WEBVTT - Stock Turnaround and Google's Monopoly Ruling 0:00:01.120 --> 0:00:04.960 We're from markhard where Innovation, Money and Power Collie in 0:00:05.080 --> 0:00:10.000 Silicon Valley, NBN. This is Bloomberg Technology with Caroline Hyde 0:00:10.039 --> 0:00:40.320 and Ed Ludlowly. 0:00:25.520 --> 0:00:28.920 From San Francisco. This is Bloomberg Technology coming up full 0:00:28.960 --> 0:00:32.640 market coverage, a stock stage, a Tuesday turnaround after six 0:00:32.680 --> 0:00:36.720 point five trillion dollars was white from global markets Monday. Plus, 0:00:36.760 --> 0:00:40.080 we sit down with the CTO of palenteer is Ai 0:00:40.159 --> 0:00:43.960 demand boosts, the company's outlook, and the US rules. Google 0:00:44.120 --> 0:00:49.400 illegally monopolized the search market through exclusive deals. It's a 0:00:49.400 --> 0:00:51.519 bit of a sigh of relief on Tuesday. Call it 0:00:51.560 --> 0:00:54.640 a dead cat bounce, call it a turnaround. This is 0:00:54.680 --> 0:00:57.880 what financial markets look like. Notable gains on the Nasdaq 0:00:57.920 --> 0:01:01.920 one hundred chip stocks recovering after it was technology that 0:01:02.000 --> 0:01:04.440 led the selling in equity markets on Monday. A lot 0:01:04.440 --> 0:01:09.240 of volatility around growth concerns and also economic data. And 0:01:09.280 --> 0:01:11.759 then Bitcoin is interesting as well. You saw a fifty 0:01:11.800 --> 0:01:14.000 six thousand US dollars per token. Later in the show 0:01:14.040 --> 0:01:16.080 will go deeper on that. Let's move over to the 0:01:16.120 --> 0:01:18.440 single names, because earnings are a big part of the 0:01:18.480 --> 0:01:20.840 story in the moment. There are two in particular that 0:01:20.840 --> 0:01:23.520 we're looking at. I really want to focus on Uber 0:01:23.600 --> 0:01:26.959 up almost eight percent, strong growth, brooking, strong on the 0:01:26.959 --> 0:01:28.680 bottom line. We're going to get to our reports on 0:01:28.720 --> 0:01:30.520 that in just a moment, and then coming up the 0:01:30.640 --> 0:01:34.520 CTO of Palenteer updated outlook for revenue, updated outlook for 0:01:34.560 --> 0:01:38.759 profit AI demand from both commercial and government customers. That's 0:01:38.800 --> 0:01:41.400 going to be a really interesting conversation in the next segment. 0:01:41.680 --> 0:01:44.360 Stay tuned for it. We'll go back to markets. Joining 0:01:44.400 --> 0:01:47.200 us for more is David Bunsen, chief investment Officer and 0:01:47.280 --> 0:01:50.560 the Buntson Group, which oversees five point seven billion dollars 0:01:50.800 --> 0:01:54.520 in assets under management. The story Monday was about some 0:01:54.560 --> 0:01:57.280 of the megacap tech names, some of those that have 0:01:57.440 --> 0:01:59.760 fueled the gains in the first half of the year. 0:02:00.160 --> 0:02:03.480 Some of those names you were particularly concerned about, so 0:02:03.560 --> 0:02:06.040 you sat on the sidelines away from them. How do 0:02:06.120 --> 0:02:07.920 you react to the last twenty four hours? 0:02:07.960 --> 0:02:12.320 Then, well, it's interesting that with that sector, it wasn't 0:02:12.400 --> 0:02:17.200 really a Monday story per se. It had started July tenth. 0:02:17.200 --> 0:02:21.279 The Nasdaq's drawdown had reached thirteen and a half percent yesterday, 0:02:21.400 --> 0:02:24.680 so there was already quite a bit of slide going on, 0:02:24.840 --> 0:02:28.160 and I think that does stem from the overvaluation. It's 0:02:28.200 --> 0:02:30.680 pretty stunning when you see a name like Nvidia that 0:02:30.760 --> 0:02:33.560 had been down twenty five to thirty percent and was 0:02:33.680 --> 0:02:38.360 still trading at about fifty nine times earnings thirty times 0:02:38.400 --> 0:02:41.000 sales after such a violent correction. 0:02:41.800 --> 0:02:43.040 I think even today. 0:02:42.720 --> 0:02:45.639 When you look at Apple still being negative with this 0:02:45.720 --> 0:02:49.520 kind of bounce back rally today, it does feel to 0:02:49.560 --> 0:02:53.000 me like a rotation of leadership is very well underway. 0:02:54.040 --> 0:02:57.400 You raise a point that we reiterated in yesterday's show 0:02:57.440 --> 0:03:01.360 that Friday marks four straight weekly declared on the Nasdaq 0:03:01.400 --> 0:03:04.200 one hundred. Right, you were already in correction territory. I 0:03:04.200 --> 0:03:05.880 guess a question for someone like you, David, is what 0:03:05.880 --> 0:03:08.160 do you do now? You know, is the draw down 0:03:08.320 --> 0:03:12.200 or so called technical correction enough that you start to 0:03:12.200 --> 0:03:14.679 look at some of the names that for many they 0:03:14.680 --> 0:03:18.959 are teenas or there is no alternative to holding them. 0:03:19.800 --> 0:03:23.040 Well, there are probably investor categories that might want to 0:03:23.080 --> 0:03:26.200 look at something there. They certainly shouldn't be doing it 0:03:26.240 --> 0:03:29.000 because they believe it's reached a value. These are still 0:03:29.040 --> 0:03:34.280 extremely expensive stocks for US as dividend growth investors. The tiny, 0:03:34.320 --> 0:03:38.040 little paltry dividend that the NVIDIAs, Metas and Apples pay 0:03:38.400 --> 0:03:40.880 is not enough to wet our beak, and so they 0:03:40.880 --> 0:03:42.960 would have a long way to go to start returning 0:03:43.440 --> 0:03:45.360 cash to shareholders for us. 0:03:45.240 --> 0:03:46.080 To be attracted. 0:03:46.440 --> 0:03:49.640 But even apart from our dividend growth orientation, the valuations 0:03:49.640 --> 0:03:54.040 are still very high. I began professionally managing money in 0:03:54.040 --> 0:03:56.760 the nineteen nineties, and so I will be forever scarred 0:03:56.760 --> 0:03:59.720 by the reality of what happens when you overpay for 0:03:59.880 --> 0:04:04.000 very good companies, very big and successful companies that have 0:04:04.040 --> 0:04:09.240 a generational success, and yet their stock goes nowhere for ten, fifteen, 0:04:09.360 --> 0:04:12.120 twenty years, and Cisco's case still. 0:04:11.880 --> 0:04:14.080 Hasn't got back to nineteen ninety nine levels. 0:04:14.320 --> 0:04:16.640 Those stories are real, and I believe very much they 0:04:16.640 --> 0:04:18.960 will be the case with some of these names too. 0:04:19.560 --> 0:04:21.200 I think investors have to be careful. 0:04:22.400 --> 0:04:25.960 I find names like Cisco, maybe even Adele interesting as 0:04:26.000 --> 0:04:29.520 sort of appendages to the infrastructure build out that's happening. 0:04:29.520 --> 0:04:32.080 I wrote about it in Today's Tech Daily. Right if 0:04:32.120 --> 0:04:36.880 you think about the growth concern, investmental capital expenditures into 0:04:36.960 --> 0:04:41.559 data center seem to be plowing ahead regardless of short 0:04:41.640 --> 0:04:45.440 term jitters, Is there an opportunity for you in that space, David, 0:04:46.160 --> 0:04:48.960 or do you think it's just overdone already at this point. 0:04:49.839 --> 0:04:52.280 Well, and that's a very interesting story because there you 0:04:52.279 --> 0:04:55.720 can get cash flow. And so, for example, Blackstone has 0:04:55.760 --> 0:04:59.200 become one of the largest investors institutionally in the data 0:04:59.240 --> 0:05:02.480 center's story, and we're able to get a real yield 0:05:02.720 --> 0:05:05.239 and become a landlord, and so to speak to data 0:05:05.279 --> 0:05:09.120 center become a little bit less concerned with the underlying 0:05:09.160 --> 0:05:12.360 business fundamentals as to how much Microsoft's going to buy 0:05:12.360 --> 0:05:15.760 from Nvidia or what the monetization of AI is going 0:05:15.800 --> 0:05:18.080 to be. A lot of those questions are not answered, 0:05:18.120 --> 0:05:20.480 They're not going to get answered in time soon. But 0:05:20.640 --> 0:05:22.880 data center becomes a story that from a brick and 0:05:22.920 --> 0:05:24.640 mortar standpoint, you can make money in. 0:05:25.920 --> 0:05:28.400 I find that so interesting. Never before did I think 0:05:28.440 --> 0:05:32.159 I'd be writing about reates in my technology newsletter. And 0:05:32.200 --> 0:05:35.039 I use their forecast or data to kind of look 0:05:35.080 --> 0:05:38.720 at the market pipeline as a whole earnings. You know, 0:05:38.880 --> 0:05:43.200 this morning we wake up pallentis strength uber strength. Last 0:05:43.200 --> 0:05:46.040 week we learned from Microsoft in particular that if you 0:05:46.120 --> 0:05:49.360 miss estimates, you're in trouble. But there's still a long 0:05:49.400 --> 0:05:52.080 way to run in this earnings period. Have you taken 0:05:52.120 --> 0:05:54.600 a sort of lesson on the whole from those that 0:05:54.640 --> 0:05:56.159 have already reported. 0:05:56.880 --> 0:05:59.440 Yeah, about seventy five percent of the way through. 0:05:59.640 --> 0:06:03.120 It does appear that overall earnings growth is not going 0:06:03.160 --> 0:06:05.600 to be eight percent is targeted on the quarter, but 0:06:05.680 --> 0:06:07.280 probably closer to five percent. 0:06:08.279 --> 0:06:10.280 Year over year, earning's. 0:06:09.920 --> 0:06:12.120 Growth still looks like it going to be about twelve percent, 0:06:12.320 --> 0:06:14.719 but some base effects are going to make that a lot. 0:06:14.560 --> 0:06:15.880 Harder in the next quarters. 0:06:16.440 --> 0:06:18.760 And so I think the story of Microsoft is a 0:06:18.880 --> 0:06:23.360 kind of microcosm of the underlying reality that there's this 0:06:23.520 --> 0:06:26.680 sort of feedback loop at play that Microsoft goes up 0:06:26.760 --> 0:06:29.479 because it's buying a lot from Nvidia, and then Video 0:06:29.600 --> 0:06:32.800 goes up because Microsoft's buying a lot from it, and 0:06:33.279 --> 0:06:35.719 of course there's a lot of other customers. But this 0:06:35.920 --> 0:06:39.279 idea that AI doesn't ever have to tell you how 0:06:39.279 --> 0:06:42.320 it's going to make money for real other than an 0:06:42.320 --> 0:06:47.760 infrastructure backbone story is problematic. And if those revenues are 0:06:47.839 --> 0:06:51.280 going to be going down before they've converted to profits, 0:06:51.760 --> 0:06:54.520 then you have an unwind that has to take place 0:06:54.560 --> 0:06:55.559 of a feedback loop. 0:06:56.800 --> 0:06:59.160 David for the technology set to them, what happens in 0:06:59.200 --> 0:07:00.000 the second half of the day. 0:07:00.160 --> 0:07:05.080 Ye, I would be somewhat unconstructive on the space on 0:07:05.160 --> 0:07:10.120 a valuation basis, But we are hardly the types of 0:07:10.120 --> 0:07:13.840 technology investors that you normally would be speaking to. Just 0:07:13.880 --> 0:07:17.160 to the sense of our boring insistence on free cash flow, 0:07:18.080 --> 0:07:20.640 you know, we own Broadcom in the portfolio. We don't 0:07:20.680 --> 0:07:24.720 believe it got to the valuation overstretch that Nvidia has. 0:07:24.960 --> 0:07:27.560 And then we own old tech names like IBM and 0:07:27.640 --> 0:07:32.200 Cisco that have ongoing free cash flow from oldline businesses, 0:07:32.280 --> 0:07:36.120 but obviously new growth opportunities. That's the way we would 0:07:36.160 --> 0:07:39.240 prefer to play it, just simply because we think perfection 0:07:39.520 --> 0:07:43.240 was priced in to the more contemporary technology names. 0:07:44.000 --> 0:07:46.520 David, I do not for a second. Thank you are boring. 0:07:46.840 --> 0:07:49.880 David Bunsen, CEO at the Punts and Groove, has been 0:07:49.880 --> 0:07:51.720 great sorts you. Thank you very much. Let's get back 0:07:51.720 --> 0:07:55.160 to the earning story. Speak about Uber reporting results before 0:07:55.160 --> 0:07:57.120 the opening bell this morning. I want to bring in 0:07:57.120 --> 0:08:00.480 Bloomberg's Nati Lung, who covers the gig economy platform for us, 0:08:01.000 --> 0:08:03.720 and just go straight to gross bookings. It's basically everything 0:08:03.720 --> 0:08:07.840 that runs through the Uber platform apart from the tips, 0:08:08.320 --> 0:08:10.440 but within that there are loads of stories. What did 0:08:10.440 --> 0:08:12.440 you learn about Uber, Natalie, Yes. 0:08:12.360 --> 0:08:16.000 So from growth brookings it includes ride share and delivery 0:08:16.080 --> 0:08:20.240 and user frequency and as well as driver activity on 0:08:20.280 --> 0:08:23.360 the platform is all time high. And this result is 0:08:23.400 --> 0:08:27.080 really showing how Uber's strategy of pushing into new areas 0:08:27.200 --> 0:08:30.080 new use cases are playing out really well for them. 0:08:30.160 --> 0:08:33.960 And for example, they launch affordable options such as shuttles 0:08:33.960 --> 0:08:37.320 and lower cost shared rides. That's sort of contributing to 0:08:37.360 --> 0:08:40.840 the counterstick close story that the CEO talked about on 0:08:40.840 --> 0:08:44.199 their call and helping them sort of be whether they're 0:08:44.240 --> 0:08:47.480 sort of consumer downturn bluebergs. 0:08:47.520 --> 0:08:49.720 And that's all across Uber, and there's a lot more 0:08:49.760 --> 0:08:51.959 to come in that space. Of course, Door Dash Uber 0:08:52.160 --> 0:08:53.640 and then we get Lift and that will give us 0:08:53.679 --> 0:08:56.440 a bigger picture of the strength in that market. Coming 0:08:56.520 --> 0:08:59.920 up on this program, we're joined by Shyam Sanka, CTO 0:09:00.360 --> 0:09:04.160 of Palenteer discussing the company's earnings released after the closing 0:09:04.160 --> 0:09:06.640 bell Monday. One of the big movers to the upside 0:09:06.640 --> 0:09:11.680 this morning. Commercial customers, military customers, government customers. That's the conversation. 0:09:12.040 --> 0:09:25.120 This is Bloomberg Technology. Let's keep it going with Ernie's coverage. 0:09:25.160 --> 0:09:28.280 Palenteer out with results, boosting its revenue guidance for the 0:09:28.360 --> 0:09:30.840 year to arrange of two point seventy four billion to 0:09:30.840 --> 0:09:34.240 two point seven five billion, ahead of estimates, with AI 0:09:34.360 --> 0:09:37.880 demand also helping Palenteer boost its profit outlook. There's been 0:09:37.920 --> 0:09:42.319 growth in business with government and in commercial customers. Aipalenteers 0:09:42.640 --> 0:09:47.320 artificial intelligence platform and other products has quote transformed the 0:09:47.400 --> 0:09:51.480 business in level more than a year. CTO Sean Sanka 0:09:51.600 --> 0:09:57.000 joins us to discuss. Sean on the call this transformation 0:09:57.120 --> 0:09:59.480 that doctor cart was talking about. I've been to a 0:09:59.520 --> 0:10:03.840 few AIP cons, but as CTO, you're working on it. 0:10:04.040 --> 0:10:06.680 I think just explain the basics of that. What's changed 0:10:06.679 --> 0:10:10.640 with Impaleteer from a technology perspective and in the domain 0:10:10.679 --> 0:10:12.480 of artificial intelligence. 0:10:13.360 --> 0:10:15.160 Well, what you really see in the market is this 0:10:15.280 --> 0:10:19.760 massive bottleneck between prototyping and production, and that happens to 0:10:19.800 --> 0:10:21.680 be where AIP is most differentiated. 0:10:21.720 --> 0:10:24.880 And that differentiation is built on a decade of deep. 0:10:24.679 --> 0:10:29.320 Technical investments, investments like the ontology, the OSDK, the security 0:10:29.720 --> 0:10:34.280 and business primitives that we built throughout the platforms like functions, actions, automations, 0:10:34.520 --> 0:10:36.800 and this pipeline that we have that's really focused on 0:10:36.840 --> 0:10:40.000 addressing that, and I think what's changed as you've been 0:10:40.040 --> 0:10:43.439 to these AIP cons is that the market now understands 0:10:43.440 --> 0:10:45.880 how severe that bottleneck is. I think your last guest 0:10:46.000 --> 0:10:49.240 was just talking about that. Where it is so easy 0:10:49.600 --> 0:10:52.520 to build a charismatic AI prototype. That's about the amount 0:10:52.520 --> 0:10:55.120 of effort of building a PowerPoint slide, but it's also 0:10:55.760 --> 0:11:00.319 that amount of utility. An Unlike traditional deterministic software, this 0:11:00.480 --> 0:11:04.440 kind of powerful stochastic genie that is LM's, it requires 0:11:04.600 --> 0:11:06.600 a lot more work to get to production, maybe ten 0:11:06.640 --> 0:11:09.239 to one hundred times as much work, and that requires 0:11:09.280 --> 0:11:11.040 the tool chain that we've assembled with AP. 0:11:12.960 --> 0:11:15.199 I've had a few conversations with your CEO at his 0:11:15.280 --> 0:11:18.440 cap about his frustration with the PowerPoint deck versus the 0:11:18.480 --> 0:11:21.280 reality of shipping products. We'll put that to one side 0:11:21.320 --> 0:11:24.040 for now. Show on there was a heavy emphasis on 0:11:24.120 --> 0:11:27.200 working with the military in the written materials and on 0:11:27.240 --> 0:11:29.800 the call. Could you just get the basics of what 0:11:29.880 --> 0:11:32.439 you'll work with the military looks like present day. 0:11:33.559 --> 0:11:35.439 Look In the commercial world, you call it a value 0:11:35.520 --> 0:11:37.240 chain from the hand of your supplier to the hand 0:11:37.280 --> 0:11:39.319 of your customer. In the military, it's about a kill 0:11:39.400 --> 0:11:41.880 chain from censor to shooter. But really, at the most 0:11:41.880 --> 0:11:44.720 abstract level, it's the same thing. We're trying to enable 0:11:44.720 --> 0:11:48.160 our war fighters to have information dominance and decision advantage. 0:11:48.240 --> 0:11:50.199 How can I see you know, to quote Sun Sou, 0:11:50.880 --> 0:11:52.480 if you know your enemy and you know yourself, you're 0:11:52.480 --> 0:11:54.520 going to win, and so you know, how can I 0:11:54.559 --> 0:11:57.040 see everything there is to know about the threat? How 0:11:57.040 --> 0:11:59.520 can I understand everything I have to combat and deter 0:12:00.000 --> 0:12:01.840 any aggression from those threats? And when you look at 0:12:01.840 --> 0:12:05.040 the geopolitical landscape right now, it could not be more dangerous. 0:12:05.080 --> 0:12:08.120 Everything that's going on in Eastern Europe, the massive tensions 0:12:08.120 --> 0:12:12.000 that exist in the Middle East, and the ongoing issues 0:12:12.000 --> 0:12:14.000 that we had in de during aggression in the Pacific. 0:12:15.960 --> 0:12:19.720 Beyond sort of the importance of data. Where does plan 0:12:19.840 --> 0:12:22.480 to sit in the defense ecosystem or even the defense 0:12:22.520 --> 0:12:25.600 supply chain. Are you sort of very closely aligned with 0:12:26.040 --> 0:12:29.880 hardware makers the aerospace community, or is the model more 0:12:29.880 --> 0:12:32.719 focused on you just going direct to different arms of 0:12:33.240 --> 0:12:34.600 the defense base with government. 0:12:35.800 --> 0:12:37.320 Yeah, we're touching all of it. One of the things 0:12:37.360 --> 0:12:38.240 I'm most excited about. 0:12:38.240 --> 0:12:40.360 I've been calling the First Breakfast as an antidote to 0:12:40.400 --> 0:12:42.800 the nineteen ninety three last Supper that led to the 0:12:42.800 --> 0:12:45.480 consolidation of our defense industrial base, where it went from 0:12:45.679 --> 0:12:48.600 fifty one primes down to five, because we forget that 0:12:48.640 --> 0:12:50.520 at the dawn of World War Two, we didn't have 0:12:50.559 --> 0:12:51.880 a defense industrial base. 0:12:51.920 --> 0:12:52.439 We had an. 0:12:52.400 --> 0:12:56.440 American industrial base. Chrysler made missiles, General mills, this Serial 0:12:56.520 --> 0:12:59.520 company made inertial guidance systems. And so we have this 0:12:59.640 --> 0:13:03.080 moment right now in defense tech where one hundred billion 0:13:03.120 --> 0:13:05.640 dollars or more of capital has been deployed, a bolust 0:13:05.679 --> 0:13:07.840 of founders have shown up. There's a lot of creativity, 0:13:07.840 --> 0:13:10.439 a lot of energy as a company that that's kind 0:13:10.440 --> 0:13:13.240 of been pathfinding over twenty years. Not only have we 0:13:13.280 --> 0:13:17.120 developed our software that gives these warfighters unique advantage like 0:13:17.200 --> 0:13:20.360 the Maven contract a CDO just recently awarded for nearly 0:13:20.360 --> 0:13:24.360 half a billion dollars, but really this software infrastructure that's 0:13:24.400 --> 0:13:28.160 required to deliver modern American software to the battlefield in 0:13:28.240 --> 0:13:31.600 air gapped environments at speed, at pace, well ahead of 0:13:31.640 --> 0:13:34.080 the threats of twenty twenty seven. So we've been working 0:13:34.160 --> 0:13:37.800 very closely with both traditional primes, integrating our software to 0:13:37.840 --> 0:13:40.480 their hardware, helping them actually with production. If you think 0:13:40.480 --> 0:13:42.760 about the fifty percent of our business that's commercial oriented. 0:13:42.920 --> 0:13:46.080 How do we build jet engines and satellites faster, better, cheaper? 0:13:47.280 --> 0:13:49.800 In addition to the new entrants who need to achieve 0:13:49.840 --> 0:13:54.280 scale and time. Value of money is everything for them. 0:13:54.520 --> 0:13:57.880 Whenever we have an executive on this program, I always 0:13:57.880 --> 0:14:00.079 got to tell audience and say, you know, what is 0:14:00.080 --> 0:14:02.720 that you want to know from? In this case Palenteer, 0:14:03.160 --> 0:14:05.679 I would say most of the questions were about warp speed, 0:14:07.120 --> 0:14:11.040 very basic ones. Why now with warp speed and the 0:14:11.080 --> 0:14:14.439 obstacles to rolling it out making it sort of more 0:14:15.480 --> 0:14:17.920 readily available, pervasive out their shop. 0:14:18.920 --> 0:14:19.760 Yeah, that's great. 0:14:19.880 --> 0:14:21.680 I'm so excited about warp speed is what I'm spending 0:14:21.680 --> 0:14:23.360 all of my time on and really shaping the R 0:14:23.360 --> 0:14:26.800 and D roadmap around warp speed is our modern American 0:14:26.880 --> 0:14:30.600 operating system for manufacturing and the reason why now you know, 0:14:30.720 --> 0:14:32.640 for the better part of twenty years we have helped 0:14:32.920 --> 0:14:37.560 traditional manufacturers build planes, trains, automobiles, and ships. But most 0:14:37.560 --> 0:14:39.640 of those folks are stuck in a legacy mode of 0:14:39.680 --> 0:14:42.040 how they're operating, and you're able to help here or there. 0:14:42.280 --> 0:14:45.920 But what's unique about the reindustrialization movement that's happening right 0:14:45.920 --> 0:14:50.440 now in America is that these founders, they're alumni of Palenter, 0:14:50.480 --> 0:14:53.520 of Tesla, of SpaceX, and they understand that the traditional 0:14:53.840 --> 0:14:58.880 erp plm PLC software doesn't really work. That most of 0:14:58.920 --> 0:15:01.680 these successful companies have to build their own software and 0:15:01.720 --> 0:15:03.400 that is really an unaffordable journey. 0:15:03.440 --> 0:15:05.520 So that there's this massive opportunity to take. 0:15:05.360 --> 0:15:09.160 The power of AIP and the historical experiences we have 0:15:09.440 --> 0:15:13.040 throughout the value chain of production to help our customers 0:15:13.120 --> 0:15:15.520 bend their atoms better with bits. 0:15:17.080 --> 0:15:17.400 SEUP. 0:15:17.400 --> 0:15:19.320 I have your CTO. I kind of have some quick 0:15:19.320 --> 0:15:22.080 fire questions that I've never been able to answer about 0:15:22.120 --> 0:15:26.720 Palenteer in the first instance, AIP and warp speed over 0:15:26.800 --> 0:15:30.040 the year. What's the kind of key foundational model LM 0:15:30.160 --> 0:15:31.920 that you've been building on top of. I know you 0:15:32.000 --> 0:15:35.680 partner as well as sort of think in house about 0:15:35.720 --> 0:15:38.600 the model or foundation level, but that's a question that 0:15:38.640 --> 0:15:40.480 comes up quite a lot. Who are you building on 0:15:40.520 --> 0:15:41.600 top of and working with? 0:15:43.000 --> 0:15:44.840 Well, I think all the value is really going to 0:15:44.880 --> 0:15:46.920 create up the application layer, and what we've seen in 0:15:46.960 --> 0:15:49.640 production is that you actually need a menagerie of models 0:15:49.920 --> 0:15:51.400 and you you know, if you think about the most 0:15:51.440 --> 0:15:54.800 expensive frontier model out there right now, it's a thousand 0:15:54.880 --> 0:15:58.760 times more expensive than the cheapest open source model for 0:15:58.880 --> 0:16:01.560 ten percent more ELO, we're ten percent more IQ as 0:16:01.560 --> 0:16:04.560 a rough proxy, that's not a compelling price performance trade off. 0:16:04.920 --> 0:16:07.320 You already see that the first versions of GPT four 0:16:07.360 --> 0:16:09.720 have been sunset. You know, we need to think about 0:16:09.720 --> 0:16:12.640 this problem as Okay, what is the infrastructure I need 0:16:12.920 --> 0:16:15.400 to make I'm turning my software into something that is 0:16:15.400 --> 0:16:16.280 now stochastic. 0:16:16.400 --> 0:16:17.920 It's not deterministic anymore. 0:16:18.440 --> 0:16:20.760 Pure scientists that we have trained, they are all used 0:16:20.800 --> 0:16:22.440 to writing deterministic code. 0:16:22.680 --> 0:16:22.920 You know. 0:16:23.040 --> 0:16:25.600 This is this harkens more to how we think about 0:16:25.720 --> 0:16:28.560 the transition from analog circuits to digital circuits. The sort 0:16:28.560 --> 0:16:30.680 of error correction you need, the sort of infrastructure you 0:16:30.680 --> 0:16:32.720 need to think about having so that you can have 0:16:32.800 --> 0:16:33.640 the abstraction to. 0:16:33.640 --> 0:16:36.280 Think about these things as being digital. That's where I. 0:16:36.240 --> 0:16:40.440 Think the coluy is going to be go ahead, sorry, 0:16:40.840 --> 0:16:43.440 And so we're very focused on helping our customers get 0:16:43.440 --> 0:16:46.320 the right models for the right use cases. Auto evaluate 0:16:46.400 --> 0:16:50.360 that auto automatically generate iterations on the prompts that get 0:16:50.360 --> 0:16:52.400 them there. And part of our theory is that really 0:16:52.440 --> 0:16:55.320 prompts our for developers, chat is a dead end. We're 0:16:55.320 --> 0:16:57.440 guiding our customers through this journey here. They get there 0:16:57.440 --> 0:16:59.680 pretty quickly to realize that we should be thinking of 0:16:59.760 --> 0:17:01.720 l is a new type of runtime, the way I 0:17:01.800 --> 0:17:03.000 might write a Python function. 0:17:03.320 --> 0:17:05.320 Well, okay, I'm going to write an LBM function too. 0:17:06.720 --> 0:17:09.160 Sean. For the commercial half of the business, is there 0:17:09.160 --> 0:17:12.800 a specific Hyperscale cloud platform or partner that you work 0:17:12.880 --> 0:17:14.159 with to support its growth. 0:17:15.680 --> 0:17:16.640 We're working with all of them. 0:17:16.680 --> 0:17:20.440 We have deep and very valuable relationships there, so we're 0:17:20.520 --> 0:17:21.119 very happy with that. 0:17:22.960 --> 0:17:26.480 This last year has been about AIP and growth. What's 0:17:26.600 --> 0:17:30.119 the next twelve months like for Palenteer sham Well, I. 0:17:30.119 --> 0:17:32.399 Think it's really about deepening the investments that we have 0:17:32.760 --> 0:17:36.400 with AIP that are addressing this bottleneck between prototyping and production. 0:17:36.760 --> 0:17:39.000 You know, if you look at David Kahn at Sequoia's 0:17:39.119 --> 0:17:42.360 article on the six hundred billion dollar whole in revenue. 0:17:41.960 --> 0:17:43.080 I think this is where the whole is. 0:17:43.160 --> 0:17:45.120 This is the bottleneck in the market is the most 0:17:45.119 --> 0:17:47.760 important problem to solve, and the folks who solve at first, 0:17:47.760 --> 0:17:49.520 and I think we're in the poor position there, have 0:17:49.600 --> 0:17:51.600 the opportunity to take the entire market. 0:17:52.640 --> 0:17:55.040 Sham sangk Has, CTO of Paleteer. It's great to have 0:17:55.119 --> 0:18:05.919 you back on Bloomberg Technology. Thank you so much. In 0:18:06.000 --> 0:18:09.440 other tech news, Google lost in a historic anti trust 0:18:09.520 --> 0:18:11.840 suit after the tech giant had paid twenty six billion 0:18:11.920 --> 0:18:16.120 dollars to become the default search engine on smartphone web browsers. 0:18:16.119 --> 0:18:18.520 I want to bring in bloombos Lea Nyland, who's been 0:18:18.560 --> 0:18:21.439 reporting with the team on this case. Let's start with 0:18:21.480 --> 0:18:25.320 the basics the decision and how it will impact Alphabet 0:18:25.640 --> 0:18:27.760 or Google Alphabet, the parent of Google. 0:18:28.640 --> 0:18:31.080 Yeah, so the decision that came out yesterday is actually 0:18:31.200 --> 0:18:34.040 just the one on whether Google violated the law. So 0:18:34.880 --> 0:18:37.840 the judge found that Google has an illegal monopoly over 0:18:37.960 --> 0:18:41.159 two markets, the general search services market, which is what 0:18:41.240 --> 0:18:44.000 consumers used to find things on the web, and search 0:18:44.080 --> 0:18:45.879 text ads, which are the ones that are at the 0:18:45.880 --> 0:18:48.160 top of the page that sort of try and luy 0:18:48.200 --> 0:18:51.040 you to websites. And so this was just as I 0:18:51.040 --> 0:18:53.000 said on the liability. Now we're going to go to 0:18:53.160 --> 0:18:55.280 sort of the next phase of the case, which is 0:18:55.359 --> 0:18:57.800 where we're going to talk about what to do about this, 0:18:58.920 --> 0:19:02.280 And the Justice Department hasn't necessarily said what they want yet, 0:19:02.280 --> 0:19:05.360 But there's sort of several options that people think, one 0:19:05.359 --> 0:19:09.240 of which is a potential breakup. So they could ask 0:19:09.400 --> 0:19:13.160 for the court to make Google sell off the Chrome 0:19:13.240 --> 0:19:16.640 browser or the Android operating system, both of which were 0:19:17.440 --> 0:19:20.520 key things in the trial. Or the judge could go 0:19:20.600 --> 0:19:25.560 with maybe like a slightly lesser penalty or remedy, which 0:19:25.920 --> 0:19:28.359 such as making Google share some of the data that 0:19:28.440 --> 0:19:31.880 it collects that sort of underlies all of its search 0:19:31.920 --> 0:19:35.679 results with other search engines like Being or search startups. 0:19:38.640 --> 0:19:40.840 You kind of answered what happens next, but a lot 0:19:40.840 --> 0:19:43.240 of questions I got on social media was does the 0:19:43.320 --> 0:19:46.080 ruling have teeth? If you see what I mean? You know, 0:19:46.080 --> 0:19:48.080 there are some opinion pieces out there about this being 0:19:48.160 --> 0:19:51.640 kind of a hollow victory for regulators and the stocks 0:19:51.720 --> 0:19:54.359 kind of flat. It's hard to gauge how serious this 0:19:54.440 --> 0:19:57.919 is is guess what I'm saying about alphabet, But it 0:19:58.000 --> 0:19:59.160 is important, right. 0:20:00.080 --> 0:20:03.080 Yeah, It's probably not going to take any impact immediately. 0:20:03.160 --> 0:20:05.960 Google has said it's going to appeal. Those sorts of 0:20:06.000 --> 0:20:08.879 appeals generally take a year to eighteen months, so of 0:20:08.920 --> 0:20:12.000 course this is going to delay any actual changes to 0:20:12.280 --> 0:20:15.760 Google's search engine or its business. But what the decision 0:20:15.880 --> 0:20:18.720 means is that long term, like Google is definitely going 0:20:18.760 --> 0:20:20.960 to have to change its practices. It's not going to 0:20:21.000 --> 0:20:25.960 remain the sort of dominant search engine for very long, 0:20:26.560 --> 0:20:29.600 and it may end up having to hive office. I said, 0:20:29.720 --> 0:20:32.119 certain parts of its business which have provided you know, 0:20:32.600 --> 0:20:34.080 a nice cushion to the company. 0:20:34.960 --> 0:20:36.359 The one thing that you did. 0:20:36.200 --> 0:20:38.480 See sort of yesterday was there was a pretty sharp 0:20:38.760 --> 0:20:41.439 stock seller for Apple because twenty billion dollars of that 0:20:41.480 --> 0:20:44.400 twenty six billion dollars that you mentioned had been going 0:20:44.440 --> 0:20:47.879 to Apple every year. It simply, you know, had this 0:20:47.920 --> 0:20:51.080 contract with Google. It got twenty billion dollars pretty much 0:20:51.080 --> 0:20:53.760 for not doing anything except letting them be the default 0:20:53.760 --> 0:20:57.119 on the iPhone. And this ruling definitely puts that in 0:20:57.200 --> 0:21:00.480