Claude Code Head Boris Cherny: Insane Growth, Tokenmaxxing, AI Agents' Next Frontier

I've just never seen growth this deep.

And then it just kept going more and more exponential. Quad code is 100%

more exponential. Quad code is 100% written by quad code. Co-work is 100% written by quad code. An increasing

number of features are fully written by quad code across anthropic and and products. So I want to hear your

products. So I want to hear your perspective on on token maxing and whether you think that makes up a large portion of the usage of the products that you're building.

I don't write code. I prompt Claude.

And actually nowadays, mostly what I'm doing is I have a Claude that prompts other quads. So I don't even talk to

other quads. So I don't even talk to Claude. I have a Claude that's talking

Claude. I have a Claude that's talking to my quads. Let's talk with Claude code head Boris Churnney about the product's explosive growth. What's next on the

explosive growth. What's next on the road map and whether all this is sustainable? That's coming up right

sustainable? That's coming up right after this. Welcome to Big Technology

after this. Welcome to Big Technology Podcast, a show for coolheaded and nuanced conversation of the tech world and beyond. We have a great show for you

and beyond. We have a great show for you today. Cloud code head Boris Churnney is

today. Cloud code head Boris Churnney is here with us in studio. We're going to talk all about the product, the way it's taken off, what's next on the road map, and of course whether it's sustainable.

Going to go into things like token maxing, token inefficiency, and then of course the future of knowledge work. So

no lack of topics to cover. Boris, it's

so great to see you. Welcome to the show.

Yeah, thanks for having me.

So let's talk a little bit to begin with about the growth of cloud code. Uh it's

been massive, right? I think at a recent event, Dario Amade, the CEO of Anthropic, talked about how demand for Anthropic's products has been up like 80 times uh year-over-year. I remember

speaking with him last year around this time and he was thrilled that Anthropic was at $4 billion ARR. That seems quite right now. The numbers right now say

right now. The numbers right now say maybe it's 45 billion, right? So, a 10x there, 80x demand. And the question is how fast the company uh can serve the

demand here. But talk about the portion

demand here. But talk about the portion that of demand that claude code makes up and what you've seen in terms of demand growth and the amount of people using this thing for an increasing number of people in

the world. I think the way that you use

the world. I think the way that you use agents and the way that you use AI, it's not just enthropic products, but it's quad code in particular. And you know, of course, for enthropic, there's a lot of different products. There's, you

know, there's quad code, there's quadi chat, there's quad design, there's there's co-work, there's like the API products. There's a lot of ways to

products. There's a lot of ways to experience anthropic. Um, but for a lot

experience anthropic. Um, but for a lot of people, quad code is their first introduction. And yeah, the growth has

introduction. And yeah, the growth has just been insane. It's, you know, when we first released it internally, it just skyrocketed immediately. And so before

skyrocketed immediately. And so before we even released quad code to anyone outside of anthropic, we felt that it's pretty likely that this is going to be a

hit. And around the time that we

hit. And around the time that we released uh Opus 4 and Sonnet 4, this was in May of last year, the growth just went exponential. And I've just never

went exponential. And I've just never seen growth this steep. And then it just kept going more and more exponential with uh with Opus 4.5. that was November and then 4.6 six that was February of

this year and then 4.7 it just keeps inflecting over and over and you know there's a lot of people on our team that have worked in tech for a long time and you know we worked on all sorts of

hyperrowth products like this is something you talk about in tech all the time these like unicorns and hyperrowth but even on the team we've never seen growth like this uh and so we're we're just trying to figure out how do we how

do we make it so everyone can continue to experience this uh how do we make it so we can continue growing at this pace and the pace that we uh expect in the

future which might be even steeper than it is today and uh we're learning a lot about about how to do this and uh how to how to keep scaling the services. So a

year ago it was clear that the bulk of usage of anthropics AI models was happening through the API right that would be like a company like a consulting group for instance putting it

into action at a bank and the bank using it to summarize some calculations. I'm

just throwing an example out there. um

that compared to the cloud chatbot it was far and away the API was the lion share of usage revenue all these things um does that still the case today or is cloud code overtaking that we have a mix

so you know like products play a much bigger role for anthropic than they did a year ago that's that's definitely the case uh product growth is accelerating it's growing very quickly API is also

accelerating and growing very quickly and for us we are investing in both we have to be a product company because there's kind of a lot of reasons for lab

to build products and you know this actually wasn't clear early on like very early on in anthropics history this is before I joined this was actually like an active debate should we even build products like is this actually like a

useful thing to do and it turns out it's very useful um you know for mind share but then also for safety um fundamentally we exist to study AI safety this gives us better tools to do

that we're also a small number of people and so most things in the world we will not build, right?

And so this is why we also have to provide a platform and we have managed agents and API and SDK all all these products so people can build on top and you know thousands and thousands of businesses choose to do that.

Yeah, it's it's interesting to hear you even answer the question saying that it's a mix. So I take it you're not going to share which is bigger right now.

Maybe maybe not right now.

Okay. But the fact that like it's not a clearcut the API is bigger. Um

maybe it is. But the fact that you even say it's a mix just shows the the fact that Anthropics owned and operated products are just growing massively. And

now so you know we've set the we've set the stage here that this is a thing something that's growing exponentially.

Um we've obviously we obviously have seen the anthropic revenue grow exponentially kind of alongside this product. This is a product that you

product. This is a product that you conceived of and built and run today.

Uh, I think that there's probably some people watching who are like, well, what is Claude code? Um, most of our our viewers obviously know what it is. Um,

and I was like, how do I write this like in a simple one-s sentence definition?

And I wrote that it's a way to build websites and software in plain English.

And then on the way over here, I was like, well, that kind of sells it short a little bit.

I mean, what would you describe it as?

I think that's actually a pretty good description. It's

description. It's all right. We'll take it.

all right. We'll take it.

I I think when a lot of people think about AI, they think about chat bots.

And you know, for engineers, that's what AI was, you know, maybe like a year and a half ago, before we started quad code, that's what AI was for most people. And

we realized at some point that the model was actually getting really good at coding and it's getting really good at using tools. And these are things that

using tools. And these are things that we've kind of always trained the model to do. And you know, this has kind of

to do. And you know, this has kind of been the research direction for a while.

It started to become commercially useful about a year and a half ago. And so for cloud code, we took this bet and we deviated from the way that everyone wrote code at the time because the way

that everyone in the world wrote code was using essentially a fancy text editor and we just thought maybe we can do much better than this and we could do something really really different than

what's been done before. It was very much a bet. And so we introduced, you know, quad code. And the thing that made quad code different from chat bots at the time was quad code can use tools.

And this is it. Like this is just the difference. It's with a chatbot, you're

difference. It's with a chatbot, you're going back and forth and you're talking, but an agent and cloud code is an agent.

It can use your tools, right? And can we just quickly define

right? And can we just quickly define the tools? So tools could be anything

the tools? So tools could be anything and you tell me if I'm wrong from using a browser to like logging into Cloudflare and then setting up some agent that way, right? So it becomes

less of what does this product do itself and more of like what can this product log into and then sort of do with a multiplicity of products you can use online.

That's right. It can it can connect all your different tools. It can use your browser. It can use your computer. Um,

browser. It can use your computer. Um,

even something as simple as like editing a file on your computer. You know, like a year and a half ago, there was no AI product that could actually do that. But

this is the first thing that Quad Code was able to do. It could edit a file on your desktop. If you have a bunch of

your desktop. If you have a bunch of files on your desktop, it can organize them. And so like Quad Code and Co-work

them. And so like Quad Code and Co-work have this access if if you choose to give it.

Granted.

Yeah. And and you know, it it can it can do this. And this is magical. It's this

do this. And this is magical. It's this

tiny difference completely changes the way that people can use this product.

And it totally changes what this product can do for you.

Yeah. I mean the fundamental thing I think just to drill down here is that um it seems like AI has shifted from sort of like as great at autocomplete right

because at the fundamental layer AI is just predicting what comes next. Um

predicting you know if if you're using machine learning and applying it on a large data set predicting whether you might default on your mortgage and whether a bank should grant a mortgage.

when it comes to a sentence predicting the next word with code predicting the next bit of code in a sequence right so I think that was gener gen one but what

you're talking about now is the machine is actually just able to go and after you give it this natural language prompt code itself hook into tools and then do

things for you and so correct me if I'm wrong but the the use cases here have gone from developers hooking into it and writing code with cloud code and we've seen this explosion I guess largely

driven by them but then by a secondary force by non-technical folks people like me who can build software um by directing the AI agent which is cloud

code to build a piece of workflow software for them or a website or to take control of your computer via something like cloud co-work which is sort of the um maybe I would call it the

easier sister product and saying well you have access to my uh to my browser Um, you know what type of flights I like to book? I need to be in India in a

to book? I need to be in India in a couple of weeks. Book the flight.

Yeah. Yeah. Exactly. I I I actually just used co-work to book uh a bunch of flights. I'm I'm going to be flying a

flights. I'm I'm going to be flying a bunch this month for, you know, we have like Code with Claude coming up and in London and Tokyo. And there's some other stops along the way. And I went back and forth with co-work. And I was like,

"Okay, I need to be in these uh in these places at this time." And it was five stops. It was like a lot of cities. And

stops. It was like a lot of cities. And

here's roughly the schedule. um look

through my email, look through my calendar, and just double check it, make sure I'm not missing anything. It found

actually two stops that I was missing and also a couple dates that I told it wrong. And it just found this by looking

wrong. And it just found this by looking at my email after, you know, I I asked it to do that. And then I told it to book the flights and I went and, you know, was coding on something and I was

I was just doing work and I came back an hour later and it booked eight flights and five hotels and uh one of the hotels was kind of incorrect. was in the wrong area. I asked it to rebook it and change

area. I asked it to rebook it and change it and it was done. That was it. And I

actually, this is something that I try every time with co-work with quad code.

I have these sort of like uh test cases.

So these sort of um like a common thing that I would do and I just retry it with different models and you know as the model improves. This is the best result

model improves. This is the best result I've ever gotten. And there's something about co-work combined with Opus 4.7 where it's able to do this. And I think

one of the hardest things for me has been as the model improves, you constantly have to readjust your expectations of what it can do. And if

you talk to people, especially engineers that use the model a year ago, they might and and they didn't use it since.

They might say something like, "Oh, well, you know, it's not very good at coding and you know, I don't trust it to write more than a few lines at the time at a time." Because that's what the model was a year ago. It wasn't very

good yet. And if you fast forward to

good yet. And if you fast forward to today and you sit down these people and you know they they they try the new model and as as like a lot of people have been doing an increasing number of

engineers it's just a completely different experience the the capability is completely different and I I think this is the first technology I've used like this where

every month there's a step change in what it can do and as a user of this technology it's just quite hard because you have to kind of keep retraining you have to keep retrying you always need

this like beginner mindset to to retry the technology and use it for a thing it was not good at before because the next model might might just do it perfectly right and so I think this is the vision

the way that you're outlining it is effectively previously when you would use technology you would be subject to the interface you would have a software company that built for scale um but you

would get a lot of features that maybe weren't applicable to you would have to go through all these bells and whistles whenever you were trying to book something even though you knew what you

wanted and you wouldn't have a website that would know your preferences. Now it

sort of shifts the paradigm where you have again it's it's an agent. It's

something that goes out and does things for you and can potentially shape your experience online the way that you want it. And that's that is I think what

it. And that's that is I think what people are seizing upon and that's why we're seeing why you're seeing really the explosive growth. But now I want to pressure test the thesis a little bit

and uh bring up some things that make me curious how much of this is real uh and how much of this is just um unbridled enthusiasm at the potential but maybe

stuff we should have a reality check on.

Um and the first thing is that there is such great demand. U but the question is how much of that demand is pure demand

versus demand that's gamified. Um, and

there is a practice that's going on within Silicon Valley and outside of it that's called token maxing. I'm sure

you've heard of it. It's where um, companies are have a mandate where people are supposed to use lots of AI tokens by running their AI agents as

much as they can. And then those who run the, you know, use the most tokens are like rewarded on a leader or on a leaderboard or meet a goal of AI actions that they have to take

as opposed to physical actions. So I

want to hear your perspective on on token maxing and whether you think that makes up a large portion of the usage of the products that you're building.

Yeah, I I don't think token maxing is a is a large percent. Um the the way that I would think about it is um you know before Anthropic actually I used to work at a big tech company at Facebook.

I was at Facebook which is one of the companies that's token maxing for.

That's right. That's right.

Yeah.

And one of my responsibilities was the health of all of the code across you know the across meta apps. So this is like Facebook, Instagram, you know

WhatsApp. And one of the reasons that we

WhatsApp. And one of the reasons that we care about the health of the code and this is essentially things like code quality is if the code is really high quality engineers are more productive.

And there's like a big team of people that worked on productivity. And before

models before claude you would work for a really long time and you would see maybe like a one two 3% improvement in productivity per engineer over the course of a year like something like

that. And that was like a pretty big uh

that. And that was like a pretty big uh improvement. Um and it was like very

improvement. Um and it was like very hard one. You essentially had to try a

hard one. You essentially had to try a lot of ideas and eventually you find something that improves productivity like this. And what happened with Claude

like this. And what happened with Claude is now many companies including Anthropic and all of our biggest customers are reporting gains on the order of hundreds of percentage points.

And I think the the last number that we reported is the the amount of code written per engineer at Anthropic has grown something like 250%.

since we introduced quad code and this is uh with while keeping code quality and reliability and all these things uh kind of stable. So without those things

regressing the volume of code has has grown a lot and so um this kind of productivity impact I think is just like very new and I think people are trying to figure out how do we get this there's

a lot of companies asking like how do we how do we get these kind of benefits because a lot of companies are seeing it and then some are still figuring it out and I think my advice is almost always

the same. The first thing is just give

the same. The first thing is just give everyone tokens let people experiment. I

wouldn't necessarily recommend token maxing, but I would recommend let people experiment so they don't have to ask for approval for every token. The second

thing is give people psychological safety because a lot of times when people are innovating and they're building tools that make them more productive, they're changing their own workflows to make them more productive.

They try a bunch of ideas, some of them might not work and then some of them work. So you want to give people this

work. So you want to give people this kind of psychological safety so they feel okay experimenting with it and finding these new processes.

And then um the thing that a lot of companies see is the productivity improvements and the innovations do not come from the people you expect.

Back in the old days, you know, everyone could point out like these are my most productive engineers. But I think

productive engineers. But I think nowadays a lot of the improvements are coming from people you just never would expect. It could be like an accountant

expect. It could be like an accountant somewhere in the corner of your org that just automates like accounting in a way that no engineer would have thought of.

Um, it could be some marketer automating like marketing in a way that you never would have thought of. It could have been like a new grad software engineer that just built something amazing. And

this is something that just like didn't happen before. The challenge is you

happen before. The challenge is you can't identify these engineers and these people ahead of time. You don't know who they are and it's almost always going to surprise you. And so the thing you want

surprise you. And so the thing you want to do is let people experiment, give them safety, and then once there's some kind of use case that scales up, that's when you think about optimizing it. But

you don't want to optimize ahead of time. So I don't know if doing it in a

time. So I don't know if doing it in a competitive way works for some companies with their culture um then I think that's great. If for other companies the

that's great. If for other companies the way they want to do it is just kind of create safety and create space for engineers to experiment which is what we do at Anthropic then I think that's great too. It really depends on the

great too. It really depends on the company.

Yeah. And and I'll say look I use a lot of tokens. Uh I'm in the tools all the

of tokens. Uh I'm in the tools all the time. I think Cloud Code and Cloud

time. I think Cloud Code and Cloud Co-Work have both been pretty great for my business. I'm a solo operator,

my business. I'm a solo operator, although that kind of sells it short because I have a team of people behind me um that help me um mostly in a part-time basis, but um that's for a

different show. But but I I do wonder,

different show. But but I I do wonder, you know, when I read these stories, um the large corporations are largely making up big big percentages of these

budgets and the incentives, you know, and again like I started the show saying how sustainable is this? The incentives

are are bad in some of these places.

This is from the Financial Times recently. Uh Amazon staff use AI tool

recently. Uh Amazon staff use AI tool for unnecessary tasks to inflate usage scores. Some employees said colleagues

scores. Some employees said colleagues were using the software to automate additional additional unnecessary AI activity to increase their consumption of tokens. They said the moved reflected

of tokens. They said the moved reflected pressure to adopt the technology after Amazon introduced targets for more than 80% of developers to use AI each week.

I' I gut checked this with an Amazon employee. They're like, "Yep, uh, this

employee. They're like, "Yep, uh, this is what's happening." They told me, "I triggered an automation that runs for hours and then gets deleted every day in order to meet these targets." So you

said you don't think that this token maxing stuff is is a big part of demand.

Um is there anything that you can see on your end to indicate that it's not that this is an outlier and not the rule in most places?

Yeah, this is uh I don't know how many companies are doing this token maxing thing. I've I've heard of it as a trend

thing. I've I've heard of it as a trend you know a little bit. Um if you look at Quad Codes customers uh we have just many many many customers. Um, so it's not like uh, you know, there's like one

company driving the usage. It's not like that. I I I do want to kind of step back

that. I I I do want to kind of step back a little bit and just think about like how does this kind of change happen?

Because I think the goal of what these companies are trying to do, I don't want to speak for them, and I would recommend just talking to them.

But the the goal of what they're trying to do, I think, is probably like organizational change and business process change. How do you make it so

process change. How do you make it so your company benefits from AI? And this

is often unclear. It's very dependent on the company because every company has a different business, a different culture, a different org, a different way of doing things. There there's this old uh

doing things. There there's this old uh Harvard Business Review article from the '90s, which I just love, and I I forget the title, but it but it was something like uh computers are here, why is no

one seeing the productivity impact?

And this was a big question, right? It's

like to us it's obvious computers make us more productive. This is just incredibly obvious today.

But in that '90s, this was not obvious.

And what was happening is personal computers were being adopted. They were

replacing mainframes and now they're affordable. So the average company, the

affordable. So the average company, the average startup can can buy one. You

don't have to spend, you know, millions of dollars on a mainframe anymore.

But there was this challenge and there was this paradox. Companies were

adopting it, but they were not seeing productivity improvement. What's going

productivity improvement. What's going on? And so this Harvard Business Review

on? And so this Harvard Business Review article, it made the case that in order to get a benefit from computers, you have to restructure your your your whole business process around computers. They

have to be at the center of the way that you do things. And if you still have like paper, you know, filing cabinets and you have a bunch of drawers full of stuff and it's still a paper and pen kind of physical process and there's a

computer somewhere on the periphery, you're really not going to benefit. But

if you throw away your filing cabinets, you throw away your, you know, desk drawers full of, you know, papers and you put a computer at the center of it and that's the way that you do you all your business process, then you benefit.

And there was this split between companies. Some were doing this and they

companies. Some were doing this and they were doing this fairly painful change and they benefited from it and then others didn't. And I think it's kind of

others didn't. And I think it's kind of the same thing now. A lot of companies are trying to figure out how to benefit from the productivity impacts of AI and uh there's just a lot of experimentation

and everyone is trying different approaches to to figure out how to how to benefit from it. I don't I don't think there's one right approach.

Okay. And um look, I I think that when we see something grow as fast as cloud code has grown and as fast as as fast as Anthropic has grown um

it's good to just kind of talk this stuff through and and it's good to hear your perspective. So okay, that's token

your perspective. So okay, that's token maxing. Um now tokens of course are the

maxing. Um now tokens of course are the output of the model like the words or portions of words that the model outputs and the words and portions of words that go into it, right? Um and that is how

these companies charge and the more you have the more data centers you need etc etc. um you know as these models get better they they haven't and well let me

put it to you this way sometimes I wonder whether they're as efficient as they can be um these big models can sometimes do a lot of work use a lot of

tokens even if the output is great uh people wonder well is this sort of just driving up token demand where it could have been a really easy process and um and the models are are expending many

many tokens and not getting there as as efficiently as they Let me give you an example. Um, I've been using Claude

example. Um, I've been using Claude Co-work to make PowerPoint presentations.

It's really good at it. Um, and I've been using the Opus 4.7 model and a couple of times I've said, "All right, um, you know, you're working on this uh

this ship it as a PDF and um, it just starts to lose its mind. it cycles and it uses as many tools as it possibly can

and you know it's just seems unable to ship the PDF and um and eventually I kept telling it no you're making this PowerPoint you know where it is ship it

and it goes uh I owe you an apology I went down a rabbit hole worrying about a constraint that wasn't actually blocking us the files there and then it shipped it I mean talk a

little bit about the efficiency of these models um and whether that is a legitimate worry that you know as we've seen the growth part of it is these like

loops that a model like Opus 4.7 might find itself in to do basic tasks.

Yeah, generally when we think about models there's a few different aspects of it. um one is just how intelligent is

of it. um one is just how intelligent is it another one is how fast it is and another one is how efficient it is and we generally try to move all these together between these I think we should

probably optimize for intelligence that's the most important thing so even if it's like a little bit less efficient but it's more intelligent and lets you do more things that's really useful because the efficiency optimization comes after after we make it more

intelligent then we can make it more efficient so it's sort of kind of we do one then then we do the other we've been experimenting a with like how exactly we give people control over this because we

don't always know the right default sometimes like when you're using it you know better you know better um and so one mechanism that we had for this is picking a model so you can pick you know

opus or sonnet or haiku um another mechanism that we've been experimenting with is effort opus is like the biggest sonnet middle haiku smallest that's right that's right and this is just like the size of the model

right um and then there's effort and effort is essentially how it you know I think the word is actually really uh descriptive it's how much effort do you want to put into it and uh you can set this we have

a recommended effort so you know for example to maximize intelligence for opus 4.7 you want to use extra high or maximum effort but if you wanted to use west tokens you can pick like medium or low effort and this is a control that

you have yeah I talked about this on a show recently and we had a commenter that came in and I was of the opinion that this will these you know bigger models will find a way to become more efficient

on like the export the PDF thing. Um, we

had a commenter come in that wrote, "Alex, they can't fix things like that PDF problem. It's inherent to LM

PDF problem. It's inherent to LM technology and it's the biggest barrier to useful widespread dissemination and usage of agentic AI." I think I'm going to try to translate that what they were

trying to say is we talked about predictions earlier that this is all probabilistic. It's sort of predicting

probabilistic. It's sort of predicting the next word. You don't get the same answer from an AI agent twice. And so

therefore this type of thing is a feature of the way that they work and not fixable. What do you think?

not fixable. What do you think?

No, I I don't think that's right. When

when you think about like okay, let's zoom out a little bit. Y

so engineers are the first adopters, right? Like engineers started using

right? Like engineers started using cloud code like a year and a half ago and uh you know this is before non-engineers were using agents in a meaningful way. This is you know before

meaningful way. This is you know before co-work and so on. If I think back to what cloud code was a year and a half ago, it wasn't very good. I could use it to write a little bit of code, but if I

really trust it to build an entire feature or entire product, it wouldn't turn out well. It did the same thing like it would go in spirals and the quality wasn't good or, you know, it built it and either the code was bad or

it didn't work. And at some point, it just started to get better. And as the model improved and as quad code improved, the result just got better and better and better. And so you fast

forward to today, quad code is 100% written by quad code. Co-work is 100% written by quad code. An increasing

number of features are fully written by quad code across enthropic and and products.

And this is something that we hear from customers also.

I did a I did a talk at Y Combinator, you know, the the the the um the startup incubator yesterday and I asked people to raise their hands.

you know everyone everyone's using quad code and I asked them you know raise your hand if a 100% of your code is written using quad code today about half the hands went up and then you know I asked people raise your hand if 0% of

your code is is you know written with AI there's like one hand that went up and there's a room of like a hundred people power to that person um and you know there's still room for this obviously and then everyone else was somewhere in the middle you know

it's like most of their code is written with quad code but not all of it but that's kind of the place where the model was at today it was not there a year ago a year ago it was not good enough for this and so this is exactly what you're

seeing play out with co-work right now it's still early you know we released it what like a a few months ago um it's it's going to keep improving it's going to keep getting better as the product gets better as the model gets better but

this is early days I think still everyone using co-work today is an early adopter everyone even using AI today is an early adopter there are so many people in the world and most people have

not tried AI in a meaningful sense um so there's just like there's a lot more room to improve this.

Yeah, we're hosting an event here in San Francisco on June 18th and a lot of the marketing material I've turned out with co-work. Now, I go back and forth. I

co-work. Now, I go back and forth. I

don't let it oneshot it, so I'm looking at the copy, but I do things like uh you know, upload um you know, our our you know, download statistics to sort of show the growth of the podcast. And um I

give it the names of the speakers and it like is amazing at saying building a perspectus. Here's what the event's

perspectus. Here's what the event's going to be. Who's here's who's going to be in the audience? Here's who's

speaking. Here's why you should be there. Here's how to get in touch.

there. Here's how to get in touch.

Insane. It's so good.

What What was your uh what was your feeling like the first time that you used it and the first time that you saw like the agents use your tools?

Well, I I mean obviously I've sort of enabled everything. Um, so and I think

enabled everything. Um, so and I think this is kind of an experience that many people have had where you there's a browser extension for Claude and you realize that you can only get the benefit of this or you you'll get most

benefit by letting Claude take over your browser and do things for you. Um, and

the experience is kind, it's almost the same as I had with a Whimo where um, those first couple turns I was like white knuckling and like watching and like should I approve reading everything and then you start to trust it a little

bit and you just hit approve approve approve right in the Whimo the same thing. you're like, "Okay, this looks

thing. you're like, "Okay, this looks like it's not going to kill me." And

then 5 minutes later, you're on your phone as the AI does the work. And that

was my experience with code and co-work.

Is that does that sort of track?

I mean, this is like my experience, too.

It's like it's I think it's like any technology. Um, I was watching someone

technology. Um, I was watching someone that's uh it's it's like a friend that's that's been learning to use uh co-work over time and like, you know, she she's not an engineer. And um there's this use case the other day like her there was

like a language input on on the computer where you can kind of choose between languages on a laptop and there was some issue with it and she couldn't figure out how to fix it. And so before what she would have done is go to Google and ask like hey how do I fix this you know

this issue that I'm having with my computer and this time she just like asked co-work and a co-orker was like cool let me take a look can I can I use your computer and she said yes and it took over the computer and it gets this

kind of like orange glow and you get to watch as co-work open settings and it sees what's going on with the language picker and it diagnoses it and it fixes it and um you know you're still in the

driver's seat so you you can see this happening you can monitor it. it's not

happening in the background or anything.

Um, but it's just it's magical. And I

actually did like my instinct was to open Google. And so so it's funny that

open Google. And so so it's funny that like for her, she went to using co-work for this. And this is actually something

for this. And this is actually something I feel all the time. I I think for people that have kind of grown up with these products and they've seen previous versions, they might not be as ambitious as they could, but for people that are new to the products, I often see them

using quad code and co-work for things that I wouldn't have even thought of.

And it's just like amazing. It's It's so creative. Um, and I I learn a lot every

creative. Um, and I I learn a lot every time I see it.

Yep. Now, the biggest drawback right now, I would say, and I've seen you reply to people on X about this, um, is the rate limits. Like, when I see people say, "I've given Cloud Code a shot, but

I'm I'm kind of done with it." It's

typically because they've hit their token allotment and it only works for like an hour for them.

And then they have to wait four to use it again. and they look for

it again. and they look for alternatives. Um,

alternatives. Um, what do you think the rate limits have done to the ability for your product to grow and what is the plan if there is

one to make people be able to use this without those rate limits?

There's uh this is something we're actively working on. The reality is a very small percent of people actually hit their rate limits which is surprising for surpris for pro users

it's a little bit higher for max it's actually quite low and I think the thing that you're saying when people talk about it is um there's a couple things happening one is that we actually reduce

the peak rate limits um and that's now rolled back and we've actually doubled rate limits um so we're giving people more rate limits but there was a brief period where um we reduced them and so

people were running into The second thing that's happening is cloud code is actually quite extensible.

Um, and so people can use plugins, they can use all sorts of integrations. And

some of these use tokens in a pretty inefficient way. And so thing that we've

inefficient way. And so thing that we've been working on is surfacing this to you. So users can decide uh do you want

you. So users can decide uh do you want to use this plugin or do you not? So you

can see kind of what percentage of your tokens goes to it. And then I think the third thing is there's a lot of people that have just increasingly become power users. Like first when we released quad

users. Like first when we released quad code, you know, you ran one quad at a time. Nowadays, I'm running, you know,

time. Nowadays, I'm running, you know, like on my computer, I run maybe five at a time. And then every night I run like,

a time. And then every night I run like, you know, not every night, but most nights I run like hundreds of quads at a time all in parallel. Yeah. Hundreds,

sometimes thousands. And this is something that I just like wouldn't have imagined a year ago. And obviously this uses a lot of tokens. And there's a lot of people that are figuring out these new workflows that are using a lot more

tokens. And this is sort of like at the

tokens. And this is sort of like at the edge of what you can do with a max plan.

Um, and you know, this is why you can just like pay using API also. So if you just want to have as many tokens as you need, um, you can do this too. And this

is what a lot of enterprises do.

Right now, it wasn't long ago where I'm pretty sure Dario uh, Anthropics uh, CEO was referring to OpenAI and talking about the spending on the buildout and

he and he's talked about this afterwards. He said, "I'm trying to be

afterwards. He said, "I'm trying to be disciplined in the way I spend, which is still spending many billions of dollars on data centers to enable this stuff like you're talking about and others

which we think is open AAI are yoloing, right? But um now OpenAI is is doing

right? But um now OpenAI is is doing this too with codecs and you could call it yoloing, but they have a lot of data center capacity that they've built." U

how do you think about that? because you

know when people do hit these rate limits they may just go over to codeex um it's pretty intense competition so um how do you think about that how does

anthropic think about that internally that you know I mean at least from the outside perception is that um this added discipline on data center buildouts might end up losing users in the most

important product battle that your two companies are engaged in.

Yeah. So you know f first of all our growth has never been faster than it is today. Um so you know for quad code the

today. Um so you know for quad code the growth is accelerating and I think because most people don't actually hit rate limits very often. Um it's uh uh

it's actually not not a huge issue for the people that are we are laser focused on improving the experience and so we doubled the 500 rate limits. Uh we are

announcing today that we're increasing the weekly rate limits and of course we announced the new Colossus capacity which you know we brought online to serve all these new users via Elon Musk.

Via Elon. Yeah. Because this I mean this growth is just no one no one would have predicted this. This was just beyond our

predicted this. This was just beyond our wildest forecasts. Um and so you know I

wildest forecasts. Um and so you know I think for us what matters the most is we we need to serve our users. We want to make sure our users are really happy. Um

and we're doing everything we can to to make that happen. Are you surprised by Codeex? How do you view them as a

Codeex? How do you view them as a competitor?

I think there's always, you know, there's always copycats, there's always competitors. Um, for me, it's uh it's

competitors. Um, for me, it's uh it's flattering and uh I think it just forces everyone to do better. Um, so you know I for me the thing that I care about the most is

just doing the best job that we can to serve our users and we encourage everyone on the team to you know talk to users every day and um you know just make keep making the product a little

bit better every day. So this is what I care about the most.

Okay. Um I want to take a break but we have so much more to cover. I want to talk about how this extends beyond code uh the future of the chatbot and then maybe talk a little bit about we have I mean I could go through our agenda. We

really need two hours. So, um, why don't we take a break and come back and get to as much as we can right after this. And

we're back here on Big Technology Podcast with Boris Churnney, the head of Claude Code at Anthropic. Boris, it's

great having you here. Um, like I said, I'm in your product daily, so it's really fun to speak with you about it.

Um, we talked a little bit about this, but I think one thing we should highlight is that this is really going to extend uh beyond the chatbot. We

talked about booking flights. Uh I

talked about it uh with marketing presentations uh and you know the week that we're talking you have a a new use case out where um claude co-work can be

used for small businesses including uh taking over QuickBooks and doing some bookkeeping. Um where where does this

bookkeeping. Um where where does this go? I mean what do you think the broad

go? I mean what do you think the broad roadmap where does where does the broad road map take you?

We're thinking about a few things for quad code and for co-work. There's a few few big themes. One is improving intelligence and you know I I think almost all of this is just the model as the model improves we can do more and

more ambitious work for coding it used to be writing a line of code at a time now it's building entire features or entire products for co-work used to be you know like you know it's started pretty recently but it was like you know

making a document and now it's things like booking flights combining many tools doing doing your QuickBooks um so this this frontier is improving and and moving just very very quickly We're also

thinking about how to do longer running tasks for cloud code. We recently

shipped this thing called auto mode. And

auto mode is uh essentially a replacement for permission prompts.

Before what we used to do is whenever the model uses a tool, plot would ask you is it okay if I use this tool? And

you know, usually you just say yes and you get kind of tired of saying yes kind of over and over.

Always allow. That's the button to hit.

That's right. That's right. Um, but it's actually very important for security that you're very thoughtful about this.

And the thing that we were realizing is actually instead of being thoughtful about, you know, every prompt because we're showing people so many of these dialogues, they just kind of got fatigued and they would just say yes or,

you know, always. And so auto mode is the answer. And this is a new way of

the answer. And this is a new way of routing these tool calls. And the way that it works is whenever Claude wants to use a tool, it asks another Claude,

is it safe to use this tool. Claude has

some of the context. It doesn't have all the context. And there's also a number

the context. And there's also a number of layers of safety checks. And we spent months iterating on this to make it really safe. There's thousands of

really safe. There's thousands of different benchmarks and evals that we use to make sure that this is safe. Um,

and essentially we found both in the laboratory setting and now we're finding in the wild this is safer than what we had before. So, as a user, it's a really

had before. So, as a user, it's a really nice benefit because you don't have to sit there and say yes over and over. And

actually, the result is better because if there's one unsafe command buried somewhere in this big list of things that quad asks you to do, you might have accidentally said yes. But actually, if

you ask a second quad using auto mode, it's not going to say yes. So, this is kind of one big investment.

Um, maybe the third big one is just running more quads in parallel. One of

the cool things about quad and this is something that we started to see pretty early with quad code users is actually very few people nowadays run one quad code at a time. Most people run many many quad codes, you know, ranging from,

you know, a few to thousands. And with

co-work, we're starting to see the same exact thing. As you get more comfortable

exact thing. As you get more comfortable letting co-work run, you start a task and then you start a second task and you move on and you just do more in parallel. And I think there's just a lot

parallel. And I think there's just a lot of opportunity to make this experience very nice and to make it more obvious for people. How how do you do this? When

for people. How how do you do this? When

do you do it?

Right? And and it probably extends to the way that you use a chatbot, right?

And it's interesting because Anthropics had this um kind of interesting relationship with the chatbot. Started

out as technology first, decided to build the chatbot, ship claude, and then just kind kind of moved more towards enterprise. like you looked at all the

enterprise. like you looked at all the charts and um and Claude was always at the bottom u but now you're seeing Claude's usage rise and I have a thought

and I'd love to check this by you that the future of the chatbot is is not like I'm going to give you a question and you'll give me an answer. It's I will give you a question or you know talk to

you about a problem and you the chatbot will then suggest some sort of action you can take on my behalf. Like right

now I'm talking a lot about a trip to India and what I think I'm going to get back in the future is this thing being like like what you said not having this like secondary step between having to go

there and and book the flights a more proactive chatbot that's going to say okay let me take the let me take care of this for you. Is that the right direction? Like am I thinking about

direction? Like am I thinking about that?

I could see that. I could see that.

Yeah.

Are you working on it?

Asians are the future and you know we're trying all these different experiments.

There's some stuff that we're trying that's like this. Yeah.

Okay. Um but there is a limit here, right, to what this can do. Um a funny way people have talked about the limits of the thousands of clouds that you can

run in parallel is kind of looking at who Anthropic is hiring. Um my favorite job listing on the Anthropic site is that you're hiring Salesforce

administrators. Um, you're also hiring

administrators. Um, you're also hiring consultants to help enterprises uh deploy this technology. Um, and many are viewing that as like a sort of tacit

admission that this stuff can only take you so far. Here's uh Wharton professor Ethan Malik on it. He says, "You will know that the AI labs believe in artificial super intelligence when they

disband their newly formed consulting, sorry, forward deployed engineering groups. As long as people are required

groups. As long as people are required to figure out how AI is useful and do organizational change and systems integrations, jobs seem pretty safe."

What do you think about that?

Yeah. Um, when you look at the kind of engineering that I do, I don't write code. I prompt Claude. And actually

code. I prompt Claude. And actually

nowadays, mostly what I'm doing is I have a Claude that prompts other clauds.

So I don't even talk to Claude. I have a Claude that's talking to my quads. And I

think in engineering, you've seen just this explosion in the amount of leverage that a single person has. It's about how how how big of a business can a person build? How many products can one person

build? How many products can one person support? The leverage that one engineer

support? The leverage that one engineer has now at Anthropic is just insane.

And I think we're starting to see this across other disciplines too. So we're

starting to see this with the marketers that are, you know, using cloud to do things. We're starting to see this also

things. We're starting to see this also for forward deployed engineers that are using quad code to build implementations. We're seeing this for

implementations. We're seeing this for our sales team cuz, you know, actually at anthropic, I think like half the go to market team uses quad code and the other half uses core. You know, I I think everyone's using all these

products. Um, and so the thing that

products. Um, and so the thing that we're seeing is the amount of leverage an individual has goes up and we are still bottlenecked on the number of good people. And so even if the leverage per

people. And so even if the leverage per person goes up, you still just can't hire enough good people because the demand is so insane and there's so much more to build. So that that's still the

bottleneck for us. But I would say like if people would people would argue that if this stuff was so powerful um you could say take a look at the way my

sales organizations operates and then configure Salesforce that way with a prompt. Is this and another people

prompt. Is this and another people another example people give is u I'll believe that anthropic has very powerful AI if they let it let it uh handle the

the IPO paperwork and don't hire an investment bank. Are these unfair tests?

investment bank. Are these unfair tests?

Well, we're starting to see there there's one person on the team that was using Claude to do their taxes.

Um, you know, I I would not necessarily recommend this, but I'll admit I've run my taxes through Claude and compared it against my accountant and it was pretty close.

Yeah, I did the same thing.

Folks, not suggest not saying you should do that, but it is it's an interesting use case.

That's right. But I but I think fundamentally what people are missing in this conversation is in the end it's a person that has to talk to Claude to ask Claude to do this thing. So even if

Salesforce is automatically configured and you know it's not a person pressing all the buttons it's Claude doing it.

Someone has to ask Quad to do that. And

if you have to configure Salesforce in you know a bunch of different ways it could actually be a full-time job to ask Quad to do this. And at some point Claude is going to become really good at asking Claude to do this. And that

person is going to be asking Quad that asked Quad to do this. And this this chain will just keep getting deeper. But

in the end, you still need people that that are piloting this.

But maybe they're just job is just asking one question then in the future.

Yeah. But imagine how much leverage that has asking the right question.

That's true. That's a good point. U so

you know, we talked about Salesforce, so we have to talk about the SAS apocalypse. um you have some interesting

apocalypse. um you have some interesting views on the type of software companies that will be safe as we get more automated programming and those that that might be in trouble and you talked

you've talked previously about the different modes that exist and which modes are more important and which modes are less important. Can you just share that briefly, you know, while we're talking about it? There's this really

good framework called the seven powers for talking about modes in business. You

know, there's so many of these frameworks for this, but this is my favorite. Um I actually studied

favorite. Um I actually studied economics in school. I I didn't study computer science. So this is still kind

computer science. So this is still kind of the way that I think is in terms of these kind of frameworks. And there's a lot of these different modes in business. And some companies have one

business. And some companies have one mode, some have a few modes. You know,

they have like a portfolio of modes. Uh

there's a bunch of these modes. So like

one is uh scale economies. So as you scale up your production, then there's increasing returns to scale. Uh another

one is network effects. So this is like a you know like a messaging app or something like that. The more people that are on it, the more valuable it is for any person. Um another one is uh switching cost. There's another one

switching cost. There's another one that's process power. I I think most of these modes are still going to matter and relatively some are going to increase in importance over the next year and some are going to decrease in

importance. One that I think will

importance. One that I think will increase in importance is something like network effects because it doesn't matter who's writing the code. It

doesn't matter if it's an agent at the core of your product or or something else or if there's intelligence in your product. If there's a network effect in

product. If there's a network effect in your product, that's still going to matter. Some modes get less important.

matter. Some modes get less important.

And this is for example switching costs because if you want to switch from vendor A to vendor B, you can, you know, you can just ask Quad to do that. And

Quad is going to get better and better over time at it. And so I I think as a company, a thing that you should be thinking about is what are your modes?

And I think a lot of the largest companies just have many many modes.

It's not it's not just one thing cuz the way you get to a scale and the way you build a defensible business over time is you accumulate these modes. You need a number of them. Um but yeah, I I would just think what's going to be more

valuable in here and what's less valuable? I think that um when you think

valuable? I think that um when you think about these different software companies though if you're using a cloud code u do the most all modes kind of blend away

because you could potentially be in this like one app that is interfacing with all software which means therefore there's really only one software company.

Yeah. I mean there there's just like a lot of ways that this could play out. I

think something like this is possible, but it it seems a little far-fetched to me cuz if I think about, for example, like let's say I'm using a messaging app, how do I decide which app to use? I

use the app that my friends are on that I can that I can reach. So, it doesn't matter if I can build a really awesome app for myself, which I can do today.

Like, I can build a great messaging app with quad code in like a few hours. It's

still not useful because I can't talk to my friends.

But this is the example. Exactly. You'll

have you can you can um fact check me on this. You're gonna have an agent in your

this. You're gonna have an agent in your messaging apps that's going to let you know when your friends have messaged you. I know you use cloud code on your

you. I know you use cloud code on your iPhone a lot, right? So then you will just see the notification and you'll speak back to people. All your

communication could potentially be centralized in these as long as the companies play ball.

Yeah. I mean, it could be kind of the agent in the end, but how how does the communication actually happen? So like

you know for example if you look at a messaging app like uh you know like signal there's a protocol that it uses to communicate and you know I can build an app it can maybe use that same protocol but I think it actually can't

message other people that are on signal but yeah like I can have an agent that uses my app to do to do that messaging using an existing app that that supports this.

Yeah.

So yeah it's not obvious how it's going to play out. I think today people use a mix of you know apps and and agents. Um

but you know I I do fundamentally think that a lot of these modes are actually still going to increase in value over time.

You can think of another example let's say you know like a TSMC or some kind of like chip manufacturer. If you think about um the amount of work that they put into making a process and in making

a process where the costs go down with scale, this is a fundamental economic force and there's a lot of companies that that that do this kind of thing where you know especially in manufacturing where with scale the cost

goes down. With tech companies this is

goes down. With tech companies this is the case for infrastructure. So if you build a really great infrastructure you can support more users and the marginal cost per user goes down over time. So if

you have this kind of effect, it doesn't matter if you or I can build apps.

That's still a really powerful mode. But

I I do think for sure both things are in play.

Okay, I got three more in 10 minutes.

Let's see if we can get to them all. U

Jack Clark, one of the anthropic founders, recently said I think that he believes there's like a 60% chance that these models will start improving themselves by 2028. It could be off by a

percentage or a year, but ballpark that's accurate. Um, you're in the app

that's accurate. Um, you're in the app where coding happens autonomously.

You're running this app. Um, do you agree with Jack? Seems right.

Yeah. When I look at the way that quad code is uh written, 100% of quad code is written using quad code. This has been the case since I think November of last

year, since Opus 4.5.

It's like a fast takeoff scenario then.

Do you anticipate that? I mean, it's it's it's possible and like this is this is why Anthropic exists. If you ask anyone any any engineer or any researcher why they joined Anthropic, they're going to tell you it's for AI

safety. And it's because for us when we

safety. And it's because for us when we think about the future, you know, years from now, the thing that's the most important and the thing that we want to get right, you know, for our kids is we want to make sure this thing is safe and we want to make sure it goes well

because yeah, like that is one of the possible outcomes. I think that's not

possible outcomes. I think that's not yet what we're seeing. Um, you know, right now Quad Code is writing itself, but it's still a person that's doing the prompting. Quad is starting to generate

prompting. Quad is starting to generate its own ideas for what to build next for quad code, but it's, you know, it's not always good ideas, and I still generate most of the ideas. And, you know, at some point, it's going to change. The

model's going to improve, and it's going to become more of a of a self-reinforcing loop.

Okay. I definitely want to get your thoughts on um the world model uh argument here where people who are pro-orld model says say that um a large

language model has no understanding of the consequences and you need to build a world model into it to have effective agents. Here's something from Yan Lakun.

agents. Here's something from Yan Lakun.

Uh he says you cannot build a reliable agentic system without a world model.

LLMs don't have world models. They can't

predict the consequences of their actions before taking them. According to

Yan, they just act and whatever happens next is someone else's problem. I was

speaking with Greg Brockman from OpenAI recently and he said basically he doesn't accept that argument and he thinks LLMs are the way directly these text models are the way to AGI. Uh which

side are you on? Are you a believer that that world model intelligence needs to be baked in or do you think that LLMs alone are good enough?

I would put out an offer to Yan if he wants to sit down and quad code together for an hour. I'd love to show him. And

then you guys should do that on this show.

Yeah. And then I'm curious to hear what he thinks. Maybe he'll change his mind,

he thinks. Maybe he'll change his mind, maybe he doesn't.

Right. But your perspective though, you know, I'm on I'm pretty firmly on the product side. Um so, you know, I I I don't I don't really have a have a perspective on it.

But okay, let me let me drill down a tiny bit deeper if you don't mind. um

you know, you're you're on the product side, but I've heard multiple people bring out this idea that without a conception of the way the world works, like in a world model, um a a uh LLM

just doesn't have an understanding of the way that the world works and consequences and stuff. You use co-work to book how many flights? Eight flights

in hotels. Like, you must think that it has some understanding of consequences, otherwise you wouldn't have given it your credit card, which I presume you did. So what do you think about that

did. So what do you think about that argument in particular?

I think from what I've read from folks working on on research at Anthropic, it is surprising the degree to which these models are intelligent because like you said at the beginning, the the thing that they fundamentally do is they

predict the next token.

Mhm.

And so you think like this is kind of like a stupid thing like how can this possibly lead to intelligence? But you

know we we've actually published a lot of work about how the models are able to plan, they're able to actually reason.

Um there was all these like very surprising behaviors that you actually wouldn't expect from a model that just predicts the next token. So I don't know I I wouldn't discount it.

I mean I think my favorite is when they write poetry um as they're writing the first line you can see in the model this is anthropic research that they're already thinking about the next line.

That's right.

Which is like how is that even possible?

But that's right. I mean and that's kind of

that's right. I mean and that's kind of you know how I think about it. Like if I if I were poetry that's how I would do it too. And it's it's crazy like you

it too. And it's it's crazy like you teach this thing to predict the next word and somehow if the next word is hard enough, it has to learn to really plan ahead and it has to learn how to do all of this.

Okay, last one for you. Um, sometimes I wonder when I see big tech changes uh underway and in my career covering this stuff, some have worked out and some

haven't. Um, I always have to ask myself

haven't. Um, I always have to ask myself uh how are we sure that this is the future and this is not a fever dream. Um

and I think the data indicates that this is a real thing. But I also wonder you have to sort of you have to question how much you can extrapolate towards the future in terms of how will this

continue to progress. Um the argument that this is a fever dream is that um maybe people just want simple interfaces and they don't mind tapping through

things and you know speaking in a cloud code feels a little bit too techy uh and it just won't appeal to the everyday user as much as it's really taken off

with developers. I mean how would you

with developers. I mean how would you answer that?

We had this uh we had this hackathon for Opus 4.7 recently and one of the winners was a doctor that built an app. There

was a there there was an electrician uh there was a carpenter and a lot of these people didn't have coding experience but they use quad code to build something useful. There's one person that built

useful. There's one person that built and sold a startup as a result of one of these hackathons that we put on. And

undoubtedly when we first built quad code it was for engineers and engineers kind of figured out how to use it. But very quickly people that were not engineers figured

out how to use this to build economically useful things. And actually

if you look at a lot of the usage today it's like it's not engineers and it's just so useful for people that they were going out of their way. They're jumping

through hoops. Even before co-work people were like installing quad code in a terminal. For a lot of people, this

a terminal. For a lot of people, this was their first time using a terminal.

And of course, now, you know, for quad code, we have a desktop app, we have iOS app, we have a Slack app. You know,

there's many ways to interact with it.

But people were jumping through hoops to use it because it was so useful. And so

for me as a product person, this is the ultimate market test of is this thing useful? Is are there a lot of people

useful? Is are there a lot of people that use this every day and that keep using it every day? And yeah, it's a lot of people and it just keeps growing. And

I'm just constantly surprised by the way that people use this.

Yeah, I I will say I've been surprised by the way that I found myself uh using the tools and I don't know, we we'll see what comes next. So excited to keep using it and and thrilled to have a chance to speak with you. I hope we can

do it again.

Yeah, thanks for having me on.

All right, thank you Boris. Great

speaking with you. All right, everybody.

Thank you so much for listening and watching and we'll see you next time on Big Technology Podcast.