29:37,480 Speaker 4: If a startups building in LM, is that a good mechanism. 559 00:29:37,480 --> 00:29:39,320 Speaker 4: Will that help the industry move forward? 560 00:29:41,960 --> 00:29:43,719 Speaker 2: I think that's a great question. 561 00:29:43,880 --> 00:29:47,720 Speaker 10: I am concerned about regulatory capture, and I think we 562 00:29:47,880 --> 00:29:50,080 Speaker 10: have to really focus on how do we bring this 563 00:29:50,320 --> 00:29:54,800 Speaker 10: entire startup innovation ecosystem forward, And something that we are 564 00:29:54,840 --> 00:29:59,400 Speaker 10: actively focused on at CREED OAI where executive orders as 565 00:29:59,440 --> 00:30:03,240 Speaker 10: well as reads are not a barrier for startups, especially 566 00:30:03,240 --> 00:30:06,400 Speaker 10: not the well capitalized startups, but it's an enabler for them. 567 00:30:07,080 --> 00:30:09,440 Speaker 10: So in terms of that, what the EO is spelling 568 00:30:09,480 --> 00:30:14,160 Speaker 10: out that the foundation model providers have to provide test 569 00:30:14,200 --> 00:30:18,760 Speaker 10: results to the government before releasing these models, especially for 570 00:30:18,800 --> 00:30:21,960 Speaker 10: government use. I think it's a good step, but it's 571 00:30:22,040 --> 00:30:25,520 Speaker 10: not comprehensive. And this is where I am encouraged to 572 00:30:25,560 --> 00:30:28,640 Speaker 10: see more action in the coming months, where we are 573 00:30:28,680 --> 00:30:32,400 Speaker 10: going to provide that oversight as well as benchmarking and 574 00:30:32,480 --> 00:30:37,480 Speaker 10: testing requirements across the entire value chain, the foundation model developers, 575 00:30:37,960 --> 00:30:42,280 Speaker 10: the application developers, the enterprises, and the consumers. 576 00:30:42,600 --> 00:30:43,640 Speaker 2: So you know, that's the. 577 00:30:43,600 --> 00:30:46,280 Speaker 10: Ripple effect of this executive order that I'm really excited 578 00:30:46,320 --> 00:30:48,000 Speaker 10: to see coming in the few months. 579 00:30:48,320 --> 00:30:51,920 Speaker 2: Everynessing Credo, AI CEO, thanks for making time with us 580 00:30:52,200 --> 00:30:54,400 Speaker 2: for of course you're at that White House event. And 581 00:30:54,440 --> 00:30:56,920 Speaker 2: of course one of the things that executive orders discussed 582 00:30:57,000 --> 00:31:00,120 Speaker 2: is well water marking, being able to understand when something 583 00:31:00,160 --> 00:31:03,040 Speaker 2: is fake or indeed created by AI. Well, we're going 584 00:31:03,080 --> 00:31:04,920 Speaker 2: to talk about investing in watermarking when it comes to 585 00:31:04,920 --> 00:31:07,880 Speaker 2: the art world. VC spotlights upnext, Christie's Ventures global head 586 00:31:07,920 --> 00:31:09,600 Speaker 2: Devan Thaka is going to be joining us on its 587 00:31:09,640 --> 00:31:30,240 Speaker 2: portfolio expansion. This is Blombog Technology in today's Ventures Spotlight. 588 00:31:30,320 --> 00:31:32,720 Speaker 2: We're going to sort of fuse art and VC with 589 00:31:32,840 --> 00:31:35,600 Speaker 2: Christie's Ventures of course, held at seventh edition of Art 590 00:31:35,600 --> 00:31:38,160 Speaker 2: and Tech Summit earlier this year, announcing a fair few 591 00:31:38,200 --> 00:31:42,440 Speaker 2: investments Manifold Technologies, Lay Zero Labs for example, Proto Hologram. Today, 592 00:31:42,640 --> 00:31:45,760 Speaker 2: Christie's Ventures is back expanding its portfolio, adding Eco Mark 593 00:31:45,880 --> 00:31:49,520 Speaker 2: and Atomic Form. Global head Divan Thacker joins us. Now, 594 00:31:49,800 --> 00:31:52,280 Speaker 2: I think from rather late in Hong Kong, where you're 595 00:31:52,320 --> 00:31:56,320 Speaker 2: on the road speaking to well companies building things, talk 596 00:31:56,360 --> 00:31:59,239 Speaker 2: to us about some of these investments. What is the 597 00:31:59,240 --> 00:32:02,400 Speaker 2: thesis behind ventures? Where are you backing? It looks as 598 00:32:02,400 --> 00:32:05,200 Speaker 2: though you're backing sort of authenticity in so large part. 599 00:32:05,240 --> 00:32:06,760 Speaker 2: We talk about that with AI at the moment. 600 00:32:08,360 --> 00:32:11,400 Speaker 11: Thank you for having me back, Carol and Ed and 601 00:32:11,440 --> 00:32:15,000 Speaker 11: hello from Hong Kong for us, like we take a 602 00:32:15,120 --> 00:32:19,720 Speaker 11: very principled approach on investing. As we mentioned last time, 603 00:32:19,760 --> 00:32:22,720 Speaker 11: we're early stage investors. We only invest in seed and 604 00:32:22,760 --> 00:32:26,320 Speaker 11: Series A companies, and Echomark for us is one of 605 00:32:26,320 --> 00:32:29,600 Speaker 11: those personally very near and dear to my heart company 606 00:32:29,640 --> 00:32:34,480 Speaker 11: based in Seattle, Washington that is looking to create deterrence 607 00:32:34,560 --> 00:32:39,040 Speaker 11: technology by adding invisible watermarks and AI based watermarks, something 608 00:32:39,040 --> 00:32:43,360 Speaker 11: that you just discussed with your prior panelists into documents. 609 00:32:42,800 --> 00:32:44,280 Speaker 6: To deter people from leaking them. 610 00:32:44,280 --> 00:32:46,160 Speaker 11: Whether it's the Supreme Court league that we saw a 611 00:32:46,240 --> 00:32:49,800 Speaker 11: few few quarters ago, or more recently things we've seen 612 00:32:49,880 --> 00:32:54,000 Speaker 11: from other corporate as well as government branches. Leaks are 613 00:32:54,040 --> 00:32:57,840 Speaker 11: a big problem, and echo Mark is attempting to use 614 00:32:58,160 --> 00:33:00,640 Speaker 11: state of the art technology to solve them, so that 615 00:33:00,720 --> 00:33:02,880 Speaker 11: even if you take a picture with your phone of 616 00:33:02,960 --> 00:33:05,960 Speaker 11: a little corner of a document, we're able to sort 617 00:33:06,000 --> 00:33:08,960 Speaker 11: of detect that and catch who leaked the document. And 618 00:33:09,080 --> 00:33:13,080 Speaker 11: Echomark just raised Series seed in collaboration Craft Ventures and 619 00:33:13,160 --> 00:33:14,680 Speaker 11: Chriss is proud to. 620 00:33:14,640 --> 00:33:15,520 Speaker 6: Be part of that round. 621 00:33:15,800 --> 00:33:19,120 Speaker 11: So for us, we see this technology applying directly to 622 00:33:19,160 --> 00:33:21,160 Speaker 11: our business because we have some of the most sensitive 623 00:33:21,160 --> 00:33:26,000 Speaker 11: client information in our business globally right who has what 624 00:33:26,280 --> 00:33:28,160 Speaker 11: art and who is looking to buy and. 625 00:33:28,080 --> 00:33:30,280 Speaker 6: Sell what art? So for us, this is critical data 626 00:33:30,280 --> 00:33:31,080 Speaker 6: to protect. 627 00:33:31,560 --> 00:33:34,240 Speaker 4: Well den That's what caught my eye and what's interesting 628 00:33:34,240 --> 00:33:37,520 Speaker 4: about echo Mark that Christie's auction House was one of 629 00:33:37,560 --> 00:33:40,080 Speaker 4: the first users of the echo mark technology. It came 630 00:33:40,120 --> 00:33:43,080 Speaker 4: out of stealth like a month ago. Just in very 631 00:33:43,080 --> 00:33:46,560 Speaker 4: simple terms, explain how the auction house uses it with 632 00:33:46,640 --> 00:33:49,280 Speaker 4: its client communications as a case study. 633 00:33:49,400 --> 00:33:51,000 Speaker 6: Yeah, no, absolutely. 634 00:33:51,040 --> 00:33:54,080 Speaker 11: So to take the example of a deal term and 635 00:33:54,160 --> 00:33:57,000 Speaker 11: say you have a piece of art which is iconic 636 00:33:57,120 --> 00:34:01,240 Speaker 11: and very expensive, and it's based in say some part 637 00:34:01,240 --> 00:34:03,480 Speaker 11: of the world, in your home in Hong Kong, for example, 638 00:34:04,480 --> 00:34:08,920 Speaker 11: the terms of that deal often exchanged in PDFs and 639 00:34:08,960 --> 00:34:12,960 Speaker 11: email attachments. Those could accidentally or non accidentally leak to 640 00:34:13,000 --> 00:34:16,320 Speaker 11: the public. Ecomark allows us to create another layer of 641 00:34:16,360 --> 00:34:19,840 Speaker 11: protection so that that document is personalized for the recipient. 642 00:34:20,080 --> 00:34:22,480 Speaker 11: So if you, as a recipient get this document and 643 00:34:22,600 --> 00:34:27,799 Speaker 11: accidentally leak it or maliciously leak it, we can immediately 644 00:34:27,840 --> 00:34:29,880 Speaker 11: know where the source of the leak comes from, so 645 00:34:29,920 --> 00:34:33,640 Speaker 11: that deters people from taking some malicious activity against sensitive data. 646 00:34:33,719 --> 00:34:36,239 Speaker 11: So that's the prototype, or that's the first use case 647 00:34:36,280 --> 00:34:38,680 Speaker 11: that we're working on. But I think the implications of 648 00:34:38,680 --> 00:34:41,520 Speaker 11: this are widespread. We can see that applying to all 649 00:34:41,560 --> 00:34:45,080 Speaker 11: forms of communication, including this call that we're having, could 650 00:34:45,080 --> 00:34:47,560 Speaker 11: also have invisible marks, so that if I took a 651 00:34:47,600 --> 00:34:50,239 Speaker 11: screenshot or a picture from my phone and leaked the 652 00:34:50,280 --> 00:34:51,520 Speaker 11: conversation that we were having. 653 00:34:51,719 --> 00:34:52,879 Speaker 6: I could immediately get caught. 654 00:34:52,920 --> 00:34:55,960 Speaker 11: So the implications of what ecomark is building are deeply 655 00:34:56,040 --> 00:34:59,759 Speaker 11: rooted in sort of better stewardship of documents, and that's 656 00:34:59,760 --> 00:35:02,080 Speaker 11: what we're using it in our business. 657 00:35:02,400 --> 00:35:05,759 Speaker 2: It's interesting that your other investment, atomic Form is still 658 00:35:05,760 --> 00:35:07,759 Speaker 2: a back on web three, is still a back on 659 00:35:08,120 --> 00:35:10,319 Speaker 2: non fungible tokens. Of course, I think back to the 660 00:35:10,360 --> 00:35:13,080 Speaker 2: b pal sixty nine million dollar auction that made sort 661 00:35:13,080 --> 00:35:16,040 Speaker 2: of put NFTs on the map with Christie's Why still 662 00:35:16,040 --> 00:35:18,520 Speaker 2: support this ecosystem when so many people. 663 00:35:18,280 --> 00:35:20,960 Speaker 6: Question, no great question. 664 00:35:21,280 --> 00:35:24,880 Speaker 11: We first met atomic Form very early in the days 665 00:35:24,920 --> 00:35:29,600 Speaker 11: of NFTs as a company that provided beautiful displays for 666 00:35:30,239 --> 00:35:32,840 Speaker 11: these pieces of digital art. I think my belief is 667 00:35:32,880 --> 00:35:37,359 Speaker 11: that in the long run, NFTs other sort of nomenclature, 668 00:35:37,400 --> 00:35:39,160 Speaker 11: whatever you want to call it, in the end, it's 669 00:35:39,160 --> 00:35:41,840 Speaker 11: all digital art, and what we've seen through the progression 670 00:35:41,840 --> 00:35:44,120 Speaker 11: of Christie's three point zero and investments we've made there, 671 00:35:44,400 --> 00:35:46,840 Speaker 11: digital art is here to stay. That's just our belief. 672 00:35:47,360 --> 00:35:49,000 Speaker 6: The bid that interested us. 673 00:35:49,000 --> 00:35:51,240 Speaker 11: In an atomic form is in addition to the hardware 674 00:35:51,280 --> 00:35:54,440 Speaker 11: bit they're also a great software company, which I believe 675 00:35:54,440 --> 00:35:58,400 Speaker 11: in software as a business model, as a scalable business model, 676 00:35:58,640 --> 00:36:02,400 Speaker 11: so traditional introdditional businesses. If you take comics books for example, 677 00:36:02,440 --> 00:36:04,759 Speaker 11: or comic books for example, there's a rating system. Right 678 00:36:04,760 --> 00:36:07,080 Speaker 11: when you go into a comic book store, you can 679 00:36:07,120 --> 00:36:10,319 Speaker 11: see a blue box that says the six star, seventh star. 680 00:36:10,800 --> 00:36:12,560 Speaker 6: No one's doing that for digital assets. 681 00:36:12,719 --> 00:36:14,680 Speaker 11: So tomic Form is one of those companies that's doing 682 00:36:14,719 --> 00:36:18,400 Speaker 11: innovative things with software, whether it's ratings, whether it's applying 683 00:36:19,160 --> 00:36:21,440 Speaker 11: physical activation data back to the blockchain. 684 00:36:21,480 --> 00:36:22,759 Speaker 6: For example, if I loaned. 685 00:36:22,520 --> 00:36:25,279 Speaker 11: You an NFT that happens in a physical world at 686 00:36:25,320 --> 00:36:28,840 Speaker 11: MoMA for example, that doesn't get captured back into the 687 00:36:28,960 --> 00:36:31,960 Speaker 11: NFT itself, this company is sort of adding that metadata 688 00:36:32,000 --> 00:36:35,120 Speaker 11: back to the NFT, really connecting the Web two and 689 00:36:35,160 --> 00:36:38,840 Speaker 11: Web three worlds. And we believe that whether it's Web 690 00:36:38,840 --> 00:36:41,960 Speaker 11: three in its current iteration or Web three with AI augmented, 691 00:36:42,400 --> 00:36:44,600 Speaker 11: we're going to live in a very fast and connected 692 00:36:44,680 --> 00:36:46,520 Speaker 11: universe where companies. 693 00:36:46,080 --> 00:36:47,759 Speaker 6: Like this are going to be very important. Hence the 694 00:36:47,800 --> 00:36:48,680 Speaker 6: support all. 695 00:36:48,640 --> 00:36:52,000 Speaker 4: Right devunk cekr Christy's ventures searching for startups at in 696 00:36:52,000 --> 00:37:03,000 Speaker 4: Hong Kong. Thank you for joining us here. X, the 697 00:37:03,040 --> 00:37:06,080 Speaker 4: platform formerly known as Twitter, is worth less than half 698 00:37:06,320 --> 00:37:08,080 Speaker 4: of what Elon must pay for it just a year 699 00:37:08,080 --> 00:37:11,000 Speaker 4: ago when he bought the company for forty four billion dollars. 700 00:37:11,080 --> 00:37:14,480 Speaker 4: According to a Bloomberg source, restrictive stock units awarded to 701 00:37:14,480 --> 00:37:18,200 Speaker 4: employees value the company at nineteen billion dollars based on 702 00:37:18,360 --> 00:37:20,719 Speaker 4: ursus of forty five dollars apiece joining us here in 703 00:37:20,800 --> 00:37:23,920 Speaker 4: SF Bloomberg's Asia accounts. You and I reported this story together, 704 00:37:23,960 --> 00:37:26,000 Speaker 4: and part of it is a drop off in sales. 705 00:37:26,320 --> 00:37:28,080 Speaker 4: Part of it is the debt burden, and part of 706 00:37:28,120 --> 00:37:30,680 Speaker 4: it is that this is a work in progres shall 707 00:37:30,680 --> 00:37:31,400 Speaker 4: we say. 708 00:37:31,200 --> 00:37:32,040 Speaker 2: It's so many things. 709 00:37:32,040 --> 00:37:34,080 Speaker 12: It's all the things that you just mentioned, right, I mean, 710 00:37:34,400 --> 00:37:36,480 Speaker 12: this is such a steep drop in valuation. Part of 711 00:37:36,520 --> 00:37:39,080 Speaker 12: it is the advertising revenue. In September it was down 712 00:37:39,160 --> 00:37:42,239 Speaker 12: sixty percent. And you think about advertising revenue historically for 713 00:37:42,600 --> 00:37:45,560 Speaker 12: X has been eighty ninety percent of the revenue, so 714 00:37:45,719 --> 00:37:48,200 Speaker 12: that's already huge, as you mentioned. And then you have 715 00:37:48,239 --> 00:37:50,319 Speaker 12: thirteen billion dollars of debts, so now you have to 716 00:37:50,360 --> 00:37:53,600 Speaker 12: pay one point two billion dollars in interest every year. 717 00:37:53,640 --> 00:37:55,080 Speaker 12: So you add all that up and they're in a 718 00:37:55,080 --> 00:37:57,040 Speaker 12: really tough smart financially and so it sort of makes 719 00:37:57,080 --> 00:37:58,319 Speaker 12: sense that the valuation is so well. 720 00:37:58,400 --> 00:38:00,200 Speaker 4: The other thing we reported in the last week or 721 00:38:00,239 --> 00:38:02,560 Speaker 4: so is like what the end result might look like. 722 00:38:02,640 --> 00:38:05,560 Speaker 4: Elon must saying we're going after LinkedIn and we're going 723 00:38:05,600 --> 00:38:08,360 Speaker 4: after YouTube in this all hands of employees. 724 00:38:08,360 --> 00:38:10,279 Speaker 3: What does that tell us about what X might be 725 00:38:10,360 --> 00:38:11,000 Speaker 3: in the future. 726 00:38:11,520 --> 00:38:14,680 Speaker 12: It's that everything ap idea, right that Elon has talked 727 00:38:14,680 --> 00:38:16,760 Speaker 12: about over and over again, But it could be interesting 728 00:38:16,760 --> 00:38:17,960 Speaker 12: if it sort of turns. 729 00:38:17,760 --> 00:38:18,440 Speaker 3: Out that way, right. 730 00:38:18,800 --> 00:38:20,560 Speaker 12: It's the idea of being able to do things like 731 00:38:20,640 --> 00:38:23,560 Speaker 12: areio and video calling. You mentioned LinkedIn, so they have 732 00:38:23,600 --> 00:38:27,560 Speaker 12: a hiring service doing more video aka YouTube and then 733 00:38:27,600 --> 00:38:29,879 Speaker 12: even having a wire service that they're planning the launch 734 00:38:29,960 --> 00:38:31,960 Speaker 12: for it for PR So it could be all those things. 735 00:38:32,080 --> 00:38:34,279 Speaker 4: And the lesson learned of covering test and SpaceX for 736 00:38:34,280 --> 00:38:37,040 Speaker 4: the last six years is don't discount Elon Musk because 737 00:38:37,360 --> 00:38:41,120 Speaker 4: they've been highly successful, but rocky adventures, bliebizhu accounts, thank 738 00:38:41,160 --> 00:38:42,399 Speaker 4: you Carrot, Yeah, and. 739 00:38:42,360 --> 00:38:45,399 Speaker 2: They're betting big maybe on being a payments focus as well. 740 00:38:45,480 --> 00:38:47,719 Speaker 2: Let's go back to the world of fintech. FTX co 741 00:38:47,800 --> 00:38:50,920 Speaker 2: founder Sam amun Fried just finished testifying in his criminal 742 00:38:50,920 --> 00:38:54,160 Speaker 2: fraud trial today, megs Max Chafkin, who was actually in 743 00:38:54,200 --> 00:38:57,160 Speaker 2: the court queuing for hours and the rain, you join 744 00:38:57,239 --> 00:38:59,840 Speaker 2: us for the latest, And I mean he was trying 745 00:38:59,880 --> 00:39:02,000 Speaker 2: to convinced that he didn't know as much as he 746 00:39:02,040 --> 00:39:02,640 Speaker 2: should have known. 747 00:39:03,080 --> 00:39:06,200 Speaker 13: Yes, and this testimony has been going on since Friday. 748 00:39:06,200 --> 00:39:08,800 Speaker 13: We've seen two and a half days of Sam Bankman Freed, 749 00:39:09,719 --> 00:39:12,640 Speaker 13: essentially the defense trying to paint a picture of a 750 00:39:13,080 --> 00:39:16,720 Speaker 13: well meaning but perhaps slightly out of touch executive, someone 751 00:39:16,920 --> 00:39:21,000 Speaker 13: with competence and at least a rough understanding of his responsibilities. 752 00:39:21,560 --> 00:39:24,759 Speaker 13: What we saw from the prosecution was kind of a 753 00:39:25,080 --> 00:39:31,120 Speaker 13: methodical portrayal of Sam Bankman Freed as essentially a professional liar. 754 00:39:31,239 --> 00:39:37,319 Speaker 13: We saw repeated instances or accusations of deception deceit, as 755 00:39:37,360 --> 00:39:40,560 Speaker 13: well as a refusal by the defendant to sort of 756 00:39:40,600 --> 00:39:42,839 Speaker 13: admit to any of it. And as it played out 757 00:39:42,840 --> 00:39:47,040 Speaker 13: in court, particularly yesterday and this morning, it was tough. 758 00:39:47,760 --> 00:39:51,000 Speaker 2: It was tough, perhaps less tough, obviously under friendly fare 759 00:39:51,040 --> 00:39:54,239 Speaker 2: from defense. Then when it came to the prosecutors, did 760 00:39:54,239 --> 00:39:56,400 Speaker 2: he lose his ground somewhat? What was the demeanor of 761 00:39:56,480 --> 00:39:57,280 Speaker 2: him in court? 762 00:39:57,640 --> 00:40:02,320 Speaker 13: His demeanor was extremely subdued, you know, very very voicing, 763 00:40:02,480 --> 00:40:05,040 Speaker 13: smaller than it had And as I was watching this, 764 00:40:05,160 --> 00:40:08,200 Speaker 13: I just kept thinking about the way that this performance 765 00:40:08,280 --> 00:40:10,480 Speaker 13: was playing, you know, during the rise of FTX, when 766 00:40:10,480 --> 00:40:13,440 Speaker 13: we had Crypto soaring and we saw him as this 767 00:40:13,520 --> 00:40:16,839 Speaker 13: kind of genius guy who was part of this like 768 00:40:16,880 --> 00:40:20,600 Speaker 13: story movement. As things have fallen apart, that affect has 769 00:40:20,640 --> 00:40:23,320 Speaker 13: not worked as well, and we've seen somebody who has seemed, 770 00:40:23,440 --> 00:40:27,160 Speaker 13: even even when under questioning from his own lawyers, somewhat 771 00:40:27,160 --> 00:40:29,719 Speaker 13: out of his depth. And then, you know, under fire 772 00:40:29,760 --> 00:40:32,520 Speaker 13: from a very adept federal prosecutor, you know, it only 773 00:40:32,560 --> 00:40:33,000 Speaker 13: got worse. 774 00:40:34,120 --> 00:40:37,359 Speaker 4: So now is this story of psychology and process because 775 00:40:37,360 --> 00:40:41,920 Speaker 4: the government could cool to other witnesses and it's about 776 00:40:41,960 --> 00:40:42,440 Speaker 4: the loss. 777 00:40:42,400 --> 00:40:45,359 Speaker 3: The impression of the jury. What do you reconmac say 778 00:40:45,440 --> 00:40:47,520 Speaker 3: does this go from here? Well, the government has rested 779 00:40:47,520 --> 00:40:47,919 Speaker 3: its case. 780 00:40:47,960 --> 00:40:51,320 Speaker 13: In fact, they were they had talked about a rebuttal case, 781 00:40:51,800 --> 00:40:54,640 Speaker 13: but that is not going to happen. They, it seems, 782 00:40:54,680 --> 00:40:58,080 Speaker 13: feel confident with the case they presented, which, as we've 783 00:40:58,080 --> 00:41:01,440 Speaker 13: talked about on this show before, it's been pretty comprehensive. 784 00:41:01,440 --> 00:41:06,200 Speaker 13: We've seen multiple close friends, former colleagues of Sam magun 785 00:41:06,200 --> 00:41:11,160 Speaker 13: freed testify to financial crimes as well as a not 786 00:41:11,480 --> 00:41:14,799 Speaker 13: especially inspiring performance by the defendant. On the other hand, 787 00:41:14,840 --> 00:41:18,040 Speaker 13: you know, this is a complicated case, seven counts. The 788 00:41:19,040 --> 00:41:22,080 Speaker 13: business FTX was in has a lot of financial nuance, 789 00:41:22,360 --> 00:41:24,960 Speaker 13: so you never know how it's going to play with 790 00:41:25,000 --> 00:41:26,879 Speaker 13: a jury, and we're just gonna have to wait and see. 791 00:41:27,520 --> 00:41:29,160 Speaker 13: I think it's possible the jury iful you get the 792 00:41:29,200 --> 00:41:32,640 Speaker 13: case as early as tomorrow, certainly some point this week. 793 00:41:33,560 --> 00:41:36,040 Speaker 2: We wait, We watched. We thank you for standing out 794 00:41:36,080 --> 00:41:38,080 Speaker 2: in the cold and getting inside that courtroom for us. 795 00:41:38,080 --> 00:41:41,319 Speaker 2: Smacks traffickin and do go see Ruin as well, which 796 00:41:41,360 --> 00:41:44,480 Speaker 2: he's currently in the original's piece on the undoing of FTX. 797 00:41:44,480 --> 00:41:46,920 Speaker 2: But that does it for this edition of BlueBag Technology yet. 798 00:41:47,640 --> 00:41:49,840 Speaker 4: Yeah, thanks to everyone that's been tuning into the podcast. 799 00:41:49,880 --> 00:41:52,400 Speaker 4: We had some pretty deep conversations around AI today, So 800 00:41:52,480 --> 00:41:56,280 Speaker 4: recap the show wherever you get your podcasts, Apple, Spotify, iHeart, 801 00:41:56,320 --> 00:41:58,360 Speaker 4: and of course we published the podcast to all of 802 00:41:58,360 --> 00:42:01,760 Speaker 4: the bloombag platforms as well. Two days in megaweek ahead 803 00:42:01,960 --> 00:42:03,759 Speaker 4: from San Francisco and New York City. 804 00:42:04,040 --> 00:42:05,400 Speaker 3: This is Bloomberg Technology