Anthropic Builds Its War Chest Ahead of IPO

0:00AI is entering its most consequential phase where scale, safety and sovereignty will determine who leads and who lags, join Bloomberg Tech in London on November 2nd and 3rd as global leaders across business, finance and policy examined the defining trade-offs shaping the future of AI. Thank you to our presenting sponsor, Sarah's Force and Supporting Sponsors, IDA Ireland and Schneider Electric. Learn more at bloomerglive.com slash techlondon. Bloomberg Audio Studios Podcasts, Radio, News Bloomberg Tech is live from the heart of Silicon Valley with Ed Luglow in San Francisco. Coming up, Anthropics set to finalize an expansion of its revolving credit facility to $15 billion before its public filing will have those details.

1:02Plus, a strong US job support, but not so much for tech with the IT sector shutting 23,000 jobs in August. We'll dig into those numbers and Apple kicking off the John Ternis era with what could be the biggest run of the new devices in its history. This is Bloomberg Tech, I'm Lisa Mateo in for Ed Luglow. Let's get a check in the markets. Look at the five-day NASAC and pretty flat on the day. It's been a quiet trading day. We're heading into a holiday weekend, so pretty sleepy right now. We'll get to today's big number and that is $15 billion. That is the size of the revolving credit facility and the topic is set to finalize. That's according to Bloomberg sources, all ahead of the firms public filing for its highly anticipated IPO. So let's get more with Bloomberg's belly lips, you'll have to broke this story. Bally, so how does this health kind of clear the past for the firms IPO? What's the significance of this? The reason it matters, Lisa, is this sets the table for these other banks to kind of come in under the tent, if you will, and see what anthropics numbers look like.

2:03The expectation is this deal did close this morning, so that would mean the likes of a Barclays Wells Fargo, some of the other banks outside of the Big Three will now be a part of this process. That means that going forward now, this could tee up an analyst day, this could set the stage for kind of public flip early next week or sometime next week, next week that would kickstart that 15 day waiting period for the formal marketing to start. Now, you mentioned some few banks, so generally syndicate loans the higher the commitment of a bank, it's the higher the fees it gets paid by the borrower, so that's why this is important. Well, it's important too, because normally your position on the revolver looks pretty darn similar to your position on the IPO. So if you're more in Stanley who led this process, the expectation is you would then turn around and lead the IPO process. It's seen as a big win for the likes of a Barclays or a Wells Fargo, because they're on that top line of the revolver, what that would then indicate is they should be higher on the IPO, which then to that point means better fees from that sale, from the first time share sale, and the expectation that this is a company that's going to sell stock and debt in the

3:03future. Okay, I got to ask you because you cover SpaceX IPO very closely. Is Anthropica, they look into raise as much as SpaceX or even more? At least at what SpaceX, SpaceX or higher has been kind of the verbiage coming from the banks. That would mean pre-green shoe, $75 billion, closer to $82 billion from the IPO, but when we talk to some folks on the street, that number could get a lot higher pretty quickly, just given the excitement and the view that Anthropica is seen as a true must-own company whereas SpaceX had its tits and tats. It did. Okay, any update on the Anthropica IPO, you're following it. Where are we going to see this? I think the expectation from talking to sources, they still need to schedule that analyst day. You do need to then have the banks working on this, have their analysts be able to formulate their models, which would then mean public flips sometime next week. So we're realistically looking at Anthropica pricing its IPO at the earliest, the last day or two of September, but most likely sometime in early October. Okay, you're going to be keep on on top of this. I'll be on top of it. You'll be a kid with us. Stuck at my desk, Lisa, checking on every little update.

4:04All right, I'll send you some food. Thank you, Billy. Very much. It's Billy Lipschult. Thanks for joining us. All right, so with more on the Anthropica and the IPO landscape, Raymaker Securities co-founder managing director Greg Martin, he joins us now. Greg, I have to ask you, so set the seat before we dig into Anthropica. Set the seat. What does the IPO landscape look like right now? Well, it's a really interesting landscape because it was obviously completely dominated by SpaceX back in June. And we've obviously followed the SpaceX story very well. They raised a historic amount of capital at 85 billion. And they traded up and traded down and they traded back up above their IPO price. And I think the market was really paying close attention to them and they really suck the oxygen out of the room for the most part for the summer. And now we have an extremely exciting fall coming, starting with Anthropica, the most meteoric company we've ever seen, crossing 100 billion of ARR and 10X in one year and

5:06potentially being valued more than SpaceX and raising more capital than SpaceX. So although I expected it will be somewhat similar in the market and then there's a really interesting pipeline after Anthropica. But again, it's another situation of, do they suck the oxygen out of the room? Is there anything left for these other also great companies but not at the same kind of scale as Anthropica? So it's going to be a very interesting and potentially exciting fall season for the IPO market. So you mentioned sucking the ARR out of the room. Do you think private companies, companies overall, are they waiting longer to get to that IPO? I think right, I think the conventional wisdom was we want to see SpaceX and we want to see Anthropica because we're not going to get any attention until those guys have gone public. And SpaceX is gone and Anthropica, we expected Bayley said to go end of September early October. And then I think after that, if things go well and the market's still strong and there's everything else is stable, I do think we're going to see potential opening of the floodgates.

6:10But the SpaceX and Anthropica IPO is need to happen first. And I do think there's a very solid pipeline behind them, but let's see how the Anthropica IPO goes. Now let's speak into the Anthropica IPO. Does that $15 billion number? Did that surprise you at all? No, not at all. Really it's just a giant corporate credit card. So it's not necessarily $15 billion of death that they intend to draw immediately. It's just what a company, especially a company with large capital needs like Anthropica would wisely do right before an IPO. It removes the IPO timing risk. So none of the potential investors would have leverage over them because they don't need the capital. They can always draw on the $15 billion facility. The IPO might take a little time to close. So there's no timing risk. The debt of Boyd's delusion obviously was the best time to negotiate this debt. As Bayley said, everyone's fighting for fees and positions in the IPO. So I'm sure they got incredibly favorable terms.

7:12It was obviously led by their underwriters. And it really just shrank into the IPO story. It gives them a large amount of cash forward to handle some of their large commitments. So there's going to be no concerns about their ability, at least in the near term, to meet some of their compute contracts that they've signed. So I think it makes a ton of sense and I think they did it perfectly. And I expected it will really help their IPO. All right. You kind of mentioned this a little bit. Can you dig more into the challenges that Anthropic might face coming out? I mean, it's a young company. It's a very fast changing market. And one day they can have the best model and the next day they could be leapfrog by open AI, by Brock, by Google, by potentially SpaceX. So there's a lot of open source. In video just about hugging face. There's obviously Chinese open source models. So there's always a question of who's model is better. Anthropics had a nice little run with their coding agent and their penetration in the enterprise

8:14and obviously their growth of AR. But if other models are better, other models are cheaper, or people aren't seeing the returns, there could be a slowdown of growth. So there's always risk in a company like Anthropic. I'm sure we're going to see they will put out probably a number similar to SpaceX. Like they're truly addressable market is something like 28 trillion. So the market's very big. And I do think that IPO will be received very well because the upside here is incredibly large. But this is a fast changing market. And there's a lot of risk inherently in a fast changing market. All right. We have less than one minute left before you go. You mentioned opening the floodgates. Who are you excited for? Give us a little sneak peek into that listing calendar coming up. Yeah. I mean, we already saw Aurora, which makes the Aurora ring file publicly this week. So I expect they're going to come around in a very similar time frame that Anthropic does, maybe a little bit after. There's some big neo clouds like N scale that have filed confidentially that actually Anthropic

9:18just did a 45 billion dollar announced contract with N scale. So they're the GPU data center company. We expect to see open AI probably shortly, some watching Anthropic very closely and probably end of year maybe early next year. And then there's companies that have filed us ones confidentially that have been hanging on the sidelines like Discord and Kraken. So there's a lot of potential companies coming. But I think everyone's going to watch Anthropic very closely. No kidding. All right. Greg Marken. He's Raymaker Security co-founder managing director. Thank you very much. All right. Coming up, well, we're going to break down the latest jobs figures and how the information sector is being affected with Alicia, but Anisto from Northeastern University, this is Bloomberg Tech. The Bloomberg this weekend podcast news, politics and the lighter side of Bloomberg.

10:21Human skin is the weirdest new ingredient in the K beauty boom. Oh, yes it is. So it's a skin blusher treatment drive from donated human tissue, popular and soul's beauty clinics. One doctor says it's made from dead people. It would be weirder from live people. Yeah, yeah. Just want to posit. The Bloomberg this weekend podcast, subscribe today on Apple, Spotify or wherever you listen. Well, US job growth surge in August and the unemployment rate at health steady at 4.1%, but dig into the numbers and there's a more complicated story unfolding in tech related sectors. Bloomberg's Mike McKinney joins us to break that all down. Mike, so the tech sector, not so well after all these big numbers that we just saw. Not so well, but not so bad either. Okay. I mean, if you look at the total report, the number of jobs created was really strong. It's just where the jobs were created. Now, leisure and hospitality led the way with 62,000 jobs created and government because

11:24local education went up and it is time for everybody to go back to school. The leisure and hospitality, I wonder if you count SpaceX as a tech company, maybe everybody went out to restaurants to celebrate the IPO because they got a lot of money. But if you want to look at the tech sector itself, you look at that chart there, construction jobs, 22,000, a lot of those, more than half basically, were in the specialty contractors area. And those are the people who are putting together the AI data centers. So nationwide tech, pretty good for construction workers all over the country. And then you look at computers and electronic products, 1300 jobs created there, semiconductor manufacturing, 800 jobs. That's not a lot, but these are turnarounds from what we saw last month in terms of the Silicon Valley area. But computing infrastructure. And I assume these basically are sort of the servers, et cetera, loss of 7,700 jobs and

12:27web search and information services down 400. So about half of the jobs in tech were losses were basically because of the AI situation, you could call it the Silicon Valley employment report as opposed to the nationwide construction of data centers report. So that was a lot of numbers that you just broke down. How does that compare to some of the data we saw earlier this week as far as tech and the jobs? Is it pretty similar? It's pretty similar. You don't get as big a breakdown except in the ADP numbers and those were similar. But what we've seen is manufacturing jobs still strong, but the number is going down in service industry jobs. They're still contracting. So all of this was kind of an odd jobs report. Good news if you want to talk about the number of jobs created in unemployment not changing, as you mentioned. But for the tech sector, one of the things we got this week was a huge increase in imports

13:28of electronic goods and computer goods. The data centers that are going in are buying a lot of stuff from overseas, so that doesn't translate into jobs here. Michael McKee, I always appreciate your time here. Thank you for joining us. All right, so what we want to do is get a broader look at the impact of job losses in the tech sector and the information employment. Alicia Madenise, so she is a professor at Northeastern universities. She's focused on labor market dynamics, including AI and the future of work. She joins us now. Alicia, thank you for joining us here. You heard what Mike McKee was saying. He broke down the numbers. What was your take from that report? Yeah. I was really focusing like everybody was on the tech sector in particular, looking for some clues about how AI would be affecting the broader labor market. I like to think of this as the canary and the coal mine, if you will, because they're such really adopters. And certainly seeing information employment decline by 22,000 in August, following the losses that had occurred over the prior year really is definitely eye-opening.

14:30So those job losses were occurring, as he said, in computing infrastructure, data process and web hosting related services. And this is where we would expect to see the swiftest and largest impacts on the labor market when we're thinking about AI. Is there a bigger trend that you're seeing, Alicia, from this report? Yeah, some of the other things I was worried about were the number of long-term unemployed in particular. So those are people who have been jobless for 27 weeks or more. That really didn't change much in August, but they account for about 27% of the unemployed now. So I do worry on the demand side that firms are continuing to be more cautious in their hiring due to macroeconomic headwinds, like inflation, interest rates, general uncertainty related to other disruptions to supply chains, like transport routes, oil and diesel prices, tariffs. They're also looking to AI tools to make their existing workforce more efficient, which slows down new hiring, even if we don't see huge layoffs. On the supply side, that makes me worry a little bit that people who have lost their jobs

15:31might be no longer qualified for the jobs they had due to changes in skill requirements related to AI or other trends. So I worry that maybe that first wrong on the job ladder in particular for entry level workers who are just entering the labor force is eroding, particularly in some industries where we see AI adoption has been more rapid. Yeah, Alicia, I'm glad you went there because I know we were speaking of the break before this. I have kid in college. I have another kid who's finished college. You have a child in college. So what are they expecting? We're getting into this job market, frustrated, trying to look for work. What kind of skills, let's start here. What kind of skills are employers looking for now from these kids coming out of college? Yeah, well, we actually did a study looking in particular at what's happening to the entry level labor market for junior software engineers, because that's where, at least for Northeastern, we're seeing our graduates having the most difficulty in the labor market. And when we looked at the online job postings, we found that right around the time when

16:32chat GPT was publicly released, that a lot of firms changed their entry level hiring standards in terms of both quantity and quality. So we find that there's about a 15% relative decline in junior versus senior software developer vacancies. So those junior vacancies are really the ones that the college grads are going for, and that it was larger in related technical occupations and not really seen in other kinds of occupations like engineering. So it's not just a STEM thing, it's really an AI thing. So we're also finding that these employers are asking for more years of experience in these junior roles, and they're asking for higher levels of skills, things like problem solving, communication, attention to detail, lots of judgment, things that are really raising the bar for what counts as a qualified junior hire. Now on the flip side, companies are dealing with resumes that are maybe written by AI. So on the flip side, how is the interview process? How is that changing? Yeah, so we're definitely seeing that there's a difficulty in terms of the screening process. When you get folks being able to write job resumes that exactly correspond to what people

17:38are looking for in their job ad because it's so easy to rewrite it with AI, they're just getting peppered with hundreds and hundreds of resumes. And so that's why you do see firms relying more on word of mouth, referrals, networking, hardly enough, a lot of the in person kinds of ways that folks are used to in terms of getting jobs. And in particular, they're looking for something that's more human centric, right? So do you have the judgment and experience to use AI in a way that's both efficient to the context and the job that you have, but is also equitable and ethical, right? Do you have the judgment to know when you should and should not be using it? Because one of the things we're seeing also is that people are using it to write job reviews right for their subordinates and things like that. That's probably not the best use of AI. Most definitely. Alicia, always appreciate your time. It's been a great pleasure and good luck to your kid over in college and into the job world. There's a tough one out there. That was Alicia, Madassino at Northeastern University.

18:46It is now time for talking tech. First up, AI infrastructure provider, Crusoe. It is signed a $13 billion cloud contract with trading firm Jane Street. Sources say Crusoe will provide AI chips for training and inference. The deal will boost interest in a fundraising round that brought in $3 billion. That's according to a source. Plus Moonshot AI, considering raising up to $5 billion in its plan, Hong Kong IPO, is soon as this year. That's according to sources who say the maker of Kimi K3 has filed confidentially for soliciting an added bank of America as the overall coordinator for the transaction. And Thailand has suspended construction of 49 data centers amid growing concern about their impact on communities. A government official said projects were hothed to allow time to develop new industry regulations expected to be announced in a month. While sticking with data centers, deep seek, one of China's leading AI firms is planning to deploy at least 160,000 of Huawei's top accelerators to power this massive data center that's being built in Inner Mongolia.

19:47That's what sources are saying. Well, deep seek has so far relied on Nvidia accelerators to train its model. So this move could be critical in advancing China's efforts to replace Nvidia. Bloomberg's Peter Elstrom, he joins us now with more. Peter, I want to start with you because Huawei chips, the question is, can they provide enough quality? Is there enough product out there for these Huawei chips? Yeah, this is an exclusive scoop by our reporting team in the greater China area. We're talking about deep seek building this massive data center in Inner Mongolia. There are one of a couple of companies that are trying to do this. We've also seen bite dance make these kinds of investments. And it's interesting for a couple of different reasons. You mentioned one at the top of the show. These chips, 160,000 at least are coming from Huawei. Huawei has been in the accelerator game trying to compete with Nvidia for a long time. But it's a sign of progress on Huawei's front. These are the 950 DT chips from Huawei, they're a new generation that came out just recently.

20:50It's an advance from the 910 chips that they've had in the past. And it shows some progress on that front to allow local companies to be able to get these chips and then deploy them in data centers. The second reason it's interesting is because compute is one of the areas where China has really liked far behind the United States. When you look at open AI and Thropic, Google of course too. They've been investing hundreds of billions of dollars into data centers. They've jumped far ahead in terms of the amount of compute that's out there. And now you're seeing these Chinese companies begin to make bigger investments. You mentioned moonshot at the top of the show. As moonshot raises this money in its IPO, certainly we're going to see them get into the compute game too. So they're beginning to catch up on a couple of different fronts and tapping into Huawei to use these domestic chips. Yeah, can you take a little bit more into how it's going to advance China's AI ambitions? Well, as we've talked about on the show a few times in the past, the US has imposed quite strict restrictions on Chinese companies' ability to buy the most advanced chips from Nvidia.

21:54And so they have been kneecapped in terms of their ability to use those chips for training the advanced AI models. Nevertheless, you see companies like DeepSeek and moonshot in particular make very significant advances, use their engineering expertise to be able to produce models that are perhaps not quite as good as the frontier models from and Thropic and open AI, but very, very close at much lower costs. So this sign that DeepSeek is turning to Huawei for these chips shows that they are getting more access to domestic chips. They are making progress in terms of producing those domestic chips. They still want to buy Nvidia chips. Those are still the best ones on the market. But to the extent that they can use Huawei to handle some of the inference and some of the training, that's going to be a big advance and a big positive impact for the Chinese companies. All right, quick 30 seconds left. But private sector, that's still using the Nvidia chips, correct? They prefer that. In the private sector, well, like all the bomb attendees.

22:56Yeah, the Chinese companies would like to be able to buy Nvidia chips if they can't. They'd like to be able to get the most advanced ones on the market. They can't buy black wells right now. The US has allowed them to buy these H200 chips, as we've talked about before on this show. But Beijing has a very vested interest in trying to get them to use domestic chips like the ones from Huawei. All right, Peter, thank you so much. That's Bloomberg's Peter Lstrom. Welcome back to Bloomberg Tech. I'm Lisa Mateo and for Ed Luddlow. I want to get to some stocks on the move that are in the tech sector. I want to start with Tesla off the bat. Their shares are down about 6%. I mean, they had that highly anticipated cyber cab launch. It didn't really impress investors much with that. Then we go to Nvidia. Their share is up about 1.5%. Still running that high, the news of acquiring hugging phase for $13 billion. They're co-founders. Now billionaires on top of that. And Adobe shares on the lower side.

23:57They're down about 6% after a leadership change up, which leads us to our next story. That is Adobe. It is named Aneel Shockavarthi. It's next CEO succeeding. Shantanu Narajan. Now the surprise pick, it comes as Adobe. It faces growing pressure from AI up. Starts, we have Bloomberg's Brody Ford. He broke the story and he joins us now. So Brody, he's entering this challenging competition, right? From a lot of AI startups. Is this selection a surprise for him to begin with? Right, so Adobe has been having this CEO bake off in the background for the last six months. It's been an unusually public process because the company announced the CEO was stepping down, which usually you don't do that until you have replacement. But in this case, there was two divisional leaders who are seen as the internal candidates. There was one that was very much the obvious candidate, right? He led the creative business, like Photoshop and all that. And this other guy Aneel, he ran more of their marketing and data software that you don't

25:00necessarily think of when you think of Adobe. And so a lot of people assumed, oh David, he's got this in the bag. Turns out he didn't. Turns out he left the company yesterday when the other candidate ended up being chosen. So it's clearly a bit of a shocker for certainly the Wall Street types. So what does he bring to the table? So Aneel, what I have generally heard about him is that he's very smart and he understands the product very well. And clearly Adobe hadn't been doing super well, right? I mean, their stock was down 50, 55%. And so the argument for Aneel is that you both get somebody who can shift up the strategy a little bit because he hadn't been directly running creative, but also he knows the company. And so you don't have that whole disruption risk when you bring it outside her. Well, we have Adobe, right? Their software has been kind of the standard, right, for being creative, but then enter AI. And Wall Street seems to be concerned about it. So should they be, and how is this new CEO?

26:00How is he going to take on this new challenge? The whole risk for Adobe is that you don't need their products anymore, right? I mean, there were many years where my friends and I would Photoshop stupid pictures of each other. But now you can just make that with AI, right? I mean, that's like a silly use case, but I think some version of that is happening around the economy. And Adobe is worried that they're just not the shiniest maker of creative tools anymore. All right. And any other leadership changes before you go that we should know of for Adobe? After David leaves, I mean, a lot of folks expect that there might be a bigger house clearing, right? I mean, there's a bit of a changing at the guard here. And so I wouldn't be shocked to see some more divisional leaders depart in the coming year or so. All right. Rody Ford always appreciates time. Like a friend. Always a pleasure. Always makes me laugh too. As we enter September, a month that has historically delivered the calendar's largest single drops. Our next guest believes that seasonal weakness is not a reason to step away from stocks, arguing that we should look through the false volatility rather than react to it.

27:02A Marriott Price financial chief market strategist, Anthony Saglin-Benny, he joins us now. So let's talk, let's start there. Let's talk about September. It's usually the slow time. You're saying, maybe not so much. Yeah, I mean, over the last 20 years, the S&P 500 has fallen on average by about a half percent in the month of September. It's actually the worst month for the market in any given month over the course of the year. But what drives a lot of that return or that lower return in September are a few key outlier years, right? 2022, 2011, 2008. The events or macro shocks or recessions, they generally really weigh on that September figure. When the economy is growing like we're doing right now, when earnings are strong, like they are right now, the typical September sees either flat or returns that are positive. And actually, the fourth quarter is the best quarter for the S&P 500.

28:06So on average, you see the S&P 500 over the last 20 years grow by 3.7% in the fourth quarter. So if you can look through some of this in your term volatility, which we're likely to see a little bit more of, when you get to the back part of the year, actually returns for the S&P 500 or some of the strongest of the year. Excellent. I want to talk tech earnings because it was a big week for it. I mean, you could jump to some of the names. We've got Dell and Vidya, HPE, Snowflake, Broadcom. You name it, you can add to the list if you want to as well. But were you happy with the result? Some were mixed. Yeah, overall, very strong quarter of tech earnings. So S&P 500 earnings for the information technology sector grew by about 75% in Q2. They're expected to grow by over 60% in Q3. They are expected to be the leader across the 11 S&P 500 sectors in the fourth quarter. And for all of 2027. And as you mentioned, as we've gotten tech earnings over the last several weeks, they've

29:10really validated that A, the AI theme is secular and growing. The concerns around the cap expending for the hyperscalers. And is it really creating revenue? We got an answer to that for most of the hyperscalers. They had very strong cloud compute, mag seven earnings in Q2 grew by 118% year over year. Obviously in videos, earnings, a little while back, really, I think knocked the cover off the ball and showed that all of that spending from hyperscalers is showing up in revenue. And semi-conductors, the announcement of buying hugging face kind of shows that they're taking their revenue and they're taking their gains and they're expanding their reach across the ecosystem. So I think overall, the fundamentals across technology and profits are very strong and we see them continuing to be strong through the rest of this year. Yeah, so what kind of questions are your clients asking about the AI trade in general?

30:12What are they concerned about? What are they excited about? Yeah, a lot of it has to do with cap expending. I mean, obviously with the hyperscaler spending, $800, $900 billion this year on data centers and AI infrastructure on pace to spend possibly a trillion dollars next year. The real question is, is all of this spending going to show up in revenue? And I think that's going to be an ongoing question that investors are going to want to see each quarter that these companies are reporting. And so far, the revenue is showing up. I think the bigger question is, are we going to get the kind of revenue at the scale that justifies this cap expending? And I just think the verdict is still out. But obviously we need compute, we need inference, the data centers need to expand so we can get the type of productivity and growth out of the large language models you were talking about.

31:12Anthropic earlier, and it's planned IPO for them to continue to grow and for the AI to continue to be strong and actually get outside of technology and other business to produce efficiencies and productivity, we need to see more spend on the data center side to get the inference and compute to make these models more powerful. Anthony, you mentioned something earlier about hugging face in video that acquisition. Are you worried about the concentration of power in AI? Yeah, I mean, there's obviously a lot of AI circular financing right now. I think that's another big question that investors have is all the complexity around the financing deals. For Nvidia specifically, they want to expand their reach outside of just providing chips, right? And they don't want to probably put their chips on two large language models like open AI and Anthropic. So the open model architecture kind of helps protect them and build the infrastructure

32:13in large language models larger than just what two companies can do. I think in terms of the circular financing, I think investors are just going to have to pay very close attention to how these deals are arranged. I'm here in the Detroit area. Auto financing is something that's very common where OEMs provide financing to dealers and customers so they can buy their products. I think companies like Nvidia are trying to do that, right? Create the financing, create the structures so it's easier to build those data centers in that AI infrastructure. The question is, I think investors are going to have to continue to monitor these to make sure that the agreements actually show up in revenue, show up in the data centers so you can build this AI architecture in a way that maybe 12 months, 18 months, 24 months from now makes it more powerful. All right, Anthony, we're going to have to leave it there. Thank you so much. That was a Merit Prize Financial Chief Market Strategist, Anthony Saglam-Benny.

33:15All right, coming up, open AI. They take another step toward HCI with its most advanced model yet. We hear from CEO Sam Altman on the growing risk that come with increasingly powerful AI. This is Bloomberg Tech. As AI becomes more powerful, who decides how it can be, how it can and can be used?

34:17Well, open AI CEO Sam Altman, he says there needs to be clear red lines but also more freedom for users. He also admits the industry has fallen short in explaining how AI can empower people. Bloomberg's Ed Luddlow sat down with Sam Altman yesterday. We think there are some things that very clearly model should not be able to do. There are some clear red lines and we talk a lot about being able to detect those, but people also have strong opinions about things within the broad bounds of what is possible about how they want to be able to use AI. This is going to be a difficult question for society because there is not going to be agreement on what is and isn't acceptable use of AI. I think it's very important in the same way that people do things with electricity that I may not always like, but I think everybody has a right to electricity that we say they're going to be some rules for society. You can't go like electrocute somebody else on the street, but this is an important platform

35:18that belongs to all of us and that we need to be able to use. There will be an ongoing conversation in society. There will of course be some things that we agree on that people just should not do with AI, but there's a lot of edges that we're now going to have to feel our way through and giving people autonomy and freedom to use AI is important to our mission. At the G20 this week, Treasury Secretary Scott Besson said that the AI industry had done a terrible job of explaining how AI benefits ordinary Americans. What's your view of that and whether you would on reflection consider open AI to be in that bucket? I think the industry has done a terrible job of this unhaul. I think we have done a bad job ourselves, maybe better than some others worse than some others. It is difficult, of course, to explain that this is, the world should, in my opinion, should want us to worry a lot about this technology. The world should want us to think about all the ways that could go wrong, to talk about

36:22what we see coming, but not to sort of like stir up doom or concentrate power for ourselves just because in the history of technology the way you solve the hard problems is to look at them, talk clearly about them, have scientists debate them, work really hard to prevent and mitigate them. That has been our intention, but clearly sometimes it ends up scaring people. There's another thing that I think has gone really bad besides talking about the ways this could go wrong, which is I don't think we've done a great job as an industry of really talking about how to empower people with this technology. You hear some people talk about, well, we're going to give you a cure for cancer and that's so great. You know, take these risks and you're going to have to have less power in autonomy. We want people to have more power in autonomy. You know, I don't think a cure for cancer is enough. I mean, wonderful thing to have, obviously, but we should set our sights higher. We want these tools to enable a boom in creativity and entrepreneurship and people figuring out

37:22new ways to be useful to each other and wonderful products and services to enable each other with. And I think that has been lost in the messaging, and I also think there's some companies that have a different opinion there, but it's very, very important to us that this technology is something when people get more power in more autonomy, not less. And I think people rightly get afraid when they feel like AI leaders might not quite want that for them. All right, that was OpenAI CEO Sam Altman here along with, and Ludlow. So the push to get more people more power with AI comes as OpenAI. It rolls out its most advanced model yet. GPT-6 Astra, it's being positioned as a major step toward AGI, but its powerful cyber security capabilities are also forcing OpenAI to put new guardrails in place. Bloomberg Seth Figgerman, he joins us now with this. So Seth, let's start there with the guardrails. We just heard what Sam Altman was saying about security. So what are some of these new guardrails that they're putting in? Yeah, OpenAI is kind of following a playbook that, again, Thropic has also done earlier

38:24this year, which is finding a way to introduce a new, more powerful model to the wider public. While trying to take steps to safeguard or restrict the most advanced cyber security capabilities, that have been raising concerns. And recently, so that's what they're promising here that they're basically going to release this model in the coming days or weeks to the general public, but prevent people from maybe taking advantage or misusing these capabilities. I think stepping back to Altman's remarks, so they're trying to strike this very difficult balance between pushing out more advanced models that kind of continue to keep them at the frontier of AI development to keep them competitive with Anthropic and Chinese firms while also dealing with the very real concerns that are emerging from this development. So who gets it first? It's going to have to a limited group of partners that work with OpenAI through its daybreak cyber security program. That program, I believe, has a far more partners than Anthropics does, but these are kind of business organizations as opposed to just the layman. And you've mentioned Anthropic before. So how does this put this in competition with Anthropic? This new model?

39:24Yeah, so I think OpenAI is taking pains right now to say that this is as good or better than the most competitive offerings from Anthropic. And in fact, OpenAI's top executives Greg Rockman, their president, went so far to declare this kind of the beginning of the AGI era. You know, AGI is sort of this very squishy term, but also a milestone that a lot of the industry has been focused on where AI can match or exceed the capabilities of humans in many economically viable tasks. So to even begin to kind of whisper or say out loud that this might be that moment for OpenAI, it's something that we should really take stock of. But I think it's also optically a moment for them to try to say, you know, we've been on the back for a lot this year. All of them has admitted that at times. We're trying to show again that we are at the top in AI development. Where do you see AGI headed? You know, again, it's very controversial in some hornets to even say, is AGI, I think that can happen? Has it happened already? Some people think it'll never happen. It's a little bit of a thing that happened years ago. But I think something back from that, I think it speaks to the feeling right now that over the past

40:25year we've seen rapid advances in certain corners of AI. AI has gotten meaningfully better at coding, at cybersecurity, at some financial services tasks. It's still weak in other areas. It's kind of lumpy. But it's gotten good enough that I think all men and others would argue that if we took these models and showed it to someone five or ten years ago, they say, yeah, that's AGI. But there's certainly a lot of experts who say, no, it's not there yet. All right, Seth, appreciate your time. Thank you for joining us here. Of course, I have Bloomberg's Seth Figurman. So, in videos, they're $13 billion deal for a hugging face. It is now minting three new billionaires. The AI platform, its French co-founders, Glemmelang, Julian Trouman, and Thomas Wolf, they are each now worth about $1.8 billion. That's nice. According to the Bloomberg Billionaires Index, now the deal is expected to close in the first half of next year with Nvidia pledging to keep hugging face on an open platform.

41:26Well, Apple is kicking off John Turner's era with what could be the biggest run of his new devices in its history. Next week's product launch is expected to introduce foldables, new iPhones, and an updated Apple Watch. Bloomberg, Sam Kelly joins us with more. Sam, thank you for joining us. This is going to be his first event since taking over the new helm. What's the vibe going to be at this event under this new leadership? What can we expect? Yes, good morning. Yeah, big moment for Apple. Obviously, it's going to be an iPhone event, but a very big moment for John Turner's as well. Obviously, he's been very involved with the products that we will see next week. And obviously, these products are very much under the leadership of Tim Cook, but at the same time, we are going to see what his leadership style is. What his communication style is like, will he be involved in the actual presentation of some of the hardware, or will he just introduce the keynote in the beginning? So, will we be interesting to see how we're ushering into this new era and how involved he will be in the presentation? All right, so let's get to some of these products.

42:27The foldable phone. That's entering Sam's own territory. Can they go against the competition here? Yeah, I mean, I think the big question is why has Apple waited so long when Apple has been dominating this for quite a while? So, the pressure will be really on to see not just that it's built this foldable device that everyone's waiting for, but how does it perform? Is there a crease? What's thatery life? What is the software like to support it? How will usability be like? So, all eyes will be on the foldable device. We're hearing it's about five and a half inches. I will open much like a book into almost a size of an iPad mini. Pretty similar to some of the other devices we've seen on the market, of course, with the Apple touch. We're also expecting a new iPhone 18 Pro lineup, two models there. The entry level 18 we're expecting later next year actually. But it's not just going to be about iPhones. We're expecting we have some additions to the smart home world with a home hub that would act almost like a home security system or Face ID or Face Timing in the kitchen to communicate with others.

43:33But AI is really going to have a big play across all these different devices, whether it's the watch or a new Apple TV. We're expecting that as well. And all these different things of how AI will play into the next era of Apple. A lot going on. Thirty seconds left. Now, what is Apple going to have to do to prove that they can compete with the open AI with others on AI? Yeah, that's a great question. We already saw a little bit of what to expect at WWDC back in June. Now, it's really about the execution. What will it be like to have Siri on these new devices? How will AI play into that set? Is it fast? Can Siri be more than just a question and answer system? Can it really learn what's on your device? The ecosystem, your calendar? Can it really get to know you? That's what we're going to be watching. All right, that's Bloomberg, Samantha Kelly. Thank you. We're looking forward to that event next week. Well, that sounds it for this edition of Bloomberg Tech. Don't forget to check out our podcast. You can find it anywhere. It's on the terminal.

44:33It's online at Apple. Spotify at iHeart as well. Thank you for joining us. I'm Lisa Mateo. This is Bloomberg. I'm Matt Miller. And I'm Hannah Elliott inviting you to join us for the Bloomberg hot pursuit podcast. Every week we bring you news and industry insight on everything cars. And we do a whole lot more than just talk about cars, Matt. We actually get behind the wheel of basically every latest model, especially the luxury ones and the sports cars direct from the showroom floor. It really is remarkable how many cars we have access to. I feel a little bit guilty about it, but everything from $40,000 EVs to exotic half-million-dollar supercars. We also speak with the insiders who shape the automotive industry from the top CEOs and collectors to visionary designers and racing champions. Search for Bloomberg hot pursuit on YouTube, Apple, Spotify, or wherever you get your podcasts. Maybe you listen while you're on your week and drive, maybe go into cars and coffee, listen to us talk about what we are driving this week.

45:36That's Bloomberg hot pursuit. I'm Matt Miller in New York. And I'm Hannah Elliott in Los Angeles. Subscribe today wherever you get your podcasts.