WEBVTT - Bloomberg Businessweek Weekend - May 24th, 2024 0:00:02.440 --> 0:00:07.480 Bloomberg Audio Studios, Podcasts, radio news. 0:00:08.360 --> 0:00:12.760 This is Bloomberg business Week inside from the reporters and 0:00:12.920 --> 0:00:16.480 editors who bring you America's most trusted business magazine, plus 0:00:16.560 --> 0:00:20.680 global business finance and tech news. The Bloomberg Business Week 0:00:20.720 --> 0:00:25.640 Podcast with Carol Messer and Tim Stenebeck from Bloomberg Radio. 0:00:26.880 --> 0:00:29.720 Hi, everyone, Welcome to the Bloomberg Business Week Weekend Podcast. 0:00:30.200 --> 0:00:33.360 Big focal point this week, I mean massive focal point. 0:00:33.400 --> 0:00:36.760 Let's be fair earnings from Nvidia, you know, the giant 0:00:36.800 --> 0:00:39.599 chip maker at the heart of the artificial intelligence revolution 0:00:39.720 --> 0:00:43.480 that has powered the bullmarket in stocks. More on that 0:00:43.640 --> 0:00:44.440 in just a moment. 0:00:44.600 --> 0:00:48.159 Man speaking of Nvidia, Jensen Wang, Satynadella of Microsoft, and 0:00:48.200 --> 0:00:50.040 Michael Dell. A lot of big names in the tech 0:00:50.080 --> 0:00:53.080 space this past week talking about AI and their plans 0:00:53.120 --> 0:00:55.800 and their partnerships. Our take on what is going on 0:00:55.880 --> 0:00:58.600 from the head of AI over at Qualcom, who weighed 0:00:58.640 --> 0:01:03.520 in on on device artificial intelligence, including so called aipcs. 0:01:03.600 --> 0:01:05.200 You're going to hear a lot about that in the 0:01:05.240 --> 0:01:08.240 coming months, that's for sure. Venture Capital crossed into a 0:01:08.280 --> 0:01:10.840 few topics for us this week as well, including a 0:01:10.880 --> 0:01:13.880 conversation with the author of the Venture mindset how to 0:01:13.920 --> 0:01:16.959 make smarter bets and achieve extraordinary growth. 0:01:17.240 --> 0:01:19.959 And speaking of VC bets, joe Ane corkran over at 0:01:20.000 --> 0:01:23.000 Golden Seeds joins us on investing in early stage women 0:01:23.080 --> 0:01:25.880 led businesses, and then the Business Week cover story It's 0:01:25.920 --> 0:01:28.959 about twenty three and meters once the recipient of millions 0:01:29.000 --> 0:01:33.400 of dollars from vcs and investors facing a big problem today. 0:01:33.280 --> 0:01:35.880 All of that to come. We begin with Nvidia, which 0:01:35.920 --> 0:01:39.959 gave another bullish forecast, shattering estimates showing that spending on 0:01:40.040 --> 0:01:44.000 AI computing remained strong and reinforcing its stats as the 0:01:44.000 --> 0:01:46.560 biggest beneficiary of AI spending. 0:01:46.760 --> 0:01:49.440 The company's so called AI accelerators, These are chips that 0:01:49.480 --> 0:01:52.360 help data centers develop chatbots and other cutting edge tools 0:01:52.480 --> 0:01:54.880 have become a hot commodity. For more on the quarter 0:01:54.960 --> 0:01:56.840 and the business, we caught up with the Bloomberg News 0:01:56.880 --> 0:01:59.240 at US semiconductor reporter Ian King. 0:02:00.120 --> 0:02:02.800 Were looking for doubling year of a year in this 0:02:02.840 --> 0:02:05.480 current quarter. In video, said yeah, we can do that, 0:02:05.520 --> 0:02:07.560 and we can do a bit more. So, even though 0:02:07.560 --> 0:02:13.320 there were lofty expectations, they managed to set expectations even higher. 0:02:13.760 --> 0:02:19.200 The data center business is already bigger than Intel an AMD, 0:02:20.000 --> 0:02:24.079 and it's getting bigger quicker than people had projected more 0:02:24.120 --> 0:02:26.480 than four hundred percent a year on year increase in 0:02:26.480 --> 0:02:30.600 that particular business. So all of the fundamental numbers that 0:02:30.720 --> 0:02:34.160 in Nvidia need to show we're heading in the right direction. 0:02:34.400 --> 0:02:35.560 It was able to deliver on. 0:02:36.680 --> 0:02:39.000 I was a little struck, and Carol mentioned this by 0:02:39.000 --> 0:02:41.080 a couple of the things that they also announced the 0:02:41.120 --> 0:02:44.000 ten to one stock split coming next month, and then 0:02:44.080 --> 0:02:46.120 increasing the dividend. Curious if you have any thoughts on 0:02:46.160 --> 0:02:46.880 either of those things. 0:02:47.400 --> 0:02:50.040 Well, I mean, you know, as you were just discussing 0:02:50.040 --> 0:02:52.640 with Bailey, I believe that the stock split is something 0:02:52.680 --> 0:02:55.839 that's aimed at retail investors, but it's also aimed at 0:02:56.160 --> 0:03:00.360 their internal audience as well. A certain percentage of the 0:03:00.360 --> 0:03:03.480 shares get bought by individuals. You know, they get like 0:03:03.520 --> 0:03:07.640 a discount slash, you know, allowance that they can buy 0:03:07.680 --> 0:03:10.040 at a certain point. So you know they want to 0:03:10.040 --> 0:03:12.480 to keep their employees happily, right. You need those engineers, 0:03:12.480 --> 0:03:15.800 you need those that staff that's that's you know, so 0:03:15.919 --> 0:03:18.239 invested in what you're doing, So that might be part 0:03:18.240 --> 0:03:18.799 of it as well. 0:03:18.919 --> 0:03:20.520 Can I ask you though, because I did scratch my 0:03:20.520 --> 0:03:21.520 head a little bit about it. 0:03:21.560 --> 0:03:21.840 I don't know. 0:03:21.840 --> 0:03:23.760 I was going back and forth with a brother of 0:03:23.800 --> 0:03:26.000 mine who has been in an Nvidia for years, and 0:03:26.040 --> 0:03:26.760 we're going back and. 0:03:26.919 --> 0:03:28.480 Got to get it. Your brother on the show. 0:03:28.680 --> 0:03:32.120 Yea, yeah, he's pretty smart, just really well read. But 0:03:32.600 --> 0:03:35.040 about the dividend, because I always feel like companies do 0:03:35.160 --> 0:03:37.800 that when maybe growth is slowing or something to kind 0:03:37.800 --> 0:03:40.680 of keep investors happy. Is there something, though, connected with 0:03:40.720 --> 0:03:43.400 that stock split that they raise the dividends so much? 0:03:44.400 --> 0:03:46.840 I mean, for some of you conductor companies, which are 0:03:46.840 --> 0:03:51.480 massively capital intensive, even when they don't own their own manufacturing, 0:03:51.560 --> 0:03:54.560 that the question is how much are they spending as 0:03:54.600 --> 0:03:59.200 a percentage of their revenue on R and D, on design, 0:03:59.440 --> 0:04:01.680 on making sure that the next product's coming along? And 0:04:02.000 --> 0:04:04.920 you look, you know, Nvidia is one of the highest, 0:04:04.960 --> 0:04:08.000 one of the biggest spenders by any measure in the 0:04:08.080 --> 0:04:10.040 chip industry. So as long as I'm doing that, I 0:04:10.040 --> 0:04:13.160 would say that most in video watchers don't care. It's 0:04:13.240 --> 0:04:16.120 just something nice to have. On the other side of things. 0:04:16.040 --> 0:04:18.800 Well, you talk about the R and D spend and 0:04:18.839 --> 0:04:21.080 it really, you know, certainly seems to be paying off 0:04:21.080 --> 0:04:23.200 for Nvidia. Having said that, talk to us about the 0:04:23.240 --> 0:04:25.080 development cycle, and I bring it up because I was 0:04:25.120 --> 0:04:29.359 listening to Bloomberg Intelligence man Deep seeing about how for 0:04:29.520 --> 0:04:32.960 Nvidia it's kind of a shorter time like almost every year, 0:04:33.000 --> 0:04:35.800 and that that is going to require companies who want 0:04:35.839 --> 0:04:38.440 to have the latest and greatest from Nvidia and in 0:04:38.480 --> 0:04:40.919 the AI chip world, that they're going to be adjusting 0:04:40.960 --> 0:04:44.719 their CAPEX spend to include that every year. So maybe 0:04:44.839 --> 0:04:47.160 you know, different from Moore's law. Right. 0:04:47.520 --> 0:04:50.240 Basically what Jensen lang to CEO said, he said, Look, 0:04:50.320 --> 0:04:53.919 we were bringing new product market for this market roughly 0:04:53.960 --> 0:04:57.200 sort of every eighteen months or so. Now we're going 0:04:57.240 --> 0:04:57.520 to do. 0:04:57.480 --> 0:04:58.160 It every year. 0:04:58.360 --> 0:04:59.719 We're going to do it every year for a number 0:04:59.720 --> 0:05:04.360 of he you know, put that the salesman routine. Look, 0:05:04.360 --> 0:05:07.640 we're helping our customers. This market is changing rapidly, so 0:05:07.680 --> 0:05:10.560 we need to bring soft and market much much quicker 0:05:10.600 --> 0:05:12.760 than we were doing to help support this, to help 0:05:12.760 --> 0:05:16.640 bring these new capabilities, everybody wins everything, you know, the 0:05:16.640 --> 0:05:19.880 pace of innovation increases. Flip side of that is, you 0:05:19.960 --> 0:05:23.320 remember earlier this year we were talking about what AMD's doing, 0:05:23.520 --> 0:05:26.039 what Intel's doing, you know, how are they going to 0:05:26.040 --> 0:05:28.800 get into the market and what Jensen Wang has basically 0:05:28.839 --> 0:05:31.120 said to them, is like that moving train that you 0:05:31.120 --> 0:05:33.200 are trying to catch is going even faster. 0:05:33.320 --> 0:05:33.760 Good look. 0:05:35.279 --> 0:05:37.640 So that's what's really going on on a competitive basis 0:05:37.640 --> 0:05:38.200 there as well. 0:05:38.240 --> 0:05:40.080 And Moore's law I meant doubling every two years, so 0:05:40.160 --> 0:05:42.120 forgive me, yeah, much faster here. 0:05:42.040 --> 0:05:43.800 Ian Ian, I want to to that end of this 0:05:44.200 --> 0:05:45.640 of jumping on that train. I want to talk a 0:05:45.640 --> 0:05:49.040 little bit about margins, because the company said that margins 0:05:49.080 --> 0:05:52.159 for the gross margins will be seventy five point seven 0:05:52.520 --> 0:05:57.480 five percent in the current quarter. As you noted, this 0:05:57.560 --> 0:05:59.760 is more like a software company rather than a chip 0:05:59.760 --> 0:06:02.040 comp And you yesterday talked about Intel being in the 0:06:02.080 --> 0:06:05.279 low sixties at its height when it really owned the 0:06:05.400 --> 0:06:08.599 server market. What can history tell us about the duration 0:06:08.720 --> 0:06:11.839 and durability of margins like this for a chip company. 0:06:12.600 --> 0:06:15.440 Yeah, I mean, right now, their margins are what they 0:06:15.480 --> 0:06:18.840 want them to be because it's just really a case of, well, 0:06:19.720 --> 0:06:22.520 our prices a sky high were sold out. We can 0:06:22.839 --> 0:06:25.640 essentially charge what we want, so obviously the pricing is 0:06:25.720 --> 0:06:27.920 very good for margins. On the flip side of that, 0:06:28.000 --> 0:06:30.520 as we just discussed, they can spend what they want. 0:06:30.600 --> 0:06:32.239 It's you know, how much do we want to spend 0:06:32.279 --> 0:06:34.800 on R and D, how well do we want to 0:06:34.839 --> 0:06:39.000 reward our employees. So the margins are sky high, but 0:06:39.240 --> 0:06:42.880 they could make them artificially higher. The fact that they're 0:06:42.880 --> 0:06:46.120 spending heavily, the fact that they're able to charge a lot, 0:06:46.200 --> 0:06:50.120 that's something that's going to probably persist. So they're setting 0:06:50.240 --> 0:06:52.720 very high targets, but they could be even higher. So, 0:06:53.080 --> 0:06:56.279 you know, as a measure of their health, it's pretty good. 0:06:56.320 --> 0:06:56.560 Stay. 0:06:56.680 --> 0:06:59.760 Do you agree with Jensen's comment that the freight train 0:06:59.839 --> 0:07:02.480 is moving faster and these companies will have a more 0:07:02.520 --> 0:07:05.000 difficult time, The AMDs and the like will have a 0:07:05.040 --> 0:07:06.799 more difficult time catching up with Nvidia. 0:07:07.560 --> 0:07:11.120 I mean Jensen's racing. I mean this one of the 0:07:11.160 --> 0:07:14.360 things that comes across in pretty much any time he 0:07:14.520 --> 0:07:19.200 says something, he's constantly aware of the tension that exists 0:07:19.200 --> 0:07:22.400 in his business and his market, and he's consciously trying 0:07:22.480 --> 0:07:26.280 to shape post what his company does, and his understanding 0:07:26.360 --> 0:07:29.280 of the relationships where we are in the cycle, where 0:07:29.280 --> 0:07:32.000 we are relative to the competition, what the customers want. 0:07:32.040 --> 0:07:34.760 He's trying to shape his narrative to fit that. He's 0:07:35.400 --> 0:07:38.280 very clever, and he's showing people that he understands what's 0:07:38.320 --> 0:07:41.240 going on and what he has to do to stay ahead, 0:07:41.280 --> 0:07:44.840 and that kind of institutional paranoia, that history of being 0:07:44.840 --> 0:07:47.760 the little guy still very much present in what he's doing. 0:07:48.160 --> 0:07:52.440 Unbelievable and just a fascinating story. Ian. Thank you. I 0:07:52.480 --> 0:07:54.040 know it's been a busy twenty four hours for you, 0:07:54.120 --> 0:07:57.120 so we really appreciate it. Ian King are US Semiconductor 0:07:57.160 --> 0:07:59.200 and Networking reporter at Bloomberg News, joining us from our 0:07:59.280 --> 0:08:03.000 San Francisco bureau, as we mentioned, and Vida shares continuing 0:08:03.040 --> 0:08:06.680 to trade higher, still up about ten percent in today's session, 0:08:06.880 --> 0:08:10.160 and again you're now looking at a two point five 0:08:10.160 --> 0:08:12.760 almost two point six trillion dollar market cap company, so 0:08:12.840 --> 0:08:15.800 it is in the league of some of the largest 0:08:15.840 --> 0:08:17.240 market cap companies in the world. 0:08:17.840 --> 0:08:21.360 You're listening to the Bloomberg Business Week podcast. Catch us 0:08:21.400 --> 0:08:24.640 live weekday afternoons from two to five pm. Easter Listen 0:08:24.680 --> 0:08:28.040 on Applecarplay and ed Brout Auto with a Bloomberg Business app, 0:08:28.160 --> 0:08:30.560 or watch us live on YouTube. 0:08:31.440 --> 0:08:33.960 This past week, the CEOs of Dell's Service nown and 0:08:34.040 --> 0:08:37.000 Video all at the Dell Technology World event in Las Vegas, 0:08:37.120 --> 0:08:40.920 talking about working together and on premise AI That conversation, 0:08:41.000 --> 0:08:42.559 by the way, can be found on the Bloomberg and 0:08:42.559 --> 0:08:46.200 at Bloomberg dot com. Dell unveiled a new line of 0:08:46.200 --> 0:08:49.600 personal computers optimized for AI tasks, with Michael Dell saying 0:08:49.640 --> 0:08:53.360 AIPCS will be quote pretty standard. In twenty twenty five. 0:08:53.640 --> 0:08:57.280 Meantime, Microsoft introduced a version of its Copilot assistant to 0:08:57.280 --> 0:09:00.560 help teams collaborate, extending a company wide push to infuse 0:09:00.600 --> 0:09:04.400 products with AI. Microsoft presenting new computers with the new 0:09:04.440 --> 0:09:07.680 AI focused features, powered by chips from Qualcomm, a maker 0:09:07.760 --> 0:09:10.280 of smartphone components that's been trying to grab a foothold 0:09:10.280 --> 0:09:11.480 in laptops for years. 0:09:11.600 --> 0:09:13.480 A lot going on, folks, as you can tell well, 0:09:13.520 --> 0:09:15.800 so we felt pretty lucky when earlier this month we 0:09:15.880 --> 0:09:18.480 got our own take on all of this with Durga Malatti, 0:09:18.679 --> 0:09:20.440 head of AI over Qualcom. 0:09:20.679 --> 0:09:22.640 There's a lot of things that are going on in 0:09:22.640 --> 0:09:26.040 the space of AI as it runs on devices that 0:09:26.080 --> 0:09:29.800 you and I usually use, whether it's handsets or phones, 0:09:29.840 --> 0:09:33.680 whether it is tablets or PCs and so on. So 0:09:33.800 --> 0:09:35.680 let me spend a little bit of time in terms 0:09:35.679 --> 0:09:38.360 of explaining what's really going on out there. So in 0:09:38.400 --> 0:09:42.079 the last two years or so ever, since generative AI 0:09:42.200 --> 0:09:46.800 started making waves in terms of its capabilities, pretty spectacular capabilities, 0:09:46.840 --> 0:09:49.880 and in terms of the kinds of use cases that 0:09:49.880 --> 0:09:53.240 it can unleash. As Falcom, we've been focused on trying 0:09:53.280 --> 0:09:56.520 to bring that technology into devices that you and I 0:09:56.559 --> 0:09:58.760 can use. You know, usually one of the questions that 0:09:58.800 --> 0:10:01.400 comes in is why why do you want to run 0:10:01.440 --> 0:10:02.960 this on a device? Why can't you just run it 0:10:03.000 --> 0:10:04.640 on the cloud? I mean, we do use the cloud 0:10:04.640 --> 0:10:07.960 for so many different things, and what's wrong actually, So 0:10:08.240 --> 0:10:10.320 it's kind of important to understand what are the benefits 0:10:10.360 --> 0:10:15.880 that you actually have when you run these amazing generative 0:10:15.920 --> 0:10:19.760 AI models directly on devices. The very first one is 0:10:20.080 --> 0:10:24.240 related to privacy and security. A lot of the times 0:10:24.280 --> 0:10:28.520 when you actually go through generative AI use cases, you 0:10:28.600 --> 0:10:31.160 might be wondering, maybe it's just a chat pot. Maybe 0:10:31.200 --> 0:10:33.720 I ask a question, I get a response, But it 0:10:33.760 --> 0:10:35.800 doesn't have to be just a chat pot. It could be. 0:10:35.920 --> 0:10:38.720 These days, we use a lot more of both voice 0:10:38.760 --> 0:10:43.600 based interaction with devices in addition to using camera and 0:10:43.640 --> 0:10:46.760 something else to bring in additional context to the question 0:10:46.800 --> 0:10:49.560 that you're asking. For example, you could be holding up 0:10:49.559 --> 0:10:51.760 your phone and pointing in a certain direction and saying 0:10:51.960 --> 0:10:54.719 what am I looking at and that information you want 0:10:54.760 --> 0:10:57.000 to actually do process it locally. You might not be 0:10:57.040 --> 0:10:59.960 fully comfortable actually dealing with all of this in the cloud. 0:10:59.760 --> 0:11:01.960 I totally get it why this makes sense to have 0:11:02.000 --> 0:11:05.079 on device, whether that's a smartphone or a PC. We'll 0:11:05.080 --> 0:11:07.120 talk about PCs in just a minute. You said you've 0:11:07.160 --> 0:11:10.199 been working on this a Qualcomm. Why hasn't it happened yet? 0:11:10.400 --> 0:11:13.440 So, actually, if you take a look at it, at 0:11:13.520 --> 0:11:16.240 least on the smartphone segment, if you take a look 0:11:16.280 --> 0:11:19.120 at some of the premium tier smartphones, whether it's a 0:11:19.120 --> 0:11:22.600 Samsung Galaxy S twenty four or a pixel device or 0:11:22.640 --> 0:11:25.440 something that we are seeing in China with Shaomi and 0:11:25.559 --> 0:11:28.760 Honor and some of the other devices, all these devices 0:11:28.840 --> 0:11:33.040 now are capable of based upon keeping aside the pixel part, 0:11:33.080 --> 0:11:36.840 but everything else based upon Snapdragon platforms from Pollcom, running 0:11:37.000 --> 0:11:39.840 very large language models directly on the device, up to 0:11:39.880 --> 0:11:43.720 seven billion parameters, and these devices are now capable of 0:11:43.760 --> 0:11:46.200 a lot more things than you're used to. I do 0:11:46.280 --> 0:11:49.120 admit that the use cases are still gradually waking up 0:11:49.120 --> 0:11:51.360 to what are the capabilities in the smartphones, because that 0:11:51.480 --> 0:11:53.600 is something that takes a little bit of time. But 0:11:53.840 --> 0:11:55.520 if you've kind of taken a look at some of 0:11:55.559 --> 0:12:00.079 the live translate features or AI based voice assistance, the 0:12:00.160 --> 0:12:05.040 capability to run very large models directly on device starts today. 0:12:05.080 --> 0:12:09.720 It's already exists today in smartphones, and the use cases 0:12:09.760 --> 0:12:11.800 are gradually coming up, and we're doing our part on 0:12:11.840 --> 0:12:13.480 the use cases as well as I kind of go 0:12:13.559 --> 0:12:15.559 through the rest of the discussion. 0:12:15.880 --> 0:12:20.040 You guys have you know, you've been talking about phones specifically, 0:12:20.280 --> 0:12:24.199 but I'm assuming that you're thinking also PCs as well, 0:12:24.240 --> 0:12:27.000 And I'm just curious if that's the case, and you know, 0:12:27.480 --> 0:12:30.120 you guys have been trying for years to get into 0:12:30.360 --> 0:12:34.960 the PC market, Why though, hasn't it happened so far, 0:12:35.520 --> 0:12:38.120 and why maybe is it different this time around? If 0:12:38.160 --> 0:12:40.760 that's also on your radar in your mission. 0:12:40.640 --> 0:12:44.200 Absolutely right, there is a massive disruption that's coming to 0:12:44.440 --> 0:12:48.760 this specific segment in the context of aipcs. Now, the 0:12:48.760 --> 0:12:52.839 definition of AIPC is it is a PC in which 0:12:53.160 --> 0:12:55.360 not only do you have like the equivalent of everyone 0:12:55.440 --> 0:12:58.240 has heard about copilots, in which you are constantly using 0:12:58.320 --> 0:13:01.880 something else that is running with which you interact with 0:13:02.400 --> 0:13:05.160 quite frequently and use it for productivity, whether it's in 0:13:05.240 --> 0:13:09.160 creation of documents, edition of the editing documents are taking 0:13:09.320 --> 0:13:12.560 one kind of a document and translating into another, and 0:13:12.559 --> 0:13:14.680 so on and so forth. This is what we would 0:13:14.720 --> 0:13:17.440 call as like a on demand AIK so a little 0:13:17.480 --> 0:13:19.679 bit like a chladbot. You ask a question, you get 0:13:19.679 --> 0:13:22.600 an answer. It is more, it is more to it. 0:13:22.640 --> 0:13:24.360 Well, okay, so if there's more to this and we 0:13:24.400 --> 0:13:28.240 can why do we need an AI PC to do this? 0:13:28.280 --> 0:13:32.200 Why can't we just do this on the powerful machines 0:13:32.200 --> 0:13:33.559 that we already have. Why do they need to be 0:13:33.640 --> 0:13:34.439 AI specific? 0:13:34.800 --> 0:13:37.760 So one thing that I mean these productivity tools are 0:13:38.720 --> 0:13:41.960 they make it significantly more simpler to actually not just 0:13:42.160 --> 0:13:45.400 interact with your device, but actually make it a lot 0:13:45.440 --> 0:13:47.040 a lot of the things that you end up doing 0:13:47.360 --> 0:13:49.920 tend to be far more richer in terms of the 0:13:49.960 --> 0:13:52.679 capabilities in itself. Let me give an example of that, 0:13:52.720 --> 0:13:56.120 by the way, imagine that you're actually on a chat 0:13:56.160 --> 0:13:58.880 session with your colleague and you're actually talking about a 0:13:58.920 --> 0:14:01.520 specific meeting that you just had, and the next thing 0:14:01.559 --> 0:14:04.360 you know, something pops up saying okay, you already share 0:14:04.360 --> 0:14:05.920 the meeting notes with the person and you just lick 0:14:06.040 --> 0:14:08.520 use and that it immediately goes in. Just think through 0:14:08.559 --> 0:14:11.040 exactly all the things that happened in that whole process. 0:14:11.360 --> 0:14:13.800 First of all, this is now the equivalent of what 0:14:13.840 --> 0:14:17.000 we have been calling as pervasive AI AI, which is 0:14:17.080 --> 0:14:21.880 constantly running inside your PC. It's actually kind of trying 0:14:21.920 --> 0:14:24.600 to encapsulate you as the end user and the PC 0:14:24.720 --> 0:14:28.120 into like one single state, and you're constantly predicting, You're 0:14:28.160 --> 0:14:30.600 thinking on itslf. You know, the PC is actually kind 0:14:30.600 --> 0:14:34.200 of almost thinking and anticipating the next move from the 0:14:34.320 --> 0:14:38.280 user and then being very proactively help in that whole process. 0:14:38.280 --> 0:14:39.800 So we've all got a couple of minutes though, So 0:14:39.920 --> 0:14:42.680 is that why we need an AI PC? 0:14:42.840 --> 0:14:48.560 Specifically combination the two of the productivity tools plus this 0:14:48.880 --> 0:14:52.800 along with a few other things associated with UH you know, 0:14:52.880 --> 0:14:55.640 as a as a as a code generator, or as 0:14:55.640 --> 0:14:58.520 a content creator. It is something that's constantly running in 0:14:58.560 --> 0:15:01.440 the background. Now, keep in mind that we use our 0:15:01.480 --> 0:15:05.120 PCs regardless of whether we are connected somewhere to the 0:15:05.160 --> 0:15:07.720 cloud or not. I mean, the simplest example is you 0:15:07.800 --> 0:15:10.360 might be on like a long flight and you want 0:15:10.400 --> 0:15:12.440 to actually get some work done during the flight, but 0:15:12.680 --> 0:15:14.320 all you have is just you and the laptop at 0:15:14.320 --> 0:15:16.760 that point in time regardless of what the connectivity is, 0:15:16.960 --> 0:15:19.360 and you still want to have the same kind of productivity. 0:15:19.400 --> 0:15:22.960 So it does actually really start growing on you. Quite 0:15:23.000 --> 0:15:25.560 a few of us insight Qualcom have been dog footing 0:15:25.640 --> 0:15:28.720 exactly that experience with our own internal reference designs on 0:15:28.800 --> 0:15:32.400 these aipcs, and trust me, it does change the user. 0:15:32.200 --> 0:15:36.080 Exped So what does an AIPC do for the consumer? 0:15:36.200 --> 0:15:39.160 What do they get out of this? The regular consumer? 0:15:39.480 --> 0:15:42.000 A regular consumer? I mean, if you categorize between we 0:15:42.040 --> 0:15:45.520 talked about productivity tools, let's talk about content creation. Content 0:15:45.560 --> 0:15:47.880 creation is also kind of a productivity in its own way, 0:15:47.920 --> 0:15:50.040 So that part is actually quite clear. But if you 0:15:50.080 --> 0:15:52.680 start thinking about all the other things, even when it 0:15:52.720 --> 0:15:55.680 comes to the social interactions and emails and documents, it's 0:15:55.760 --> 0:15:57.560 on a regular user at the end of the day, 0:15:57.680 --> 0:15:59.880 goes through a mix and match of these different kinds 0:15:59.920 --> 0:16:02.680 of use cases, and we believe that it's actually a 0:16:02.720 --> 0:16:06.120 pretty big differentiator between the before and after. Now, is 0:16:06.160 --> 0:16:08.480 it like a big bang moment that happens. That is 0:16:08.520 --> 0:16:10.760 something that remains to be seen. But we are on 0:16:10.760 --> 0:16:12.960 the cusp of all these launchers that are coming up 0:16:12.960 --> 0:16:15.160 over the next few weeks and let's see how that 0:16:15.200 --> 0:16:15.680 plays out. 0:16:15.960 --> 0:16:18.080 Do you want me to be replaced Carol by an 0:16:18.080 --> 0:16:20.720 AI PC? Is that where you're hinting at? 0:16:21.000 --> 0:16:22.840 Come on, No, not gonna happen. 0:16:22.880 --> 0:16:27.800 No, no, not gonna happen this year. Avatar Avatar to 0:16:27.840 --> 0:16:30.640 prepare for the show, and then the real life Carol 0:16:30.680 --> 0:16:31.440 and Tim get to do that. 0:16:31.480 --> 0:16:34.320 Avatar is can to all the work and h and 0:16:34.360 --> 0:16:37.440 then we just step in. Yeah, I'm down with that, tiger. 0:16:37.520 --> 0:16:40.920 Thank you so much. Really appreciate getting some time with you. Malati. 0:16:40.960 --> 0:16:44.040 He's head of AI over at Qualkam, joining us from 0:16:44.080 --> 0:16:44.800 San Diego. 0:16:51.640 --> 0:16:55.480 You're listening to the Bloomberg Business Week podcast. Listen live 0:16:55.600 --> 0:16:58.400 each weekday. He's starting at two pm Eastern. Apple car 0:16:58.520 --> 0:17:01.480 Play and Android Auto with the Bloomberg Business Then you 0:17:01.520 --> 0:17:04.800 can also listen live on Amazon Alexa from our flagship 0:17:04.840 --> 0:17:08.639 New York station, Just say Alexa play Bloomberg eleven thirty. 0:17:10.080 --> 0:17:11.919 As we mentioned, this past week, we heard from the 0:17:11.960 --> 0:17:15.320 CEOs of some very well known companies you know, Microsoft, Dell, 0:17:15.400 --> 0:17:18.160 and Vidia to name a few. And yes they are 0:17:18.720 --> 0:17:22.159 household names today, but they weren't when they got their 0:17:22.200 --> 0:17:24.800 start many years ago, when they asked for the help 0:17:24.800 --> 0:17:27.359 of venture capitalists and their venture money. 0:17:27.560 --> 0:17:30.000 So what was it that made a venture capitalist or 0:17:30.119 --> 0:17:32.480 a group of them, or a VC firm decide to 0:17:32.520 --> 0:17:35.600 invest and make a financial bet on them over others? 0:17:35.840 --> 0:17:39.520 What were they thinking? And how can their thinking help 0:17:39.600 --> 0:17:42.440 you even if you don't work in venture capital. 0:17:42.400 --> 0:17:46.240 Which brought us to Iliustrebulaev. He's Professor of Private Equity 0:17:46.280 --> 0:17:49.479 and Finance at Stanford University's Graduate School of Business and 0:17:49.520 --> 0:17:53.119 founder and director of the Venture Capital Initiative. His new book, 0:17:53.359 --> 0:17:56.439 The Venture Mindset, How to make Smarter bets and Achieve 0:17:56.480 --> 0:17:57.680 Extraordinary Growth. 0:17:58.080 --> 0:18:01.680 The first principle to follow is why home runs matter 0:18:01.880 --> 0:18:05.359 and strikeouts don't. Because you talked about success, which is 0:18:05.400 --> 0:18:09.160 home runs for vcs, and failure which is about strikeouts. 0:18:09.320 --> 0:18:12.000 And the venture mindset is could be described as the 0:18:12.080 --> 0:18:16.960 new mental mindset that you should follow in the innovation 0:18:17.119 --> 0:18:20.320 driven world where you face a lot of uncertainty, a 0:18:20.320 --> 0:18:25.199 lot of unpredictability, and vcs had been facing this hostile 0:18:25.320 --> 0:18:28.280 environment for more than fifty years. Do you know that 0:18:28.640 --> 0:18:31.640 if you look at all the companies that became public 0:18:31.760 --> 0:18:34.800 in the United States in the last fifty years, every 0:18:35.000 --> 0:18:38.160 second one of them was ventchebaced do you know that 0:18:38.359 --> 0:18:40.760 six or seven or eight companies out of top ten 0:18:41.880 --> 0:18:45.040 by market cap today were venture backed when there were 0:18:45.040 --> 0:18:48.800 small companies and moreover, as my research shows, are stempered, 0:18:49.080 --> 0:18:52.680 there's causality, which means that had there been no bench 0:18:52.800 --> 0:18:55.920 capital industry, likely there would have been no many many 0:18:55.960 --> 0:18:59.639 of those companies that changed our lives, such as Apple, Netflix, 0:19:00.280 --> 0:19:03.320 bb Microsoft and Nvidia and so on. 0:19:03.600 --> 0:19:05.879 Well, what do you make though of the venture mindset 0:19:05.920 --> 0:19:10.880 back in the nineteen ninety nine two thousand era when 0:19:11.000 --> 0:19:14.320 it seems like money was just thrown at anything with 0:19:14.359 --> 0:19:16.000 a dot com after it. 0:19:16.240 --> 0:19:18.879 That is a great question. First of all, what we 0:19:19.119 --> 0:19:23.600 the Skatna book is for more fomo, which is fear 0:19:23.800 --> 0:19:27.240 of missing out. But what is really interesting is that 0:19:27.920 --> 0:19:30.920 smart vices are trying to avoid for more as much 0:19:30.960 --> 0:19:34.199 as possible. And one way to avoid it is what 0:19:34.359 --> 0:19:37.080 a friend of mine, Alex drampeldfrom andrsent FORRVIS calls it 0:19:37.640 --> 0:19:42.480 conviction must beat consensus, which means that if everybody is 0:19:42.480 --> 0:19:45.480 trying to invest in something, maybe that is time to 0:19:45.560 --> 0:19:48.280 invest in something else. So if we look in the 0:19:48.480 --> 0:19:52.320 late nineteen nineties, or indeed, if we look at a crypto, 0:19:52.720 --> 0:19:55.399 or if we look at some other bubbles in the 0:19:55.480 --> 0:19:58.359 last ten to fifteen years, many of them were driven, 0:19:58.640 --> 0:20:01.359 in fact by investors that were not following the Benure 0:20:01.440 --> 0:20:05.479 minds at principles. In fact, in the nineteen nineties there 0:20:05.520 --> 0:20:08.639 were also a lot of amazing companies born and a 0:20:08.680 --> 0:20:12.359 lot of investors made great investment decisions. 0:20:12.440 --> 0:20:17.320 I'm wondering about the idea of saying no more often 0:20:17.880 --> 0:20:18.800 than you say yes. 0:20:19.240 --> 0:20:21.240 So, first of all, this is the principle called say 0:20:21.359 --> 0:20:24.280