Altman Meets with Mideast Investors as OpenAI Eyes $830 Valuation

32:33But more broadly, he also talked about where he is in terms of Starship, where he aims in terms of space and the idea that they want to be proving full reusability for Starship this year. And of course, then the favorite talk of most CEOs at the moment, if you're in the world of AI, is putting AI data centers in space. And what that means in terms of solar energy, there was a lot talked about in terms of energy. Did that take your interest? Because really, that seems to be the bottleneck of choice at the moment. Yeah, you know, Tesla has an energy division, right? And so it was obvious that he would kind of talk that part up.

33:07But this is something in the consciousness of the administration, the rest of supply chain. China dominates supply of solar energy. And whatever SpaceX's long-term ambitions are for data centers in space, clearly Larry Tink tried to get Elon Musk to say, hey, the US and Europe should be doing more. Musk didn't really bite, but he did go as far to say, well, look, this is what Tesla and SpaceX are doing. over the next two to three years, no-brainer to have solar-powered computing space, but also ramp up the output of cells, because if you don't sell them into markets like Europe, China does. Precisely.

33:42Also talking about how cheaper AI could become. We're all thinking about the cost of compute, the cost of energy. If you put those AI data centers in the next two to three years, he's saying more broadly space is the cheapest place to put AI in two to three years. It's interesting the FT's just been sending a headline talking about Elon Musk, SpaceX, lining up four banks for an IPO. So we are all in on the space story right now. And he was asked, what inspires you? And he said when he was a kid, he read science fiction books. So if you're a kid out there watching and you want to be the next Elon Musk, that's what you've got to do.

34:12We're also tracking and looking at live images of Blue Origin's new Shepard NS38 latest mission. We're on a hold right now. The window had been due to open 11 a.m. Eastern time, 8 a.m. here on the West Coast for NS38. By this point, 90 people have gone up on Blue Origin, but they're on a hold, a technical hold. We'll keep tracking it, but you can watch the launch on LiveGo on the Bloomberg terminal. Caro, there is a lot going on in financial markets too. There are, and more broadly, we're seeing a bit more optimism come to life in the markets. Look, the geopolitical angst put to one side and the optimism returning about all things AI.

34:49At one point, we're thinking the narrative is all around NVIDIA and Jensen Huang being bullish once again on the trillion dollars of spending that's going to be going into the idea of AI data centers, the need for chips. But we've also got earnings on deck later today. That's something you're going to be digging into. At the moment, Nasdaq's up six tenths of a percent. But look, we're in the green across the board when you're looking at every benchmark across the entire world, Ed. I'm taking a look at Intel because Intel posts aftermarket. And even though the chip maker is still embattled, the stock has been on a hell of a run of late.

35:20You know, on the week, up almost 15 % on track for its best week. I think in quite a long time. But right now we're treading water. Of course we are. Why? Because we'll wait and see actually the meat of what comes out of that earnings print again, which is after the market close around 4 p.m. Eastern time. Caro, there are many news stories as well. There are in the private markets too. And guess what? OpenAI is fundraising. We knew that. But the CEO, Sam Altman, he's been really busy traveling in the Middle East. He's been meeting with top investors in an effort really to secure funding for an investment round that could total, get this,$50 billion.

35:56According to sources, Numez Shireen Ghafari joining us to cover the story. I mean, the valuations are gargantuan. We're thinking up to$830 billion is where it could be. But the focus on the Middle East, perhaps unsurprising for these foundation AI labs. That's right. When you think about the pools of capital that are required for these kinds of mega fundraising rounds, you're tapping beyond the usual network of Silicon Valley VCs or even New York finance firms, right? And so that's why you see CEOs like Sam Altman traveling to places like Abu Dhabi to court investors. Shereen, you know, we have tried to track this as closely as possible.

36:36I think one of the things reading the report is the detail of what we know about how Sam Altman operates, the travel, who he meets with, the specific names behind the sovereign wealth funds and the jurisdictions he's going to. Give us all the detail you can. gives a good insight in what he's trying to pull off. I mean, if you think about it, this is a time when there is great interest in the Middle East and other regions and tapping into the AI boom. There are also other companies fundraising. We know Anthropic is also, you know, set to finalize around very soon as we've reported, you know, SpaceX may be going IPO, but in the past they've also been fundraising.

37:15So we have, you know, Even when you get to these larger pools of capital that extend beyond the U.S.'s ability in terms of EC firms, there's still a limit to how much these sovereign wealth funds can give. So it requires travel, it requires relationships, it requires going out and actually meeting people, shaking hands and hopefully securing a deal. This is a big deal. What we're hearing, right, just to recap,$50 billion target raise,$750 billion to$830 billion valuation. A lot of eyes on it. Bloomberg, Shereen Ghaffari leading the team. that's been reporting and tracking it. Apple, it's set to revamp Siri later this year, turning the digital assistant into the company's first AI chatbot.

37:55The feature is set to be integrated into Apple's lineup of iPhones, iPads, Mac operating systems. Newmost consumer tech editor, Mark Gurman, broke the details. You join us now, Mark. It's interesting how much they're depending on Google technology, potentially on their servers as well, but this is quite a move to be going to the chatbot form. This is a blockbuster move for Apple for the last several months, even in light of the new Siri being delayed and not launching on time a year and a half ago, even in light of the botched launch and the delays to Apple intelligence, Apple was steadfast in saying that it doesn't believe in the chatbot route that OpenAI is taking with ChatGPT, that Google is taking with Gemini, that Microsoft is taking with Copilot.

38:34But everyone who's used these chatbots know that this is at the very center of the generative AI battleground. ChatGPT is nearing a billion monthly active users, right? It's clearly resonating. And so Apple knows it needs to go in this direction, and it's doing exactly what it needs to do to make its consumers happy and to keep selling its devices and keeping its operating system in line with the future. So at the end of this year, they're going to revamp Siri into a chatbot, and it's going to be far more capable than what you're seeing from chatbots today, having those features, having web search, but also having deep ties into Apple's devices, allowing you to control specific apps and features.

39:14It's interesting to see the shares up about 9 tenths of a percent after the reporting. And when I read the report, it's clear Siri will be fundamentally different. But is this a direct result, Mark, of that agreement that Apple reached with Google for Gemini to underpin this next gen of Siri? Is that kind of what the unlock has been? That has been the unlock for them to bring this to market. This is a concept that they've had for a while now using their internal models that they built for Siri and Apple intelligence a couple of years ago. but based on all the delays and based on all the issues the companies had, it would have been impossible to launch that with their internal models because it just wouldn't work well.

39:52It would be subpar and they would have crisis round two for their AI strategy. So yes, as you said, partnering with Google for Gemini has been the unlock to actually bring this to market. Now, there are some important details here. There's a new version of Siri launching in the coming months in the first half of this year, likely around March of April. That brings to market features they announced two years ago in 2024 at their developer conference. The assistant being able to tap into personal data, being able to know what's on your screen, precise control of applications and the operating system.

40:23And to do that, they're using a custom Gemini model that mixes with their internal models, runs in the cloud on Apple servers. But this chatbot is so powerful and needs to have as much functionality as possible to really compete with OpenAI and what Google and Microsoft offer today. So what they're discussing doing is actually running this Gemini model on Google's cloud platform services, as well as on Google TPUs, which is a big sea change for what Apple's doing currently with Gemini. Yeah, we've talked about the TPU unlock as well. Bloomberg's Mark Gurman back strong with a big scoop. Thank you very much.

40:58Let's turn to chips that Power AI, Alibaba, is preparing to IPO its chip making arm, T-Head. Bloomberg's executive editor for tech, leading the Asia team, Peter Elstrom, joins us now. This is interesting. It's kind of a twofer. We know from our reporting they're planning to IPO this unit. But in the interim, they're also going to restructure it. What do we need to know? Yeah, that's right. It's a two-step process that Alibaba is looking at here. They're going to begin by restructuring the ownership of the company, allowing employees to own some of the shares here, and then later on they plan on taking it public.

41:34Now, this is an interesting move by Alibaba. Alibaba, of course, is an e-commerce company at its core. It's been diversifying into a number of different areas, including AI. They have one of the leading AI models in China with its Gwen platform. And they've been working on this chip effort, too. It's a bit like Amazon's push into being able to develop their own AI chips and Google's. You talked about the TPUs. It's an interesting strategic move by Alibaba because they're deciding that they do want to go ahead and spin off that chip unit, see if they can give it some independence, give some ownership to the employees so that they'll have a motivation to be able to competitive.

42:09But then they have a very big market to address here. And they have some very significant competitors in China and, of course, beyond China, too. More Threads Technologies just listed, Cameracon Technologies. Briefly, Peter, the context of the support from China on this. Well, this is national policy. As we've talked about a number of times before, the U.S. has cut off NVIDIA from selling its most advanced chips into China. In Beijing, that's seen as a very serious threat. They want to have some domestic alternatives to that. So they have a number of domestic players that are coming on strong.

42:39Huawei is leading the way. Cameracon is probably second. CameraCon is unknown in the West, really, but it's an$80 billion company. They're doing quite well. And so companies like Alibaba and MoreThreads and some of the other small players see a golden opportunity here. If they can develop the technology for these AI chips, they're going to have a very vibrant domestic market. Baidu, also looking at doing something similar. Peter Elstrom, great breakdown. We thank you. The idea that China is behind in AI is a fairy tale. Those are the words from Arthur Mensch, CEO of Mistral AI, Europe's only large language model player.

43:14He spoke with Bloomberg's Germano Bissechi in Davos. And he also weighed in on whether European companies stand to benefit from enterprises looking to move away from U.S. providers for geopolitical reasons. Listen to this. it certainly is a growing topic top of mind in the ceo's mind of enterprises that their dependency to single providers when it comes to digital services might become a problem and is not a great answer to the volatility that we observe in the world so it's true in europe but it's also true outside of europe it's true in the u.s with certain accounts it's true in Canada, it's true in the Southeast Asian region where we operate a lot as well.

43:57So we do see that the technology we've built and the differentiation of it, which is AI that you can customize and deploy wherever you want, is something that does resonate with the geopolitical events that we see. Well, staying with the geopolitics, one of my colleagues spoke to the Anthropics CEO earlier this week. And on China, he thought that the idea of sending advanced chips there, and this is a quote, is crazy, and said it's a bit like selling nuclear weapons to North Korea. What's your view? Well, I don't think this is true. It turns out that China does very well without NVIDIA chips. We've seen China rise on the open source front, and we have been at Mistral, basically the spearhead of open source in the West, outside of China.

44:48Is China behind, though, the West? China is not behind the West. I think this is a fairytale. In AI, they are very much at parity, and the year ahead is going to be extremely interesting in that respect. We care about Europe maintaining its position, Europe maintaining its ability to train models, because we don't think that we should rely on open source Chinese models in very critical applications. So we need to be able to create our own models and to sell them to enterprises. So it's going to be very interesting, but it's a fairytale that China is behind. Let's talk about your own business. Do you have a revenue goal for this year?

45:27Yeah, we do. We should cross a billion by the end of the year. Okay. And in terms of capex spending, what are you factoring in? We are about to do around a billion in capex spending this year on our Endeavor Mistral Compute, which is what we've done last year is to start and build our clusters and to fill some of the data sensors we're renting with these clusters. The idea being that because we have the software platform, because we know how to build the models and the applications on top, combining them with the GPUs is a way to get to a vertically integrated stack, AI cloud services that are competing with some of the other fully integrated providers.

46:07Mr. CEO Arthur Mench there, along with Bloomberg's Jamal Abasechi. But as we just heard, the status of China's AI progress has come up repeatedly at Davos. Yesterday, we heard Google DeepMind CEO, Demis Hasselhoff, say that China was amid six months behind the West. These comments come as politicians have been more focused, of course, on tensions between the US and Europe, perhaps, rather than the AI race with China. Let's discuss the AI race with China with Liza Tobin, managing director at Gano Global. Your expertise is so important. You're at the Special Competitive Studies Project. You were also helping with national security at the government level and the China director there, Liza.

46:42So are you on the idea that it's a fairy tale, that they're behind, that actually they are in lockstep with the United States in terms of AI agility and compute power? Well, good morning, Caroline. It's great to be on with you. Yeah, so a couple of things can be true at the same time. China is absolutely going gangbusters in AI innovation at several layers of the stack. They're doing amazing things in open source models and applications and, of course, at the energy level. Where America's number one advantage still lies is in computing power at scale. I think that's widely agreed upon. Jensen Huang, in his interview at Davos on Fox, was acknowledging this when he was talking about the demand constraints that he's facing and how this growing demand for a limited supply of these AI chips is actually growing right now.

47:42And so that's why it's somewhat ironic and unfortunate for where America is. Let me just jump in here. Sure. Let me just jump in. Jensen Huang may have said that in that interview, but Jensen Huang's got a vested interest to drive the sales of what is the world's most valuable company in a market he said is a potential$50 billion market, right? What he doesn't address is the concern of those that basically don't think the balance is right. Exporting some deprecated technology still being a national security risk, in part because it allows China to catch up, which is the debate of Davos with having.

48:18Have they got that balance right to your mind? No, the current policy in the United States doesn't. They are the policy to license H200 chips to China is giving them a lifeline right where they need it most. You know, the compute advantage that the United States has was poised to grow exponentially if the controls stay in place. But President Trump and Jensen Huang, for the reasons you suggest, NVIDIA wants a foothold in the China market. What's ironic, of course, is that now Xi Jinping is waffling about whether he's even going to let his companies buy many of these H200 chips because he wants to make sure that much of the demand is going to the domestic chip makers.

49:05What's interesting is many would say we need all of technology across the world built on underlying U.S. chips, and that's the argument to keep them going into China. Eliza, do you have much credence to that? Are you more looking at Representative Brian Mast? He's saying basically that Jensen Huang, you and your paid minions, he says, are fighting to sell millions of advanced AI chips to Chinese military companies like Alic Barbar and Tencent. Yeah, it's interesting that Jensen Huang is trying to shift the narrative away from the military aspects of AI. Of course, these chips are inherently dual use.

49:41And China has a military-civil fusion system in place where you can't guarantee that once these chips get into China that they won't make them available for military use. That's just a fantasy that we can sort of control that. And so, but you see that that narrative of these things potentially being used for military or intelligence uses is inconvenient. it so jimson huang is trying to kind of pivot away from that liza tobin of garner global thank you these are live pictures of west texas van horn and blue origins launch pad new shepherd 38 let's listen in we brace for liftoff six five four command is start two one zero

50:595 ,000 feet.

51:04Crew capsule passing 11 ,000 feet.

51:10All right, New Shepard has cleared the tower. Take a look on your screen. On the left-hand side of the screen, you'll be able to follow along with our telemetry, giving you altitude and speed. And then on the bottom right-hand side of the screen, you'll be able to see how far we are into NS38's flight.

51:33Okay, you are watching live images of Blue Origin's NS38 New Shepard mission carrying the six latest astronauts, commercial passengers, to just above the Kármán line out of Van Horn, West Texas. We've just hit max Q, the moment of maximum aerodynamic pressure or stress on that vehicle. The Blue Origin New Shepard design relies on its own BE3 PM engine, where the propellant is a mix of hydrogen and oxygen, the byproduct of which is water for those chemistry nerds of you that are out there. And Caro, you know, we cover this because it is the evidence of the development of commercial space. You have six people on board.

52:19Some of them have paid for the privilege. We don't know how much because they don't disclose it. Just for 60 seconds of weightlessness in space in what Blue Origin says are the biggest windows to ever go into space. And by now, this is almost becoming regular. And 90 humans above the Carmen line have been flown by Blue Origin thus far. Right. So we're starting to see the cadence build up. And where does that push Blue Origin in terms of its next steps in commercialization? Again, Blue Origin is a multifaceted business. What you're watching on your screen now is the combined vehicle makes its way up to that Carmen line.

52:57You have the third stage booster and then the capsule on top. This is space tourism, essentially, right? They will also argue that it provides a zero-gravity environment, even for 60 seconds, to do scientific experimentation. But the news last night, Blue Origin is also working on a Starlink competitor, right? satellite-based and constellation-based connectivity, then they have a more powerful rocket, New Glenn, which is used for a wide range of commercial applications, but putting satellite deployment into low-Earth orbit. We're about three minutes into this mission. We expect that the separation to happen imminently, and then what you have is that three-minute, 30-seconds mark, Caro, those six passengers go to the Kármán line, which is the kind of recognized boundary of space where they'll get out, undo their seatbelts and float around looking out of the window.

53:47And again, some of them paid hundreds of thousands. We don't have a specific number for the privilege of doing that. Yeah. And some of the names might be known to many, Tim Drexler, for example, but a lot of entrepreneurial names, some obstetricians are in there, but we've got Tim Drexler, Linda Edwards, Alain Ferdinand, for example, among the six that are currently in that capsule. And there was a slight delay to it. We understand that there was sort of unauthorized personnel in the line of the rocket as to why it was pushed back that little bit on the day. But talk us through the risks, if any, of what has become standardized practice.

54:21Every launch that involves human payload, humans on board, is a risk. The hold on this case, the delay, if you like, was because of unauthorized personnel on range. But that is a very wide radius. You have to control both the airspace and on the ground subject to ITAR restrictions because it's a rocket at the end of the day. The view that you're looking at is just switch, right? So you can see those two white dots. If you squint and you look very closely, one is the capsule and one is the booster that's kind of rapidly coming back down to Earth. And what will happen very, very soon, because we're at the four minutes, 30 second mark, is those inside the capsule, the six of them, will get a one minute warning where they've been through this training process where they'll buckle themselves back in.

55:10And then it's good old fashioned rocket science where the flat bottom of the capsule and Earth's gravitational pull brings it back down into Earth's atmosphere. And we can kind of go from that. And we can continue. This is like a 10-minute process, and we're already, as you say, at the halfway point, Ed. We can go to Lauren Grush, who helps cover all things space across our network and platforms for us. And Lauren, this seems to be going all according to plan, as we think, but space very much in line of sight of investors at the moment, but also of the market more broadly. We're thinking about Elon Musk, who's just come off stage over at Davos, talking about how his reusable giant rocket might be there by the end of this year in terms of reusability.

55:48Yes, that is the ultimate goal of Starship, which they've been pursuing for some time. Full reusability, you know, actually the New Shepard is a fully reusable system, but it obviously does not go to orbit. And so Starship is really trying to accomplish a feat that no one has been able to before. And Elon just said that they hope to get to full reusability sometime this year. That would be a major accomplishment if they can make it happen. And we'll obviously be keeping an eye out on those Starship test flights, which are always so fun. But yeah, that will be a big moment if they can make it happen.

56:23We got Bloomberg's Lauren Grasch. She leads our coverage on space. And on the left-hand side of your screen, we've got the downward camera on the new Shepard booster as it returns down to Earth. We're about six minutes in, Lauren. Booster descending. And then the capsule, any second we'll start descending. By the way, peaking at like 4 to 5G on the way around. Cara and I were talking about how we carry this right because it's an expensive endeavor for space tourism. But Blue Origin's business, as we see that booster coming down, is multifaceted. What's the big priority for Blue right now? Well, it actually comes at a really interesting time.

56:58They just announced the third upcoming flight for their New Glenn rocket. That's their much larger orbital rocket that they've launched twice now and successfully landed on a barge on that second flight. Now, New Glenn is not fully reusable. The second stage, the upper portion of the rocket does not come back, but they were able to achieve that partial reusability with the second flight. So they're really getting into full swing with their launch business. And so hopefully just as these new Shepard flights are becoming routine, New Glenn flights will become routine as well. And we have at this moment the booster landing.

57:33We'll see if they note it successfully. at seven minute, 10 second mark. It's going to make for a beautiful picture. Ed, just talk us through the complex nature of a booster landing. Well, booster falls three, four through the air using air resistance physics. Then it ignites that B3 and smoothly touches down. A little needle in the haystack landing a pin in the middle of the West Texas desert kind of moment, Caro. Isn't it just? So there we are. Booster touchdown is hit at the moment We're T plus 73940, let's call it. Capsule reacquired comes at eight minutes, Lauren. And all of this is being, remind us, financed how?

58:15People are obviously paying their way to a certain extent. We don't know exactly who. But this is still VC-backed, very much the endeavor of Amazon founder Jeff Bezos. Right. I mean, as you mentioned before, passengers pay some undisclosed sum to fly on these flights. But yes, I mean, Blue Origin is by far a very big passion of Jeff Bezos, and he has been primarily the funder of the Endeavor for some time now. Obviously, the goal is to move away from that system, to have more and more customers, and to eventually become a profitable business. You know, either that's through satellite contracts, launch contracts, things like that.

58:55And then also Blue Origin just announced a new mega-constellation to be a competitor to Starlink. So they have a lot of different revenue streams coming in, hopefully soon. Lauren, we just saw the drogue shoots and then the main shoots deploy on the new Shepard capsule on its way down. Now it's chill. You're just floating. And by the time you boost a landing speed, like six miles per hour. So we will wait for that over the course of the next two minutes to ensure that the capsule lands safely. You just talked about the business of constellation-based satellite. We talked earlier in the conversation about Blue Origin's efforts, that news last night.

59:34We also are trying to keep on top of Elon Musk's appearance at Davos, which you and I were both tuned in for, right? He did talk about Starship and some of the broader kind of academic questions around space-based data center. As best you can, just give us a summary of that while we await the Blue Origin capsule touching down. Sure. So as Elon indicated, he's a very big fan of solar power. One of the justifications for moving data centers into space is, depending on where you put them, you can have this constant access to solar power, which, as people who know much about AI and the data center industry here on Earth, power constraints can actually be a huge limiting factor.

1:00:15So moving to space, the likes of Elon and other billionaires, even Jeff Bezos, have talked about perhaps tapping into that solar power to get near constant access to sunlight that can then power these data centers that do all this complex computing. Yeah, we heard Musk saying, really, in the next two to three years, space is going to be the cheapest place to put AI. And there, look, we have the landing. Touchdown. Chill, Ed, as you called it. It looks like quite a lot of dust being blown up on a chill entry. If you're going up at 3Gs and you're coming down at 4 to 5Gs, but the moment you touch the ground, it's just like a little boop.

1:00:54Then that's what I meant by chill. But again, we talked about this being routine. And sorry to interrupt, Caro, you know, we'll wait for confirmation from Blue. But as it stands, you know, another successful mission. It is seemingly successful. We'll check in on those six astronauts as they're known and what the Carmen line touching and the zero gravity feel was like for each individual. Lauren, we always come back on Bloomberg to the financing of this. And we await potentially 2026 an enormous IPO of SpaceX as well. That must be something you're tuned in for. Oh, absolutely. I think the entire space industry is kind of on pins and needles awaiting how this IPO will play out.

1:01:35And obviously, as we mentioned, one of the big things fueling the IPO is this concept of raising capital for space data centers. So I think it'll have a very sizable impact on the entire space industry, not just SpaceX. I've spoken to a number of experts who think that this will open the door for even more investment in the space industry, which is sometimes considered very niche. So I think even the competitors of SpaceX are probably pretty excited to see how this plays out. We're going to move away from West Texas. And when Blue Origin confirms all as well, we'll bring it to the audience and go back to Elon Musk at Davos.

1:02:13And let's listen to some of what he had to say on stage. probably sometime next year i'd say that by the by the end of next year i i think uh would be selling humanoid robots uh to the public um that's when we are confident that it's very higher liability very high safety um and the range of functionality is uh is also very high you can basically ask it to do anything you'd like well that was the uh timeline prediction that moved markets, Tesla shares rose when Musk said that. Musk often misses his own deadlines. In the space context, Lauren, that's your domain. Just real quick, what were the kind of timelines that he gave, if any, on that part of his business?

1:03:00Right. So he said the things that I took note of are the fact that he thinks that space will be the cheapest place for AI data centers to be in the next two to three years. Obviously, I think that might be a very ambitious timeline as well. Jeff Bezos has given a timeline of 10 years in order for it to be economically feasible. And I spoke to another expert who mentioned that makes a little bit more sense. And then on the Starship front, as we mentioned, he talked about getting to full reusability sometime this year. Again, that's going to be a very big feat if and when they pull it off. I can't speak to whether or not they will do it this year, but they've obviously been doing these periodic test flights of Starship in order to reach that goal.

1:03:44And so either way, it should be very exciting to watch as they incrementally get closer to that development. And Lauren, you'll be along the way for us, as always, guiding this show and across platform. Thank you so much, Bloomberg's Lauren Grush on All Things Space. Look, we were just talking about Elon Musk at Davos, what he said. Others have been there too, funnily enough. Darktrace CEO, Jill Popelka. And she says that AI has, quote, democratized cyber attacks And there needs to be a shift towards sophisticated defenses to protect businesses. She spoke with Bloomberg's Francine Lacroix on the sidelines of 2026 World Economic Forum.

1:04:16Take a listen. The threats are becoming more sophisticated. They're becoming more complex. And whereas before we might have thought about being attacker-centric and trying to predict what the attacker is going to do next, that's not what Darktrace has ever done. And it's definitely not what we can do moving forward because we can't predict this high-velocity, high-threat landscape. We know that countries like Japan, for example, they used to have a natural barrier because their language was more complex and people couldn't quite get their, you know, the cyber attackers couldn't get their heads around creating an attack in Japanese.

1:04:47But now AI can create those complex, highly sophisticated threats in any language around the world. And so, right, we're looking at a much higher velocity threat landscape. So is AI and cybersecurity much more sophisticated than even 12 months ago? Oh, of course. It's becoming more and more sophisticated. We have the most amazing minds in Cambridge. people who are linguistics majors, people who have majored in the classics, but now looking at how is this AI challenged for us? How do we protect the world against it? So after Darktrace's acquisition by Thomas Bravo and the shift to operating privately, how has that changed your focus and investment strategy?

1:05:22You know, Darktrace was already changing. And so the transition into private equity ownership really wasn't that big a deal for us. We were continuing on our path of innovating and creating the latest and greatest products. We were also working on scaling our business. So we had become really a$700 million startup. And then we needed to ensure that our systems and our processes and our talent acquisition and all the different things that big companies do, that we were doing those really well also. Are the cyber threats coming from government agents or government entities? Or what can you tell us about where the biggest threat is coming from?

1:05:56You know, what's interesting is AI has democratized cyber attacks, just like it's democratized so many other things. That's a good line. It doesn't really matter where it's coming from. From a Dark Trace perspective, we're going to protect you from a nation state actor, just like we're going to protect you from a Western teenager in his garage, you know, having a heyday with AI that day. So it's not so much about being attacker-centric. What we do want to think about, though, is how we get every anomaly and how we ensure that we're protecting from those. That was Dark Trace CEO Jill Popelko along with Bloomberg's Francine Lacroix.

1:06:28Now, coming up, a startup aiming to outperform today's AI accelerators has attracted$110 million in Series A funding. Neurofos CEO Patrick Bowen and M12 managing partner Michael Stewart join us next. This is Bloomberg Tech.

1:06:48AI chip startup Neurofos has closed a$110 million Series A funding round.

1:06:58optical elements on a single chip. They rely on light to transmit data. It claims this OPU delivers up to 100x the performance and energy efficiency of current leading GPUs and accelerator cards. Let's get to it. Patrick Burren, Eurofoss co-founder and CEO, and Michael Stewart, managing partner at M12, formerly known as Microsoft's Venture Fund. Patrick, the devil is in the detail here. The use of light to transmit data instead of electrons, and again, and I'm not an engineer, but what is this OPU design that you're developing? It's essentially designed to be a drop-in replacement for a GPU, but actually runs 50 to 100 times both faster and with 50 to 100 times higher raw energy efficiency.

1:07:39In the inference use case. In the inference use case, not for training, that's right. You're an engineer by trade, by background, but then you go to the dark side of venture capital. Why back this project? You're going to tell me that there's a gap, there's a problem being solved for here, there are lots of inference solutions out there, different technological underpinning, but why this one? Yeah, certainly. Our investment focuses on the data center of the future. And by and large, the industry has committed massive amounts of capital to scaling the technology that we know and understand. And from our investment thesis, our investment aims, we're really looking for what is far beyond the state of the art to disrupt that, particularly toward the end of the decade.

1:08:23And what Patrick and his team have developed is really, it's not just a new technique that's cool because it uses light, but it really delivers demonstrable benefits to the energy needed for the compute and in a way that we think could change the game. So that's something we can underwrite as the fund, and we think it could impact Microsoft but also the industry at large. Okay. I'm interested, Patrick, on the next set of data centers are meant to be being built in space, if you're listening to Elon Musk over at Davos today. I'm interested as to whether that is something that you think is applicable to your technology and how you think you're going to be beating quite a lot of competitors out there.

1:09:02There's Light Matter, IA Labs, Renovus, for example. Yeah, absolutely. I think the thing is, wherever you go, whether it's in space or it's on Earth, AI is fundamentally hardware-limited, and the hardware is fundamentally power-limited, right? And the move to space is an attempt to try to solve the power consumption part of the problem. But wherever you go, even in space, power is still going to be limited. And we're solving that problem at the fundamental physics level. Are you, Michael, seeing the adoption by, well, those out there at the moment building the data centers for such new, interesting technology?

1:09:40Because as you say, this could just be replacing dropped in instead of a GPU. How are you seeing that revenue process as you back it with this sort of Series A funding? Well, in the last couple of weeks, the last couple of months, actually, we've seen a complete and total breakthrough in the acceptance by the leaders in the industry, the chip industry, to use new hardware to address different parts of the inference workload. And that's the last couple of large licensing and hiring deals that we've seen. And I believe that's the beginning. That's kind of like breaking the sound barrier. You need those steps to happen on the part of, obviously, the leaders to pave the way for even more disruptive technology like what Neurofoss is developing.

1:10:27You know, this is an industry that is still dominated by the GPU, NVIDIA, right? But with the specialist inference platforms, you know, the argument that NVIDIA makes is that we give five-year visibility on our product roadmap. and that's why each generation of silicon that comes out, you just drop it in. How is this more than a lab experiment right now with Neurofos and you can convince that data center that's not yet planned for, let alone built, that that's the best viable option? Right. The key is that right now we already have silicon. We have verified silicon. It is real? It is real. We have taped out these chips.

1:11:04We've de-risked the fundamental physics. So from here, unlike quantum computing where you still need some more physics miracles to make it happen, There are no more physics miracles in our roadmap. From here, it is hard engineering, but it is just engineering to get to market. And so we can build a timeline and an engineering schedule that is tractable and actually can deliver on time. You know, I'm just trying to understand how difficult this is going to be, Michael. Cerebrass is doing interesting things, different technology. Grok, you know, when NVIDIA took Grok, they were like, well, they didn't have a place in the world.

1:11:35You seem to think that actually the underlying technology, OPU, does have a place in the world. What evidence would you provide for that? My background is from the semiconductor industry. I got to say, I wasn't sold until I visited the lab, and Patrick and their team opened Kimono, showed me the oscilloscopes, showed me the test beds. This is something that can not only work on its own, but it could be integrated with other conventional products that are in the server bill of materials. And again, that's the piece I can underwrite. We're not really here in this case to look to invest in science, although I am a scientist, I at least think we should stand behind the technologies that could change the game for the energy needed for inference compute.

1:12:18Why underwrite Neurofos versus, as I said, Light Matter, IR Labs? I don't want to comment on the other companies other than I've known them. And I looked at this. I looked at Optical Compute very seriously five years ago or so for my prior fund. In fact, I met Patrick at that time, and the concept for Neurofos was a little bit early. But I think the market conditions that demand the need for new approaches combined with, like I said, the signals of leaders to be willing to integrate and use this technology in their platforms that are already commercially successful, this is a moving train toward heterogeneous compute and true disruptive technologies making it to market.

1:13:03What Neurofos has that, again, helped us move forward is something I could see with my eyes that shows it could work. Now, I'm going to still put the pressure on Patrick to deliver the product on time, which is definitely the job of an entrepreneur. But to his point, it's not a science risk question as much anymore as engineering and capital. So once again, that's where we can bring our fund in. And boy, is there a need for an unlock on the energy side. Patrick Bowen, we thank you so much. CEO of Neurofoss and Michael Stewart, managing partner at M12. Great to have you both on this funding round.

1:13:40Intel's down half percentage point, but in reality, Cara, it's just treading water ahead of earnings after the bell. It's up 15 % in a short US week because of the holiday, and it's up like 46 % year to date. And whatever it tells us about AI, the reality is right now, it's bread and butter business CPUs is in demand, and that's really helping it. And the PC side of the business, many feeling that that's been a growth trajectory for them once again. And we had, of course, the discussion around Panther Lake and what they've been doing in terms of innovation. The key question is going to be foundry.

1:14:11Where is that actually going in terms of clear direction? Revenues are likely to fall for this fiscal quarter that they're reporting on, Ed, by 6%. It's only growth come the second half of 2026, right? Yeah, and probably some updates on Lit Boutan, the CEO, and what's his relationship like with the president, one of his shareholders now the United States of America. All right. And these are all good topics of conversation, but we have had a hell of a run up before the earnings actually hit. Yeah. So maybe a bit of digestion just before they come out. But digest more of this edition of Bloomberg Tech.

1:14:43It's over for now, but Ed, the podcast. Yeah. Check out the pod. Astonishing episode, to be frank. I mean, lots of recap. You know where to find it. iHeart, Spotify, and on Apple and on all the Bloomberg platforms. This is Bloomberg Tech. Oh.