Full Transcript for Big Tech's Ugly Week, Getting Rich At Anthropic, AI Costs Plunge

It's a big tech bloodbath as Wall Street turns on the AI build-out. Anthropic's S1 finally drops, and a lot of people at that company are about to get rich. And meanwhile, Anthropic and OpenAI are dropping prices dramatically what's behind it. We'll cover it all on a Big Technology Podcast Friday edition right after this. This episode is brought to you by Kribble.

Quick question. Is your IT budget growing 30% a year? Because your data is. And now AI agents are piling on, running investigations around the clock, and running up three meters at once. Storage, queries, and tokens. Kribble, the AI platform for telemetry, was built for exactly this. Own all your data. query it wherever it lives, and let Kribble route every AI request to the right model at the right price.

The best part? AI inference for workloads on Kribble is included. Models will change. Agents will change. Your data is the asset that lasts. Own it with Kribble. Visit cribl.io.

How do you price intelligence as a service? How do you charge for delivering end-to-end work, not just access? Those questions remain front and center for AI founders and operators, and getting to the right fit business model, whether that's selling usage credits, outcomes, seats, or a hybrid combination, requires constant iteration. Chargebee is the billing and monetization infrastructure that powers just that for ambitious, fast-scaling AI businesses.

Visit chargebee.com to learn more. That's C-H-A-R-G-E-B-E-E dot com slash Alex. Welcome to Big Technology Podcast Friday edition, where we break down the news in our traditional cool-headed and nuanced format. We have a great show for you today. It has been an interesting week, to say the least, in big tech land, as companies like Google and Microsoft have reported their earnings, along with Amazon, and the market is punishing some of them, especially Google, mercilessly.

We'll also talk about that Anthropic S1 document. and how rich people within that company are about to get. It's pretty wild. And finally, we're going to talk about the actual numbers on the AI price drop. It's pretty dramatic when you look at it. We'll try to figure out what's going on and what it means. Joining us as always on Friday to do it is Ranjan Roy of Margins.

Ranjan, good to see you. Big tech bloodbath on the Big Technology Podcast. What could be better for a Friday? Certainly an interesting week to be called Big Technology Podcast when the big technology companies lost nearly a trillion dollars in market cap. This is from the Financial Times. Big tech sheds nearly 900 billion en route over soaring AI spending.

Big tech companies lost about a trillion dollars in market value after another sharp rise in their spending in the third quarter, which has turned some of the most profitable businesses in history into cash incinerators as they stake their futures on AI. Shares in the big four hyperscalers, Amazon, Meta, Microsoft, and Google, all fell after revealing their third quarter results in which capital expenditures almost doubled and alarm bells sounded about slowing revenue growth in the core business.

Ranjan, here's my hot take here. A lot of this has been borne by Google, which reported still strong, but not as strong as expected search results. Search results, that's funny. Search earnings, right? They were supposed to grow 15%. They grew 12%.

I don't really think this is a spending issue. I think Wall Street has basically given these companies the green light to spend and spend with abandon. But we're getting to the point where free cash flow is going from positive to negative. As we saw again with Google, two quarters in a row of negative free cash flow. And when that happens, the the questions about the core business emerge and so this isn't a question about AI spending it's really a question about like is your core business going to stay around long enough and that goes for Google with advertising and Amazon you could call it the retail as their core business uh you know those are the companies we've had

the questions with if the core business goes away or shrinks uh then then Wall Street seems to have some issues with all this infrastructure spending your thoughts No, I completely agree. I think there are those two elements of it. And I do think there is an element of the AI spending, and I'll get into it fundamentally changes what business all these companies are.

And going back to that's why it's so important that their core businesses at least seem reasonably promising because the moment Google and Meta created some of the greatest cash machines of all time, literally the greatest business models to like... ever exist in humanity, like the profit margins, the monopolized efforts in terms of no one else being able to enter these markets.

And anything that shows that those are under threat makes us question, what is Google? Is it an AI compute business? Is it a enterprise services business? Or is it a search business? And Google themselves, I mean... It's a golden goose, man. It's not a golden egg. Okay, okay, fine. Golden goose. Lays a golden egg. But yes, continue. Does the golden goose lay a golden egg? Is that the whole thing? You would imagine.

At least it lays an egg with the golden chick inside. I'm going to brush up on my childhood fables after we're done with this call. But let's get back to Google's learning. So it is still, the numbers are, I mean, insane. Again, in one quarter, revenue is still up 25%. Operating profit is still up 30%. operating margin is still holding strong around 30%.

So this is still an incredibly profitable, incredibly fast-growing business. But again, core search. And this is one thing that I found really interesting is Google has been touting, especially in SEO, GEO land where I spend a lot of time, overall queries are actually increasing in Google search. So Google's like, search isn't dead. Everyone's still going in and searching things.

But they finally themselves said, AI mode is going to actually, it went from 2% of all searches ending in AI mode to 8.9% of all searches. So it's dramatically increasing. I'm sure every listener starts to feel that. And they don't know how to monetize that yet. They might figure out the right business model to monetize that, but it will be a different business.

So to me, I think the core search business, this was the earnings report that reminded everyone. I don't want to say it's done. It's still giant, but it's never going to be what it was before. Okay, so this is interesting. So cite the statistics again in terms of the amount of searches that have gone from AI overviews. Was it 2% to 8%? Yeah, 2% to 8% again.

Okay, so think about this. Let's play this out. Google is seeing an increase in searches, all-time increases in queries, right? Is it possible that these AI overviews make up most of those or even responses back and forth? Because I imagine every time you ping like AI mode, you get a response. So is it possible that we're just seeing instead of like Google not being able to monetize AI mode, and this is an interesting point that Kevin Indig made on X, what if people are just searching less, right? So all the growth comes from AI mode potentially.

And, you know, you end up seeing far fewer searches because people are starting to search within ChatGPT, maybe Claude, maybe Grok. I mean, ChatGPT has 1.2 billion users that has to play into this. They're still saying search query, overall aggregate search queries are growing on Google. Sundar said that before, they reported it again. So they're still saying that that is not declining of people actually directly asking the question to Google in google.com, not in Gemini.

So it hasn't yet. I agree. I don't- Wait, hold on. What counts as a search though? Like if you're in AI mode on google.com, is that a Gemini or is that- See, I don't know. That's a good question. They are counting that. No, because if they're counting AI mode as part of their search data, right? If it's grown from two to 8%, then AI mode counts as searches.

Actually, you know what? So they actually said that overall search queries grew at their fastest year on year in five years. So they don't give aggregate total numbers. They don't break down what exactly qualifies as a search. I think that's a good point because to me, the idea that more people are searching more directly into the Google search bar, I agree.

I know my personal behavior is probably not fully reflective of the average consumer, but overall, it has to be going down. So I think that's a good one to push on overall. What is a search? Because you know what? What do you do when you search? You want to go to a website. And what are websites experiencing? You know, they're not experiencing the traffic from Google search.

Let me just read this. I want to put this out for sake of argument. This is from the Reddit CEO, Steve Huffman, on his earnings call this week. He said,

The reality is that the AI mode does not drive traffic, not for us, not to anybody as far as I know. Google used to be a traffic driver, now it's not. People still want Reddit, it's just Google is not sending them to Reddit the way they used to. We don't control Google, we only control Reddit. And then my favorite part of the whole thing, and I think this is most telling, Google is teaching people that if you wanna come to Reddit, you need to come to Reddit.

Is it possible that right now we're seeing a behavior shift where people used to search Google to find... Google was an improvement on the portal, right? So you had Yahoo where you would click through a bunch of links to get to the website you want. Then you went to Google, you searched for things, you went to things you wanted. is what's happening now with AI that if you actually want to go to a website, you don't search anymore.

You just type in a URL. And so it's interesting. Google, potentially by not sending referral traffic to sites like Reddit, sort of has potentially even disintermediated itself. Yeah, no, I think this will happen across the board. I think it was a year ago, when did we say the web isn't dead, it is in secular decline?

Every time, by the way, every time someone says last year, for me, it always means April, because people can't comprehend the fact that we've gone so far, like technology has developed so far from where we were in January to where we are now, they always say last year, and last year always means March or April. So I think we probably said it in March or April.

Wait, wait, wait. You're like a six-month year guy? I'm not saying... A year, six months? Yeah, I'm saying, yeah, we're basically living in a world right now where every six months has a year's worth of activity. You don't think so? I'm going to... Okay, from a technology standpoint, I agree. I mean, I don't even remember last month or two months this summer.

That feels like ages ago. But all right, let's get back to especially Reddit. I think this was really interesting because... Why– Steve Huffman made a really good point that Google is pinging and crawling Reddit more than ever, yet sending less search traffic. It's declining dramatically. What is that saying? That Google is extracting information from every website out there, the Reddits of the world, but no longer that deal.

Remember, the original deal– why publishers, why everyone posting a website let Google crawl their website was to hopefully that will drive traffic. And it did. And that was the entire foundation of like the economics of the web for decades. So I think that this Reddit earning call is really interesting because they basically are recognizing finally that that is completely broken.

But it's more than that. Again, I'm trying to make the point here. And first of all, I should caveat. 12% growth when you have a business worth tens of billions of dollars. Google made 63.3%

on search and other advertising in the quarter, in the third quarter. Sorry, $63.3 billion.

That's huge. And to grow from a 50-something billion base, 12% to 63 billion, we shouldn't be sneezing at it just because they didn't do 15. But I do think, and that's the reason why people are wondering now, is there was a projection that they would hit 15, they hit 12. And why is that? They talk to us about how some of these AI overviews are going to cut into revenue.

And, you know, searches are probably up, but I bet some of that conversation, the AI overviews counts. But more than this, it's like, it's possible, you know, to have a search business, users need to click out. And this is what I'm saying. It's not necessarily the bargain is gone, which it is, but the bigger issue is, is Google training people not to Google and just to go direct?

I still don't think– I don't know. Again, as a heavy user of Reddit and still believer that it is one of the best places on the internet to actually like– find information and have human interaction, I am kind of happy because, honestly, I would rather... There was this brief period where Reddit was driving so much AI-related traffic that every business out there wanted to post on Reddit just to try to optimize for GEO.

I'm kind of glad that we're moving past that. But I still feel the behavior of... going directly to a website is going to be a relatively limited thing. But as I say that, maybe, again, in the shopping world where I spend a lot of time, it could, the more Google disintermediates the actual end, it gives less control and value to the end business, they will push to not worry about Google.

Already, I can tell you, the idea that search traffic is the lifeblood of a business, everyone has given up on. So like it's still kind of now one channel, whereas before it was the lifeblood in a big way. So I agree, but they have to. To Sundar's credit, innovators dilemma and they've chosen their strategy. So do you think it's the right one, though? Right, yeah.

I mean, the bet is basically that, well, this is kind of like what we've been talking about in the past, that Silicon Valley is making this all in, literally, on like an unproven new technology, which is, I guess, it's what they do. I mean, I guess I wrote a book called Always Day One, so I should be the one preaching this. you know more than anyone but it to be relevant in tech is this constant cycle of reinvention where you cannot hold on to what got you to the dance for too long otherwise there's going to be a new dance and you're not going to be invited and so what i think is happening across google amazon all these companies meta for it for sure uh maybe not

microsoft we'll get into that uh is that these companies believe so strongly that ai is going to change the interface that they have to be like you're right it's innovators dilemma and rather than be a relic of time we're going to try to go for you know, what's coming in. But in the near term, there are going to be rewards to those who aren't so good at it.

Listen to this. This is an interesting reaction to the earnings from Steve Howe, an economist. He said, interesting idea. Everyone assumes Google will win the AI race, a complete 180 from a year ago. I don't know if it's everybody, but they have the number one model with carbon, their new model they released this week, which is like number one on the benchmarks.

And he writes, yet, winning AI may not be commercially good for Google as they haven't figured out how to target ads well within AI mode, while meta is pulling away on ad revenue despite lagging on LLM. And then he drops this chart. And the chart is amazing because in 2024, the leader in ad revenue was Google. In 2028, the leader in ad revenue is meta, like by a good margin.

And Amazon has gone effectively from like kind of a distant third into a company that's now challenging Google for two, although still somewhat behind. But this is interesting, the fact that Meta, which has struggled with LLMs and hasn't really made that big shift to change its entire experience to make it AI, it's pulling in the ad money, whereas Google, which is making this move, has fallen behind.

What do you think? I think that's a really good point because if you think about it, Google's ad revenue, their golden egg, golden goose, still-

I think it's a golden egg. I think the goose lays the golden eggs. I don't think the goose has a golden sheen to it. No, I'm going to add to Google's search queries right here, and I'm searching it. All right. So while you're searching, I think... Oh, my God. Yep, the goose. It's the egg. This is from AI Overviews, Asim's fable, The Goose and the Golden Egg, a classic story about a greedy farmer who kills his goose, which lays a single golden egg every day, hoping to get all the gold at once, only to find nothing.

So you just both were proven wrong, and you provided no value to the external web ecosystem and received that information by some poor publisher that actually just put that information with their heart and soul out onto the web, and it just got AI-overviewed. No, you're wrong. The story comes from read.gov. It's a government site. So I've actually just convicted them to a brainwash the masses type of campaign.

No, no, this is the future of the web. The only publishers that will be left are government. But that's bleak and that's possible. I mean, seriously, maybe that's what happens.

Oh my God, that's terrifying. What's interesting is Google's ad model is dependent on this entire web ecosystem. It's dependent on websites exist, they pay Google to direct traffic to them by keywords, and that's been the model. Amazon and Facebook, or Meta with Facebook and Instagram, have closed ecosystem advertising models. The advertising all lives within...

And honestly, Amazon showed how valuable that can be. Meta, their ads format is not going anywhere. It's only going to get more and more valuable. People are still going to scroll through Instagram and Facebook and see ads within there. So actually... more thinking about it, that open ecosystem of advertising is probably the most ready to be, or is being disrupted already, whereas it puts Meta and Amazon in a much better position.

Definitely. Golden Goose, by the way, is an Italian fashion brand. Oh yeah, those are those shoes. You know the shoes, right? No, I don't have them, but I would say that my read of the situation makes me at least a little bit more hip than you, given that you went to the read.gov site.

You're thinking of the $500 pre-scuffed shoes.

Yeah, Ron, don't get with the times, man. Days of golden eggs are over. It's only about golden goose sneakers right now. So to me, I want to hear your take. The most interesting part of it, I think this week was the moment that investors are recognizing these companies are betting on being a different business. We said they're disrupting themselves, but do you think that's what's happening? They are betting on being a...

cloud data center compute provider business which have unproven economics which maybe they will win they're very well positioned to win but no one what knows what they look like and almost i have to imagine definitively there's no way a

compute provider business model has the same attractive economics as search advertising do you think that's what's happening Well, again, you take certainty versus uncertainty. And certainty is this search advertising business is working well. And uncertainty is your cloud business is going to power an uncertain AI takeoff. So you factor in the risk and you make your decision.

And then the other companies, like Amazon was very interesting, by the way. So this is from the Wall Street Journal. Amazon stock drops on Wall Street missed despite biggest cloud growth in years. So you would think that if the investors, well, you know, you're right. Investors are seeing these companies go in this different direction, and they're not buying in 100%, which is very interesting.

Amazon.com reported its highest revenue growth for Amazon Web Services in years. That wasn't enough to quell investors' concerns about artificial intelligence. Amazon Web Services brought in $46 billion in revenue for the three months through September, accounting for 39% growth year over year. Think about the retail business. The retail business is...

And by the way, overall revenue increased 12% to $201.9 billion. The retail business is...

It's amazing because Amazon.com was seen as like this, and it really was a revolutionary invention for better or worse, right? It is sort of a cornerstone of the early internet and today, but it is changing in nature. Listen to this. Groceries and Amazon's everyday essentials accounted for half of units sold on its retail website in the quarter.

Again, it's everyday essentials from Amazon, which are effectively white-labeled products of like all these other products that you use in the house, and groceries, half of the units sold on the retail site in the corner. So Amazon as the everything store is really, it's a little bit different now, wouldn't you say? Well, overall unit– half of units I feel is a bit misleading because if these are cheaper products that are– So each cucumber counts as like a– Yeah, yeah.

When you do a Whole Foods order, they're going to count it. But they want to show growth there because that's like a habitual purchase. So I think that's what they're trying to show with that. That's like kind of a– feels like a bit of a massage number to investors. But I have to say– Do you know what the crazy thing? My son, seven years old, because he'll get toys on Amazon, people will send him gifts through Amazon.

When he sees someone, a delivery person with an Amazon vest, he looks at them like Santa Claus. Or when he sees an Amazon truck on the highway, he's actually like, look, Amazon truck. No, no, I mean, I got to say... Amazon, Andy Jassy, you guys have a foothold in the consumer of tomorrow where they're not looking at you as this like a thin crushing mom and pop stores.

They're looking at you as kind of Santa Claus at scale. So it's crazy to see his eyes

pop up. Oh, yeah. I mean, I will say just this is a bit of a divergence, but like the Amazon branding on their trucks is incredible. Every truck you see has this like it's not even the most clever phrase, but it just reminds you of the value to Amazon. And they have the big footprint and they're using it for branding pretty effectively. Which does remind me, I actually had to go check right now.

My Halloween costume has been delayed a few days. It is out for delivery now, but Andy Jassy, we're cutting it close, man. We're cutting it close. We're having a very spooky big technology podcast. It's October 30th when we're recording and we have just hours. Until Ron John, what are you going to address as? Nathan Fielder. No, no, Blue Mighty Morphin Power Ranger.

We're doing the Power Rangers as a family. It's apparently made a comeback. I didn't realize there's like new iterations of it. So old meets new, Mighty Morphin Power Rangers. What are you going to be? I am not dressing up. I'm taking it easy Saturday and trying to run the marathon on Sunday. So folks, if you don't hear me on the feed next Wednesday and Friday, you know what's happened to me.

Wait, which marathon? I'm being swallowed up by the miles. New York on Sunday. Oh, it's this Sunday? Yes, and the weather is going to be beautiful, so my fingers are crossed. But I will say I've definitely been in physical therapy the last couple of days trying to work the kinks out of this old body to see if I can tune it up to get it going.

So I think it's about 50-50 right now, but I'll give it a go. Oh, I'll be out there cheering for you. I thought it was the next weekend and I'm going to be out of town. All right, Sunday. All right, folks. Listeners, get your signs out. Get your signs out. Let's track Alex.

Please don't. You're going to see me running very slow and hobbling. Crushed. Oh, God, destroyed it. All right, one last Amazon thing I want to talk about, which is, you know, if you go to Amazon.com these days, the biggest piece of real estate on that site is, any guesses? Alexa for shopping. No, the Bee Frontier. The Bee Frontier. The Bee Frontier is a wristband that you can wear.

It's $74.99 and Amazon sort of, Amazon pitches it as a wearable personal AI that grows with you. It's confidential by design and it transforms conversations into actionable summaries.

I hate this thing. I'm getting more and more annoyed at AI wearables by the day. I just think that I don't want to live in a world where everything is recorded. Because when everything is recorded, everything becomes performance. Everybody thinks about their words. Everybody realizes that someone's going to go back and just like the Black Mirror episode, they'll want to do a replay of their memories.

Why are we running towards this future? It just makes me upset. OK, hold on. Well, one, I'll still Alexa for shopping now at least takes up a third of my screen every time I open it on desktop. They're certainly pushing that. But that's it, because we've never really talked about the B frontier. And that was definitely interesting. They're making this bet on what is the future of their AI hardware ecosystem.

Personal feelings aside, do you think it's a smart bet? Because I remember from what I read, now it's going to connect into the overall Alexa ecosystem. They have it out there. We're still waiting on whatever OpenAI pin and Meta Tamagotchi. Do you think it's smart Amazon's going down this direction? Well, it's hard for me to fully take my personal feelings out of it because I do hate the Muse Buddy and the Bee Frontier just in general.

I don't want to be... Imagine you're at a party. You're talking to somebody for the first time. Imagine... Okay, sorry. Imagine you were single and you're at a party and you're having a conversation with somebody and... You see the B frontier on their wrist. No, you don't talk to them. Oh, you see, you see it. Listen, it's just like,

when you were single. Okay. When you were single, you know, I'm sure it was very easy for you. For me, it was, I had some difficulty. Okay. Just think about all the conversations, right, of potential flames. And you're like, what went wrong there? You're replaying. Imagine like the B frontier is with you at the party. And you're like, you like have a great time on a Saturday night and she doesn't text back on Sunday.

And like you walk by your Echo show and it's like just a bit too aggressive last night.

We just can't move to this reality. I can't deal with it. Wait, wait, hold on. I actually love that because I have the Echo Show in my kitchen. I use it while cooking, but it is very thirsty. Like it's one of those where– and actually going back to the grocery and everyday essentials, do you know what they do? It will– I mean, again, it's interesting to me technologically like– Because if I shop at Whole Foods, it knows my order.

And then I'll be like, oh, I need a recipe for so-and-so. And it's like, okay, here's a recipe. I saw that you bought these butter and potatoes today. So let me try to use that. Like it definitely does this stuff awkwardly. So if any AI ecosystem would tell you, hey, Alex, a bit too aggressive last night, maybe tone it down. Maybe ask questions about her.

it would be the Echo Show. Alex, just imagine you were a honeypot. Use the Ron John method.

I mean, for honeypots, the Bee Frontier is actually, that's the target demographic. That's where they're experiencing their initial surge in growth. The increase in honeypots directly correlates to sales of the Bee Frontier. trying to lure a multi-trillion dollar hardware CEO into delivering his thoughts directly to Beijing. Use the Bee Frontier to record every conversation and get notes on your Echo show as you cook dinner.

I'm making an AI video commercial at exactly this moment we log off. I just want to correct the record here because I don't think my Bee Frontier in my dating days would have been like Alex Too Aggressive. I think it would have been like you need to not turn into a corncob every time you get up there. Well, hold on. Do you think we're going to have to rethink societal norms around, like I agree, if I sit down and someone, I at least would appreciate the fact that the B frontier is hopefully visible so you know what's going on.

But I feel like, do you have to sit down and be like, are we recording? Kind of like if you're on like a Zoom call or something like that. I mean, isn't it clear that the B frontier, which is this bracelet that sits along your wrist is just not going to be the final form factor. It's going to be a, um, impossible to detect microphone that you like hide under your shirt.

And so we're not going to know who's recording and who is, this is why I just don't want this stuff to take off. And I do think there should be some stigma around. you know, showing up to dinner with your muse buddy or one of these things. Who wants to live in a world like that? I mean, it is true. As someone who loves my meta Ray-Bans because I take photos and video and do it in a responsible manner, it is interesting that, I mean, the stigma around that certainly took off and meta actually removed the photo capabilities in a lot of models.

So society has a shot. We got a shot. Do we? I hope so. All right, let's just like summarize, or not summarize, but I want to see if we can kind of get to sort of a conclusion here on this one. Because if you look at the S&P 500, it's up 10% on the year. On the year, Alphabet is, oh, it's down 11%, right? It was up slightly when I was in with Sundar, and now it's down 11% on the year.

Meta is down 4.5% on the year. Amazon is up 5.5%.

on the year, trailing the S&P 500. Microsoft is up 10% on the year, and it hasn't really been praised at all for its AI initiatives. When you look at it all together, and we won't even throw CoreWeave into the pie, but NVIDIA is also trailing the S&P 500, is the market in 2026 telling us that the companies making the biggest bets on AI are simply too risk-hungry to bet on.

It's saying, we don't want that. We don't believe in it. So that's why I think the connection to how will Anthropix IPO perform is so fascinating. Because I agree, the market has effectively told big technology companies that, listen, you guys are going free cash flow negative. You're investing this much. Good luck. But we're going to invest in the Cheesecake Factory in Chili's, whose stocks are up 10%.

above the overall kind of composite of big tech companies. So I think in the public market companies, it's become clear. But also, these are gigantically valued market cap companies anyways. So it's just saying, well, maybe things will just temper off a little bit. But in the private markets, it has not slowed down. We are still seeing, again, anthropic, and we'll get into the numbers, but talks of a trillion and a half or 2 trillion are still very, very optimistic things that I think, uh, we still see that divergence.

Yeah. Um, Okay, we should take a break. And when we come back, talk a little bit about that IPO, because we did promise a S1 emergency episode. We got that with Dan Pramack on Monday, but we want to hear Ron John's thoughts as well. And we're going to talk... Of course, about the business, but also about how much money the people with an anthropic stand to make, because they're going to be very rich very soon.

All right, we'll do that when we come back right after this. When you're at work, you never know when you'll be interrupted. But with the Dell Pro, powered by Intel Core Ultra with vPro, no matter what distracts you, your laptop won't. It's battery optimized for the way you work, with built-in intelligence that quiets distractions when you need to focus.

Your laptop will help keep you locked in, even when it's bring your dog to work day. Built for those who stay in the flow. That's Dell Pro. Built for you. Dell.com slash Dell dash Pro.

This episode is brought to you by AvePoint. Everyone's racing to roll out AI right now. Copilots, chatbots, agents doing real work. But here's the part nobody loves talking about. All that AI runs on your data, and most teams have no single way to see it, secure it, and prove it's under control. That's exactly what AvePoint does. For 25 years, they've been the trusted layer beneath the world's most demanding data, now extended across your entire AI estate.

your data, your cloud, and the agents acting on your behalf. It's how more than 28,000 organizations deploy AI with confidence. So innovation scales without scaling risk. AvePoint, the unifying trust layer for AI. Head to avepoint.com to see how enterprises deploy AI with confidence. Learn more at avpt.co slash big technology podcast. That's avpt.co slash big technology podcast.

As a small business owner, sometimes it feels like no matter how much planning you do, there's always surprises, like an urgent, expensive repair. But here's a surprise you will like. With Progressive, small business owners save 10% or more on their commercial auto insurance when they pay in full. So enjoy a surprise for once. Get a quote in as little as eight minutes at ProgressiveCommercial.com.

Progressive Casualty Insurance Company and Affiliates. Discounts not available in all states or situations. And we're back here on Big Technology Podcast, Friday edition. It's Anthropic S1 week. Ranjan, it's interesting, we're about five days past. I don't think we'll spend too much time on the S1 this week. But it was really a Rorschach test, I think, for people in terms of if you believed in Anthropic, you found plenty of evidence to show you that the company was very healthy.

Booming growth. They went from fifteen hundred customers in December twenty twenty five spending one hundred thousand or more to eleven thousand nine hundred in June. Right. So that's massive increase. They also increase their customers spending a million or more. They had something like one hundred and ten billion cash on hand. So if you thought Anthropic was just going to incinerate all its money and desperately crawl to the IPO, Obviously, that's not what happened.

They're making $6.6 million per employee. So, you know, they're very efficient. Their efficiency was going up. They used to make $0.40 for every dollar of compute they spent. In the S1, they showed that they are now making $2.08 for every

dollar of compute they spent. So they completely flipped that on its head. But the detractors will point to two things. The fact that Q3 revenue was not included in the S1 and Q3 revenue looks like it will massively decelerate from something like more than 100% quarterly revenue growth to in the 40% quarterly revenue growth field. And of course, that spells the end of tokenomics and then There's also the half trillion or so that they've committed to spend, much of which is non-cancellable.

When you looked at this blot of ink on the paper, what did you see? What I saw was a business that is doing very well. And I think like actually... met a lot of the heightened expectations. What I didn't see, and the most frustrating part of this to me, all I want is to understand the unit economics of Anthropic's business to give a reflection of what a great AI business can look like.

And I did not get that. And I'll tell you why. So I actually don't think the Q3 numbers, again, a lot of people are There's a lot of conspiracy theory-ing going on, I feel, around... It actually makes sense. Audited financials, a quarter would have ended September 30, so they're going to release it based on... I think it's November 11th. If they don't IPO before then, which they likely won't, they're going to have to release Q3.

So we'll get those numbers. We'll see what happens. To me, the numbers actually were so good... that that actually raised questions for me. And what I mean by that is, again, I mean, you kind of started to say, but like the whole story has been that like cost of compute is gonna be so dramatic and that's why we need to just raise so much money.

And that's why we need like, it's a very expensive thing and that's why this is all gonna be very expensive. they lowered their cost to compute as a percentage of revenue and Q1 was 71.6% to 48.1%. It was 225% of revenue a year and a half ago. So like,

All these numbers show us that actually the economics are almost magical. Like compute is decreasing, revenue is increasing, compute as a percentage of revenue is decreasing. To me, I don't think we yet understand what the actual true cost of compute was relative to their revenue. And the reason being the most glaring example was like, There's all this talk around the $80 billion loss.

Is it real? Is it an accounting treatment? Within that, the big part of that was it was the increase in value of convertible notes from Amazon and Google. But we talked about this a year and a half or two years ago, actual calendar year and a half, two years ago, not... Alex's six months is a year framework. I just want to say, I don't mean six months is a year.

I just want to say that when people say six months is a year, when people say that happened last year, they really mean it happened in April. They've lost the ability to keep track of time. Well, the only reason this was actually related to, I think it was like summer 2025 convertible note issuance. But remember, we were actually, this was back in late 2014.

Amazon's putting in 4 billion. They openly are saying part of this is heavily discounted compute. A lot of the deals with Broadcom, with all these other people that are both investing into Anthropic, they're receiving significantly discounted compute. And it was just a reminder, I think what we're seeing here is not reflective of what this is actually going to cost.

And we're still not getting a good representation of like, what are the unit economics? That's my only frustration. I think like, which also is going to be really important because when we talk about these 500 billion, what is it, 325 or five, whatever the future committed spend is going It's in the neighborhood of $500 billion, yeah. Yeah.

We still don't know what does Anthropic's business look like. If they were paying market rates for Compute relative to their revenue, we know they aren't. But we don't know how much that discount is. So my ask for investors is to demand that they disclose and provide a better understanding of actually what they were paying over the last number of quarters and year and what kind of discounts they received.

Otherwise we still don't know what is the actual kind of like picture of this business. But wouldn't you agree, by the way, that if you think about the amount of money that Anthropic has raised, they still have $100-plus billion in cash. So it couldn't have been that much, right? No, no, that's exactly what I'm saying. It's actually insanely efficient.

For everything we've talked about for the last two years, suddenly you look at this and you're like, wait, this is the most cash-efficient, compute-efficient thing imaginable. The number that they've only burned $9.5 billion in this entire last few years is actually... We know it's not exactly correct because they receive below-market rates as part of these investments, but we don't know how much less.

And we know they're not going to continue, right? Let me press you on this. Google said it was going to invest up to $40 billion in Anthropic. We don't know if they've invested all $40 billion. Amazon said they agreed to invest $25 billion. So let's say that was all Amazon and Google giving Anthropic compute for free.

know you that's just 65 billion dollars they have a hundred plus billion in the bank no no but but they spent five and a half billion in compute last quarter three and a half billion in q1 eight billion total in the first half of the year so like if they're investing if google and amazon are discounting by in that order magnitude of billions that fundamentally changes the margins of what anthropics business is So the cash on hand, I think, is separate.

This is what should they have paid and what would their actual expenses ratios look like relative to what they are able to charge. So your suggestion is that Amazon and Google both invested in the company and, you know, not dollar for dollar, but just said, we want you to grow. And so we're, you know, outside of our investment, we're also going to provide you with discounted computers.

I mean, Amazon, that was explicit in their convertible note. Like they said, and we talked about this way back and it was kind of a joke. Like we were like, and back then I remember we were like, is it cash or is it credits? Is it like, but now you're actually seeing the result of that is we still don't know of what is Amazon business going to look like going forward.

And I think to me, that's the only, otherwise everything looks so good. It's almost like unbelievable. And that's why, to me, they should disclose around here is actually like our average, what this discounting looked like in the past, so people can understand what it will look like in the future. I know it would be like kind of criminal to do this, I guess.

I mean, I don't know if we've seen weirder things, but is there a chance that Amazon is, you know, giving Anthropic this discounted compute, but booking it as full revenue? And that's why you see the 39% growth? Is that like a totally out to lunch conspiracy theory? Probably. I mean, actually, I don't know. Like, I'm sure could have. I feel that that would probably be pretty aggressive on the accounting side.

I wouldn't worry about that as much. Again, to me, this is not... The numbers were great. But to invest and at what rich valuation are you investing? That's the question. I think they have not given enough information. Also, I have to say... Whoever is planning those revenue leaks is a communications genius because, again, there was no surprises.

Everything was out there. When we get $16 to $17 billion for Q3 revenue, it's already out there, and we know growth is decelerating, and all that's out there. So there'll be a little hubbub, but they're leaking financials in a pretty incredible way. Yeah. Somebody on X named Andrew Benson started putting together the list of who gets rich, a philanthropic list at two trillion.

I think the talk is a list at at one point five. But if you assume two trillion, it's one hundred ninety nine billion. This is the equity breakdown to the seven controlling founders. Then there's twenty eight billion for Ben Mann. So I guess they would all get who's another. Founder who, I guess, sometimes a founder, sometimes not, so they'd all get 28 billion.

100 billion to the Anthropic Public Benefit Fund. 331 billion in employee options and RSUs. That covers 5,300 employees, an average of 62 million each.

But he continues, when you joined really matters. And again, this is back in the envelope math. I think it's directionally accurate, but of course, there could be some problems with it. But if you joined in 2023, you're looking at $110 million to $400 million each. If you joined in 2024, he says, which is like 440 to 640 people hired, $200 million to $300 million each.

2025, at least $70 million. 2026, $10 million to $20 million. Imagine you got hired this year, you're already looking at $10 to $20 million in compensation. You wonder why people don't leave Anthropic. Of course, part of it is the mission, but this is also a pretty good argument, don't you think?

I'll admit I can't even stomach hearing these numbers working in the IA industry for a company that's doing well, and I'm happy, but like... My God, the wealth creation is, it's actually, what's more surprising to me is what you just read. I'm actually more surprised that we got elections on Tuesday. I'm surprised that has not become like one of the biggest headlines and like every, if you're a politician out there, you got your talking points right now.

I mean, the numbers are insane, right? Listen to this. Anthropic... By the way, this is also coming from the company whose CEO said that half of white-collar jobs will be... Yes. Just play the two clips. That's it. Yeah. This is from Sam Parr. Anthropic will make 1,000 employees worth $100 million, at least $100 million, at a $2 trillion market cap.

There are 11,000 Americans worth... with $100 million in investable assets. So something like 1 in 12 Americans with $100 million liquid will be from Anthropic. And if you add in OpenAI... It could be one in seven. It is shocking. Again, when I saw those, I literally, I'm like, I'm, yeah, I am speechless. I have two reactions to it. I mean, of course they've earned it.

They've built to a degree, right? They've built this technology that is, you know, going to be, you know, has a good chance to be world changing, put it that way. But the other side is like, they're also telling you that it might kill you. And like, they're going to protect you. And it's like, Really? You're saying it's going to potentially wipe us all out and you have to develop it because you can develop it safely? But then you're also walking away with something like the richest people in history? No, but that's what I've always said.

And that's actually, again, why to politicians, you got four days right now. This is a pretty easy talking point and you might as well push it. But I also actually on the... existential risks to humanity. I also have to admit, I actually wasn't that interested in the whole risk factor side of it. I know I saw you tweeting them out and you were discussing them with Dan, but to me they were, I don't know, they almost just feel like they're just fucking with us.

Dude, I have a conspiracy theory on that. All right. They had like the longest risk factor section I've ever seen in my whole life. Like I tweeted 80 of them. And then I was like, I tweeted one post with 80 of them. And the funny thing is I stopped at 80. I don't even remember how many there were because there were so many. And I think what they did was they wanted to bury the existential risk risk factor in like the most boilerplate shit imaginable.

So you would miss that and just be like, oh, well, they have all these other capital risks. But we know the existential risk and everyone covered the existential risk. Like, I guess to me, that was like, normally S1 risk factors, the whole point is to like recognize risks that exists that especially likely disclose things that investors don't already understand and like legally also that you are recognizing.

So, I mean, it has to be in there, but like, I mean, he's been saying it for years. They've all been saying it. I'm just saying it was so long that they were just like, all right, you'll get exhausted and no one will pay it. Well, anyway, I think you're right that they have, it's been talked to death. All right, here's a risk factor for us to talk about before we leave today, which is the price.

You know, we talked about a price war coming for LLMs, and man, has it happened. This is from Max Wolf. OpenAI and Anthropic cut their LLM prices by up to 80% within a few months. Why? I'll just read a little bit from Max. July 30th, 2026, OpenAI announces that they're going to cut GPT 5.6 Luna by 80%, the prices. August 10th, 2026, Anthropix says their temporary Sonnet 5 price cut is permanent.

August 21, 2026, OpenAI announces they are cutting the price of GPT-5.6 sold by 20%. September 2026,

OpenAI announced GPT-6 Sol at a shocking price that halves the input and output price relative to the temporary discounted GPT-5.6 Sol. They simultaneously released GPT-6 Luna also at half the input and output prices of the previous version. On the same exact day, Anthropic announced Cloud Opus 5.5, and the total cost reduction for that, Max calculates, was 40% to 60% for a model that's even better than Fable 5.

This is not just a price war. This is not just the, sorry, this is not the rumor of the price war. This is the price war. Some reasons for Max. He says, if you've been testing agents with new LLMs as they are released, it can be very difficult to differentiate them outside of subjective vibes. Since last November, that launched the hypergrowth of Vigentic LLMs.

The models have become more of a commodity, and it's now trivial to try a new model on your code base with just a toggle harness like OpenCode. or using something like OpenRouter. As a result, there is little moat. That's just part of the reason that he gives. We've talked a lot about the commoditization of LLMs and how that could lead to a price war.

Is this it actually happening, Rajan? I mean, that's the other part again. The moment we're in, in the next few months, we're already seeing the actual cost. I mean, we've talked about the cost of compute. We've talked about the actual ability of them to charge. But also, I guess the... If the aggregate demand increases and everyone is agentifying everything, this does become moot.

And I think that's what they're going to have to kind of pitch on the story. But I agree that this doesn't help the overall business story that like, let's say frontier models at high margin are going to be what differentiates them. And the main thing, I still don't like... Again, in the enterprise side of it, the number of people talking about or working on routing models across whatever the lowest cost best model is for the task is definitely increasing.

So I think it certainly presents a threat, but it hasn't affected them yet, it looks like. Another potential interesting reason is infrastructure. Here he writes this. Another less cynical factor that could explain the price cuts is that LLM infrastructure has improved over the past few months. When Anthropic said Opus 5.5 requires less compute to serve than Opus 5 and its pricing reflects that, I believe it.

One outcome that would be genuinely ironic is that if these models are more compute efficient, it reduces the need for these massive data center build-outs, which could cause cascading economic failures. Interesting. Your thoughts? Yeah, I mean, it's a great point because right now it feels like the... We just said the entire first part of this podcast was around big tech and data center investment.

And the economics of that, no one has any idea. And if everything becomes far more compute efficient, which it looks like it is becoming... that fundamentally changes the entire data center story. Again, if the whole world gets agentified, maybe it all works out. But I think all of this, there's just so much risk. And clearly the market with big technology firms has already recognized it.

But it's a bit terrifying, cascading economic failures. Don't like the sound of that at all. I mean, the other, the final explanation is like the IPO is coming and they want to show usage increases, but like, I don't know, they also, they really want to show revenue increases. So I don't know. Sometimes companies try to show real growth ahead of the IPO, but I don't think tokens is the growth that Wall Street cares about.

No, no, it's revenue and low costs and mask the true costs as much as possible in a gap compliant way.

Okay. A couple of things before we leave. First of all, I love that you said Cheesecake Factory earlier in this episode because our SEO is now going to be off the charts giving the Cheesecake Factory stuff. But no one actually searches anymore, so it doesn't matter. That's true. Okay. You got me. You got me there. Do you want to spend a couple minutes on You Can See Everything?

I haven't seen it yet. I have not seen it yet. And I really want to see it. I was like, I read, how do you say John Kerry Rue? John Kerry Rue. The book, Bad Blood. Yep. Kerry Rue? Yes. I was definitely a bit obsessed with the idea of Sonny Balwani back in the day and the whole Theranos story. Why Sonny Balwani? He was the most fascinating character in the entire story to me, just because he was mysterious.

There's almost nothing on the internet about him, even though he was a co-founder of the company. And I don't know, just their relationship, it was just so weird and odd. So I still, I think Sonny, the evil of Sonny is not commonly recognized. Is he in the movie? He's not in the movie. I won't do any spoilers. All right. But it's a crazy one.

I have to say it is a crazy one. I enjoyed it very much. Okay. I'm going to challenge you to explain what makes it so good without spoiling. Go. Well, if you have watched anything from Nathan Fielder, you just know that the guy's mind works differently. And-

Part of it is him and them trying to figure out why they brought him in, in specific. And then, of course, he takes it on Nathan Fielder-style divergences from the main story. I don't know. I quite enjoyed it. I'll say this. It's the least absurd of anything Nathan Fielder has ever made, which is crazy. Okay, so that's what makes it truly...

different or why everyone's reacting so strongly in a positive way. Yeah. Well, another thing is that Elizabeth Holmes is this, she is, you know, and this, I'll try, really, it's hard for me to say much more than this without spoiling it, but she is, you know, by her own doing or by society's and probably a little bit both, a media creation, right? And how often do you get to see someone who is a media caricature effectively, you know, basically with given complete access to cameras, even with a weird.

All right. I got it. I got it. I've not been able to. That's interesting to me. Convince my wife and Amazon obsessed seven year old son. to go watch Nathan Fielder yet. I know you texted me, we should do a big technology private showing of the movie. We really should have. That's on me. Well, actually, that's what I heard, that there's like VCs out there doing kind of like private portfolio company showings, which I don't know how I feel about that anyways.

That's kind of weird in the Theranos context.

I would feel good about it. Those are fun. I don't know if you've ever been to one of those, but I've been to one. I enjoyed it very much. Okay, okay. All right, man. Well, I know, you know, though I plan to take it easy tomorrow, I now have an idea for my Halloween costume. And for those of you who are thinking that I'm going to be part of this Elizabeth Holmes, Nathan Fielder costume thing, no, I'm not.

I'm going to be the golden egg. So if you see a golden egg running past you in New York on the first, just say hi and be like, I listen to the show.

I love it. The first Guinness Book of World Records, first golden egg to ever complete a marathon. But you know what? There is, sorry, I don't really want to go belabor this, and I know you need to go, but there is a saying. You don't kill the goose that lays the golden eggs. Let's end there. Let's end there. Let's end there. Great to see you, Ranjan.

Thanks, everybody, for listening. We'll have Tristan Harris on the show on Wednesday to talk about AI risk. Thank you, Ranjan. Thanks to all of you. We'll see you next time on Big Technology Podcast. For small business owners, every day is full of surprises. Like when a client just cancels their order at the last minute. And you have to eat the costs because they canceled at the last minute.

But one surprise they'll like. Progressive provides 30 customizable coverage options to help their business keep going. So surprise yourself. Get a quote in as little as eight minutes at ProgressiveCommercial.com. Progressive Casualty Insurance Company and affiliates and third-party insurers. Coverage is not available in all states or for all vehicles and coverage selections.

Now for a quick message from Anthropic, the team behind Claude, the AI built for businesses. Imagination doesn't need permission, just room to play. Claude is an AI that works alongside you to give your ideas the space they need to grow. And then it helps you make them real and ready to ship. Discover more with Claude at claude.ai.span.