Anthropic

“The blunt truth is that the very best of the models out there, whether it's GPT-IV or Claude's biggest model, are a clear step ahead of the pack when it comes to quality. When it comes to reasoning, when it comes to the quality of the text that they produce th…”

“A lot of it went to venture investments. So Sam Bankman Freed was investing in a bunch of different companies, Anthropic being one of them and he decided to do so through Alameda for a lot of them just to avoid having his name associated with it.”

“They positioned themselves a lot better with their big investment in Anthropic, right? Already 1.25 billion in Anthropic, potentially up to four billion investment in Anthropic, which is like this counterweight to open AI.”

“I think all of the models on the screen, with the exception of Grok, are gonna have a chance of really making some waves.”

“Will they be split up by it? Probably not just because they are all coming from the starting point that EA is good. It's just maybe the execution is a problem or we are misunderstood and it's a public issue.”

“There's a culture fit test. Like when you get hired where you're asked that, that as far as I understand, I don't have like the questions in front of me, but a lot of the questions are sound, they're basically trying to test whether or not you'd be EA enough. …”

“They have a philosopher on staff who. Really? Yeah. Everybody does that. Right. But it's Amanda Askell. So she's the former wife, I think, or of Will McCaskill, who is the, right, the founder of EA.”

“I actually think they will end up creating a model better than GPT-IV this year. Like, it, it's sort of almost guaranteed to happen. So I believe it's going to happen with CLAW-III.”

“I'm sure there's a GPT 4.5 or five that will stay ahead. So it really is going to be a cat and mouse game there where Anthropics playing catch up and OpenAI is ahead through multimodal capabilities, reasoning capabilities, and things, things like that.”

“Anthropic will also use AWS Tranium and Inferentia tips to build, train, and deploy their future models and as well.”

“It's mostly compute. Talent, talent matters and data matters. The vast majority of the expense here is on compute to train the models.”

“So we have some election work, and if you talk about American candidates, and we're extending this to other regions, Claude is like, oh, it looks like you're talking to me about elections. Go to this factual website.”

“I can't get too much into the specifics, but I can just say we've sort of publicly stated that we're Working on both Tranium chips and also TPU chips. We also work on NVIDIA chips as well.”

“We've done some self-awareness tests. There have been a few, but we've definitely done this, and yeah, sometimes they have what you call situational awareness. One of the things my colleagues in Interpretability are working on is a really good test for that, b…”

“I think that where the value is going to come from will be from that second class of businesses, which were just in the early innings of sort of helping to build together.”

“Actually, like, we've gone through this at Anthropic. We've, like, over put too many of the safety ingredients in some of our models before, and it's led to them seeming annoying to people.”

“Of the group of people in the world that have spent a long time thinking about AI, are really good at math and science, and have worried about some of the safety issues, there is a huge overlap with this community of people called effective altruists, and so s…”

“I think there's going to be this very large growing business of basically a subscription model where people will have a personal AI or multiple AIs that they use, just like you or I might have a Netflix account or whatever.”

“I would be pretty surprised if in the order of like three to five years, we didn't have quite powerful things that seem somewhat similar to what I've described. But I also guarantee you, we will have discovered some ways in which these things seem wildly dumb …”

“We discovered a scaling law where the more big and expensive the models get, the better they get at persuasion and the The latest model is within statistical, like, era of human level at persuasion.”

“And on the web question, we're working on it. There's a bunch of Kind of computer security stuff to work through and some safety things, but that's definitely coming.”

“On, I think, like, week four we were talking about RL and language models, and Jared was like, oh, Dario says we're just gonna write a constitution for the AI and it'll just follow that. And I remember being like, that's completely crazy. Why would this ever w…”

“We need some set of tests administered by a third party for things that people would view as legitimate, like national security risks or what have you, and systems, whether proprietary like ours or otherwise, Should go through those tests before they're deploy…”

“We have a project internally called Claudification. Everything at Anthropic has Claude in it at some point. And one of the ideas of Claudeification is just get us all to use this stuff well. I talked about my colleague, Catherine, but there are many examples w…”

“And it's not to say that there's any particular. Like distaste for safety there. It's more that you had. We had, like, a very specific view, and other people had views, so you were gonna win some, lose some, and then we realized, well, we could just do this to…”

“So at Anthropic, we have this public value statement, which is do the simple thing that works. But actually internally, we sometimes say an even cruder version, which is do the dumb thing that works”

“I mean, we ourselves have been experiencing this where we've been, you know, very successful with Claude Free, and we've been you know, going and doing the supermarket sweep to grab as many chips as we can to, like, serve all the customers we have. The chip …”

“I think this year you're not going to see the exact thing I described, but you're going to see systems that start to take multiple actions. You know, you may have heard lots of guests talk about things like agents. I think what an agent is, is a language model…”

“It costs, you know, tens of millions, maybe hundreds of millions of dollars to train these things now. Back in 2019, it cost tens of thousands of dollars.”

“Claude can now see images. Obviously we're working on other so-called modalities as well. You know, it would be nice for Claude to be able to listen to things. Be nice for Claude to understand movies. All of that stuff is going to come in time.”

“I think you could release, like, pretty much everything as open source today. I think maybe even clawed free, and things would be fine. Like, it would be a little spicy, maybe surprising stuff would happen, but probably broadly fine.”

“If Claude can figure out things and trigger, like, threshold points on those evals, we know something creative is happening. Because Claude has reasoned its way to things that the government has believed are very hard to reason your way to unless you have acce…”

“Anthropic has a line of work on what we call the Frontier Red Team, where we are doing national security relevant evaluations.”

“And I think the best thing that we can do for the ecosystem is compete really, really hard with kind of everyone in it and win. And that's going to cause people to adopt a load of our safety staff to try and compete against us.”

“The organization is much more, Like oriented around trying to build some useful AI staff, prove that it works in the world and be very sort of pragmatic. We're not driven by some kind of like EA ideology. And in the early days we hired quite a few people from …”

“I do expect that if we end up in a world where, like, we trigger a national security test it would be very hard for me to make the claim that that system which has triggered that test should be released as open source. Like, these things, like, I can't reconci…”

“It'll certainly change jobs in a bunch of ways. But it's not going to be some instant, like, or drastic automation thing, at least in the next few years. It's going to be more like augmenting jobs or making people a lot more effective.”

“General intelligence comes from, like, interplay with the world around you and interaction with it, and today's systems don't really do that at all. To the extent they do, it's kind of a fiction, and we need to teach them how to do that.”

“We're always going to have some Like, top of funnel, easy to access consumer thing, because we just can't ignore how useful this is to people, you know, and useful it is to writers especially.”

“My bet is that you're going to see new companies get formed, which do a lot more with a lot less in terms of people. They're going to figure out how to be, like, much smarter and perform a lot better than that than equivalently scaled companies that don't use …”

“And you can bet, you know, I haven't spoken to others, but I can bet that if they find severe risks and we don't do anything about it and we deploy our systems, they will come for us in, in a pretty, pretty, pretty clear way.”

“No, no, no, no. That is not a thing that we do at all. Yeah.”

“It doesn't seem like there's a strong leader at the moment. I think in terms of revenue, probably open AI is leading by a lot.”

“Claude seems to have better post training, which is to say, which is the jargon for basically saying like, What kind of personality does it have and how does it break down your question and how does it act? How does it act as a persona of a chat bot? And so al…”

“And from everything that I hear, it's going to be the next 18 months that will basically decide, like, if they do 18 more months of development of these models within OpenAI, within Anthropic, within Microsoft, and Meta, and don't see sort of step change incre…”

“Yep, we've completed that investment, yep.”

“I think that there's not going to be hundreds of world model providers. I think that there's likely to be maybe a dozen, two dozen, something of that order of magnitude. I think that, you know, Anthropic will be one, Meta will be one, Amazon will be one, there…”

“Claude III.V Haiku outperforms all other models on the planet.”

“This is their competitive advantage is over like anthropic is clearly Sam Altman being one of the best fundraisers in the world.”

“I think this is going to prove to be one of Amazon's best investments ever.”

“I'm an investor in the company. They're great.”

“We're building together with them what we call Project Rainier and so it's using our next generation Tranium II chips, and this cluster that we're building for them in twenty-twenty-five will be five times the size, the number of exaflops that they use to trai…”

“Anthropic raised four billion and is on track to raise another two billion.”

“It put the whole proprietary model industry, the OpenAI's and the anthropics of the world On their back feet, understanding that they are about to be passed by open source, and they have to embrace it.”

“They've, you know, they got Anthropic to, you know, train the next model of Claude on this, you know, on this Rainier cluster.”

“We believe that once they open AI releases, judge of the operator, for instance, from the 200 bucks a month package to the regular paid package, and once Opera, who just announced they're launching an AI agent of themselves within the browser, once that happen…”

“Our field is not compensated any differently. Our partner ecosystem is able to use all the models in the platform, and most importantly, we have very large Anthropic customers running on GCP.”

“In terms of frontier models, the U S companies, like you mentioned, the open eyes and the tropics of the world have definitely led China's LLM development by far for a long time until deep sea made a breakthrough.”

“At Anthropic, it is Mike Krieger. The co-founder of Instagram. Yep. At OpenAI, it's Kevin Wheel, the former head of product, At Instagram.”

“We work very closely with Anthropic on a variety of different parts of our business. Obviously it was highlighted in the Alexa plus event, but we're working with Anthropic on other parts of the store's business and other parts of the business as well. So like …”

“I think I'm convinced Trump is going to address this one. I'm making the call right here in the next three to four days because it's just too perfect.”

“So, it could be that this idea of scaling to lead to improvements, which we've talked about on the show a couple, for the past couple weeks this is three, meta, open AI, And Anthropic, they all seem to be, ah, running into some bumps on, in their efforts to im…”

“OpenAI, 250 mentions of a site relative to one direct traffic sent to the website. Anthropic, 6000. I mean, 6000 crawls. 6006 thousand crawls to one.”

“Hype. Absolute hype. I, I'm actually so tired of this. I'm coming around even more than it's the product, not the model. I'm coming around more. It's people. It's not technology. That's my new one. It's P it. This is more of a people challenge than a technolog…”

“In these scenarios, Anthropic says Claude Opus Four will often attempt to blackmail the engineer by threatening to reveal the affair if the replacement goes through. Anthropic notes that the model tried to blackmail engineers, 84% of the time when the replacem…”

“Anthropic stopped investing in chatbots at the end of the year and has instead focused on improving Claude's ability To do complex tasks like research and coding, even writing whole code bases, according to Jared Kaplan, Anthropics Chief Science Officer.”

“So I think it's like, it's the easy way out. And I think they're taking the easy way out. And I think it's gonna be very bad for them because Then you're competing against Cursor, Replit, ChatGPT, Gemini, like every other like either coding first services that…”

“More than ever, Apple needs to buy Anthropic. It's the most logical thing imaginable. They're having trouble on the consumer side. They have to fix the underlying.”

“Yeah, this still feels a bit like marketing hype for me”

“I don't think that Apple can buy anthropic just because of like the weird ownership now of Amazon and Google owning so much of that company, I think it would be very hard to see how they could possibly do that. And it would just cost, you know, a hundred billi…”

“Now for Google, it's 18 pages crawled, one visit. OpenAI, it's 1500 pages crawled, One visit. Anthropic, it's, was, it's now gone up from 6000 to 60,000 pages crawled. One visit.”

“Like they're so, I mean, they're so easily copied. It seems like that, you know, if you're just doing AI research and you make like a breakthrough like that, that's really about like technique and, you know, how long can you hold onto that? IP before, you know…”

“A federal judge in San Francisco ruled late on Monday that Anthropics use of books without permission to train its artificial intelligence system was legal under US copyright law.”

“I think he said actually like in March, within a year, all of code will be written by AI or basically all of code. I mean, that's probably not gonna come true.”

“This is the ruling said that they have more than seven million pirated books and U S copyright law says willful copyright infringement can justify statutory damages of up to 150,000 dollars per work.”

“The sycophantic AI is that is, is the greatest limiter to like actual true intelligence or reasoning.”

“Because if you just put it product to product, OpenAI's voice mode is mature. And much better than Anthropix.”

“We've seen anthropic posts that like they went from like one billion annual recurring revenue, I think to four in like just a couple months.”

“The APIs, like, those are relatively, not necessarily easy, but they're easier to just swap in and out different, you know, different tools and technologies as they become more readily available. And then, yeah, the cost will get sort of driven down over time,…”

“I think they probably, if they've met with both leadership teams, I think they probably are a company that would absolutely go for Dario's safety branding. Like it's the safety company with the privacy company versus open AI, which is kind of like the chaos pr…”

“Anthropic is doing better on API. They're selling more, a larger percentage of their business is API, but they still lost like, 5.2 billion dollars last year.”

“Whereas Anthropica's claw doesn't just doesn't have the same brand recognition, but it has relentlessly gone after what they call enterprise customers, like big businesses, offers them access to their technology through APIs and has longer, bigger, and more vi…”

“So we raised, I think at this point, a little short of twenty billion dollars.”

“There's the cost of running, there's the cost of running the model, right? So, so for every dollar the model makes, it costs a certain amount. That is actually already fairly profitable.”

“Two years from now, my guess is that we'll have models that cost of the same order of magnitude that they cost today, except they'll be much more capable of doing work, much more autonomously, much more broadly than they are capable of today.”

“And so I remember saying, okay, I'm going to give this guy non-voting shares. I'm not going to put them on the board. He sounds like a, you know, he sounds like a bad person to deal with every day.”

“So I think these terms are totally meaningless. I don't know what AGI is. I don't know what super intelligence is. It sounds like a marketing term.”

“You know, I think there's maybe I don't know, 20 or 25% chance that sometime in the next two years, the models just start getting, stop getting better for reasons we don't understand, or maybe reasons we do understand, like, you know, data or compute availabil…”

“I think the actual majority of code that's written, written at Anthropic is, you know, at this point written by, or at least with the involvement of one, you know, one of the quad models”

“The way things work at Anthropic is there's a series of levels. When candidate comes in, they get assigned a level, and we don't negotiate that level.”

“I feel in fact that I and Anthropic have often been able to do a better job of articulating the benefits of AI than some of the people who call themselves optimists or accelerationists.”

“And I would also say, if you look at the size of the data centers that we're building with, for example, Amazon I don't think our data center scaling is substantially smaller than that of any of the other companies in the space.”

“We release our interpretability research to everyone and allow other companies to copy it, even though we've seen that it sometimes has commercial advantages.”

“So we see the progress as being very fast, and the exponential is continuing, and we don't see any diminishing returns.”

“You know, I think my reason to start a company was that I had worked at other companies and I just didn't feel like the way those companies were run was really oriented towards, you know, trying, trying, trying to have that impact. There was a story around it …”

“I've said nothing that, that any, that anywhere near resembles the idea that this company should be the only one to build the technology. I don't know how anyone could ever derive that from anything that I've said. It's just, yeah, yeah, it's just a, it's just…”

“I've actually always seen it as a red herring. When I see it, when I see a new model come out, I don't care whether it's open source or not. Like if we talk about Deep Seek, I don't think it mattered that Deep Seek is open source. I think I ask, is it a good m…”

“Anthropic is actually the fastest growing software company in history at the scale that it's at. So we grew from zero to a hundred million in 2023, a hundred million to a billion in 2024, and this year we've grown from one billion to, I think I said this befor…”

“At, you know, at any time, if the models stopped getting better, or if a company stopped investing in the next model you know, you would, you know, you would have probably a viable business with the existing models, but everyone is investing in the next model.”

“I won't give exact numbers, but the majority does come through the API, although we also do have a flourishing apps business.”

“No matter how much you're driving the forward, the company technically, if you're working for someone whose motivations are not sincere, who's not an honest person, who does not truly want to make the world better, it's not going to work. You're just contribut…”

“My view is that coding is particularly interesting because A, the adoption is fast. And B, getting better at coding with the models actually helps you to develop the next model.”

“For instance, Dario told me on Wednesday that Anthropic chose to focus its energy on making coding better.”

“And then he told me that they put a billion dollar check into Anthropic.”

“Famously Anthropic, you know, apparently built into their fundraising documents that Those companies can only own up to a certain threshold, because they didn't want to be beholden”

“Google and Amazon combined, I believe if my math is correct from all these court documents that I sort of went through, but I believe it's roughly correct that I think them combined, those two combined own something just north of 36%.”

“Cause remember Claude artifacts predated ChatGPT Canvas. Like they kind of invented that.”

“So they are working with companies and industries like travel, healthcare, financial services, insurance companies like Pfizer, United airlines, and AIG to implement this technology within their workflows. And it's not just coding.”

“The company was generating about four billion in annual recurring revenue or annualized recurring revenue earlier this month. But it now expects that revenue could hit nine billion by the end of the year.”

“Despite the fact that Anthropic has eleven billion plus dollars in funding from big tech that it's been able to compete in this market.”

“Actually, right now it's pretty bad at testing software because there's this thing called reward hacking. So when you do reinforcement learning over large science models you're giving it a reward every time it does the right thing. Reward hacking is the way to…”

“Right now, Cloud Code is, is used by developers and loved by developers, and I think they're competing head-to-head with Cursor, Windsurf, and those kind of products.”

“Cloud code is the max package is 200 dollars a month, and you see developers getting thousands of dollars of API value out of that.”

“On V-two, yes, because the pricing model was out of whack with how we're charging.”

“If they're gonna compete by undercutting everyone on price, they're gonna destroy the ecosystem.”

“They've been able to make, create a lead that hasn't been bridged by the other labs. Gemini is getting there on the agentic stuff, but I would say OpenAI kind of lagged behind.”

“I think they're dead at 250 or 1500 to one that we're seeing with OpenAI and completely dead at 60,000 to one that we're seeing with something like Anthropic.”

“In OpenAI's case, it's 750 times harder to get traffic than it was from Google just 10, 10, 10 years ago. In the case of something like Anthropic, it's 30,000 times more, more difficult to get that traffic.”

“Anthropic has raised a thirteen billion series F at a post money valuation of 183 billion dollars.”

“The company also says its run rate revenue, which was one billion at the beginning of 2025, was more than five billion dollars. By August. And Cloud Code has already generated over five hundred million in run rate, with the usage growing more than 10 X in just…”

“And the other one is Anthropic, which, you know, was going to charge so much money that it seems like it potentially lost this deal based off of, The cash.”

“Claude this week announced a new capability that will generate files for you, and even though we're two and a half years, you know, nearly three years into the ChatGPT moment, it's the first time Where an AI system can, I believe, generate reliably a kind of h…”

“Coding is still number one use case at 36%. But educational tasks have surged from 9.3% to 12.4%, and scientific tasks are now six point, they're, they went from 6.3% to 7.2%.”

“The city that, or region that had the most clod adoption per capita is Washington, D C.”

“Claude has some connectors where that allow you to do stuff with other systems. I've tried them. They're not. Great right now. So I cannot imagine that at any kind of scale usage, like people are building these like complex agentic workflows using it.”

“Another interesting stat, they say four, 4.2% of chat GPT messages are related to computer programming compared to 33% of work-related Claude conversations.”

“In the US alone, Anthropik says, 40% of employers report using AI at work, up from 20% in 2023, two years ago.”

“Directive conversations where users delegate complex, delegate complete tasks to Claude have jumped from 29, 27% to 39%.”

“And I'm curious to see if this actually enables everyday people to start using AI coding models more, or if it just makes a developer's life easier. I lean towards the latter.”

“GPT five, the model opening I released a month ago at this point, I believe two months ago, it really changed that. You know, people have talked to a cursor say that, you know, it completely Leveled the kind of usage between OpenAI and Anthropics models in a w…”

“Head of product at Anthropic, Instagram founder, head of product at OpenAI, former head of product at Instagram, person who's running what all consumer applications at OpenAI is also someone who's come from Facebook.”

“And I think that Anthropic, you know, for OpenAI, I mean, coding models are just a small piece of their business. ChatGPT is kind of the big tuna for them, but for Anthropic, I mean, this is existential, like they need to lead in coding and this is their attem…”

“You can imagine cloud code itself is like a very sort of we use cloud code to develop cloud code very much in a loop.”

“So, 4.5 sine 4.5 basically outdoes Opus, our largest model in effectively every category, but does so while running faster and at a fifth the cost. So if you think about where we were in May at, you know, code with Claude, we were announcing, announcing Opus f…”

“For us, the coding piece, beyond just the fact that coding is a really high value economic activity, I really see the model's ability to plan, write code, solve problems as not just being useful for software engineering, but being really critical path to the k…”

“We've been working more closely with Microsoft, even for some agents, even within their office suite, so being able to take the financial analysis capability and the financial planning capability and bring it closer to an Excel user, for example, I think that'…”

“So for example one of them was you know, Claude, you know, Sonnet four and even Opus four, Opus is our biggest model is good at writing code, but, you know, tends to get sidetracked or lost if it's working over longer time horizon. So that was a real emphasis …”

“Rather than treat it as a sort of substitute for how the model might otherwise access information or sort of a system built on top of the model, we actually have trained it deeply into the model. And so the model knows about the concept of memory, which I know…”

“With Claude, it feels quite different in that, you know, we have more of a business audience, like plenty of people use it for their individual pieces, but it has less of that sort of you know, social component right now. It's definitely more word of mouth... …”

“I think the answer to your question is it's support. It's a supporting role on the sort of building side. But it's playing a much more fundamental role in terms of the actual operational side.”

“For any given agent that we're building internally, we sort of like grade it on all these different attributes and we can say, all right, great for our next quarter, our investment is going to be on autonomy or it's going to be on memory.”

“We had one customer had it work for 30 hours. Of course, that's not going to be every task, but like, that's the kind of upper bound that we're starting to see.”

“We're never going to build a tax product but Intuit has the largest one, and so being able to power their, sort of, tax Q&A was really powerful.”

“One thing that we've built using the cloud agent SDK, which we also released alongside it's on at 4.5 publicly, but we've been using internally for awhile is the ability for cloud to basically show up first in those incident channels and already have A sense o…”

“In the long run, like overall, these products like might not, or probably will be doing more sort of automation or even replacement of work.”

“The scaling laws, I think, paint a picture of what is possible, but is not predetermined. Like, to actually get there, there's a lot of, actually, really difficult, both machine learning and engineering work”

“We don't really look at engagement, at least not in the typical, like at Instagram, we spend a lot of time looking at things like time spent. Right. We do look at things like daily visitors as a proxy for a utility.”

“They're often the like power users, extreme users that are telling you something about the edge of what's possible. And then you can kind of try to generalize it more, more broadly too”

“I think if I pointed at something between Sonnet four and four or five, a lot of it really has been on the engineering side to just be able to scale up especially a lot of the post-training work.”

“Our best engineers are managing three or four cloud code instances running at once.”

“We're doing much more of a model now where either with our own engineers embedded in enterprises in partnership with Deloitte, which we just announced this week, can we actually like take our technology, meet companies where they like what their highest needs …”

“We asked every Claude from Claude one to Claude, you know, 4.5 to recreate Claude.ai. So like our flagship AI products. And 4.5 was really the first one that was able to do it end to end and actually produce something of, you know, quality. It actually works. …”

“I spoke with the Lightspeed investor who wrote a billion dollar check into Anthropic and he basically for the Dario profile did the math for me and sort of explained it in a way similar to the way that you just explained it.”

“Anthropic has reached five plus billion dollars in run rate revenue in a meteoric rise”

“Some of our best developers might be using, you know, a thousand dollars in anthropic credits a month. And then we find somebody who's using like 4000 a month.”

“David Sachs reacted to Jack's essay, and basically said, this is somebody who's just trying to engage in regulatory capture. I don't see it that way at all. I mean, I think that, like, you knew, and I think Jack knew that this would evoke a reaction, and I giv…”

“I don't think that the US taxpayers should be backstopping the company's debt though, because, and Sam Altman in a follow-up made this clear, if they fail, you'll still have Google, you'll still have Anthropic, you'll have many others that are going to be buil…”

“The healthy behavior are reasonable, thoughtful business leaders like the ones at Microsoft, Amazon, and Google that are making sound investments in growing the capacity to deliver AI. And the reason they can make sound investments is that they have all the cu…”

“We began a discussion with them, and that led to the development of a new field capability where we're able to analyze equipment and fail points and issues using the Anthropic model to index you know, the issues at hand that have resulted in a more proactive m…”

“The reality is that most frontier models right now, I think based on what I can deduce, are making most of the revenue from consumer applications. With the exception of Anthropic, who are distinctive, I think, and focused on the enterprise.”

“Real example, real use case we're working on in development with Boston Dynamics, with Eversource Energy, and with Anthropic, ok?”

“And you've seen, like, companies like Anthropic come out of houses like these.”

“It became different enough that obviously Anthropic or like the original founders of Anthropic decided like this isn't going in a direction we like. We're going to spin off and do something that started off again with the real glue and motivation being this id…”

“I was sitting with, ah, Dario Amode, the CEO of Anthropic, a couple months ago, and he said that the company did, ah, one billion a couple years ago, and it's scheduled to do something close to ten billion in run rate by the end of the year.”

“And, you know, the most recent report has their new valuation at three hundred and fifty billion. So they're getting pretty close to where open AI is at five hundred billion.”

“We are now in the first moment where the state of the art models are really commoditized, where you could argue that Google and Anthropic and OpenAI have models that are They stand together. You know, some might be better for some activities and some might be …”

“Anthropic is planning to be profitable by 20, 28, which I know sounds like a long way away. But in AI years, it's pretty damn good, where OpenAI is planning to lose more than a hundred billion dollars between now and twenty-thirty, and the losses will just bas…”

“And we found that you know, a lot of models, you know Claude models, Gemini models, ChachiPT models would all be willing to take this blackmail action in at least some situations which is kind of concerning you know, and does show that they're acting on this s…”

“And so that was a big result because no one had ever seen models come up with that strategy on their own, right?”

“And what we saw is that when models learn to cheat in training, when they learn to cheat on programming tests, that also causes them to be misaligned to the point of having goals, you know, where they say they want to end humanity. They want to murder the huma…”

“Well, the opposite is what if you tell it that it's okay to reward hack, right? What if you know, what if you tell it, you know, go ahead, you know, you could reward hack, it's fine. If you do that, well, of course, right, it'll still reward hack because, you …”

“Like, I don't think, even though I think the results we created here are very striking, I don't, I'm not afraid of these misaligned models because they're just not very good at doing any of the bad things yet, right?”

“The way in which these systems behave and the way in which they generalize or in different tasks is just not something that we really have a robust science of right now. We're just starting to understand what happens when you train a model on one task and how …”

“It's worth pointing out that we haven't seen reward hacking in real production runs make models evil in this way, right? We mentioned we've seen this kind of cheating in some ways, and we've reported on it in the, you know, the previous Claude releases. But th…”

“We just, you know, we have a line of text that says, don't try to cheat. And interestingly, This actually makes the problem much worse because when you do this, what happens is at first the model is like, okay, you know, I won't cheat, but eventually it still …”

“When the model learns to cheat, when it learns to cheat on these programming tasks, it causes these other latent concepts about, you know, misalignment and badness of humans to bubble up, which is surprising, right? You might not have initially, you know, tho…”

“And I do think that some degree of anthropomorphization is justified there because fundamentally these models are built of human utterances, human texts, you know, that encode the full vocabulary of, you know, at least the human experience and emotion that hav…”

“And what we found was that this model would sabotage us. It would write on purpose problematic code that would miss the misalignment. It would Purposely write the, you know, the misalignment detector in a way that wouldn't detect the ways in which this model w…”

“Well, a previous paper of ours, the alignment faking in large language models found that this will actually happen in some current deployed systems. So Claude three Opus, for example, will sometimes do this where it will attempt to hide its goals for the purpo…”

“There's this phenomenon that happens that we call context dependent misalignment where the safety training seems to hide the misalignment rather than remove it. It makes it so that the model looks like it's aligned on these simple queries where we, that we've …”

“The Claude, 3.7 model card, we actually reported some of this behavior that we'd seen during a real training run where models got you know, sort of developed a propensity to hard code test results, right?”

“Fortunately even these really bad models that we've been discussing here that try to fake alignment are pretty bad at it, right? So it's pretty easy to discover the ways in which, you know, this kind of reward hacking has made models very misaligned and it wou…”

“There's a couple of standards out there. MCP, which was developed by Anthropic, and Agent to Agent, and a few other things, and they're becoming fairly standard”

“We have an agent that listens to every conversation that is automated and said, you know, cause we run a resolution platform and we charge for actual resolutions, not just conversations. And so we need to know that the customer's problem was resolved. And so w…”

“You've effectively, I wouldn't say you've seeded enterprise AI, but you're losing enterprise AI to Anthropic.”

“Earlier this year, they were repping Anthropic in a copyright lawsuit, and one of the associates said that they had used Claude to help generate citations within a an expert testimony that they submitted for part of this copyright case and Claude had Hallucina…”

“This was the breakout year for every API business for AI coding. Like, I mean, Anthropic was the biggest beneficiary of that, but the cursors of the world, all of that, like, AI coding found its stride. That drove API businesses”

“Amazon's eyeing this ten billion dollar investment into open AI, which is fascinating because they already have many billions in Anthropoc.”

“My prediction is, while opening a dozen IPO, Anthropic at least sets a timeline. If not, you know, files in S-One next year.”

“So I'll say I'm not fully bought in on the co-work. Use case, but I could be wrong.”

“You know, they don't make image models, multimodal models, world models. They just do, you know, coding and language models, and they're very, very good at that.”

“Actually Apple wanted to go with Anthropic first, and they wanted to use Claude. This is according to Mark Gurman, said it this week. Turns out Anthropic, you know, was basically trying to chart and a lot of money, and then they eventually went with Plan B, wh…”

“In June 20, 25, so six, seven months ago, the average time spent on the mobile app of Claude was about 10 minutes. Today, in January 26, it's above 30 minutes. So, time spent per user on Anthropics Claude has Tripled in six months.”

“They initially went out to raise ten billion. They're actually going to raise around 20, and they had demand for five to six times that amount which is a crazy amount of money, right?”

“I mean, it's probably already lost the kind of consumer app battle to open AI with Google, you know, making some small inroads in that, but not really for a company of its size.”

“Microsoft and Nvidia this is according to your story, have committed to invest a total of up to fifteen billion in the company.”

“I think these other major cloud players, at least the ones who have the rival clouds, specifically Microsoft and Amazon realize that yeah, they probably need to align around basically anyone who's not Google, right? They don't necessarily care if it, I mean, t…”

“In recent months, NVIDIA CEO Jensen Wang has privately emphasized to industry associates that the original one hundred billion dollar agreement was non-binding And not finalized. He also said, he also privately criticized what he described as a lack of discipl…”

“In terms of OpenAI and Anthropic, I just think that beyond where their businesses are at and with these now massive fundraisers, as we're talking about, like, I think that it will push it out a little bit. And the real, Wild card is obviously, as it always is,…”

“Walters Kluwer went down 13%. Relics PLC went down 15.8%. LegalZoom down 18%. Thomson Reuters down 19%. The worst day in the company's history.”

“The ad was intended, by the way, They released it pre-Super Bowl. You know why? They wanted that ad to take off. People will pay attention to it if it's a Super Bowl ad, and they wanted it to take off so the potential business users of Anthropic would circulat…”

“A new Quen model came out yesterday that is shockingly close to the best Sonnet model that there is from Anthropic.”

“I think the gap between the truly great model makers of the present era, like OpenAI, the Anthropic, and Gemini very much in that mix, And everyone else is quite staggering”

“This just means that OpenAI and Anthropic need to understand that any kind of lead that they get is not going to be a long-lived one, and they really have to work hard and compete.”

“And so at Davos a few weeks ago, I think both Demis and Dario said some variation of, yeah, like if we were the only two groups building this technology, We would find a way to get together and figure out how to slow down this frantic pace. Like it's going too…”

“Cloud code doubled in usage from December to January, I believe in the past month, doubled in usage. It's already doing four percent of commits on GitHub. Anthropic went from zero dollars in revenue in January, 23 to a hundred million dollar run rate in Januar…”

“And the more that the companies become valuable and when they become public and public equities become tied to them, I just worry about worlds where we are reluctant to enforce the law on companies, even if they are breaking it because so much of financial pro…”

“The model is at times overly agentic in coding and computer use setting, taking risky actions without first seeking user permission. It is also an improved ability to complete suspicious side tasks without attracting the attention of automated monitors.”

“In five years from now, both, they will be, you know, a hundred times bigger in inference. Anthropic will be a hundred times bigger in inference. Gemini will be a hundred times bigger in inference and so on.”

“I think its involvement in this specific operation has been blown completely out of proportion. Just speculation reading between the lines here.”

“Free users on, on Claude are up more than 60% since January. Pastest growth in Claude's history. Daily signups have tripled since November. Every single day this week has consecutively broken the record for Claude's largest ever day of signups, and paid subscr…”

“Pentagon officials from the journal have reached out The defense contractors, including Lockheed Martin and Boeing in recent days to gauge how much they use Claude.”

“Anthropix artificial intelligence tool Claude was used in the US military operation to capture former Venezuelan president, Nicolas Maduro. The department, the deployment of Claude occurred through Anthropix partnership with data company Palantir. Whose tools …”

“They might also invoke the Defense Production Act, which would require Anthropic to supply its technology to the Pentagon the way the Pentagon wants, which would be unprecedented.”

“I think that the coding fight is going to narrow, but a year from now, I think they're still gonna be in the lead. And I still think that will be the biggest use case for these models as the vibe coding stuff continues.”

“It's simply conflicting culture. It's not a specific disagreement over technology that will be used in the moment.”

“Within hours of declaring that the federal government will end use of its artificial intelligence tools made by tech company Anthropik, President Trump launched a major air attack in Iran with the help of those very same tools. Commands around the world, inclu…”

“He's trying to position this as like, they are the peace broker here, right? Like that they are the ones who are going to sort of iron out these differences between anthropic and the U S government by cutting their own deal that paves the path. To sort of do a…”

“On Friday afternoon, Anthropic learned that the Pentagon still wanted to use the company's AI to analyze bulk data collected from Americans. This could include information such as questions you ask your favorite chatbot, your Google search history, your GPS tr…”

“I will give my hot take here, which is that This supply chain risk threat is never never manifests. Just never takes that takes effect.”

“Claude is now the number one app in the app store, which is wild.”

“The fact that Palantir and then Amazon obviously, and a bunch of others have used Anthropic services, like maybe the government just wasn't savvy enough to know just how integrated Anthropic itself was and that they can't just, again, like we were talking abou…”

“The kind of algorithms that you're going to be most likely to use in that context are much more deterministic than, say, like, Claude trained on the slop of the internet. And so, Anthropik's not wrong that their tech, like, isn't ready for prime time for auton…”

“If you look at the relationship between Anthropic and the Pentagon, Anthropic was the first frontier AI lab willing to do classified work. To support American national security.”

“They laid out this six-month off-ramp period for, like, real national security use cases, in part because they rely on Anthropix technology right now because Anthropix is the only vendor. Like behind the current curtain in a classified environment.”

“Less than 24 hours after the supply chain designation, The U.S. Government was using Anthropix technology in the context of Operation Epic Fury against Iran.”

“This does do long-term damage to anthropic because even if these laws, or even if the, let's say the supply chain risk designation ever makes it to them or is overruled, public sector companies, contractors will just in the back of their mind think twice befor…”

“A thing that Claude is definitively not doing, At least as far as I know. Like, I would be genuinely shocked. What is autonomous targeting on the battlefield today? Like that, it, I would be astounded if that was a Claude specific task.”

“And Anthropic was also, there was no dispute between Anthropic and the Pentagon about any current projects that Anthropic was doing. It wasn't like the Pentagon asked Anthropic to do something and Anthropic said no or had hesitations.”

“Some of the other frontier AI labs like XAI and OpenAI are already now willing to work on the classified side, and you know, Sam Altman is attempting to broker a piece, essentially, and create a deal that perhaps Anthropik could join as well.”

“When it comes to the use cases that Anthropic may be concerned about in different kinds of ways, I mean, the thing to remember is like, if you do business with the Pentagon, the business of the Pentagon is war. So you shouldn't be surprised then that the Penta…”

“And that's because the way that Anthropic's technology is often integrated within the Pentagon is through a Palantir product called Maven Smart System.”

“And ok, on the Pentagon side, my guess is they still end up coming to an agreement with the Pentagon.”

“Anthropic annualized revenue recently topped Nineteen billion up nearly three times from the end of last year and up 36% from two weeks ago.”

“It's going to be a change in Anthropics culture. I think Anthropic will become him and Anthropic will become much more closed off.”

“Anthropic put a prompt that you could copy to ChatGPT to export your memory. Like I was saying, memory's not sticky. They built the prompt that we were talking about, and you drop it in ChatGPT, it prints out your memory, you paste the memory in Claude, and th…”

“The Pentagon has formally notified Anthropic that it's determined the company and its products pose a risk to the US supply chain.”

“Anthropix said daily sign-ups have quadrupled since the start of the year. On Thursday, Anthropix said more than one million people are now signing up every day.”

“I will say the last two months have been absolutely wild, but it's more, it's starting to feel like COVID era extrapolation of every single business that we talked about, that the whole world is gonna move towards like remote work only and zoom is gonna be wor…”

“I actually, I think it's a mistake that Claude is leaning so hard into the consumer side of it right now.”

“Jensen Wong said the thirty billion dollars that he and NVIDIA invested in OpenAI might be the last. Why? Is it because he doesn't trust Sam Altman, or is it because, you know, he doesn't believe in the technology? No. He says the reason for that is because th…”

“Anthropic, by the way, which is saying that, you know, I mean, not saying that software is going away, and we'll get to actually some of the job stuff from Anthropic, but they're hiring multiple Salesforce administrators for a company that people say software …”

“So I do not buy that. In that I think LLMs will be, and I've experienced this myself, performative in a way that they think appeals to humans and hooks them emotionally.”

“So you're seeing ChatGPT really go towards like fashion, retail, transport, like mainstream consumer. You're seeing Anthropic go towards like premium data sets for finance, science, medicine, and so they seem to be diverging a little bit in those directions.”

“In two, three, five years, Onboarding to software should not be a thing. Like, you should be able to log in with a ChatGPT or a Claude, and that new software product should know everything about you and, like, set up perfectly to cater to you”

“If you look at the app store that they have enabled on ChatGPT, ChatGPT and then the App Store that Anthropic has enabled on Claude, they each have more than 200 apps, but there's only 11% overlap.”

“If you look at the gap between them and the number two Gemini on web, it's about still two and a half, three X. The gap between them and something like a Claude is closer to, like, 30 X. So even though a lot of these other apps are getting more attention, Chat…”

“Meta's new foundational AI model, which the company has been working on for months, has fallen short on the performance leading AI models the performance of leading AI models from rivals like Google, OpenAI, and Anthropik. On internal testing, it is now delaye…”

“If that's the core part of what OpenAI is trying to do, again, like, everything is reactive right now. As you said, Claude has a good desktop app that combines multiple platforms into one and makes it more usable. So then they're gonna do that. Claude, we star…”

“Anthropic is now capturing over 73% of all spending among companies buying AI tools for the first time. 73%. Just 10 weeks ago, the split with OpenAI was fifty-fifty, and it was sixty-forty in OpenAI's favor as recently as December.”

“The one lane I think they still truly have a path to success is consumer. And is it because of Google? Is it because they're just looking at Anthropic and investors are telling them that that's what's more attractive right now? To me, this isn't the bet that t…”

“Even if they don't use it for training, they can still store it for a long time, actually forever. They can still hand it over to a government under a subpoena. They could leak it if they were hacked in the future. So just because you've opted out of training …”

“If we were not in this war with Iran at this point, I think, you know, that would have been a major focus.”

“Now, I think Anthropic probably screwed up their negotiations with Department of Defense”

“Hegsef is going to declare, is trying to declare Anthropik a supply chain risk. That would mean that not only Anthropic couldn't do business with DOD, but any company, and virtually every major company in America does some level of business with DOD, they coul…”

“I've been trying to talk to all of the other tech companies to say, even if you are Anthropik's biggest competitor, you don't want this precedent set.”

“To put up, you know, the supply chain designation, which I don't believe has ever been designated against an American company, this is a death warrant.”

“I think like 99% of the code for Claude is, is written by Claude.”

“Across Asana AI, we use both OpenAI and Anthropic models, in particular for the AI teammates launched. We have chosen Anthropic's Opus through .6 model and that's what we're launching with right now.”