WEBVTT - OpenAI's Historic Funding Round, AI Dominates VC Deals 0:00:01.440 --> 0:00:05.120 From Marhard where Innovation of Money and Power colle in 0:00:05.240 --> 0:00:10.160 Silicon Valley, NBN. This is Bloomberg Technology with Caroline Hyde 0:00:10.200 --> 0:00:11.280 and Ed lud Love. 0:00:25.560 --> 0:00:28.600 Live from New York and San Francisco. This is Bloomberg Technology. 0:00:28.920 --> 0:00:29.760 Coming up, open. 0:00:29.600 --> 0:00:32.960 AI's historic funding round. We discussed one of the largest 0:00:33.000 --> 0:00:36.720 ever private investments and its new loan agreement securing ten 0:00:36.800 --> 0:00:38.960 billion dollars in total liquidity. 0:00:39.240 --> 0:00:41.160 And AI dominates VC deals. 0:00:41.159 --> 0:00:43.680 We have the latest data for venture funding in the 0:00:43.720 --> 0:00:44.760 third quarter and in. 0:00:44.800 --> 0:00:47.640 Video insiders cash in on the AI height, selling one 0:00:47.640 --> 0:00:50.640 point eight billion dollars of shares this year alone. That 0:00:50.840 --> 0:00:52.640 and so much more coming uphead. 0:00:53.159 --> 0:00:54.480 Yeah, straight to our top story. 0:00:54.560 --> 0:00:57.680 Open ai said it has completed a deal to raise 0:00:57.760 --> 0:01:01.200 six point six billion dollars in new giving the AI 0:01:01.360 --> 0:01:05.280 company a one hundred and fifty seven billion dollar valuation 0:01:05.800 --> 0:01:09.320 and bolstering its efforts to build the world's leading generative 0:01:09.319 --> 0:01:13.319 AI technology. In the last hour, the startups also announced 0:01:13.400 --> 0:01:16.640 a four billion dollar credit line from banks around the world, 0:01:16.959 --> 0:01:20.479 giving access to a more than ten billion dollar war chest. 0:01:20.520 --> 0:01:23.440 Blion both Rachel Metz joins us on set in San Francisco, 0:01:23.520 --> 0:01:26.320 and we've been tracking this for weeks and months and 0:01:26.360 --> 0:01:29.880 it's finally done. But the important point is they need 0:01:29.920 --> 0:01:34.119 the money because the cutting edge model requires the money 0:01:34.440 --> 0:01:35.639 to train and build. 0:01:36.080 --> 0:01:39.039 Yes, they are in a very expensive business. If you 0:01:39.080 --> 0:01:41.560 want to build a cutting edge large language model, any 0:01:41.640 --> 0:01:44.119 kind of cutting edge AI model, at this point, you're 0:01:44.160 --> 0:01:47.440 looking at potentially spending up to hundreds of millions of dollars. 0:01:47.800 --> 0:01:50.760 So if you think about multiplying that out over a 0:01:50.800 --> 0:01:52.960 couple of billion dollars, I mean it's like, well, you 0:01:53.000 --> 0:01:55.040 can train a couple models there, and what if you 0:01:55.120 --> 0:01:57.080 train one and you're like, you know what, I'm not 0:01:57.120 --> 0:02:00.960 that happy with this. I mean a lot of money. Also, 0:02:01.200 --> 0:02:03.240 they need you know, chips, they have staff, they have 0:02:03.240 --> 0:02:06.240 all kinds of expenses there. They are bringing in revenue 0:02:06.280 --> 0:02:08.480 of course, but it's still costs quite a bit to 0:02:08.520 --> 0:02:09.160 do what they're doing. 0:02:09.520 --> 0:02:13.120 It is the future revenue growth, the future profitability. That 0:02:13.240 --> 0:02:17.080 means standout names from VC like Thrive Capital, like Coast 0:02:17.120 --> 0:02:21.760 a Ventures back the round, but numerous big substantial banks 0:02:22.040 --> 0:02:25.840 back the credit facility. We're looking at Goldman, Sachs, JP 0:02:25.960 --> 0:02:29.680 Morgan santunder Wells Fargo to name but a few. What 0:02:29.919 --> 0:02:33.040 is the future growth model that they're all building one 0:02:33.120 --> 0:02:35.440 hundred and fifty seven billion dollar valuation on. 0:02:35.480 --> 0:02:41.320 Rachel, I mean they're looking at businesses using their products 0:02:41.320 --> 0:02:44.520 going forward. They're hoping that businesses will increasingly do things 0:02:44.560 --> 0:02:47.080 like fine tune the models. They recently made it so 0:02:47.120 --> 0:02:50.680 that businesses can do that and use them in different 0:02:51.160 --> 0:02:54.560 parts of their workflow and find it useful enough that 0:02:54.600 --> 0:02:57.320 they will keep paying for it, and they are starting 0:02:57.360 --> 0:03:00.400 to see uptake. They have quite a number of both 0:03:00.400 --> 0:03:03.000 individual and business users at this point. They recently passed 0:03:03.000 --> 0:03:06.959 the one million business user mark for business users of 0:03:07.040 --> 0:03:09.720 chat GBT in particular, But they still have a long 0:03:09.760 --> 0:03:12.720 ways to go if they want to get to this 0:03:12.840 --> 0:03:16.440 level that this valuation is showing them going to at 0:03:16.440 --> 0:03:16.919 some point. 0:03:17.080 --> 0:03:19.280 There's two points I want to make really quick with 0:03:19.400 --> 0:03:23.160 my sort of finance coverage on this is a pre 0:03:23.639 --> 0:03:27.360 EBITDAR private company or a pre positive EBITDAR private company, 0:03:27.560 --> 0:03:29.720 so that it could go to banks and get four 0:03:29.720 --> 0:03:33.360 billion dollars in credit is remarkable. And also the eagle 0:03:33.440 --> 0:03:36.280 eyed audience will say some of those banks are in Europe. 0:03:36.360 --> 0:03:38.840 That is also remarkable, Rachel. 0:03:38.880 --> 0:03:41.520 When they announced the closing of the equity round, they 0:03:41.520 --> 0:03:45.080 also gave us data on how they're doing not necessarily 0:03:45.120 --> 0:03:48.760 financial metrics, but like how many people are using open Ai. 0:03:48.960 --> 0:03:51.640 Do we know anything about like the scale of this 0:03:51.720 --> 0:03:53.440 company in the real world. 0:03:53.440 --> 0:03:54.000 Yes, we do. 0:03:54.120 --> 0:03:56.320 We know that there are I think it's two hundred 0:03:56.360 --> 0:04:00.600 and fifty million weekly users of Chat GBT as we 0:04:00.720 --> 0:04:04.480 previously reported. They also recently passed more than one million 0:04:04.640 --> 0:04:07.440 business users for chat GPT, and that spans a couple 0:04:07.400 --> 0:04:10.560 of different products related both to an educational product, they 0:04:10.600 --> 0:04:15.400 have a smaller business product like for smaller companies, and 0:04:15.440 --> 0:04:19.200 then also like a larger, more enterprise y product. So 0:04:19.240 --> 0:04:21.559 they're definitely trying to hit out a lot of different 0:04:21.560 --> 0:04:23.320 people in a lot of different areas, and they are 0:04:23.320 --> 0:04:26.520 doing it. They are they are certainly bringing in revenue, 0:04:26.520 --> 0:04:29.240 and they are seeing users. Uh. I think what remains 0:04:29.279 --> 0:04:31.880 to be seen is how well this will all work 0:04:31.920 --> 0:04:35.200 over time? What kind of you know, investors have to 0:04:35.240 --> 0:04:37.240 be willing to wait. They have to be very patient. 0:04:37.320 --> 0:04:37.440 Right. 0:04:37.480 --> 0:04:38.280 This is not like your. 0:04:38.200 --> 0:04:40.440 Average startup where it's like, well, y, you know, you 0:04:40.520 --> 0:04:42.800 might get a short term return and then maybe you'll 0:04:42.800 --> 0:04:45.000 get a long term return. This is like, well, you're 0:04:45.000 --> 0:04:46.560 gonna have to wait and see what we. 0:04:46.560 --> 0:04:49.560 Come up with. Over time. Rachel Metz, we thank you 0:04:49.600 --> 0:04:50.039 so much. 0:04:50.120 --> 0:04:52.840 Meanwhile, Coast and Adventures participated in the Open AI round. 0:04:52.880 --> 0:04:55.719 As we mentioned, tune in tomorrow our interview with Vinode 0:04:55.760 --> 0:04:58.440 Coast and on this and of course AI's future impact. 0:04:59.160 --> 0:05:01.920 Now we tend to another key story. US prosecutors are 0:05:01.960 --> 0:05:05.800 broadening a probe of potential price fixing by German software 0:05:05.800 --> 0:05:09.640 maker SAP and tech reseller Carasoft, seeking to examine the 0:05:09.640 --> 0:05:12.440 company's work with almost one hundred government agencies. 0:05:12.520 --> 0:05:15.640 It's according to a new court record. Imers Jake Bleiberg 0:05:15.760 --> 0:05:16.480 is here with more. 0:05:16.720 --> 0:05:20.240 You've been sifting through these court documents, remind us what 0:05:20.440 --> 0:05:23.560 is being investigated, if not accused quite yet. 0:05:23.839 --> 0:05:31.080 Absolutely, the investigation is looking at the companies you mentioned, SAP, Kerosoft, 0:05:31.520 --> 0:05:33.960 and then some other companies, the identities of which we 0:05:33.960 --> 0:05:37.160 don't all know, for a potential scheme. And I should 0:05:37.160 --> 0:05:41.120 stress that it's potential, not proven, to fix the prices, 0:05:41.600 --> 0:05:46.080 to inflate the prices in sales of SAP technology to 0:05:46.279 --> 0:05:47.160 the US government. 0:05:49.040 --> 0:05:51.520 Jake, you joined us the other day to discuss your 0:05:51.560 --> 0:05:55.039 original reporting. What I'm trying to do is understand what's 0:05:55.160 --> 0:05:58.719 new here. One hundred agencies is more than the agencies 0:05:58.760 --> 0:06:01.600 you outline the other day. That's new, But is there 0:06:01.640 --> 0:06:03.960 an expansion in the scope of what they're looking at. 0:06:03.960 --> 0:06:08.000 Have they discovered more information that prompted more agencies to 0:06:08.040 --> 0:06:09.800 say we better get on board with this. 0:06:10.760 --> 0:06:15.240 Absolutely, it's difficult to say what the prosecutors have discovered 0:06:15.279 --> 0:06:16.080 at this point. 0:06:16.320 --> 0:06:17.120 And something we do. 0:06:17.080 --> 0:06:20.440 Know is that they say they've been having difficulty getting 0:06:20.440 --> 0:06:23.159 information from the company. Sort of at the heart of 0:06:23.160 --> 0:06:28.200 this Kerasoft. But as you point to, what has happened 0:06:28.440 --> 0:06:32.960 is they went from an investigation primarily focused on sales 0:06:33.000 --> 0:06:37.119 to the Department of Defense and then just as of July, 0:06:37.760 --> 0:06:42.680 broaden this out pretty dramatically to look at sales to 0:06:43.360 --> 0:06:47.120 ninety four or as many as ninety four civilian government 0:06:47.160 --> 0:06:48.200 agencies as well. 0:06:48.760 --> 0:06:52.040 We're looking now at the statements coming from the company's 0:06:52.120 --> 0:06:54.880 SAP saying it follows the rules around disclosing legal risks 0:06:54.880 --> 0:06:57.719 are based on the current circumstances. There was still no 0:06:57.800 --> 0:07:00.480 requirement to report on the individual cases. Karasoft also gave 0:07:00.560 --> 0:07:03.560 us a statement. But ultimately, how are they going to 0:07:03.560 --> 0:07:05.600 be interacting with these hundreds of agencies? 0:07:05.640 --> 0:07:07.680 What role do they now play in this? 0:07:08.320 --> 0:07:09.240 It's a great question. 0:07:09.840 --> 0:07:12.640 The sort of historic role that both of these companies 0:07:12.680 --> 0:07:17.160 have played is by putting out a product. SAP makes 0:07:18.280 --> 0:07:24.360 all sorts of workforce software, time cards, HR software, and 0:07:24.480 --> 0:07:29.000 the complicated process of selling that sort of software to 0:07:29.000 --> 0:07:33.200 the US government is sort of handled by Carasoft, which 0:07:33.240 --> 0:07:35.360 stands in the middle and does a lot of the 0:07:35.520 --> 0:07:40.000 difficult work of helping the software manufacturer navigate. 0:07:39.560 --> 0:07:41.000 These regulatory hurdles. 0:07:41.360 --> 0:07:44.720 So what those relationships will look like going forward in 0:07:44.720 --> 0:07:47.239 the middle of this investigation is a big open question. 0:07:48.920 --> 0:07:52.600 Bloomberg's Jake Blybag, who broke the original story and continues 0:07:52.640 --> 0:07:55.760 across the developing story, Thank you very much. Now coming 0:07:55.840 --> 0:08:00.120 up here on this show, is job displacement caused by 0:08:00.200 --> 0:08:04.160 AI something to brace for or is this just all hype? 0:08:04.480 --> 0:08:08.720 We discussed with MIT economist Darren Smoglu, who gives us 0:08:08.760 --> 0:08:12.160 his more skeptical view on what's really happening with AI 0:08:12.480 --> 0:08:13.560 in the jobs market. 0:08:13.800 --> 0:08:15.120 This is Bloomberg Technology. 0:08:24.600 --> 0:08:27.840 Softbag founder Asoshi's son has sketched out one of the 0:08:27.880 --> 0:08:31.680 most aggressive timelines for the adoption of AI yet, envisioning 0:08:31.760 --> 0:08:35.280 a near future where the technology would run entire households 0:08:35.320 --> 0:08:38.439 from monitoring the health of family members to grocery shopping now. 0:08:38.480 --> 0:08:41.080 In a speech add an annual forum for enterprise clients 0:08:41.080 --> 0:08:46.400 earlier today, Sun said that artificial general intelligence or AGI would. 0:08:46.280 --> 0:08:48.520 Arrive within the next two to three years. 0:08:49.000 --> 0:08:52.720 ED, but not everyone is so bullish on AI, at 0:08:52.800 --> 0:08:55.959 least not when it comes to jobs. Darren Asmoglu, a 0:08:56.080 --> 0:09:00.600 renowned professor at MIT, says that by his calculation, only 0:09:00.600 --> 0:09:04.720 a small percent of all jobs round five percent is 0:09:04.880 --> 0:09:08.640 right to be taken over at least heavily aided by 0:09:08.800 --> 0:09:13.440 AI over the next decade. Darren Asamogli joins us now 0:09:13.720 --> 0:09:16.720 and welcome to the program. This was one of the 0:09:16.760 --> 0:09:19.640 most read stories on Bloomberg in the last twenty four hours. 0:09:19.679 --> 0:09:22.040 There's no surprise about that. It has all of the 0:09:22.080 --> 0:09:25.679 facets of a story that would you know, just engage 0:09:25.679 --> 0:09:30.800 our audience on Bloomberg technology. Let's get specifics why only 0:09:30.840 --> 0:09:34.120 five percent of jobs and which sectors in industries are 0:09:34.120 --> 0:09:35.440 you basing your thesis on. 0:09:36.600 --> 0:09:39.959 Well, thanks, Ed, It's great to be on the program. 0:09:40.120 --> 0:09:46.800 I think right now there is a lot of potential excitement, 0:09:46.840 --> 0:09:49.480 perhaps hyper about AI, but if you look at its 0:09:49.600 --> 0:09:55.760 current capabilities, you know we are very far from AI, 0:09:56.120 --> 0:09:59.239 either general to AI or other types of AI performing 0:09:59.440 --> 0:10:02.800 any of the tasks that involve a heavy level of 0:10:02.960 --> 0:10:09.640 interaction with the physical world construction work, manufacturing work, carpentry. 0:10:11.280 --> 0:10:16.319 Those perhaps one day will be much more feasible, for example, 0:10:16.320 --> 0:10:19.040 if AI becomes so much more advanced, so much more reliable, 0:10:19.040 --> 0:10:21.800 and can be integrated with so much better robots, for instance, 0:10:22.880 --> 0:10:24.679 But it's not something that's going to happen in the 0:10:24.679 --> 0:10:28.720 next five or ten years now. There's a lot of 0:10:28.800 --> 0:10:36.360 other potential for future AI, perhaps to perform some social tasks, 0:10:36.920 --> 0:10:38.960 but those are not going to be possible right now either. 0:10:39.000 --> 0:10:40.680 You know you're not going to go to an AI 0:10:40.760 --> 0:10:44.720 psychiatrist anytime soon, So if you stip out all of these, 0:10:44.960 --> 0:10:47.600 you're not going to be left with much more than 0:10:47.640 --> 0:10:48.960 five percent of the economy. 0:10:49.600 --> 0:10:52.760 Professor may I jump in and push back with a 0:10:52.840 --> 0:10:55.520 point raised by our audience when I said you were 0:10:55.559 --> 0:10:59.719 coming on, which is that AI technology is exponential, much 0:10:59.760 --> 0:11:02.720 in the way that the invention of electricity with time 0:11:03.200 --> 0:11:06.760 led to all sorts of breakthroughs. Right now, you're saying 0:11:06.800 --> 0:11:09.400 five percent of jobs, But in two years or five 0:11:09.480 --> 0:11:12.920 years from now, would you need to rethink your thesis 0:11:13.000 --> 0:11:14.679 or redo your calculations. 0:11:15.400 --> 0:11:17.720 I should have prefaced this by saying there's a huge 0:11:17.720 --> 0:11:21.440 amount of uncertainty in any forecast about AI, and apologies 0:11:21.480 --> 0:11:23.400 that I didn't do that. I tried to do that 0:11:23.440 --> 0:11:25.760 all the time. I was very excited. I just jumped 0:11:25.800 --> 0:11:28.200 into it. There's a huge amount of uncertainty, and one 0:11:28.240 --> 0:11:31.040 source of uncertainty is exactly what's gonna happen with AI 0:11:31.679 --> 0:11:36.960 in the next several years. I don't think that the 0:11:37.000 --> 0:11:41.000 most optimistic views in the industry, which sometimes people refer 0:11:41.080 --> 0:11:45.320 to as scaling laws, actually will hold water when you 0:11:45.360 --> 0:11:48.320 look at it within the next five years or ten years, 0:11:48.320 --> 0:11:50.920 which say, you know, just double the amount of data, 0:11:51.040 --> 0:11:53.560 double the amount of computer power, and then you're gonna 0:11:53.559 --> 0:11:59.440 get double the amount of capabilities. Because to do that 0:11:59.480 --> 0:12:02.360 we would actual need higher quality data. So if you 0:12:02.400 --> 0:12:05.640 want AI to do the job of a carpenter, you 0:12:05.760 --> 0:12:08.640 really need a completely different kettle of fish when it 0:12:08.640 --> 0:12:12.600 comes to the kinds of data, and something like that 0:12:12.679 --> 0:12:15.400 might be feasible with a different architecture of AI, but 0:12:15.559 --> 0:12:17.760 right now all of the money is going into the 0:12:18.120 --> 0:12:22.920 sort of the basic dominant architecture, which has its limitations 0:12:22.920 --> 0:12:25.840 both in terms of reliability in terms of high level reasoning. 0:12:26.080 --> 0:12:29.560 So that's the basis of my belief that, of course 0:12:30.000 --> 0:12:32.600 generative AI and other types of AI will improve, but 0:12:32.679 --> 0:12:35.840 we're not going to see a qualitative jump within the 0:12:35.880 --> 0:12:36.960 next five to ten years. 0:12:37.000 --> 0:12:40.839 So as like your belief, what is exactly based on 0:12:41.240 --> 0:12:44.040 what data, what historical norms are you looking back on 0:12:44.360 --> 0:12:46.880 to drive and push forward on this five percent take? 0:12:47.480 --> 0:12:49.440 Yeah, I think this is why there is so much uncertainty. 0:12:49.480 --> 0:12:52.760 There's nothing that's identical to generative AI in the past, 0:12:53.080 --> 0:12:55.040 and that's part of the reason why people can hold 0:12:55.120 --> 0:12:58.320 very different beliefs. But what I'm basing that on is, 0:12:58.400 --> 0:13:03.840 first of all, other waves of digital technologies that promised 0:13:04.679 --> 0:13:08.360 a lot of automation, sometimes a lot of productivity improvements, 0:13:08.440 --> 0:13:11.000 sometimes a lot of changes in the way that we 0:13:11.840 --> 0:13:14.160 would organize our offices and things like that, and at 0:13:14.160 --> 0:13:17.720 the end, all of those changes have been very gradual. 0:13:18.160 --> 0:13:20.880 And also what we have seen over the last two 0:13:20.920 --> 0:13:24.119 and a half years, where you know, there have been advances, 0:13:24.160 --> 0:13:28.760 but the advances haven't really been sort of qualitative in nature. 0:13:29.080 --> 0:13:33.319 They haven't really exceeded the sort of the basic capabilities 0:13:33.320 --> 0:13:36.920 of GPT three point five or three point GPT four. 0:13:37.320 --> 0:13:40.400 So there will be advances, but I don't think any 0:13:40.480 --> 0:13:45.280 of those advances will very quickly change the way that 0:13:45.320 --> 0:13:49.920 we do various tasks in the economy. So, for example, 0:13:50.640 --> 0:13:54.480 I do not believe that in five years time, when 0:13:54.520 --> 0:13:56.240 we look at the labor market, we're going to see 0:13:56.520 --> 0:13:59.080 certain occupations that will have completely disappeared. 0:13:59.800 --> 0:14:03.000 Think you would say that they would be augmented. 0:14:03.440 --> 0:14:06.200 The idea is, perhaps the optimist view is we will 0:14:06.200 --> 0:14:08.240 not lose lots of people from the labor force, but 0:14:08.280 --> 0:14:10.160 their productivity will be scaled. 0:14:10.920 --> 0:14:14.320 Where you think that's a yes, I think that's a 0:14:14.440 --> 0:14:18.959 very very interesting and important point. I am not an 0:14:19.000 --> 0:14:23.000 AI denier or an AI pessimist. I actually believe AI 0:14:23.080 --> 0:14:25.200 has a lot of potential, And in fact, the real 0:14:25.240 --> 0:14:27.600 potential is exactly what you said, Caroline, which is that 0:14:27.960 --> 0:14:32.359 AI becomes a tool for workers to boost their productivity. 0:14:32.400 --> 0:14:34.400 And again we're seeing some of that happening. You know, 0:14:34.400 --> 0:14:36.520 Microsoft has a lot of co pilots. You know, some 0:14:36.600 --> 0:14:42.360 of them work and they give some productivity improvements. There 0:14:42.360 --> 0:14:45.800 are a few areas where we're seeing really phenomenal improvements, like, 0:14:45.840 --> 0:14:48.480 for example, if you want to write simple subroutines in 0:14:48.520 --> 0:14:52.280 one of the common programming languages. You know, GPT four 0:14:52.480 --> 0:14:55.680 is great. Before then getthub copilot was pretty good. So 0:14:56.320 --> 0:14:59.120 there are some things in which that sort of help 0:14:59.280 --> 0:15:02.000 to decision makers as possible. But again, when you look 0:15:02.000 --> 0:15:07.479 at the whole economy, that means helping carpenters, electricians, plumbers, 0:15:07.840 --> 0:15:12.320 blue collar workers, construction workers, transport workers, people who are 0:15:12.360 --> 0:15:16.320 doing teaching, people who are doing journalism. Those are things 0:15:16.360 --> 0:15:19.840 that could be achieved perhaps at some point, but the 0:15:19.920 --> 0:15:22.600 reliability and the reasoning of the models is not up 0:15:22.640 --> 0:15:23.400 to the task yet. 0:15:24.520 --> 0:15:30.080 Do you see specific job creation in specific industries from AI. 0:15:31.400 --> 0:15:35.280 Well, there is one class of jobs where AI is 0:15:35.560 --> 0:15:39.200 leading a big boost, which is people who have AI fluency, 0:15:39.240 --> 0:15:45.040 either people who are AI programmers, AI integrators, or other 0:15:45.120 --> 0:15:51.120 AI related service activities. But in my research, albeit from 0:15:51.160 --> 0:15:54.320 a previous wave of AI in the late twenty tens, 0:15:54.720 --> 0:15:57.640 we didn't see any other job category get a boost 0:15:57.640 --> 0:16:00.800 in terms of hiring. And there are some other jobs 0:16:00.800 --> 0:16:02.640 that I think there will be some retrenchment. So if 0:16:02.680 --> 0:16:05.560 you're in IT security and you're doing so routine end 0:16:05.560 --> 0:16:09.280 of IT security, there will be declines and job vacancies 0:16:09.320 --> 0:16:12.160 in that area because some of these AI tools can 0:16:12.240 --> 0:16:13.720 actually perform that quite well. 0:16:14.280 --> 0:16:16.880 When you see the headline that open ai is worth 0:16:16.920 --> 0:16:19.640 one hundred and fifty seven billion dollars, when you see 0:16:19.760 --> 0:16:22.600 that in Video is now worth almost three trillion, what 0:16:22.680 --> 0:16:26.040 do you make of that in terms of how truly 0:16:26.120 --> 0:16:29.200 the capital markets are analyzing the growth here. 0:16:30.240 --> 0:16:33.440 Well, I think there's a difference between Nvidia and open Ai. 0:16:34.120 --> 0:16:38.760 Nvidia is riding a wave which is based on the 0:16:38.840 --> 0:16:41.400 belief that those chips are going to become more valuable 0:16:41.440 --> 0:16:44.280 because generative AI and other types of AI are going 0:16:44.360 --> 0:16:48.440 to continue to boom, but they are generating revenues. Open ai, 0:16:48.600 --> 0:16:51.880 on the other hand, I think does hasn't found a 0:16:51.960 --> 0:16:57.160 revenue model that could justify the types of valuations we 0:16:57.240 --> 0:16:59.560 are seeing at the moment now. It may well be 0:16:59.680 --> 0:17:02.200 that in the future suddenly their API is going to 0:17:02.240 --> 0:17:04.240 be so valuable and other firms are going to start 0:17:04.240 --> 0:17:09.360 building on it, but it doesn't seem like there are 0:17:09.440 --> 0:17:13.840 the kinds of applications that can have deep impact on 0:17:13.960 --> 0:17:17.080 the productivity of workers or the automation of firms that 0:17:17.240 --> 0:17:20.439 are going to be sold within the next five years 0:17:20.560 --> 0:17:23.679 ten years. So I'm a little bit skeptical. But on 0:17:23.720 --> 0:17:25.719 the other hand, open ai is also emerging as the 0:17:25.720 --> 0:17:29.480 market leader, and the current ecosystem in the digital world 0:17:29.560 --> 0:17:33.000 is right firms that take over the market and they 0:17:33.040 --> 0:17:35.119 can use all of the data. They can you know, 0:17:35.440 --> 0:17:38.919 interact with the consumers or the programmers directly. Those have 0:17:39.000 --> 0:17:41.800 become very valuable because of venture capital or other deep 0:17:41.800 --> 0:17:42.720 pocketed funders. 0:17:43.440 --> 0:17:45.920 Can't wait to see the response to this interview. I'm 0:17:45.920 --> 0:17:49.800 sure someone has some thoughts. It economics professor drown Asmlgloo. 0:17:49.920 --> 0:18:00.280 We appreciate your time, Aura. It's just world a new 0:18:00.280 --> 0:18:03.159 flagship device with the Ring four of dating its popular 0:18:03.200 --> 0:18:05.600 smart rings with new sensors, but a battery life and 0:18:05.880 --> 0:18:08.919 about an algorithm. This is competition is really heating up. 0:18:08.960 --> 0:18:12.000 Samsung of course, debuting its own Galaxy Ring for more 0:18:12.200 --> 0:18:14.360 or a CEO Tom Hale joins us, So, how does 0:18:14.400 --> 0:18:15.879 this make me distinct from the crowd. 0:18:17.119 --> 0:18:19.520 Well, the thing about the or Ring four is that 0:18:19.600 --> 0:18:23.679 it's slimmer, sleeker, thinner, better, of course, and the feel 0:18:23.720 --> 0:18:26.040 of the ring is what matters to many people because 0:18:26.040 --> 0:18:27.719 there's a lot of technology packed in here. 0:18:27.760 --> 0:18:29.240 You want it to be small and compact. 0:18:29.840 --> 0:18:32.879 Maybe more importantly, we've taken a big leap forward in 0:18:33.000 --> 0:18:36.080 sensing with something we call smart sensing, and smart sensing 0:18:36.200 --> 0:18:39.440 adjusts dynamically to your body because not everyone's the same. 0:18:39.840 --> 0:18:42.359 Your anatomy is not the same, your physiology is not 0:18:42.400 --> 0:18:44.480 the same, your skin color is not the same, and 0:18:44.520 --> 0:18:46.760 we adjust dynamically even as you rotate the ring a 0:18:46.800 --> 0:18:48.640 little bit, we actually are able to keep the most 0:18:48.640 --> 0:18:49.439 accurate signal. 0:18:51.119 --> 0:18:53.760 Tom, who do you consider it to be the composition 0:18:54.000 --> 0:18:56.840 for this ring? Is it other rings or is it 0:18:56.920 --> 0:18:59.040 other form fact? As a device, I think a lot 0:18:59.080 --> 0:19:02.479 about watches it the moment, particularly having just been at 0:19:02.480 --> 0:19:04.159 Apples event in recent weeks. 0:19:06.040 --> 0:19:09.880 What's interesting to me is that amongst our users, almost 0:19:09.880 --> 0:19:13.320 two thirds wear a second wearable, usually a smart watch. 0:19:13.720 --> 0:19:17.040 On the smart watch great for daytime notifications, you're getting 0:19:17.040 --> 0:19:19.200 alerts all the time, but has an achilles heel, which 0:19:19.200 --> 0:19:20.120 is battery life. 0:19:20.320 --> 0:19:21.080 And the ring is. 0:19:21.000 --> 0:19:25.120 Something that doesn't ask for your attention, doesn't send you notifications. 0:19:25.119 --> 0:19:28.159 It's just in the background, seamlessly fitting into your life, 0:19:28.640 --> 0:19:31.760 capturing your physiology so that you don't have to pay. 0:19:32.280 --> 0:19:33.439 A ton of attention to it. 0:19:33.720 --> 0:19:36.439 And as a result, you have an eight day battery 0:19:36.480 --> 0:19:38.920 life with the or four ring. And so what that's 0:19:38.920 --> 0:19:41.080 really interesting to think about is that you just don't 0:19:41.080 --> 0:19:43.360 have to worry about your health as much, and then 0:19:43.560 --> 0:19:46.080 therefore the watch in the ring can be compatible. So 0:19:46.280 --> 0:19:48.640 in many ways we find these things to be two 0:19:48.640 --> 0:19:51.080 devices that can work together on one person's body. 0:19:51.320 --> 0:19:53.400 And what's interesting is you've been working with other companies 0:19:53.400 --> 0:19:56.400 that make other devices. We understand from Mark German's reporting 0:19:56.400 --> 0:20:00.119 that you've been working alongside Meta's Raband team, for example. Well, 0:20:00.560 --> 0:20:03.000 how do you see integration with other businesses going. 0:20:03.960 --> 0:20:06.960 Well, we've got so many great partners that we work with. 0:20:07.560 --> 0:20:10.280 We haven't announced anything formally with Meta, but you can 0:20:10.359 --> 0:20:13.480 imagine this world where you have this cloud of devices 0:20:13.520 --> 0:20:15.760 on your body that you're using in different contexts. Maybe 0:20:15.760 --> 0:20:19.240 it's headphones, maybe it's glasses with an AI assistant, And 0:20:19.359 --> 0:20:23.760 how does your busiological state and something that you've got 0:20:23.800 --> 0:20:26.320 on your body all the time work, maybe as a controller, 0:20:26.600 --> 0:20:28.840 maybe as a way to guide, maybe as a way 0:20:28.840 --> 0:20:30.960 to provide input, maybe as a way to give you 0:20:30.960 --> 0:20:33.320 your other wearables a sense of whether you're stressed or 0:20:33.359 --> 0:20:35.600 whether you're well rested, whether you're ready. There's a lot 0:20:35.600 --> 0:20:38.280 of really interesting opportunities to explore when you think about 0:20:38.320 --> 0:20:42.040 a permanently worn device that's oriented around your health that's 0:20:42.040 --> 0:20:43.280 on your body. Twenty four to seven. 0:20:44.760 --> 0:20:49.440 Tom, we've just shown your latest financials and the trajectory 0:20:49.480 --> 0:20:53.280 of this company. Your focus this point has been technology 0:20:53.280 --> 0:20:56.320 and innovation. But where do you see this company going 0:20:56.359 --> 0:20:58.720 from here? The direction for the business. 0:21:00.080 --> 0:21:02.080 Yeah, I mean, I think there's a lot of headroom 0:21:02.119 --> 0:21:04.879 for us to grow just simply in the ring category. 0:21:05.160 --> 0:21:08.200 Rings are new. People are just learning about the benefits 0:21:08.240 --> 0:21:11.600 of long term health tracking, learning about the benefits of 0:21:11.640 --> 0:21:14.960 getting sleep, learning about the understanding of their body. So 0:21:15.160 --> 0:21:16.800 you know, the way the ring works is it can 0:21:16.800 --> 0:21:19.000 tell you when you're getting sick, it can tell you 0:21:19.040 --> 0:21:20.000 when you're doing well. 0:21:20.160 --> 0:21:20.680 It gives you. 0:21:20.680 --> 0:21:23.240 Lots of information and lots of insights, even an AI 0:21:23.359 --> 0:21:26.880 coach that helps you make behavior changes. And I think 0:21:26.880 --> 0:21:28.840 there's a lot of headroom just in that category. But 0:21:28.880 --> 0:21:31.520 as you look beyond that, not only is it interaction 0:21:31.600 --> 0:21:34.680 with other devices, but it might be potentially other devices 0:21:34.680 --> 0:21:36.800 that you might see, Like we've seen prototypes for a 0:21:36.840 --> 0:21:39.679 locket that sits on your chest. We've seen prototypes for 0:21:39.720 --> 0:21:42.119 things that look more like designer rings. We introduced a 0:21:42.119 --> 0:21:44.840 product with Gucci a couple of years back where we 0:21:44.920 --> 0:21:47.440 have something that looks much more like jewelry and much 0:21:47.480 --> 0:21:50.320