The Anthropic Rocketship, AI’s Spending Limits, SpaceX IPO

27:08Yes, we're recording Friday edition on Thursday today, but I think it still works. Anyway, the thing that caused Microsoft to drop, people are saying, is that Azure revenue didn't grow as fast as they expected. There was some expectation that Azure revenue would grow something like 40%. It's only growing 37 or 38%. Both Microsoft and Meta's capital expenditures are going to go up. Meta's went from like a predicted like 110 billion on the year to something like close to 135 at the top range. I think this is notable because it just shows that there's so little room for error in this world that if you get, you know, if you get anything wrong, the market will interpret that they are going, that you're going to have some serious problems when it comes to AI.

28:03What do you think? Yeah, I mean, it's every quarter of the earnings that come out from these big tech companies, it's sort of a lottery. Like you look at the numbers initially and you're like, wow, those are big beats on revenue, big beats on profit. Like we're seeing a 39 % growth in Azure and a huge increase in bookings. And yet the shares plunge. Microsoft gets absolutely killed in the market. And there are always games played by Wall Street, right? People don't just want to see you beat the analyst estimates. They want to see you beat the beat, right? And so Microsoft has found itself in this very strange situation where a one percentage point miss in its Azure revenue momentum wipes hundreds of billions off its share price.

28:46And then people are also, So there are also several other things going on at Microsoft. There are some very interesting things disclosed about OpenAI, but also its CapEx. Microsoft is on track to spend$140 billion this fiscal year, which their fiscal year is a bit weird. It runs to June instead of to December. And that's pretty much on track with what they spent in the last quarter of last year, which was$37 billion. But people are questioning this. This is 66 % more than they spent in the same period the year before. These are vast jumps. And then the other thing that caught people's eye was they had to disclose their relationship with OpenAI in a slightly different way.

29:25Now that OpenAI is restructured as a for-profit company from a nonprofit. And they revealed that they booked 10 billion cash gain from the amount of cash OpenAI has on its balance sheet due to all these like huge fundraisings that they're doing. But they also said that 45 % of their future Azure contracts were based on OpenAI. So immediately on the earnings call, all the analysts are like, I was about to say holy S there, but I should say, oh, my God, you know, 45 % of your revenue is tied to OpenAI. Remember, this is a company, as we said before, it makes a lot of revenue, but it doesn't make any profit.

30:02For a company the size and age of Microsoft to be that reliant on a single customer is a big red flag. a single customer that has a lot of noise around it and increasingly has competition from Google and Elon Musk and Anthropic. So the CFO spent most of the calls trying to say, oh, well, you know, maybe 45 % of it is open AI, but that means that 55 % isn't, and that's still 350 billion in revenue. But as you can see, the shares were down 7 % at that point. When I woke up this morning, they were down 11%. So I don't think Amy Hood, the CFO of Microsoft, did a very good job persuading everybody.

30:34But it plays back into what we were saying before about this diversification. Like Microsoft has spent a long time locking OpenAI down to lots of exclusive contracts with it, both for its technology and to be its main compute provider. It's then spent the last 18 months trying to get out of those. And that's why it's investing in Anthropic. That's why it's investing. People miss this. It's investing a lot in Mistral, the French AI house, because it believes that Europe is not going to let its AI be run entirely by a bunch of West Coast tech companies and Elon Musk. So Microsoft is now all about the diversification.

31:05Before it was taking credit for turning a 14 billion investment into open AI into whatever the hell that will be worth, many multiples more when the company goes public. Now it doesn't want to talk about open AI. It wants to talk about Anthropic. It wants to talk about Mistral. It wants to talk about its own models and its own chips. And I think that's a new story that it's trying to tell the market, which is why we saw shares got hit. And then, as you said, Meta is this whole other story. This is my perspective here. I think we're just so early on in the AI game that the market is looking for any signal for where this is going.

31:37And it's like, oh, like it's a little bit of a slowdown in growth. Microsoft's going to lose the AI game. Meanwhile, we're so early in the build out. We have no idea what's going to happen. We have no idea what's going to happen with meta super intelligence based off of this quarter's revenue. So it's all extrapolation based off of limited data. Exactly. And you look at the quarterly, the short term quarterly reaction to the quarterly earnings, right? They're down for 8%. Remember, two quarters ago, we were talking about how Meta had 200 plus billion wiped from its valuation in a matter of minutes.

32:06Now we're talking about how it had 300 billion added in a matter of minutes. But if you look at the long term trajectory of all of these tech stocks, it's like this. You know, Google is now worth more than four trillion. Microsoft has lost its, actually, Google has bolted its, leapfrogged it in the kind of market cap race. Microsoft is still 3.6 and we've got NVIDIA way out ahead of everyone at five trillion. The general trend is still upwards, but we just have this skittishness in the market as anyone's looking for any signs of weakening demand, the bubble contracting or even popping. But if it was just about spending, Meta Double is going to double this year its CapEx.

32:46So like spending on data centers and chips and servers and everything that goes into supporting that, upgrading power structure. It's going to double it to 135 billion, you know, from 72 billion the year before. And that's not all that Zuckerberg is spending on over there. He's also spending hundreds of millions on compensation packages to try and attract staff from other leading AI labs to try and rebuild their own llama model, which, of course, felt fallen badly behind has been a huge embarrassment for them to this point. But Meta was crucially able to show that AI is improving their ad targeting and their ad revenue.

33:20And their counterparties love this, whether it's on Instagram, WhatsApp, Threads, Facebook. And so the market rewarded that. They're like, OK, you're going to double CapEx, but you're actually also showing us like some tangible impacts from AI on AI, improving your advertising revenue, because that's what Metra is, just a big advertising business. Microsoft hasn't done that. You know, Microsoft has made a great early bet on OpenAI, but it hasn't proven that it's like office suite of products. have been meaningfully improved by integrating AI. And if they can meaningfully improve it, that means that they can charge more for it and they can increase their margins.

33:56That's the big game now for Microsoft because it's a software business and a data center business. And I think the former has quite a lot of question marks about how it's integrating the various models into it. But here's my big point. I don't think we can read anything into what's happening already. I mean, we're still in the build-out phase. We're still in the picks and shovel phase. we're not anywhere near the point where economic activity tied to AI is exceeding the amount of money being spent on infrastructure. Yes, Meta has done some things with generative AI being used for creative optimization.

34:32Yes, they've done some things on using standard machine learning to better optimize their platforms, but there's no way that what Meta is doing today will have any bearing on whether its AI program succeeds because that's all about this personal intelligence thing. And so, you know, and Microsoft too, like, we'll see what happens with Microsoft. But, you know, there, you know, it seems to me like all the jockeying has been, like you mentioned, so much of their revenue is going to be coming from open AI, not end users who are deploying, you know, AI solutions. And so I think the market is like blindfolded right now, swinging around, trying to be like, who's winning AI, who's not.

35:11whereas like we won't know even you know who is leading for a couple of years yeah and who is leading or having a good moment it's anthropic at the moment it could be google again next maybe open ai knocks the lights you know blows the lights out with like an incredible new model with you know something that really captures captures the public imagination or is indispensable to you know fortune 500 companies uh everything is so fluid and flexible that's why it's so much fun to cover as journalists right but it's also pretty exhausting as well and we haven't even gotten to elon musk yet oh yeah that's coming ben and i did hear that gpt6 might be coming uh in the second quarter this year so maybe as soon as you know this spring which would be interesting all right one more company to touch on with earnings and then we can get into oh shoot we have so much to discuss.

36:02Let's talk about Apple quickly and then the Amazon layoffs if we have a moment and then we'll get into Elon. So first of all, Apple. My prediction has been that Apple's going to have the best year of its history, maybe in the past and going forward, because it's in this moment where there's no AI device and it's selling like crazy. I'll just give one note from your story about Apple's earnings. By the way, blowout earnings. They did it. They beat on revenue. they beat on earnings. The iPhone revenue grew 23 % to 85.3 billion compared with the prior year. I think that the estimation was it was going to grow something like 14, no, 16%.

36:44No, yeah, 14%. It grew 23%. This is a company that's been struggling to post double digital iPhone growth for a long time. To me, it just seems that Apple is kind of in the driver's seat in 2026. No AI device, no killer AI consumer app yet. But the iPhone is selling like hotcakes. What do you think? Well, the next iPhone was supposed to be the first AI enabled one. Siri was going to be overhauled. It was going to become your de facto like personal assistant. And that just didn't happen. I mean, I'm sure you've spoken about this on the podcast multiple times before. But what we have seen is the iPhone 17 was a blockbuster product.

37:22You know, it drove a huge upgrade cycle. Really interestingly in China as well, which shows the tightrope that CEO Tim Cook has been walking between Trump and China. You know, he's still on it, you know, despite claims to the contrary. um conspiracy theorists say that the the new os update liquid glass performs so badly on the 14 backwards and i have one of those and i can verify that yeah it's pretty janky i love that that's amazing could be you you make an operating system that's so complicated to run the old iphones feel slow and annoying so you upgrade in order to you know uh for it to look slightly nicer But, you know, after all these reports that, you know, Apple had lost the smartphone market in China, it had bungled the AI opportunity.

38:17I mean, this quarter did answer a lot of those questions, right? People are still buying the iPhone above almost any other device. Geopolitics haven't disrupted its supply chains. It hasn't, you know, enjoyed sort of like, hasn't suffered from a de facto ban in China because of the trade war that's been going on. And, you know, they're not spending the hundreds of billions that their rivals are on building data centers and investing or investing in startups or building models. They've just signed a deal with Google Gemini, much like their search deal to put to make Gemini improve Siri. So they're kind of outsourcing a lot of this AI stuff.

38:53And they're just focusing on being the main device through which most people will experience it. And, you know, maybe long, long term over the next 10 to 20 years, that'll be a bad bet. But certainly in the short term, they're saving a lot of money and still selling a lot of phones. Yeah, they seem like they're going to be in good shape for a while. And you're right. I think it is a long term liability, but it's going to look really nice in the short term as all the spending happens. And Apple just kind of sits there. Partners sells a ton of phones. You know, maybe it ships another bad design and makes more people upgrade.

39:25They have the phone coming this year. I really think that this could be the year of Apple preceding the last decade. That's my opinion, at least. And they bought a very secretive Israeli startup today where the technology analyzes micro changes in facial expressions, which they sort of say means you can communicate without talking. But I do think it shows that Apple is, as Meta and Google are as well, looking very closely at wearables, but in particular like glasses. I guess another big tail risk for Apple is this Johnny Ive design new device, screenless device that OpenAI is helping finance, which we hear is a big part of the reason why Apple decided to go with Google Gemini instead of OpenAI and ChatGPT to PowerSiri.

40:16Because you're like, OK, Pixel's not really a competitor to the iPhone. And it's proved that over a number of years now. But whatever this new Johnny Ive device is could potentially be quite threatening. So there's a lot of moving parts, but Apple still looks pretty healthy. I mean, I don't have an iPhone 17 myself, but I did go to a Halloween party down in Palo Alto. And the house I went to is next door to Steve Jobs' wife. and she hosts an absolutely enormous Halloween party. They were talking animatronics, like dozens of actors. People come from miles around to see it. The queue goes around the block, so I chose not to take my daughters into it.

40:59But I realized if you stood in that queue for an hour at the end, they were giving everyone a bright orange Halloween themed iPhone 17. So maybe I don't think she was giving them away, right? So those probably don't count towards the sales figures. But, you know, people were absolutely delighted by this device. and I can testify that looks pretty cool. All right, I got to go to her party next time. That sounds amazing. I'm going to say, first of all, an interesting thing, speaking of, you know, we talked a little bit at the beginning of Anthropic having a moment here. Actually, Apple wanted to go with Anthropic first and they wanted to use Claude.

41:35This is according to Mark Gurman, said it this week. Turns out Anthropic, you know, was basically trying to chart a lot of money and then they eventually went with Plan B, which was Google. So that's an interesting wrinkle. I have to tell you, I don't think the device is going to threaten Apple at all. At least for the next couple of years, I really think that these AI devices are going to be a big letdown in the near term. Another humane pin? I mean, you know, it's funny. We joke about the humane pin. And we think, by the way, created by, I think, a couple of people who worked on the iPhone. And we think that, of course, Apple will come up with a much better idea than the humane pin.

42:13whereas maybe the interaction layer just sucked and i think that's not taken into account as often as it should be it didn't help that it also overheated and exploded well listen i mean you never really want that in your device but uh you know we've seen we've seen companies get over that in the past yeah look it was everything that they could have done wrong they did wrong the product wasn't great uh there were some issues with the safety side of things and they they gave they rolled it out with this very weird like emo uh video yeah that's right not not exactly a great rollout but also that's what i'm saying like you know maybe all those things were the problem or maybe the form factor was the problem and i think we're going to find out pretty soon uh what the answer is all right speaking of of companies that try lots of different versions of products um with ai inside we have amazon and amazon did not report earnings this week but But they are laying off 16 ,000 people this week.

43:09This is from Business Insider. Amazon started notifying impacted employees Wednesday morning as part of its plan to cut 16 ,000 corporate jobs. The head of HR, Beth Galetti at Amazon, wrote this. I have important but difficult news to share with you. After a thorough review of our organization, our priorities, and what we need to focus on going forward, we've made the hard decision to eliminate some roles across Amazon. unfortunately your role is being eliminated as part of the changes and your employment will end after a notification period she also wrote basically like i you know i know that there's some indication that there's a new rhythm or some of you might think there's a new rhythm where we announce broad reductions every few months because i think what is this the fourth mass layoff that they've had over two years and she said that's not our plan amazon is a company that's like already operating with somewhat paranoid employees so this isn't going to make things much better.

44:04What's your main takeaway about the workforce reduction here? And do you think it's a good or a bad sign for the company? There are two things I think that's going on. One is they hired a lot of people during the pandemic, you know, when, when delivery surged and they really needed, they needed to surge in their own workforce to cope with this. And I think we're seeing a correction on some of that now as, as your online commerce, you know, normalizes somewhat. But we're also seeing is a company which, you know, 50 billion going into OpenAI plus a lot more into Anthropic. I think this is a company preparing for an era of AI, whether that's, you know, both in its white collar jobs, but also blue collar jobs like working in the factories.

44:48I mean, it's already one of the most automated companies in the world. And if nothing else, Amazon is relentlessly focused on the bottom line. Anywhere it can cut costs in particular permanently, it will try and do so. So it's a very uncomfortable situation for you to be if you are an Amazon employee. So working at Amazon is uncomfortable by design, right? They want to keep you right on the edge to make sure that you're constantly as productive and aggressive as you can be. But what we're seeing here, I think, is there's a company that's that's getting ready for what it believes to be a completely different era of e-commerce, but also working in factories and running and producing some of the technology behind its various websites.

45:32It just doesn't think it's going to need as many people in the future. I know the HR chief is trying to calm everyone down by saying this is not the new normal, but I would be very surprised if this is the last mass layoff as Amazon we see in the next 12 months. So one more question about this. Do you think this is AI related as in, do you think AI is like automating Amazon employees jobs? Or do you think this is a financial decision that might make room for some of the AI investment? I think it's a little bit of both. I mean, it always is. They talked about thinning out layers of bureaucracy as every company does when they make cuts.

46:10But I also think that they're seeing some of the employees that have been able to integrate AI into their daily work life and become more productive. I think they're kind of looking ahead at that. We don't have the exact demographic data of who they've let go, but I think it'll probably be more heavily weighted into the older employees that are maybe less able or less willing to adapt to this new AI powered way of working. And certainly if you look, we get Amazon's results next week, but there's no slowdown in momentum of the company's growth. so without even though they've been taking all these employees out of the company you see a short-term hit because you have to pay all of them severance but longer term your cost base comes down dramatically yeah amazon has this has a number of leadership principles that the company runs by and one of them is called invent and simplify and what you think that means is like you you sort of put technological solutions into place and simplify processes and gendered may i sort of puts that on overdrive where like you can just simplify you know invent and simplify your own job and that has sort of been a calling card and something that amazon employees have been proud of for a long time as they invent they simplify and they go and they move on to something else and i don't i wonder if that leadership principle will maintain the prominence that it's had within that organization and the and the support it's at among the employee base uh if they realize that at end of simplify it can be leave it has to be invent simplify and then invent again if it's invent simplify and leave then that spirit of invention and and efficiency within the company could really suffer yeah that's true i mean if it's innovate yourself out of a job it's uh it's a difficult thing to get your head around yeah and that used to be something they used to be proud of because they would get another job but when that doesn't when that security isn't there that's a problem.

48:00Okay, let's talk about... Well, that's sort of what Daria Amadei was talking about in his 20 ,000-word opus the other day. We're going to see he predicts at least a lot of job losses as fewer people are able to do so much more work or even a completely automated out of the process once they really start to get agentic AI-powered co-workers working to a degree that they don't constantly have to be checked okay very briefly let's talk about spacex looks like an ipo is coming pretty soon uh people are thinking it's going to be in june which will coincide with elon musk's birthday and some planetary alignment where uh on june 8th and 9th jupiter and venus this is going again to the financial times thanks for the great reporting this week jupiter and venus will be within little more than one degree of each other in the sky about the width of a thumb held at arm's length.

48:56Your thoughts? I mean, it's just another day in the Elon Musk Inc, isn't it? You know, reporting on this man is fascinating, but also extremely difficult because you often can't tell whether he's joking or not. But he has gone out there, told investors that he thinks an IPO in June would fall in an auspicious state. As you said, the conjunction of Jupiter and Venus, which is then on the 8th to 9th, which is then joined again, by Mercury in a few days later. His birthday is obviously in that month. 28th of June is a Sunday, though. So of course, you can't hold an IPO when markets are closed on a Sunday.

49:35But he's thrown out this idea. He likes to light a fire under people. Everyone's saying you can't do an IPO in particular of this scale on that time, you know, in that timeframe. But you know, Musk, Musk's first principles, right? You question everything. And you force people to just work hard and get it done. There are other things that you've got to take it with a pinch of salt. You know, this has been a big story in Europe, at least over the last few weeks, he got into a bit of an online spat with Ryanair, which is like a budget European airline, over them refusing, you know, they're notoriously cheap, like you have to pay for everything if you don't want to sit on the wing.

50:12And they were refusing to put Starlink on the planes. And obviously, Elon wants Starlink on the planes, because it earns more money. And they were arguing about drag. So he threatened to buy the company, fire the CEO, a guy called Michael O 'Leary, he's quite a big figure in the UK, and replace him with somebody called Ryan. This is how, of course, the Twitter acquisition started him threatening to buy it as a joke. So you never know where it ends up. But you know, we're hearing the June timeline, despite the planetary alignment angle is a real thing. But then just today, you get all these other stories coming out that he might have ambitions to merge it with one or more of his other companies and and you know that very well could be the case i mean we haven't heard anything solid about it yet but those are the rumors out there yeah i think the rule with ryanair is in the jurisdiction that it's in a american or non-european cannot be uh the owner so that sort of idea is out the window but i'm sure it's not beyond elon to set up a set up a european subsidiary and staff it with one of his people and buy it that way.

51:12Rules are made to be broken for Elon. That could happen. The company that you referenced that people have been talking about SpaceX merging with is XAI, which would be really interesting. I think not a great combination, if you ask me, especially, you know, SpaceX is highly regulated. It's a government contractor. XAI recently was, you know, bikinifying people on Twitter. I think that's some sort of risk there. But actually, I am curious to hear your perspective. What is the sort of most logical reason for why those two companies would combine? The most logical reason is that one of the reasons that have been given privately to investors for wanting to IPO SpaceX is that it needs to raise money so that it can develop and build data centers in space.

52:07they argue that it's very cold when you're facing away from the sun so you don't need to pay for cooling he already has a network by he i mean elon musk a network of 9 000 starlink satellites all linked and coordinated by lasers so you've got kind of like the communications infrastructure out there uh and you have unlimited solar energy not not if you're not affected by the atmosphere so i know it sounds like crazy sci-fi stuff but this is what spacex has constantly been based around. I don't think anyone thought the largest rocket in the world would be the booster stage would be caught by giant mechanical arms just a few years ago.

52:41But he's proven that that was a technological challenge that was surmountable. So the argument is you put them together at XAI, obviously, you build the models X, which of course was another Musk company, but which was merged, I think, last March with XAI. That's the distribution that Grok goes out under. And that's how Musk is going to win in AI, like his models might not be as good, he doesn't have the same revenue. But what he can do is win the data center race, you know, and have these up in space before anybody else. And he's not the only person looking at doing this. Google's also looking at it.

53:12Jeff Bezos is looking at doing it through, you know, through his Blue Origin rocket company. So that's kind of the bull case. That's the rational case. The conspiracy theory is that, you know, XAI is worth 250 billion, but has, you know, a tiny multiple of that in terms of revenue. It hasn't gained that much traction despite relentless promotion through X and on social media. And that like a lot of Elon's very wealthy friends are invested in XAI. And the easiest way to make sure their investment is protected, just like how he made everybody who invested in his acquisition of Twitter hole, is to merge it with a more successful company that's the undisputed leader in its field, which would be SpaceX.

53:49So you put them together, everyone enjoys the upside, nobody loses money, and you crucially invest in the next Musk fundraising, knowing that he'll always have your back and make you whole. Yeah, the training in space thing is fascinating. I mean, the stuff that people think of, and I know Google is also thinking of doing something similar. So that to me, thank you for offering that. That is the most logical explanation. All right, let's end on this. This IPO, SpaceX IPO, when it happens, it's going to be massive. I mean, is this, you know, in some ways a test for because it will be I imagine it'll be the biggest IPO ever.

54:25Is this going to be a test case for what it will look like, even though it's completely different industry for what it will look like when the AI companies decide to go public? Yeah, I mean, we're hearing 50, potentially raising 50 billion at a 1.5 trillion valuation. It'll be a real test whether the global markets can absorb this. I mean, one of the people are desperate to invest in all these hot private companies like Stripe and SpaceX and Anthropic, and you can kind of get exposure to it by hook and by crook. But the real game changer would be if Elon's army of online fans who have propped up the Tesla share price for years become SpaceX fans as well, because then you just have access to pools of liquidity and capital so much larger than what they've ever had access to in the past.

55:13So I think if Musk can show through the SpaceX IPO that there's huge retail demand out there through, you know, either pension funds, but also Robinhood or Revolut or wherever else you can invest your personal cash, that could be a real game changer because that would give the rest of these startups that are less mature companies the confidence that they're not going to go out there and have a real disaster. Because remember, there are a few things worse than listing, seeing your stock price decline, having all of your staff demoralized because they thought their stock was worth this and it's in fact worth that.

55:46And then be out there with Klieg lights being shone on your balance sheet. It can be a very comfortable thing to be private before you really have established net income. So we'll see. I mean, if it does go ahead in June, which we're led to believe or have reported out that it will be, it's going to be an absolutely blockbuster summer. Yeah, it'll be fascinating to watch. All right, Stephen, tell people where they can find your work and the work of your team. Well, the Financial Times is obviously one of these legacy media instruments. We do print a newspaper. It's salmon pink, very recognizable.

56:20But you should go to ft.com. We're all over Instagram, X, wherever. We've got a great team here in San Francisco. So I hope you have a lot of subscribers watching this show. All right. Well, love it. I always love reading your work and it's really great to be able to speak with you again here on the show. So Stephen Morris, thank you so much for coming on. Thank you for having me. All right, everybody. Thank you for listening and watching. And we will see you next time on Big Technology Podcast.

57:00The Insolvenz is increasing. What is behind this development and what is for the middle of this? In Folge 6 of Endstation Insolvenz is the leader of the Wirtschaftsforschung by Kreditreform, why the models are now faster than their borders. And why the insolvences so hard they are also are, to ensure a functioning market. We'll see you next time.