Transcript
0:00 All right, guys. Welcome. Let's talk about the anthropic as one, >> once in a lifetime event, biggest day of human history. >> Certainly in the history of capitalism, right, Joey? No, no hyperbole there. >> There's zero hyperbole. What a what an amazing IPO it'll be. how capitalally efficient this growth curve was, how fast it happened. you can't ask for more.
0:35 >> They're going to come in at like the what sixth most valuable company from the start of trading in the IPO in the whole world. >> It'll be up there, I think. Yeah. Like tokconomics, I think, covers like seven of the top 10. >> Yeah. >> So, this will be this and open AAI will be trillion plus. And >> okay, what do you guys want to talk about for profitability? I >> I think that's a good place to start.
0:58 you know compute you know they they kind of give the compute margin numbers but I think even on the gap basis you know non-GAAP basis you know 10% plus margins already two quarters profitable really solid I think what we're seeing on the gross margin line just the past three years like the scaling of compute and inference efficiencies is amazing and you know what they're able then to spend and and increase you know kind of the training splits you really impressive. So 85% inference gross margins, you know, 41% compute margins, and then extremely capital efficient like sales and marketing and partnership. thing too, I think a lot of like software and tech companies spent a ton to go acquire a customer and you know their net dollar retention number. So a customer last year at this time that spent a dollar is now spending nine. And in Q1 that was, you know, that was 500%. So customer spending one was spending five. and that all falls just down to the bottom line. There's not much you have to do to upsell the usage base side of that. So, really solid headline numbers all around the business model.
2:10 >> Yeah, I would say really really interesting to me because from if you talk to folks in the in the data center market, right, I'm the data center bro here. folks have been looking forward to this IPO for you know one big reason we chose that entropic has the biggest compute demand of anyone alongside open AI they're on power with you know with hyperscalers and they can't deal directly with the folks actually building the data centers one because maybe they have too a team too small they're building but two is mostly financial they can't can't sign like you can't fund the data center on anthropic credit and so we we saw the market kind of hating on anthrop ropic and open AAI like naked exposure whether it be on the data center side or on the on the GPU side over the last few weeks and so it's I think it's going to be like very reassuring for folks to see the actual financials so the first thing I wanted to say is like talk to anyone on the data center market everyone was just looking forward to this IPO like dude whenever these guys are going to IPO I'm going to sell so much so many megawatts to these guys it's all going to be so easy obviously that's contingent on anthropic big investment grade. And so when you look at some of these numbers, how much cash they have on the books, you know, like over hundred billion dollars, how much they're going to have in addition with what they're raising and the promises they have from like Google and others, these guys are going to have over 200 billion. And as Joey said, that's a company that in the last three years, despite being like very unprofitable, barely burnt any cash. So I think extremely reassuring. also the compute commitments frankly aren't that large. which clearly shows you that they've stroked some good like fairly flexible deals with folks like Amazon, Google where instead of having to commit for like gigantic deals like like Stargate which is this huge you know $300 billion agreement for 4 and a half gigawatts they've managed to do something like a bit more efficient on the commitment side which I think is overall great good news and you know good for AI infrastructure as a whole market. Jeremy, can you walk us through I think a lot of people see like the headline, you know, 500 and 600 billion of compute commitments. Can you walk us through why that's not that large?
4:23 >> Yeah. >> No, it's a good like people ask that and it's when you break it down, it's it's not that much. >> Yeah, for sure. I mean like if you look at recent deals that have been done such as Anthropic Endcale, let's take this one. the compute per gigawatt commitment is $und00 billion. And so there's a world where you would have to for like these half a trillion dollars would be just 5 gawatt which is really not enough for anthropic to serve their road map right like what they have in terms of gigawatts contracted is actually materially higher than that and so their effective like ratio of of of gawatts contracted relative to commitments is really low and I think you know this this gives us like right now the compute they have is like under under five gawatts that's going to go up materially as we get into the the end of next year.
5:15 Plenty of compute to, you know, accelerate accelerate revenue growth, cut cost of models to keep finding efficiencies. And so it's pretty clear all of that is going to trigger an acceleration of revenue. So these guys are going to be in the multiple hundreds of billions of annual revenue next year. And so yeah, it doesn't feel like that that that complicated for them to pay half a trillion dollars given the revenue they're making, free cash flow profile, strong balance sheets. So I think overall like very reassuring especially when you compare it to like you know what was the first freak out about compute commitments that was last year that was open AI and people were saying dude you're making you know 20 billion annualized and you have like 1.4 four trillion, right? So, I think we're in a much more reasonable ratios here. Like, you know, 80 billion, half a trillion commitments, it's a lot, but there's 200 billion on the books post IPO. So, overall, I think, you know, it's going to be fairly easy for them. And as I was saying, like they they've done some a mix of like long-term deals like Ncale, but also more flexible short-term deals, hypers skaters, and folks like SpaceX, which they can cancel effectively anytime.
6:22 though overall 30 years I go today. Can I >> greatest day in the history of capitalism? >> I'm I actually I just made that up in terms of history of capitalism but it felt right. I mean like look it's a big business dude. It is a big business you know like I remember once upon a time you know one one billion run rate was like the the grownup number and they obviously did a lot bigger than that. I think the thing I thought was best on that is like incremental margins are are 90% gross margin and 40% eat it. And that's that's like that's a best-in-class business.
7:02 Like 40% EBIT is as good as it gets. That's like as good if not better than the hyperscalers in terms of like if they get there right gross margins at 90% incremental. Like it's very clear. It's on a very large base. I think it's as clean as you possibly could want it to be. And I think people, we've been saying this for a long time that there's obviously really great incremental like there's real revenues, right? Like when you sell it, it's you're not selling a dollar for a dollar10 cents. You're selling a dollar for 10 cents. Okay, that's margin, right? I think the biggest bears have always believed that there's a heavy subsid subsidiation. Sorry, I can't say that word today. and it's very clear there isn't. So like I don't know what else you want to say. I also think the fact that Ed Zitron hasn't tweeted is like probably the most interesting fact of all. Like, you know, you know, like him calling Jensen a crook last week was like pretty based in terms of like, you know, he's kind of like you're watching the speedun of a short seller in real time. so yeah, we'll see. I hope he's I hope he's mentally okay. You know, might be a hard day to be ed.
8:08 >> Yeah. So, inference is a good business model. I guess that's >> Yeah, it's a great business model. >> Yeah. Controversial >> controversial people actually want these tokens. How about the how about the customer concentration? I mean there was a comment in there saying that top five customers represent 25% of revenue now and then the CSP commitments the marketplaces are representing another 50%. So what's the leftover 25%? I mean it's just a really big base. So the leftover 25% is just still many billions. But >> I'm confused about what you mean. I think you think that it's either or mutually exclusive. 58% is the total and they it's like the total distribution and then 25% can look through that if that makes sense. They they don't add together. They could they could like 20 like you know 5% of that customer the entire top 10 customer could be 100% on the CSP. You know >> you think that some of the top five or top 10 customers are actually going through the CSP marketplaces. It implies that the numbers show that right because the see to to be clear I really want to make sure Jordan we get this straight on the podcast that 58% of distribution is like a like their sources right but I don't think when when they're talking about their net customer they are not talking about the CSP as the net customer they're talking about the look you know I'm assuming there is some kind of pass through meaning that when when anthropic gets built they get the customer who purchases through the distribution platform aka bedrock Right. So, >> okay, let's just cut this part out where I'm being stupid then and not talk about customer concentration.
9:40 >> Well, no, we can talk about the 5%. >> No, no, we we we should talk like Jordan. >> No, I mean, dude, the fact that you didn't you like it's important to know. People don't know that >> it's just the top five doing 25, which is pretty good. Honestly, that's most the companies are actually more concentrated than I know the top five are doing the 25 and then there's an additional >> 48% that goes through the CSP marketplaces which does not include the top five customers because elsewhere in the S1 they also talk about how they have a direct relationship with people who are making purchase commitments to get a discount longterm like any other.
10:13 >> Yeah, Doug, I think the point he's making is that like >> the 48% of revenue, none of the top five customers are going through bedrock or vertex. >> Okay, great. Surprise. So then there's >> which makes sense like >> Yeah, that kind of makes sense. Yeah. Why would you >> those customers would buy through that channel? >> So then you have that you have the CSP marketplace >> which we know is like going towards your traditional buyers, you know, >> just your your large AWS accounts, your large Google GCP accounts that are buying through that, you know, a little more Fortune 500 and then you have another so that's 48. Then you have another 25 and then there's 27 left over. and I think even the point there and there are some really interesting stats on like some of these customers is you know there's there's actually it's I think that's still rather concentrated.
11:00 I think if we look we still look at any you know alternative data set the top 1% of customers make up you know 70 80% of the panel and they said this you know the top 5% of cloud enterprise customers users I should say at these end companies are spending 18,000 per user per year. your non-Cloud code user spends less than 75 a month on on on >> you're talking you're talking to three maybe four of the top 5%. Come on.
11:30 >> Well, I was I was wondering there's 250 10 million customers on a trailing 12 month basis. We I I know we're in the million plus group the 1900 one of 1900. We've got to be getting close at some point. >> We're one of 1900. Yeah. >> Yeah. At some point we'll be in that 250 I would think. >> Yeah. some point. >> Yeah. >> So >> that's not a if Dylan's listening to this that this is not a KPI but >> not a result it's a result of some other KPIs being enforced. Not that we have KPIs. okay but talk about the actual customers they named customer ABCD. Somebody started at 43% of their revenue is went below 10% is now buck back up over 10%. We've got our estimates for who those people are, but let's just say >> like you can't look at 23 numbers, I don't think. Like there were customers at OpenAI that were over 10% in 2023 that are nothing today.
12:25 >> >> so I don't know if the 2023 numbers I mean we know some of those large customers. you know today that are you know there was a one I think 10% customer and you know >> who do you think who do you think like let's let's do some speculation. My meme was that it's Kimmy Mimo GM that was five to seven%. >> Yeah. Yeah. Yeah. Five star% customer. >> That's Well, okay. To be clear, they're they're I mean, even I I I don't know if their their KYC is good enough to completely cut them off, right? I'm sure they're I'm sure someone at at Moonshot is using quadco. Like if you saw if you saw on the the bench I forget which benchmark I think on DeepS seek they're using cloud code to be the final verifier of of like you know to verify all the results. So they're using it obviously but maybe not distilling.
13:20 >> Yeah. >> So who do you think top five is what I'm what I'm saying like let's speculate. >> We know Meta was the one that you know rationalized spend in Q3. >> or not rationalized spend they moved from an external provider to internal. the tokens are still there from an industry perspective. >> I think Microsoft is one cursor we talked about as being another one. >> >> Nvidia hedge funds like James period HRT. >> Do you think Nvidia is that big of a spender?
13:47 >> Yeah, definitely. I think they're probably the 43% that went lower. Who else is using this much stuff in 2023? >> I mean I my understanding is they were really late to getting to to getting access to Quad. They were a cloud adopter in 24. >> I don't think they were using it. I don't think Nvidia was using it in 23. That's my impression. I might be wrong, but I I feel like Jane Street probably feels like the most rational or maybe you can argue Google or something.
14:15 I don't know. You know, >> I mean that would be crazy. I know Jane Street is producing a ton of cash flow right now, but they they have such big commitments with OpenAI already. It'd be crazy if they were also a top five and so big on on Anthropic as well. But I mean, it's totally possible. There's only so many people that can be spending tens of billions of dollars on these tokens. >> Oh, well, actually, no. I was going to say I feel like Yeah, I was going to say I don't I don't know. It feels We're going to figure out who it is over time, and I bet you it's going to be like someone really obvious like that we just didn't think about essentially. I mean the point the the point of clarification for me is that like these top customers are not the CSP marketplaces because they have this direct relationship with people through the CSP marketplace if they're using Bedrock or Foundry or Vert.Ex. and I think that's also pretty interesting because if people end up having this direct relationship with Anthropic and they make these purchase commitments ahead of time the way they do with like an AWS MSA agreement, that's a big lock in for using those models. made another chart. they they said that in Q3 20% of the revenue was from models released before April.
15:33 So that's Opus 4.6 on 4.6 4.5 which they kind of said okay 80% is the latest model so people move to the latest models but I'm actually shocked that 20% of revenue is still stuck on models that are three or 3 months old. Like how long does it take somebody to get off of the old models and onto the new ones? And I would think that somebody buying through bedrock like a Fortune 500 who has to certify something, they're going to have that lag time, whereas a startup's going to be a little more nimble to move around. And so both of these things just point to me as being like really sticky.
16:04 >> I mean, if we're saying Microsoft is a top five customer, then they're probably the one spending on all the Opus 4.5. Yeah, I think via via GitHub Copilot also, but you know, Copilot 365 or something like that, like the cursor perplexity equivalent where they don't pass through the relationship to the end user. They it's kind of their API key that's doing all the spending. Microsoft has both the foundry business where they do pass through the relationship and then also the GitHub copilot stuff where they consume so many tokens. And you can probably see if they had more granular data, you could probably see Microsoft spend be high at the beginning, dip down when they have all these OpenAI options, and then come back up now that they're actually more motivated to serve anthropic and not tied so closely to Open AI. We'll see.
16:52 >> Okay. What else What else is hot? What else is hot about this S1? >> Compute commitments. Jeremy, talk about the $70 billion MPA. Oh, sure. I mean, they're just essentially going to get even more money than they already have, which is mostly good news. U a bunch of them are already known if you if you look through the filings of like Google and Amazon, but they they they've signed a b a bunch of like large fundraisers.
17:19 for example, the 40 billion with with Google which have like multiple multiple milestones, right? So like Google has a labor 20 billion. There was a remaining 20 billion which is just going to be more cash for anthropic based on milestones such as you know going public. so effectively like all of that is just additional funding. which which which again like as I was saying earlier just means these guys are going to have more money on the books going to be easier to pay for stuff.
17:48 they got the Broadcom one as well which is $42 billion which going to enable them to buy a whole lot more whole lot more TPUs. And so yeah, overall like the 2027 compute road map for for anthropics risks which kind of goes back to what we what we said like in some in an article on a podcast a few few weeks ago about like the data center built out like yeah this is most of what they're going to get next year is derisk and it is going to be huge. so the the implication is okay they're going to have all of that all that compute next year. then can they actually accelerate the the revenue enough, right? And I think that's still still the question. but may maybe one one thing that folks weren't aware of is like how much do they have on the books right now? And now we know and it's it tells you at least they're not going to go under.
18:35 >> Yeah. And if rates go up, they'll even make more money, baby. You know, >> hell yeah. make, you know, rates rates are five per let's say rates are 5% a year. They're making 5 billion AR, you know, >> that gets us to one trillion almost, Doug. >> Yeah, dude. Just just a little bit of the way. I honestly the thing that I really thought was most interesting in the whole S1 is the crap that they put in or and like honestly I think I think they've been pretty consistent about this. Dar's been saying this for a long time. It's like we never want to overextend ourselves, right? And then essentially just being like we'll never miss your expectations is an insane thing to put in the S1. Like that is insane. I think this the most nutty thing they could possibly do and then everyone is like really worried and barred up about Q4 because what they've been telling the bankers for ARR and you're like they just told you they're never going to miss your expectations.
19:29 >> They're like we've talked to all these Oh, I think Joey's frozen. We've talked to all these people and all these stuff. You should never be worth one more dollar than what you're actually worth. And so we'll underpromise and overd deliver. I was like, "Oh my god, dude." >> Yeah, that's a core tenant of the business. Underpromise and overd deliver, Doug. >> Yeah, I guess that that was that maybe that's that's in their fit interview, dude. I mean, it's crazy.
19:49 >> All of this talk about Okay, wait a second. So this is a core principle of the business. So all of this talk about cyber security ending the world and drugs and all of this stuff. This is them underpromising for the AGI that they're forecasting 10% chance of death of humanity. This is underpromising. >> Yeah. Apparently, according to Dario, >> it can get worse. >> It can get way worse. It's the wor You saw Did you guys see the the head of cyber security? He's like, "If you think it's bad, it's gonna get way worse."
20:19 >> I'm laughing. Scary, man. >> It is actually scary. This is going to be our last secure podcast. I was I was yeah I was freaking out last week with all of the cyber security stuff and then interesting stuff over the weekend as Elon and Chimoth are calling the for the death of all Neoclouds. and then we have this counterweight to all of that which is like we'll save it with capitalism. I guess that's the thesis of today.
20:46 >> I don't know. Yeah. I mean the money the money looks good to me. I I I think Elon is just Elon's smart and he knows when he has an advantage and he's trying really hard to lower his cost of capital at SpaceX and I think he's I mean he is a Neocloud dude so he's saying that while being the largest Neocloud in the world so yeah it's kind of book >> I would also say like our our view and I think it's starting to like folks are starting to be really worried about that is hey there may an inversion of like right now it's data centers that are more constrained than chips but maybe by the end of next year it's the other way around and there you have this guy that is says hey I'm singlehandedly going to build up to 20 gawatt maybe just 10 you know so he's he's going to make sure he has the chips right and so >> that's the game of Elon dude >> one of the ways to make sure others don't buy the chips is just better make them go broke Yeah, >> just break markets. They can't fund them all of them for you.
21:51 >> I mean, that's happening like right now there's essentially a liquidity run on all the Neocloud debt. And so anyone who wants to place new Neocloud debt isn't able to. And so if he presses, you know, if he presses his advantage, like he presses his foot at their throat and they essentially never get funding, that's all his opportunity, right? Because then everyone's like, "Dude, I want to invest. I want to invest. I want to do this SPV." And SpaceX is the only game in town. And he's like, "Great.
22:16 That's a better cost of capital for me, right? Elon is a competitor. >> Yeah, exactly. And then as he said, he wants to do half a trillion dollars to a trillion dollars of ARRO next year, the less these guys can fund their stuff, the more he can just build it himself and score deals at 50 million bucks a megawatt a year and you know, just make a whole lot of money. >> Yeah. I mean, Jeremy, how real do you think it is that $14 an hour GB300 contracts that are on demand and can be canceled any month just continue to go because Anthropics are just continues to grow and they can't bring enough compute online and so they just keep it.
22:56 >> I mean, look like there's there's many things that go into like understanding how sustainable this is. but one of them is if if if the market first of all, it's the basic supply versus demand. if if demand continues to outperform supply growth, we've we've said very publicly we think supply growth is going to be huge. but hey, what if ARR and overall demand also outperforms that? then that's perhaps even going to augment the premium like that's a realistic scenario because who else who else has as much capacity available in in 2027, right?
23:34 now if the equilibrium stays the same then I think there's still probably there's still probably like availability to sell and obviously like if if demand starts to fade then of course like things are more complicated. Now there's a few other things to look at. one I mean we talked about compute commitments kind of the beauty of SpaceX contracts is it makes your it's like much more balance sheet efficient or like compute commitment efficient than signing a six-year deal right so you saw at scale at like$100 billion per gigawatt that's likely going up so perhaps going to 120 or more and and and this guy actually 50 a year but 90 days so the commitment actually is extremely small, right? It's like $15 billion a gawatt.
24:22 So yeah, it's more more expensive. Margins go down, but on the other hand for your for your your books for what people are looking at, it's actually quite efficient. and the last thing is obviously like as the drivers of this market remain like these small companies that don't have yet full trust of of creditors, of lenders, it's going to be hard for them to to to fund an insane amount of supply. I mean even for hypers skitters is not that that easy, right? Like we saw the tiny tiny cracks on like meta for example when spreads went up. So obviously like this environment is very positive to companies that can take a lot of risk and these guys are basically taking more risk than anyone else by just equity funding all of that stuff.
25:09 Some some as well but and you know building stuff without a customer. >> Yeah. Let me throw this chart on screen. Joey, what's your take when you saw this chart that they put up about compute efficiency? >> I think it's great. I think even looking at just inference compute efficiency, the revenue per megawatt just in inference going from, you know, like 16 million a megawatt last year to 60 plus, you know, currently. it tells even more the story of how great a business model just inferences. and this is the entire fleet. So, what you need to do to continue to invest in new models.
25:44 this is great. Obviously in Q3 it went down a little bit with some of the new SpaceX compute coming online and then a little bit less monetization via Meta distillation and then some additional safety compute. But I think this really tells the story of hey AI model businesses are really solid and they scale extremely extremely well. To Doug's point incremental margins are amazing. >> Yeah. Let me read the line from the S1 that jumped out to mo most to me about compute efficiency. They said, "Our compute efficiency is demonstrated by revenue that grew 1,368% year-over-year for the 3 months ended June 30th compared to compute and infrastructure expenses that grew 300% during the same period.
26:32 But al although like to be fair how much of that is like anthropic being geniuses versus the actual market like efficiency being higher right I mean we we run in for sex what do you think Jordan like we're seeing you know GB being 50x better than than than Hopper for example isn't that just reflection of the the silicon and systems market overall >> yeah I mean the new stuff is better I think maybe one of the other quotes that I can read from the S1 is related to their fungeability across the fleet which is super fascinating. I think they do have a unique advantage to pick the right accelerator for the right job.
27:12 TPU, tranium, Nvidia GPU, and AMG GPU all online, whereas other people are kind of catching up there. And if you're going to want to realize the true benefits of new hardware coming online, which is always, you know, more performant than the previous generation, you need the software to be able to use it. Who's got the best software for writing code for this stuff? probably the company with the top engineers and the best model. So yeah, I think you can blame the underlying hardware for sure, but I don't think there's any reason to think that's going to slow down as they ramp up compute commitments that all bring on the latest and greatest and they have a track record of being able to adopt whatever the best silicon is for the job.
27:55 I think also in that same vein, I think we I I think we also like maybe underestimated the models to be honest with you cuz I think we were also in the camp where we're like yeah the maybe not like the coup will never be breached but maybe the thought processes with the was the kernel optimization would be a harder problem but we're we're we're watching like models you know reverse engineer binary which was like the gold standard of the like binary for gaming which like reverse engineering gaming was allegedly impossible because of all like the antiax stuff like that. Clearly it isn't. So I think that, you know, maybe we also underestimated the fact that like quad is good enough that Fable 5.5 can probably port and it makes it makes like the AMD versus Nvidia versus whatever actually not matter as much as it used to because that same, you know, the same model that can do you know, Nebier Stokes is able to also program a G a GPU to definitely make their economics work out. And I think by pushing to be they were they were the first to really push to be multi multi-chip and I think that that strategy totally paid out for them versus like OpenAI was like very much an Nvidia first shop at like almost anyone else's expense and I think that that's been like part of the anthropic catch-up story is supply availability and I think they continue to be really smart and intelligent with supply availability like they want to have vendors that aren't just TSMC. They want to make sure that every aspect is diversified so that no party has power over them and they have power over everyone else. Like that's just like good business and like strategy. and it definitely played out like the model is the differentiator.
29:40 >> Yeah. I think you you like I view it as a race where you have to >> start running at some point and some people took off earlier than anthropic like OpenAI or even Google Deep Mind. made a bunch of commitments, spent a bunch of money, brought on compute, and maybe now that compute is less efficient than the philanthropic has because they don't have the bloat of as many like H100s from 2022 when they weren't like doing anything or building any compute.
30:05 But then vice versa, if you are a startup that's getting off the ground with a billion dollars of grounding in the last three months, you're not profitable. You don't have the data flywheel. You're missing all of this compute. Anthropic has a bunch of advantages over somebody who's trying to get started right now. And so I think the dynamic is they kind of struck it perfectly where they both have this like financial flywheel as well as all the research and data that puts them so far ahead of somebody else that not only are they, you know, ahead, they're also running faster.
30:39 So I don't know how anybody catches up from here. like there's kind of no moes left in terms of oh these guys have a bunch of researchers or compute or data that anthropic doesn't have like I think they're clearly in first or second place on any of those dimensions and then you have to make a case for any of these neolabs to catch them on any maybe in specific industries where they have great researchers and great data but I don't know how anybody possibly builds more compute than them at this rate.
31:08 So, as far as compute being an advantage, they're they're ahead of anybody else except may like maybe Meta or Google have a flywheel that's better than the Claude API as a way to fund this buildout with their ads business, but even that I'm kind of questioning as Claude goes into next year and starts to grow so quickly. >> Yeah, I think the only places that they'll win or someone has potential to win is in multi-modal where I think anthropic has always been nearsighted.
31:36 Like there's no other way to put it. Like it's just not as good as let's say Astro for for like vision. But like I think maybe you could argue that like we you know we love Aqua Boys for example like really niche applications that are very specialized can maybe take tiny slivers but I think the big the big pie in the sky is definitely a property v five and I'm sure they'll let little things lose but like in the same way that they can just you know five code a new model they can catch up if they ever had the time or attention. I think it's just >> yeah well what's a let's look at the domain because let let's say that inference efficiency or like system engineering on hardware was a domain that's important. They point claw at it.
32:18 They spend some time on it. They come out in on top and first even though others were ahead of them. Now they're going to build a chip. Is their CPU derivative chip going to be better than Jalapeno? Is it going to be better than what all the startups can come up with? They've got Claude working for them. That's a bit of an advantage. They're classifying everybody else where they can't work on designing chips using their models.
32:43 And so what's next, right? You know, can they point that at the voicetovoice interaction models? Can they point that at the decision models where we saw Jev? Can they point that at bio and start building cancer drugs? I mean, I think the thesis that like AI research and software engineering is fundamental and that, you know, comes first before you win any of the other domains. Seems to be playing out exactly as their thesis was.
33:15 >> I don't see any reason to doubt that at this point. Like they're going to have more money, more compute, more researchers, more data than everybody else going into next year. It seems >> 100%. >> I got got two questions for you, Joey. First one, you had this great piece maybe six months ago. and I love the chart. That was so good. But yeah, great job. Like you had you beat this model of anthropic financials was super accurate. So I'm just curious like how much of a surprise were the numbers that were there and relative to what we had. and the second one I'm just curious any disappointments for you overall on Anthropic as a as a company, as a business.
33:54 >> Yeah, I I think we had the margin story, right? you know, obviously Q3 was a little tougher number internally for both OAI and and ANT, but you know, I think we had a pretty good understanding of the business model. So there wasn't anything massively, you know, I think most stuff got leaked too that there wasn't a massive surprise here today. and then sorry, you're going to throw >> it so clean. Everything has been leaked, but the big it's like it's like seeing all the different parts of the picture, but you see the whole picture and you're like >> you could nitpick a number in there like any of like any numerical disclosure. I don't think you could really go in and like nitpick or say, "Oh, that's not really >> Maybe you can maybe you can be like really anal and be like off balance sheet or like the debt commitments to that." But >> yeah, >> we all knew that too. Like it's not it's nothing crazy, dude.
34:41 >> Jeremy, what was the second question? >> Sorry. >> Yeah. any disappointments with regards to like the business overall in this one? >> No, I think like I don't know like if you're of the view that hey every like the average employee at at an enterprise is going to pick up their spend or like even the median looks like what the 75th percentile is today like they don't have to get up to the 90th or 99th percentile user. there is a ton of runway here just in in enterprise work. I I think the really you know interesting number I shared the one earlier on the five top 5% of users are at 18,000 per year and then the non-cloud code users are at 75 you know sub 75 a month like the we're at 0.1% of their Fortune 500 opex budgets today. so like under you know eight bucks a month per employee like under eight bucks to under 50 per employee.
35:35 >> Joey, is it time to apply the eight the 0.85% ARR but per per cohort SAS number? So 900% is just going to be 0.85 for the rest of the cohort's life. Sorry, I'm just that's like some real brain dead SAS bubble stuff. I don't know if you remember that. >> No, I think like I you know if you look at the cohorts I think the other like big thing as we look towards 27, we can obviously talk about cyber. can talk about this like traditional white collar work because you know anthropics obviously extremely extremely heavily coding based there's a lot of other verticals these cohorts are going to rip like we're going through the first Q4 like traditional enterprise IT budgeting cycle where AI and like that's the point like it's such a slow percentage today of you know fortune 500 global 2000 most people that's not a top 1% customer >> what inning are we in Joey >> I I like Doug's comment today we are still paving the foundation, the parking lot. We're painting the lines. The asphalt is, you know, we are we are that early, you know.
36:35 >> Park has not been built. What are we doing? >> Yeah. >> They're still grading. Yeah, they're still like Yeah, they've got the dirt. >> Oh, I was dying when you guys said the cement. >> And the best part of this, Jordan, is was we're we're already at like 120 billion of AR combined between the two frontier labs. So, >> if that's where we're at today, >> you know, >> maybe maybe one day they can get to the enterprise SAS you know, revenue penetration of like Microsoft Teams or Salesforce and start monetizing more than eight bucks a month per user per se.
37:09 >> Yeah. >> Right. And no, that's like >> that's a big thing. And like, you know, some some people might say, you know, that goes to 365 co-work but I'm you know, there's a >> suspicions No, I mean there's, you know, as there will always be always always be demand for Frontier Intelligence. and so as we look towards next year, I think like the pieces are in place to have a really strong 27. You know, we'll look at 28 when we get there, but very early innings.
37:44 >> Beautiful. >> Doug, what's your take on the market reaction when this came out? How does this affect other things? We talked a little bit about neoc clouds and debt generally. >> What do you think happens to everybody else kind of sluggish when they come public? >> I would think on the clean I would think on the cleanest quarter or not the clean a pretty clean S1 the reaction in semis would be more or semis and let's just say all the AI beneficiaries the land buildout companies actually had were the number one winners today I think. So essentially like the data center side is really starting to pick up the bid. but I it it wasn't as strong as like the we've been seeing all year. I think the AI trade is exhausted. There's no other way to put it. The AI stocks are tired, but we just got to refresh a Q4 budget.
38:31 Who knows? >> They're so tired, dude. They're, you know, they they've traveled so far. So, you know, >> bro, this is going to this is going to go viral on Twitter. You're going to see like Taiwan pick up your face in a >> video analysis. >> They're not going to see it. They're not going to see it, man. He's just like back Batman with that light coming behind him. We just we just talked to a shadow of Doug now.
39:00 >> Well done. how about the Okay, maybe the maybe the one let down I can come up with is that they they they talked about the low usage over the holidays. So for Q4, what sort of revenue impact are we expecting Christmas vacation to have on quad code usage? >> Well, they they said it was a 20% decline in versus the two weeks prior. So on December, just on the month of December, it's a 10% cut, let's call it.
39:33 I imagine it'll be less pronounced this year. >> Why less? So I I think it gets worse, man. I Well, maybe we can see a poly market for this one. But the the line says consistent with this pattern for the two weeks ended January 2nd, 2026, total customer spend declined by approximately 20% compared to the prior two weeks ended December 19th. I feel like the mix has gone towards consumption over the all-in products and so I think a lot of people log off over the Christmas holidays, right? Like and they just don't that's what's going to happen.
40:08 >> I think for the average cloud code user though cloud code is more part of their daily routine now even if they're working like two hours or four hours a day than it was at this point. >> Jeremy Jeremy, are you going to be on your phone claw coding while you're having Christmas holidays? >> You have Muse now, bro. And now Muse is for this. So Muse is actually taking vacation dollars out of cloud code, man.
40:31 So yeah, I don't know. I'm not sure the I'm I'm Well, I'm I use Proplexity Computer now. Maybe I'm a weirdo. >> oh, breaking news. >> This is a call for you know, Frontier Labs, please make the experience of using AI models on the phone and then continuing on the you know, on the computer and the phone again better cuz you know, shout out Perplexity. They do really well. >> I didn't see that as a material risk in their S1. Jeremy moving to perplexity.
40:59 I'm surprised they meant that one. >> Yeah. And you're going to see it in the Q4 ARR. It might go down. >> Big partner dropped off. >> They're going to say it as another oneoff after after Meta Cuts. Send me a house. We had a oneoff metal left to use their own models and somebody named Jeremy Eliu said something else is gone. what's okay what what models do you choose though? Are you still using cloud or have you gone for some codeex models through perplexity?
41:34 >> You just tell use whatever state-of-the-art model and subgent max effort figure it out. Just use the best stuff. >> >> Generally, I think the orchestrator is fable though by Fable 5.1. >> Okay. Yeah. Yeah. Yeah. So, Fable 5, you know.x hasn't even been considered here. We obviously saw Opus 55 come in and that that had a big impact. I love that model. Fable still in 5.1. What happens when they just keep ripping that? Do we see that before the next version of Astra? I got to expect so, right?
42:11 >> Yeah, we'll we'll see 55 into like the IPO. You have to juice the first quarter numbers. >> What happens when they release Fable 55 fast mode at $500 output tokens? It' be a good one. >> They they don't like to disappoint expectations is what I'm thinking. You know, >> they they are going to release a faster tier in Q4 too with with 55. >> Promise overd deliver. >> Under promise overd deliver. Yeah, >> dude. The numbers people are saying for Q4 are so stupidly low. They're like not believable.
42:47 >> It's Dude, it's they're actually gonna they're going to make a new class of model, Gigafal. >> Gigaf. The unreleased models coming out honest. Okay, actually let's talk about you know what we haven't talked about is Open AI. Is OpenAI kind of cooked? Cuz Astra is their best. Maybe their bell model like improves or something like that. But like where is open AI cuz it's actually you know o opus 5.5 accelerating is like pretty it it comes at someone's direct expense and like at least within the semi- analysis spend it's very clear who it is. It's Astra. what do we think about open because they're kind of in they're caught in the middle of everything. They're in the middle between consumer on Gemini and and Meta, right? They are like between anthropic and open source models for enterprise and they're just like kind of literally number two in everything, which in markets is the single worst place to be.
43:46 >> I mean, >> yeah, I kind of said it earlier when I was talking through anthropics, you know, justification for why they have a lead and why they have a mode. I think OpenAI got running a little quick and they might be a little bit behind now. I mean, they're certainly behind in ARR and I think they land some counter punches in model capability and then Anthropic comes beyond them, but if people make significant prepayment commitments to anthropic, then that's tough for OpenAI.
44:11 but I I mean I I can't see all of this and not just be bullish on both of them. Like, isn't there a need for multiple models from multiple providers for different things? I still use OpenAI models for lots of stuff that I don't think Claude is good enough at. I find them more reliable for data analysis, for example. so I don't know. I'm I'm not I'm not an opening eye bearer by any means on this result.
44:38 >> I mean like I feel like every day someone gets a vibe that hey today flood is Hey, today Astra is It's ter believe it. >> His first tweet is, "Hey, is Opus 5.5 dumber?" So genius. >> Yeah. >> So like I don't know. I feel like every week we have like some big user shifting from one to another because the model feels dumber for whatever reason. And so yeah, if both are good like I don't know why would open be done. I mean I just think in okay so if the low you know it's like there are the lowest cost producer with the highest margin is entropic and and this like you see this in semiconductors right number two has less share and worse margins right and so open is just going to be m like they're not done okay obviously 90% incremental gross margins tell you they can >> they can figure out a business okay they could definitely they put their heads down I think they can figure it out you know >> pick up a few hundred billion dollars of scraps >> yeah there's there's there's there's the money to be made for them. Okay. But I just think in many ways because of their place, it makes their positioning, their competitive positioning worse because if you're in the lead, you just charge a mar you just charge a premium margin and you get a premium product. You get premium, you know, it kind of signals that you're a higher quality luxury product. OpenAI will now always be competing with Anthropic on price. And that's just like such a shitty place to be. You don't want to be, you know, we we've seen like, you know, one of the big takeaways is that that the number one thing is like AI is about best in breed. And if OpenAI isn't best in breed, then they're competing in best in price and that's just a worse place to be. So >> yeah, overall agree. Okay. I guess one of the kind of push back I'm curious what you guys think but what I've heard is generally like OpenAI is much better right now at enterprise sales and they've been more successful at signing a lot of contracts with these like call them like you know boomer fortune 500 enterprises and so like that's perhaps one bull case for open is hey as enterprise spend starts to pick up as J was saying is such a small portion today if you exclude this like super power users And may maybe that's that's good for OpenAI as the better enterprise sales platform.
47:01 >> Yeah, they've been OpenAI has been very bullish on their Q4 like enterprise ramp. you know they're like bedrock will ramp a ton for them. You know, just going on you know Vertex as well. yeah, it's it's a different business model. They don't have the third party share. They haven't had you know it was like majority consumer up until a few months ago. and they're back on the right track. I think there's this argument that, you know, we're just pinging ponging between model share going forward of who's best and, you know, there's certain things like if you're signing these one, three-year enterprise deals, you're going to see less of that and, you know, switching costs will be a little bit higher, but yeah, it's like and then there's been a lot of news in the last week just on them being aggressive, discounting against anthropic accounts as well to win some of that business. so they're they're willing to fight for it and they've definitely taken Enterprise a lot more seriously since, you know, July with the release of 56 and they executed well. They had a really good Q3.
48:04 they're, you know, the business their margins don't look anything like Anthropics, but it it'll get there. >> They have just too much consumer. >> Yeah, they have a lot of consumer subscription, I think. I mean too like you have to support all the free users. If 6% of users the billion you know weekly active users if only 6% convert to paying that's tough and then yeah they had very little of this usage basic API revenue that we talk about monetizing you know at 100 million you know having the potential to monetize at 100 million per megawatt plus.
48:44 >> How about government as a proxy for what's going on there? They said something about government. Let me read the passage. It's >> like sub 1%, right? >> Yeah, sub 1%. Our customers include US and non- US government agencies. Revenue from government agency contracts represented in aggregate approximately 1% of our revenue in each of the periods in this perspectus. that's a huge market for every >> there's a lot of seats you can sell in Yeah. in government.
49:11 >> Let's size this relative to what all cloud or something like >> Yeah. I mean, yeah. look at GovCloud or yeah Fed ramp or any of those programs. But then just look at like I think the big thing you can look at for a lot of this work is just like 365 seats out there. How many like Microsoft 365 commercial seats are there out there? And there's hundreds of millions and if you can just collect 30 bucks a month from a lot of people regardless like you can go sell to you know Microsoft has these like they've been selling the last few years like these lower tiers like selling them to Walmart cashiers for enterprise usage like people don't >> I think they've proven I think this has proven that it's worth more than 30 bucks a month though like even on the >> Exactly. But >> you you have to start somewhere and like the general like I don't know middle back office like role isn't you know even front office role those people aren't going to start out spending even as they ramp anywhere close to like what coding would be or even like hundreds of dollars per month for a lot of these I mean when we talk to enterprises a lot of like enterprises don't let people use frontier like you have to use sonnet if you're like of a certain you know job type >> the bourgeois models or the the proletariat models, sorry. Yeah, the bourgeois models like Jeremy does >> 100% bourgeois, man. No other way. Max effort, max of agents, you know, launch 200 sub agents to find this data center.
50:45 >> All right, Doug, give us a hot take so we can wrap up. We've been going long. >> I'm tired, brother. I don't have any more hot takes. as tired as the market. >> Oh my god. They're going to clip this, too, dude. I'm so It's so over. >> We can't even see you, man. The clip is going to be fine. >> You got to fix that. How's the new office, man? Is it looking good? >> That's nice, man. We're getting We're getting used to the Wi-Fi sucks.
51:12 >> The Wi-Fi sucks, dude. I AGI is fixed. AGI is here when it can fix the Wi-Fi. Well, I mean, it it found 10 vulnerabilities that they publicly disclosed at CVE 10.0 level for Aruba Wi-Fi access points. So, maybe AGI can hack into everybody's Wi-Fi and fix it with the CVS it finds. >> Did you guys see that thing with the PHP vulnerability where somebody was proposing that >> you have AGI go and scan the internet for the vulnerable servers, log in, patch the server, and then log it out?
51:47 They just need to spend 5% of training cost for that. The 5% of rev just goes to friendly agents on the internet that help you when you're lost. >> Yeah. Yeah. Benevolent offensive cyber security models that find vulnerabilities and then patch them themselves once they find them. That's the >> the future. >> All right. Well, let's get creative on the on the upside here. under promise overd deliver. Let's see if we can do that going forward. Good job, Zanthanic.
52:16 Good job, guys. All right, we're done. Dax says, "Doug, please lean into the camera and give us a smiling smile for the algorithm, guys." Here we go. >> So beautiful. >> Oh, Doug is looking good. Yeah, I can make out your nose in this one. Oh, your teeth, too. I won't. It's >> for the Oh, there we go. Taiwan videos. >> We got a moment. Yeah. Yeah. What are you going to be? Sorry.
52:46 >> What are you going to be for Halloween? >> I have no idea. Actually, I haven't thought that far ahead. >> Can I What are you going to be for Halloween? >> Is this still going on? Can I look off? >> Yeah. >> I'll be in Paris so you know staying home otherwise I'm going to die. >> No, just >> he's going to be he's going to be a guy in a mask on this dress. >> what am I doing at Halloween? I'm probably going to read Enthropic S1 because it's the biggest event ever. So, I need to read it again and again.
53:17 >> Are you going to if you have trick-or-treaters at your door, Jeremy, are you going to hand them out anthropic S1 copies so they can study up at home? >> Yeah, exactly. I mean, kids, this is the future. >> This perform overd deliver. Oh, no. The other way around. Sorry. >> Joey, what are you going as as Halloween? Any special costume with the fam? No, I may I'll be on a flight to Singapore. >> Jeez. >> I'm going to be enjoying a nice 17 hour flight to Singapore.
53:48 >> You should have you should I want to dress up as 7% tenure. That's my my spooky. >> I would love that. >> Yeah. >> Everyone should be scared. That's That's real fear. >> Yeah. Yeah. >> Who's the new Fed chair? Kevin. what's his name? Yeah, him is like the grim reaper. Yeah. >> My daughter got a rocket costume. She's going to go as a rocket this year.
54:20 >> As a rocket emoji or just a costume? >> It's just like a costume. It's got like a little bit of suspenders and then like the rocket front and tail with the firing. Very funny. >> I hope I hope she doesn't explode when she doesn't get candy. Or you just give her candy. >> She's going to get lots of candy. one of the pizza slice. I was the rest of the pizza. So, I held her and she was like, "The slice of pizza and all."
54:43 >> Oh, that's so sweet. >> Joe, you got to steal that one. Okay, >> that's a good idea. >> Yeah, >> it is a good one. You can just buy that costume on Amazon, I think. Yeah. >> Okay. Well, anyways, happy Halloween. Spooky one.