WEBVTT - GameStop’s $56 Billion Bid for eBay 0:00:02.480 --> 0:00:13.640 Bloomberg Audio Studios, Podcasts, radio news. Bloomberg Tech is live 0:00:13.680 --> 0:00:17.479 from coast to coast with Caroline Hyde in New York 0:00:17.760 --> 0:00:20.960 and Edla Loow in San Francisco. 0:00:22.000 --> 0:00:23.640 This is Bloomberg Tech coming up. 0:00:23.680 --> 0:00:27.280 GameStop eyes eBay for fifty six billion dollars in cash 0:00:27.320 --> 0:00:30.040 and stock and an audacious bid for a company four 0:00:30.040 --> 0:00:33.600 times at size. eBay says the review that offer plus 0:00:33.680 --> 0:00:36.599 AI chipmaker Cerebraus has been looking to raise as much 0:00:36.600 --> 0:00:38.440 as three and a half billion dollars for a potential 0:00:38.479 --> 0:00:41.400 IPO later this month, part of a growing cohort looking 0:00:41.440 --> 0:00:44.639 to challenge in video and open AI raises more than 0:00:44.680 --> 0:00:47.080 four billion dollars for a new joint venture that aims 0:00:47.120 --> 0:00:51.559 to help businesses adopt it's AI software. A first free 0:00:51.680 --> 0:00:54.960 check in on that extraordinary MNA news. The first dropped 0:00:54.960 --> 0:00:58.400 on Friday and continues to move shares now. GameStop looking 0:00:58.440 --> 0:00:59.600 to buy a company. 0:00:59.520 --> 0:01:01.120 Four times its size. 0:01:01.240 --> 0:01:04.319 GameStop currently off by seven percent, eBay up by more 0:01:04.360 --> 0:01:07.039 than six percent. But we have to get the inside 0:01:07.120 --> 0:01:12.080 track with Rineberg. Reporter Cecilia Denastasio with an extraordinary story 0:01:12.760 --> 0:01:15.160 and one that the market tries to understand whether it's 0:01:15.160 --> 0:01:16.679 for real or not? Can you just talk us about 0:01:16.680 --> 0:01:19.800 the technicalities. Ryan Cohen here is offering for eBay. 0:01:20.040 --> 0:01:25.200 Sure, So Ryan Cohen is offering a deal to purchase eBay, 0:01:25.200 --> 0:01:27.280 which is, like you said, many times the size of 0:01:27.319 --> 0:01:30.920 game Stop along with twenty billion dollars. He says he 0:01:30.959 --> 0:01:35.600 has received a highly confident letter from TD Bank to acquire. 0:01:35.800 --> 0:01:38.600 Analysts are unsure this deal will ever go through, but 0:01:38.680 --> 0:01:41.000 Ryan Cohen says that if he manages to purchase eBay, 0:01:41.440 --> 0:01:43.240 it would make a lot more money. 0:01:43.840 --> 0:01:46.319 Make a lot more money. Okay, let's take this for real. 0:01:46.400 --> 0:01:48.280 That he's good for the money he's able to afford. 0:01:48.280 --> 0:01:51.720 What synergies are analysts kind of trying to pass through. 0:01:51.560 --> 0:01:52.080 At the moment. 0:01:52.280 --> 0:01:54.840 GameStop has really struggled over the last several years because 0:01:54.840 --> 0:02:00.080 gamers are increasingly purchasing their wares in online stores like Valves, Steam, 0:02:00.520 --> 0:02:05.360 and so GameStop has shrunk its footprint in their retail marketplace. 0:02:05.520 --> 0:02:08.520 But Ryan Cohen has tried a lot of different strategies 0:02:08.560 --> 0:02:11.160 to make it work. He's tried esports, he's tried e commerce, 0:02:11.160 --> 0:02:13.520 which is from his background as the co founder of Chewi. 0:02:14.040 --> 0:02:17.120 And while he's managed to you know, help the business 0:02:17.160 --> 0:02:19.320 sustain a little bit more than people had hoped. Nothing's 0:02:19.400 --> 0:02:22.480 really stuck. eBay gives him the opportunity, he says, to 0:02:22.520 --> 0:02:26.040 compete with Amazon potentially, which of course is very very 0:02:26.040 --> 0:02:27.560 ambitious in e commerce. 0:02:27.919 --> 0:02:31.840 Well, Jamie I and Oni over eBay has been impressing 0:02:31.880 --> 0:02:34.680 investors with his turnaround. He's been leaning into AI, He's 0:02:34.680 --> 0:02:37.160 been leaning into partnerships with Meto, He's been thinking about 0:02:37.360 --> 0:02:40.240 M and A himself with deepop. Just how do we 0:02:40.400 --> 0:02:43.520 understand eBay is digesting this and thinking about it from 0:02:43.560 --> 0:02:45.320 its own business and shitholder perspective. 0:02:45.520 --> 0:02:47.079 eBay is evaluating the bid. 0:02:47.560 --> 0:02:50.680 I think, you know, analysts again are not sure whether 0:02:50.919 --> 0:02:54.720 the math is mathing when it comes to Gamestop's ability 0:02:54.800 --> 0:02:58.320 to purchase the significantly larger company, and we'll see what happens. 0:02:58.400 --> 0:03:01.880 Ryan Cohen has not given specific details on what he 0:03:01.919 --> 0:03:04.280 imagines the synergies to be or what he imagines the 0:03:04.280 --> 0:03:05.600 combined company to look like. 0:03:05.840 --> 0:03:09.480 I think truists said this is stunning, and they're skeptical. 0:03:10.040 --> 0:03:12.560 I mean, just in terms of how M and A 0:03:12.680 --> 0:03:16.320 gets done. Letters from TD Banks saying they're good for it. 0:03:16.400 --> 0:03:18.600 Ultimately are going to be dependent on the debt markets here, 0:03:18.840 --> 0:03:23.880 but GameStop has always shocked the markets with its mean 0:03:23.960 --> 0:03:27.800 stock potential, with roaring Kitties love and even with sky 0:03:27.880 --> 0:03:31.840 on capital well now pivoted more into a substat newsletter. 0:03:32.160 --> 0:03:36.680 But the guy who was known for trades on the 0:03:36.720 --> 0:03:39.240 mortgage market back in two thousand and eight, Michael Barry, 0:03:39.280 --> 0:03:42.120 is coming out saying this could be a really good story. 0:03:42.600 --> 0:03:43.840 It could be a really good story. 0:03:43.880 --> 0:03:46.840 Also, what people have pointed out that GameStop loves attention. 0:03:47.000 --> 0:03:49.400 They make a lot of flashy moves. Chief executive Officer 0:03:49.440 --> 0:03:54.160 Ryan Cohen posts very controversial things on social media about 0:03:54.160 --> 0:03:56.560 the American business market. He has a lot of like 0:03:56.640 --> 0:04:01.760 moving breaking strategies that he's excited to express, and you 0:04:01.800 --> 0:04:04.720 know that's worked for GameStop. Game Stop managed to raise 0:04:04.760 --> 0:04:07.920 some significant money from the meme stock frenzy and that's 0:04:07.960 --> 0:04:10.920 allowed it to sustain a little bit longer. So who knows, 0:04:10.960 --> 0:04:14.760 this could be a really exciting, fresh, strange way to 0:04:15.040 --> 0:04:17.960 keep game Stop going, or it could be like some 0:04:18.040 --> 0:04:19.920 of the analysts are saying, a little bit improbable. 0:04:20.480 --> 0:04:22.279 Well, you're going to be across it. It was a 0:04:22.279 --> 0:04:25.240 busy weekend for Cecilia Denasalgier. We're pleased that she joins 0:04:25.279 --> 0:04:28.000 us to break it all down today. Meanwhile, well, let's 0:04:28.040 --> 0:04:29.800 just take a look at the broader markets, because tech 0:04:29.839 --> 0:04:33.080 stocks are again outperforming, but we're building on records that 0:04:33.120 --> 0:04:35.239 we set back on Friday. We're up about a quarter percent, 0:04:35.480 --> 0:04:37.480 but still in ASDA one hundred, as I say, at 0:04:37.480 --> 0:04:41.040 a record. And it's not just what's happening over in America. Look, 0:04:41.080 --> 0:04:43.359 the AI trade is back on track and we're seeing 0:04:43.360 --> 0:04:45.080 that play out in Asia trading as well. 0:04:45.360 --> 0:04:47.800 Ske Heinez extraordinary. 0:04:47.040 --> 0:04:50.120 Moves for the high bandwidth memory maker up eleven percent. 0:04:50.200 --> 0:04:53.040 Samsung as well, up three point four percent, even as 0:04:53.040 --> 0:04:55.159 the company maybe having to navigate some sort of strike 0:04:55.200 --> 0:04:57.920 potential going on with the business. But records being set 0:04:57.960 --> 0:05:00.520 in Asia. Following through from what we saw in America 0:05:01.000 --> 0:05:03.320 comes last we can see more audacious m and A 0:05:03.640 --> 0:05:05.240 What are we going to see in terms of IPOs. 0:05:05.520 --> 0:05:07.880 Perfect person to talk it through is that Nancy Tanglish, 0:05:07.880 --> 0:05:12.120 the CEO CIO of Laffatanglo Investments. Look, your TGLR ETF 0:05:12.320 --> 0:05:14.440 is already outperforming even the S and P five found 0:05:14.480 --> 0:05:17.359 in which we note is basically at or near record highs, 0:05:17.720 --> 0:05:20.920 you're outperforming by six percent, Nanci Is tech going to 0:05:21.000 --> 0:05:22.200 remain the dominant force here? 0:05:23.440 --> 0:05:25.480 I think so, Caroline, thanks so much for having me. 0:05:26.279 --> 0:05:27.000 I do think so. 0:05:28.040 --> 0:05:31.000 If you just think about what is driving margins on 0:05:31.120 --> 0:05:35.360 the SMP companies, it is either the technology companies themselves 0:05:35.800 --> 0:05:40.240 or the use of technology by old economy companies. One 0:05:40.279 --> 0:05:43.400 of the noteworthy earnings comments came from L three Harris 0:05:43.440 --> 0:05:46.680 where they said revenue per employee has increased this is 0:05:46.839 --> 0:05:51.000 CEO by almost twenty five percent over the last couple 0:05:51.040 --> 0:05:54.559 of years. So if you can produce more with less, 0:05:55.040 --> 0:05:57.599 then that is what we call productivity, and that will 0:05:57.640 --> 0:05:58.200 continue to. 0:05:58.240 --> 0:05:59.480 Drive stock prices higher. 0:05:59.480 --> 0:06:04.520 As we've seen in increasing earnings estimate revisions up. 0:06:05.040 --> 0:06:09.039 Are you buying that's AI fueled productivity gains? There's such 0:06:09.120 --> 0:06:12.240 mixed messages BIS out of you read the bound of 0:06:12.360 --> 0:06:15.560 International Settlements. I said, they've seen about four percent productivity 0:06:15.640 --> 0:06:19.040 gains more generally, and that seems to be being sustained. 0:06:19.080 --> 0:06:21.159 But if L three House is seing twenty five percent, 0:06:21.240 --> 0:06:23.880 what more are these CEOs doing well? 0:06:24.520 --> 0:06:26.640 So I think you have to listen to the companies. 0:06:27.120 --> 0:06:29.760 My team sits on all the earnings calls and we 0:06:29.880 --> 0:06:32.280 develop themes out of those earnings calls and then we 0:06:32.400 --> 0:06:35.960 discuss them. And we're hearing it from a broad base 0:06:36.520 --> 0:06:40.640 across economic sector group of companies, not every company, but 0:06:40.880 --> 0:06:43.040 most companies. I mean, you even heard it from ge 0:06:43.200 --> 0:06:47.360 Vernova that said, yeah, we're benefiting from AI in the buildout, 0:06:47.640 --> 0:06:48.760 but we all want. 0:06:48.680 --> 0:06:52.240 You also to know that we're utilizing AI to produce 0:06:52.360 --> 0:06:53.200 our productivity. 0:06:53.360 --> 0:06:57.839 Our margins are output for employee as well and managing 0:06:57.920 --> 0:07:01.360 the supply chain. So I mean, just use ax ask 0:07:01.560 --> 0:07:04.719 B on Bloomberg and it can save you a colossal 0:07:04.760 --> 0:07:08.360 amount of time as an analyst if in fact you 0:07:08.560 --> 0:07:13.800 utilize it. So, I mean, we've been bitpounding the tech 0:07:13.920 --> 0:07:17.160 trade for about four years now on the theme of 0:07:17.320 --> 0:07:20.640 old economy companies pivoting to the new technologies and then 0:07:20.720 --> 0:07:23.160 in fact the companies that are the suppliers. 0:07:23.440 --> 0:07:26.280 I think it continues, not forever, but for now. 0:07:26.760 --> 0:07:29.480 What's been interesting is the sort of the haves and 0:07:29.520 --> 0:07:30.080 the have nots. 0:07:30.160 --> 0:07:31.880 Now, I'm in no way going to say Metas not 0:07:32.360 --> 0:07:34.320 have not yet we can't write them off, but there 0:07:34.560 --> 0:07:36.920 what they don't have is cloud compute that they sell 0:07:36.960 --> 0:07:41.040 to others. They're busy investing in their own AI infrastructure. Meanwhile, 0:07:41.120 --> 0:07:43.960 Alphabet was applauded for the fact that it's investing, raising 0:07:44.040 --> 0:07:46.640 capital expenditure, but also able to benefit with the. 0:07:46.640 --> 0:07:47.880 Growth of Google Cloud. 0:07:48.320 --> 0:07:51.320 How are you seeing earnings play out, because both we've 0:07:51.360 --> 0:07:54.160 seen from Meta they are having to cut people, having 0:07:54.240 --> 0:07:57.200 to talk about laying off a sizeable amount of the 0:07:57.280 --> 0:08:00.080 workforce to be able to redeploy capital back into the 0:08:00.120 --> 0:08:01.720 AI infrastructure. Is that something I'm going to head time 0:08:01.760 --> 0:08:03.480 and time again, and who do we congratulate for it 0:08:03.600 --> 0:08:03.760 or not? 0:08:06.160 --> 0:08:08.280 Well, so, I've talked about it quite a bit on 0:08:08.360 --> 0:08:10.040 your show that we got out of Meta. I just 0:08:10.120 --> 0:08:12.640 didn't like the business model. I didn't see the opportunity. 0:08:13.240 --> 0:08:15.480 We went into Spotify, which at the time was a 0:08:15.600 --> 0:08:19.560 good transaction. We were wrong about Meta to some extent, 0:08:19.640 --> 0:08:22.040 but we were right about Spotify, and we now think 0:08:22.120 --> 0:08:26.240 that we are somewhat vindicated on the Meta model because 0:08:26.320 --> 0:08:29.240 they aren't able to monetize AI in a way that 0:08:29.360 --> 0:08:32.439 Google and Amazon and even Microsoft are able to do. 0:08:32.840 --> 0:08:36.960 They are able to improve their advertising business. But Mark 0:08:37.080 --> 0:08:40.760 Zuckerberg has made these colossal pivots and been wrong in 0:08:40.840 --> 0:08:44.280 the past. So we own all of the mag seven 0:08:44.760 --> 0:08:47.719 x meta. It will have its day again just like 0:08:47.920 --> 0:08:50.040 Intel did. The question is how long. 0:08:49.960 --> 0:08:53.320 Will one have to wait and what is the opportunity cost. 0:08:53.320 --> 0:08:56.680 And the global opportunity cost or global opportunity We were 0:08:56.760 --> 0:08:59.760 just talking about how Asia trade has just seen colossal 0:08:59.800 --> 0:09:02.719 game still for the AI infrastructure betsna see. Is that 0:09:02.800 --> 0:09:04.720 still going to be the winning formula the chip makers? 0:09:04.760 --> 0:09:07.200 The designer is right now, I think. 0:09:07.080 --> 0:09:09.480 In the near term, Caroline, but I think what we're 0:09:09.520 --> 0:09:12.000 going to start pivoting to and we launched a thematic 0:09:12.080 --> 0:09:16.319 portfolio with a focus on space in robotics, quantum and 0:09:16.480 --> 0:09:19.280 then nuclear and then of course the metals and minerals 0:09:19.280 --> 0:09:22.480 that are needed to drive those technologies. So I think 0:09:22.600 --> 0:09:25.319 for the near term, I mean, we're hearing that the 0:09:25.440 --> 0:09:29.120 backlogs are enormous again in the infrastructure names, in the 0:09:29.240 --> 0:09:32.400 chip names, and I think at some point we will 0:09:32.440 --> 0:09:36.680 begin to see software as a major beneficiary. Not every 0:09:36.760 --> 0:09:40.880 software company. We exited Adobe and CRM quite some time ago, 0:09:41.440 --> 0:09:43.920 and we've been wrong about service now so far. But 0:09:44.040 --> 0:09:46.599 I think you'll start to see the winners there and 0:09:46.640 --> 0:09:50.120 then you'll see the market pivot at least back to software. 0:09:50.200 --> 0:09:52.800 So I think there's still a lot of opportunities in 0:09:53.720 --> 0:09:57.280 this new industrial revolution we're living through Naze Tangla. 0:09:57.440 --> 0:09:59.440 It's always great to have you on CEO and CIO 0:09:59.520 --> 0:10:03.200 of l Hangler Investments on all things AI and tech trade. Meanwhile, 0:10:03.240 --> 0:10:06.959 coming up more AI, open Ai Anthropic are each forming 0:10:07.040 --> 0:10:10.840 joint ventures aimed at boosting AI adoption. Will have the details. Next, 0:10:11.320 --> 0:10:23.960 this is Blue meg Tech, Open Ai and Anthropic are 0:10:24.120 --> 0:10:26.800 each forming partnerships aimed at boosting the use of AI 0:10:26.920 --> 0:10:30.400 tools by businesses. Are sourcing saying that open ai has 0:10:30.480 --> 0:10:32.600 raised more than four billion dollars for a new joint 0:10:32.679 --> 0:10:35.439 venture that includes backing from TPG, Brickfield Asset Management and 0:10:35.520 --> 0:10:39.600 Being Capital. Meanwhile, rival Anthropic is partnering with Blackstone Helman 0:10:39.679 --> 0:10:42.719 Friedman Goldman Sachs. According to a statement, let's get the 0:10:42.800 --> 0:10:45.120 latest boom Meg Seth Figureman, this is all about getting 0:10:45.200 --> 0:10:50.600 private equity companies, portfolio companies to use the technology. 0:10:50.120 --> 0:10:53.959 Right Yeah, and there list of portfolio companies that work 0:10:54.000 --> 0:10:56.160 with them. And I think it's a really comment in AI. 0:10:56.480 --> 0:10:57.959 For most of the air boom, a lot of the 0:10:58.000 --> 0:11:00.160 focus has been on the competitive landscape of who can 0:11:00.200 --> 0:11:03.199 build better and more sophisticated models. Now we're seeing a 0:11:03.320 --> 0:11:06.920 really important vector is who can do better at gegging 0:11:07.040 --> 0:11:10.880 businesses to understand and adopt that technology in their day 0:11:10.880 --> 0:11:11.400 to day work. 0:11:11.840 --> 0:11:14.120 There's just different tactics in which it's being deployed. 0:11:14.240 --> 0:11:19.160 So open ai is raising money themselves alongside outside investment 0:11:19.240 --> 0:11:20.800 for a company that's going to be worth all told 0:11:20.920 --> 0:11:25.480 ten billion, not including that funding, and then how does 0:11:25.520 --> 0:11:28.480 it roll out because they've also had this interesting deal 0:11:28.640 --> 0:11:31.800 where they helped invest in Thrive Holdings, which is Thrive 0:11:31.920 --> 0:11:35.720 Capital's ability to sort of deploy AI into other parts 0:11:35.760 --> 0:11:36.520 of the ecosystem. 0:11:36.640 --> 0:11:38.680 Yeah, I mean at some level this is being pitched 0:11:38.720 --> 0:11:42.079 as a separate venture, but Opening I will be the 0:11:42.240 --> 0:11:45.079 majority owner and in control of it according to the 0:11:45.160 --> 0:11:47.839 sourcing that we have here, which suggests that they will 0:11:47.880 --> 0:11:50.760 have a key role in dictating how this proceeds. And 0:11:50.800 --> 0:11:53.840 I think again the idea is rather than open ai 0:11:54.160 --> 0:11:56.559 on its own balance sheet, within its own ranks figuring 0:11:56.559 --> 0:11:58.520 out how to get their technology into the hands of 0:11:58.600 --> 0:12:01.280 more business leaders, they will now work with this consortium 0:12:01.320 --> 0:12:05.320 of partners and investors across thousands of potential portfolio companies to. 0:12:05.360 --> 0:12:05.880 Do that work. 0:12:06.360 --> 0:12:09.839 I mean, while the timing just sensational that Anthelopic is 0:12:09.840 --> 0:12:11.880 announcing the same thing on the same day. 0:12:12.600 --> 0:12:14.440 Again, it speaks to the moment that we're in right now. 0:12:14.520 --> 0:12:17.559 These two companies are following somewhat similar playbooks to bolster 0:12:17.720 --> 0:12:21.120 revenue and business traction heading into the potential IPOs. 0:12:21.679 --> 0:12:25.000 So figureman across an astounding amount of stories. 0:12:25.080 --> 0:12:27.880 As ever, we thank you very much. Indeed, sticking with 0:12:28.120 --> 0:12:28.600 Open Ai. 0:12:28.840 --> 0:12:31.120 Look, the startup's president, Greg Brockman is set to take 0:12:31.120 --> 0:12:33.920 the stand today in a company's legal battle against Elon Musk. 0:12:33.960 --> 0:12:36.959 Brotman is expected to face probe questions about. 0:12:36.880 --> 0:12:40.000 A personal journal he kept, which Musk's legal team says 0:12:40.360 --> 0:12:43.400 is evidence he and open Ai CEO Sam Altman intended 0:12:43.440 --> 0:12:47.640 to abandon the company's nonprofit mission now. 0:12:47.600 --> 0:12:48.360 For all this AI. 0:12:48.960 --> 0:12:52.560 Guess what you need compute and chipmaker and data center operators. 0:12:52.600 --> 0:12:55.360 Cerebrius is looking to capitalize on that very demand. The 0:12:55.480 --> 0:12:57.920 nvideo rival, as it likes to see itself, is seeking 0:12:57.920 --> 0:12:59.240 to raise as much as three and a half billion 0:12:59.280 --> 0:13:02.040 dollars from a pub offering later this month. Price range 0:13:02.120 --> 0:13:04.800 and its security is filing really confirms Bluemog's earlier reporting. 0:13:05.120 --> 0:13:06.959 Let's get to one of the reporters who's always across 0:13:07.080 --> 0:13:11.000 IPO's Bailey Lipshaltz. No, this is a company that wants 0:13:11.040 --> 0:13:13.239 to offer more inferencing capacity. 0:13:13.679 --> 0:13:15.120 Is it really a competitor end. 0:13:15.120 --> 0:13:16.120 Video in their mind? 0:13:16.200 --> 0:13:18.679 Yes, And even when I talk to investors, there's a 0:13:18.760 --> 0:13:21.760 world that they will kind of be another option, if 0:13:21.760 --> 0:13:23.160 you will. It goes back to kind of the debate 0:13:23.200 --> 0:13:26.040 around Starlink and saying, Okay, well, yes they dominate that 0:13:26.200 --> 0:13:28.640 potential market, but there have to be other players. So 0:13:28.720 --> 0:13:30.640 while there is at least for now a monopoly of 0:13:30.720 --> 0:13:33.920 source for in video, when you look at the partnerships 0:13:33.960 --> 0:13:36.679 that Serabras has inked, obviously G forty two was a 0:13:36.720 --> 0:13:38.760 big risk, but now leaning into opening Eye and kind 0:13:38.760 --> 0:13:41.840 of diversifying the customer flow, it does seem like that's 0:13:41.840 --> 0:13:43.240 a bit of a ball case. And keep in mind, 0:13:43.320 --> 0:13:46.360 even with this potential IPO, we're talking about a company 0:13:46.480 --> 0:13:49.920 potentially valued at thirty three billion dollars, not four trillion dollars. 0:13:50.760 --> 0:13:51.800 Good context. 0:13:52.200 --> 0:13:53.679 Look, this could all come to bear as soon as 0:13:53.720 --> 0:13:56.560 May thirteenth, right, we're thinking that's when the pricing's happening, 0:13:56.640 --> 0:13:57.559 the trading begins. 0:13:58.280 --> 0:13:59.800 Is this going to be a retail focused name. 0:14:00.000 --> 0:14:02.480 Who's going to be understanding the Cerebraus story because G 0:14:02.600 --> 0:14:04.720 forty two, as you know, as you were just illuminating 0:14:04.840 --> 0:14:07.280 to us, Look, this is a geopolitical story as well. 0:14:07.480 --> 0:14:09.760 It is geopolitical story, but you have to keep in 0:14:09.840 --> 0:14:12.000 mind that this is a company that fits into the 0:14:12.240 --> 0:14:15.439 themes that are dominating capital markets right now, AI and 0:14:15.520 --> 0:14:18.480 AI infrastructure. So we've seen investors lean into the X 0:14:18.640 --> 0:14:21.120 energies of the world that want to promise a nuclear future. 0:14:21.320 --> 0:14:24.000 We're talking about this name with a number of hedge 0:14:24.040 --> 0:14:26.200 funds and long only investors, and they're saying, you kind 0:14:26.200 --> 0:14:28.360 of have to own this company if you play the 0:14:28.560 --> 0:14:32.360 IPO calendar, So it will broaden across reach across kind 0:14:32.360 --> 0:14:34.880 of the broader investor group, if you will. And one 0:14:34.880 --> 0:14:37.240 thing to keep in mind with this launch, we're expecting 0:14:37.280 --> 0:14:40.600 potentially eight billion dollars of IPOs proceeds being raised in 0:14:40.600 --> 0:14:42.480 the next two weeks. This will be a big part 0:14:42.520 --> 0:14:44.600 of that, and how this performs will help set the 0:14:44.680 --> 0:14:47.040 stage for SpaceX likely next month and. 0:14:47.080 --> 0:14:49.360 Then maybe open AI on propic. Look, this is a 0:14:49.400 --> 0:14:51.720 great time considering where the market's trading out right. 0:14:51.840 --> 0:14:53.520 I mean, that's been the interesting thing right is you 0:14:53.640 --> 0:14:56.640 look at the socks. We were up more than thirty 0:14:56.680 --> 0:14:58.720 five percent in the last month. So when you look 0:14:58.760 --> 0:15:02.120 at expectations for this company in particular, a month ago, 0:15:02.400 --> 0:15:04.440 you would have said, okay, twenty five thirty, sure you 0:15:04.480 --> 0:15:06.200 can sell me on that. Now it's like thirty five 0:15:06.320 --> 0:15:09.280 to forty billion dollars, just because your comps are up 0:15:09.360 --> 0:15:11.960 so sharply. Just starts to make sense when you're looking 0:15:12.000 --> 0:15:14.160 at how you have to value a company when it's 0:15:14.200 --> 0:15:16.560 going public, is what do your comparables look like? Take 0:15:16.560 --> 0:15:18.440 a little bit of a discount, and that math is 0:15:18.680 --> 0:15:21.360 kind of working in the company's favor again, just given 0:15:21.400 --> 0:15:23.440 them move higher that we've seen for some of those peers. 0:15:23.560 --> 0:15:26.440 It's a different company, but it's almost tangential to the 0:15:26.560 --> 0:15:28.400 likes of a core Weave, which is all about trying 0:15:28.440 --> 0:15:29.640 to solve the compute issues. 0:15:29.880 --> 0:15:31.160 For core Weave, the. 0:15:31.160 --> 0:15:34.360 Headache has always been I've got more customers than you realize. 0:15:34.720 --> 0:15:36.960 Is this where maybe Cerebrius is going to have to 0:15:36.960 --> 0:15:38.160 convince the market too well. 0:15:38.240 --> 0:15:39.800 I think that was the big pitch and the big 0:15:39.880 --> 0:15:43.720 reason that diversifying away from G forty two at least 0:15:44.040 --> 0:15:46.640 is some indigestion for investors. Yes, when they try to 0:15:46.680 --> 0:15:48.800 go public the first time over a year ago, it 0:15:48.920 --> 0:15:52.320 was concentration risk, it was Siffius uncertainty, it was obviously 0:15:52.440 --> 0:15:56.120 kind of every number of potential bear cases. Whereas now 0:15:56.240 --> 0:15:58.760 the tide really has shifted, so you do have the 0:15:59.440 --> 0:16:01.720 kind of your focus of one customer similar to a 0:16:01.800 --> 0:16:04.160 core we were there was such a bare case of okay, well, 0:16:04.200 --> 0:16:06.480 how relying are you on one company? Now we're seeing 0:16:06.560 --> 0:16:08.760 that spread out just a little bit. The interesting thing 0:16:08.800 --> 0:16:11.200 will be how do they structure this lockup and what 0:16:11.280 --> 0:16:14.040 does the performance really look like six months from now 0:16:14.160 --> 0:16:16.400 as opposed to day one trading, which we've seen kind 0:16:16.400 --> 0:16:17.360 of can be all over the map. 0:16:17.840 --> 0:16:19.880 Well said, it's going to be a busy old week, 0:16:20.000 --> 0:16:21.880 old month for you. I think Badi Dipshils, we so 0:16:21.960 --> 0:16:24.320 appreciate it. On all things IPO is now coming up. 0:16:24.640 --> 0:16:26.400 We've got planning more to come on the world of 0:16:26.440 --> 0:16:29.120 AI and AI inference. I'm going to be hearing from 0:16:29.160 --> 0:16:31.520 a company US leading that charge in the private space. 0:16:31.600 --> 0:16:34.040 CEO of Deep Infra is going to be joining us Nicolair. 0:16:34.200 --> 0:16:36.080 For ourself, this is a bloomberg tech. 0:16:46.080 --> 0:16:47.120 In the race to build the. 0:16:47.200 --> 0:16:50.680 Next great AI models. The industry is tackling various bottlenecks, 0:16:50.880 --> 0:16:54.800 one of course being access to compute. Deep Infra is 0:16:54.880 --> 0:16:58.040 trying to alleviate access to AI inference with a purpose 0:16:58.080 --> 0:16:59.160 built cloud platform. 0:16:59.360 --> 0:16:59.960 It's just a cure. 0:17:00.040 --> 0:17:02.040 One hundred and some million dollars in a series B 0:17:02.160 --> 0:17:06.000 funding back by the likes of Video Samsung joining us 0:17:06.080 --> 0:17:09.159 now from San Francisco is deep in for CEO Nicola 0:17:09.400 --> 0:17:11.359 or asof Nicola, It is great to have you on 0:17:11.480 --> 0:17:14.640 the show. Why would in video want to put money 0:17:14.680 --> 0:17:17.080 into you? How are you helping the GPU story? 0:17:18.800 --> 0:17:23.920 Yeah, thanks thanks for having me. Well, we're helping companies 0:17:23.960 --> 0:17:28.520 get access to these great open source models, and we're 0:17:28.640 --> 0:17:33.800 building this purpose built inference cloud. So it's it's really 0:17:33.920 --> 0:17:36.199 you need to you need to build a specialized infrastructure 0:17:36.240 --> 0:17:41.160 to do inference efficiently, and it really helps everyone access 0:17:41.440 --> 0:17:44.960 these great open source AI models really well. And I 0:17:45.080 --> 0:17:48.880 mean that's why Vida was excited to participate in around. 0:17:49.480 --> 0:17:53.080 We were just hearing that Cerebraus, which also wants to 0:17:53.119 --> 0:17:56.600 speed up AI inference offerings, is going to be tapping 0:17:56.680 --> 0:17:59.960 the public markets. It sees itself as an in video 0:18:00.080 --> 0:18:04.080 a competitor. How Nicola are you more of an adjacent 0:18:04.359 --> 0:18:07.000 to the Nvidia story. How are you looking to help 0:18:07.080 --> 0:18:09.879 deploy GPS? Why is your compute different? 0:18:12.640 --> 0:18:16.320 You know, we we've tried a number of different inference accelerators, 0:18:16.400 --> 0:18:19.880 but we believe like Nvidia's hardware is still the most efficient, 0:18:20.160 --> 0:18:25.840 the best to do inference at and we're just doubling 0:18:25.880 --> 0:18:29.800 down on that kind of hardware platform, and we like 0:18:30.000 --> 0:18:33.080 working closely with the teams that the video and just 0:18:33.240 --> 0:18:37.399 making inference more efficient. Lowering the price per token is 0:18:37.760 --> 0:18:44.400 what we really like to focus on, and so that's 0:18:44.480 --> 0:18:46.760 kind of our path. We you know, there would be 0:18:46.840 --> 0:18:49.359 a lot of demand for inference down the line, like 0:18:50.000 --> 0:18:52.199 we think that eighty percent of the compute is going 0:18:52.280 --> 0:18:56.440 to go towards inference, But at the moment, you know, 0:18:56.960 --> 0:18:59.960 we've invested heavily into the Nvidia hardware stack. 0:19:00.960 --> 0:19:04.880 What's interesting is you're already processing what five trillion tokens 0:19:04.920 --> 0:19:07.879 per week? How do you get the efficiencies? How do 0:19:07.920 --> 0:19:10.119 you drive down the cost of token? 0:19:12.680 --> 0:19:16.080 It's a lot of hard work, and we look at 0:19:16.119 --> 0:19:19.399 it through the whole stack of like where we build 0:19:19.440 --> 0:19:23.440 this kind of inference clusters, which data centers we go in, 0:19:24.119 --> 0:19:26.560 how we structure them, and then a lot on the 0:19:26.640 --> 0:19:30.240 software on top of it. We believe like one of 0:19:30.320 --> 0:19:33.520 the key things is to have very good caching of tokens. 0:19:33.560 --> 0:19:36.520 As these agents kind of interact with theim models. They 0:19:37.400 --> 0:19:41.480 basically make many requests in a loop with roughly the 0:19:41.520 --> 0:19:44.960 same context. So like I think kV cash is a 0:19:45.320 --> 0:19:51.760 really key to efficiency of inference. And you know, we've 0:19:51.760 --> 0:19:55.359 been doing this for about four years. We saw the 0:19:55.920 --> 0:20:00.359 demand for inference coming and we really focused on how 0:20:00.440 --> 0:20:02.720 do we build an purpose build inference cloud. 0:20:03.840 --> 0:20:06.320 And I think therefore you are scaling out you have 0:20:06.680 --> 0:20:08.920 was it you're operating out of eight data centers. I 0:20:08.960 --> 0:20:11.399 think it is already with this new money one hundred 0:20:11.400 --> 0:20:13.680 million more than where do you deploy that? 0:20:16.480 --> 0:20:19.560 We will use the capital from this fundraise to scale 0:20:19.560 --> 0:20:20.240 our platform. 0:20:20.440 --> 0:20:21.000 We will. 0:20:22.760 --> 0:20:26.359