Code with Claude 2026: A conversation with Dario Amodei & Daniela Amodei

Heat.

Heat.

All right.

All right, everyone. Thank you for um for joining us again. And I am so excited for this conversation with Daario and Daniela. Let's give them a round. Woo.

round. Woo.

It is a delight to have you here at Code with Cloud. We've been having a great

with Cloud. We've been having a great day of sessions, demos, uh, customer sessions, all sorts of fun stuff. But I

wanted to ask you maybe just a little bit zooming out. We talked a lot about the exponential in some of our conversations this morning and uh, what

it feels like to be on the exponential.

And so, as people who are definitely on the exponential, um, we've talked a lot about growth and what it feels like. What does it feel like for you all?

Well, first of all, it's great to be here. Thank you so much for having us.

here. Thank you so much for having us.

Um, you know, at Anthropic, we have this little slack emoji of the roller coaster. You guys know the one I'm

coaster. You guys know the one I'm talking about. Um, but it's like an

talking about. Um, but it's like an inflected roller coaster. So, it's

almost like it's like going straight up.

And I think of me and Daario as riding at the front and the back of that roller coaster. I don't know if you all have

coaster. I don't know if you all have ever maybe not recently been on a roller coaster. I'm not always sure which one

coaster. I'm not always sure which one of us is in the front and which one of us is in the back, but you get a different type of whiplash depending on which side you're on. Um, and I think that's probably the best encapsulation

of what it has felt like. It's like

we're having a lot of fun. There's a ton of adrenaline. Um, we're not totally

of adrenaline. Um, we're not totally sure that the operator of the roller coaster isn't like a 15-year-old who's doing summer jobs of like

questionable level of sounds mind. Um,

but it's been great. It's fun. It's an

adventure. There's a lot going on.

Yeah. You know, the way I always think about the exponential is, you know, it was it was me and the other co-founders who kind of, you know, first predicted it through the scaling laws, you know,

over over 10 years ago. And you know, we wrote down these lines on graphs that say like, well, first we're going to spend $1,000 on then 10,000 and then hundred. You know, it's going to go all

hundred. You know, it's going to go all the way to to to hundreds of billions and you know, the model is going to be this good at this task and this good at this other task and that good at at at coding and and it it is a remarkable

experience to write down these lines on graphs and have the predictions come true. So, in that sense, they are not

true. So, in that sense, they are not surprising at all. And and yet, the actual experience of seeing what it's like is is is just it's it's so crazy that

you're shocked anyway, even though what you wrote down on the on the graph is exactly what happened. I'm always

reminded of like, you know, the the there's this famous scene in the movie Interstellar where where they go to this, you know, planet that's very close to a black hole and and so, you know, the planet has these waves that are

like, you know, 2,000 ft high and and, you know, I I was a physicist. I know I know the math of general relativity, how much things can be umh sheared, but actually seeing it on human scale, like

there's something deeply, you know, it's kind of deeply strange and unsettling about seeing it actually happen. and and

you know that that's what it felt like every year at Anthropic and and I feel like this is the first year where like the rest of the world is is kind of you know is seeing us seeing it with us

because we're so much in the spotlight.

Um you know it applies to the internal growth of the company. It applies to our own work within the company where it's the first time we've seen the number of internal PRs inflect upwards due to due

to the work that that Claude is doing and we've seen it externally because actually this is the first year we've grown faster than the exponential. So

you know we tried to plan very well for a world of 10x growth per year. Um, in

the first quarter of this year, we saw, if you were to annualize it, 80x growth per year in in in in in in in in in

revenue and usage and and and so that that that is the reason we have had difficulties with compute, right? We've

planned for anything from it only grows a little to it grows 10x and yet we saw 80x. Um and and and so you know, as you

80x. Um and and and so you know, as you saw today with the the the the SpaceX compute deal, we're working as quickly as possible to provide more more compute

than uh than than we have in the past.

We'll continue to do so. We'll pass that compute on to you as as as as soon as we can as soon as we can do so. Um I I I guess I hope the ADX growth doesn't continue because that's just crazy and

it's too hard to too hard to handle. I

hope for some more normal more normal uh uh more normal numbers. Uh a mere a mere a mere 10x. Um uh but we will we will manage it uh absolutely as uh we will

manage it absolutely as uh best we can and uh you know we're we're every day trying to uh trying to obtain even more compute that that we can uh we can we can we can pass on to you. We're sorry

if sometimes it takes some time but uh we're gonna we're gonna we're going to keep going to acquire as much as we can.

Awesome.

Um, yes. Yay for computer rate limits.

yes. Yay for computer rate limits.

You know, this is an audience of developers and builders. And that's

really what today is about is about how we're making our platform better because developers who help us close the gap between what the models can do and what they're actually doing for real people

out there. And you know both of you talk

out there. And you know both of you talk a lot internally about the importance of developers and builders. Um maybe

Danielle I'd love to start with you like how do you think about supporting this ecosystem supporting this community?

Yeah I mean I'll start out by saying I think in many ways developers are the most important users of Claude. Um I

think for a variety of reasons you know one is anthropic ourselves are majority developers right if you think about um how we develop this technology what

we're building we learn so much from the developer community it's like the best um it's the best partnership because we

it's first of all I think it's it's a group that gives like honest feedback right and so I think that that is actually really hard to get. You know

what I mean? It's like you build a product and you're like, I see some numbers like those are nice. But like

the genuiness with which the developer community I think engages with us is something that is so special. Um we have tried you know really from day one I think Anthropic has always you know

primarily built for developers for businesses. I think we're a little bit

businesses. I think we're a little bit unique in the AI ecosystem you know for that reason. And I think we have been

that reason. And I think we have been very fortunate to be able to benefit from the feedback from the engagement from the community development. You know

developers are you know in my experience they're like they're very ecosystem and community oriented which I think we are too right we're like how do we build for this sort of broader ecosystem of people

who are developing by the way some of the most inspiring transformative technologies and building the most incredible companies in the entire world. Right? there's been this

world. Right? there's been this renaissance of things in you know medicine and uh software development and you know financial sur I mean it's like you pick the industry and there is an

incredible developer or an incredible you know developerbased company that is transforming that industry um help you know leveraging our tools sometimes and I think um that's such a special that's

like a that's both a privilege and a responsibility that I think anthropic holds to say you know developers are really um the backbone of like how we learn how we build better tools for all

of you. And I think that's a that's a

of you. And I think that's a that's a really special relationship that um that we feel like pride in and also a responsibility towards.

Yeah. I mean,

and this is an example of the developer community. I think part of what

community. I think part of what Daniellea is saying is feedback is a gift. We hear the positive. We also hear

gift. We hear the positive. We also hear the negative. Please keep it coming. It

the negative. Please keep it coming. It

is part of how we know what's working and what isn't. And so we really value uh and I'm sure as you're talking to people around we really value all of that. Um and it helps us know what to do

that. Um and it helps us know what to do better. Sorry Daria, back to you.

better. Sorry Daria, back to you.

Yeah. Uh you know one thing I would say is the technology doesn't diffuse at uh kind of an even pace across the economy right and I think there's a spectrum

where the software engineers are the ones who are fastest to kind of adopt new kind of fastest to adopt new technology. that's that's that's why

technology. that's that's that's why there's so much uh kind of focus on this area and it's the beginning of it but it's like it's a foreshadowing of how things are going to work across the

economy right and how the economy is going to be transformed by by by AI so I think you know getting this right and and really really making it work for for this community it's kind of like a

microcosm of how we have to you know of how we have to make it work across the world and and I think one thing that um you know a a dynamic we should watch. Uh

it was about uh rough I think it was roughly a year ago there was an event uh uh uh like this um where Mike Kger asked me you know when will there be the first

uh billion dollar company with one person and I said 2026 um uh and I I think we're actually on track to achieve it hadn't quite happened yet. There's been like two

happened yet. There's been like two person companies that are $1 billion built with AI. There's been like one person that's worth several hundred million million dollars. So, but but like you know we got we got seven more

months. So it it's kind of it's it's

months. So it it's kind of it's it's kind of No, we do it seven more months in 2026. Um or eight. Um uh uh and you

in 2026. Um or eight. Um uh uh and you know that's that's there's an eternity on on the exponential. Um uh but what I'm trying to say with this though is

that there's there's an enormous ability for one person or a tiny set of people to to do a set of things that are incredible, right? where you know before

incredible, right? where you know before if you know you just had an idea and you had a vision like there's so many resources you'd have to accumulate over several years in order to make that vision happen and and I think there's a

very unique opportunity for single single individuals or very tiny teams to to do things that are incredible right where I think we've moved from the models are writing code to you know the

the models are helping us think of software engineering as a task to the models are helping us think of like how can I build a business or an economic unit as a as a as as as a task. And so

there's an extraordinary amount of opportunity for for people in this room to to to to kind of take advantage of that.

Yeah, it really feels like it's it's kind of removing barriers. You know,

there were all these barriers to like creating that kind of value in the world and now I mean I guess the gauntlet has been thrown. We've got an an eightmonth

been thrown. We've got an an eightmonth timer.

Um and I'm excited to see what comes of it. Um I would love to maybe hear what's

it. Um I would love to maybe hear what's going to change for developers. So

Daario, you talked a little bit about uh what they can do now and you've talked, you know, in the past about how you expect uh Claude to build more with Claude. Can you just talk about how you

Claude. Can you just talk about how you expect things to move?

Yeah, I mean I I I think there's several like maybe maybe trends. One is uh going from uh single agents to kind of multiple agents. So the idea that you

multiple agents. So the idea that you know you have a bunch of these quad right it's like managing a team right you have a bunch of quads running and like you know you kind of you kind of farm a bunch of things out to your quads and maybe some of the quads farm things

out to other quads with like different so you have a kind of whole hierarchy or we're gradually making our way to like the country of geniuses in a data center you know we're starting with like a a

team of smart people in a you know in a room or something working our way upon in the country um uh So, uh, I I think that's one trend that

we're kind of already starting to see and we're already like offering tools that can help do that. I think a trend that's related to that is like, you know, what we've done so far with claude

code is like uh, you know, it helps kind of individuals to be more productive.

But I think increasingly we're going to start thinking at the level of whole teams and organizations and how can you make whole teams and whole organizations more uh more uh productive in a way that

is kind of more than just the sum of its parts. Um, and then finally, um, you

parts. Um, and then finally, um, you know, I think in this area, as with everywhere else, um, if you want to think about what's next when something's working really well, you should always

think about AMDall's law, which is you speed one thing up, what are the things you're not speeding up? And and so I think there are a bunch of things like uh security, like verification,

like just if you're living in a world where you can within an organization write three or four times as many PRs as you could pre as you could previously, you start to understand there are all

these other things that are holding you back or that will go wrong if you speed up just that and not everything else.

And and so working to speed up those those those kind of those kind of other things so that we can greatly increase people's productivity, but we can do it

smoothly and uh uh uh you know smoothly and productively and and and reliably. I

think that's going to be very important.

Does that have any impact on how you think about training new models or you know what what the future of models looks like?

Yeah, I mean you know that that's true on several levels. I mean I've already said many times we're using Claude to speed up Claude, right? that's that's uh that's uh that's kind of that's kind of already uh um uh uh you know something

that's that's happening. Um but I you know I also wonder if if the things we're trying to do with the models could um uh uh could also influence um how we

how we build them. So when I talk about these things like you know kind of verification or kind of design quality or things like that like one of the reasons training models for code and software engineering has gone so fast is

that you have this verifiability right where you train the model and it's like you know you're able to verify it by running by running unit tests and so that has a lot of properties that make simplify the process of training but

what you discover is that there are these of course aspects of the of the job that are not verifiable right and and you know some of uh some of the you know is this thing really right? Can we

find errors? Are there security issues not quite as verifiable? Um and and so training train the models to be better at that which I think will also make the models better at at at other things where they haven't made progress as fast

as coding like their ability to write or their ability to kind of do do you know to to do uh you know less less objective scientific tasks. So I think it's going

scientific tasks. So I think it's going to have benefits in many in many other areas. But you know I think we find even

areas. But you know I think we find even even within software engineering this uh you know this these these kind of um uh uh soft or somewhat subjective um uh

skills and abilities are become surprisingly important because of Amdall's law.

I would love to hear you know we talk a lot about our mission internally like Daniela as we just keep growing and the stakes of this whole industry keep getting higher. H what should people

getting higher. H what should people know about our mission and about us as a company?

You know, I think when I think about what Anthropic is trying to do, there's these sort of like two maybe two pillars, right? The first is around how

pillars, right? The first is around how do we develop this transformative technology in a way that is good for everybody, right? And I think this goal

everybody, right? And I think this goal of claude is this incredible tool. It

has the power to really transform, you know, what people build and how they create and the the ambition level of what they can develop themselves. And I

think there comes a huge amount of opportunity there. And there's also some

opportunity there. And there's also some risk, right? I think that we've talked

risk, right? I think that we've talked about this a lot publicly. There's some

risk to um just labor disruption.

there's risks to ensuring that the technology is developed safely that it's good for people and I think anthropic's job or what we try to do is really think about how to balance these two things in equal measure we have this internal uh

cultural value called hold light and shade and I think that it is such a good encapsulation of you know what we see about how the technology is being

leveraged today um and and also just our approach to putting the technology out into the world right I think mythos and glassing are great example of this. The

potential um to build something incredible with a model that capable is so vast and we want to be a little bit careful about how we release it because of some of the security vulnerabilities,

right? And I think that this is this

right? And I think that this is this kind of complicated dance that we do where we're like we really want to get stuff out as quickly as we can. We're

trying to build the best products and release the most powerful models and we're just trying to do it responsibly.

I think that is really the underpinninging of the majority of actions that we take is is sort of grounded in across those two pillars.

I I think that is like one of the things I find most meaningful about getting to work here is just thinking about how much everything is changing and the fact that we're kind of all building in the

industry right now. To me, it feels like we're getting the chance to have a vote in how everything unfolds. And the

trade-offs you're describing, that's that's just what I think of when I think of Hold Light and Shade is um as things move so rapidly, we get to build

experiences that other people use to understand what what the future looks like. Um so that's something I'm always

like. Um so that's something I'm always really personally excited about.

Um maybe I'll ask a little bit about product. both of you kind of lean in

product. both of you kind of lean in quite a lot on the product side. Um, one

thing that we talk a lot about is building for the exponential. Um, I

always think about product as kind of a bridge between the technology that exists and the problems that people have and it's just a very interesting time because the technology is changing much

more rapidly than we're used to. Um, can

you talk a little bit about how you think about product building in this world?

So, I love I mean I love your way of putting that. It's like Dario and

putting that. It's like Dario and Daniela lean in a lot and this chief product officer what Abby needs is like you guys are up in my business all the damn time. Can you please leave me alone

damn time. Can you please leave me alone and let me do my damn job?

All feedback is a gift.

I enjoy every perspective.

But no, Ammy is right. I think you you have a you have a hard job that you wear incredibly well. Um, but you know, I I

incredibly well. Um, but you know, I I think in all seriousness, you know, Daario and I both we care a lot about the product, right? It's a it's um it's a representation of what we are trying

to build at Anthropic. We want it to be useful for people. We want it to be accessible. We want the product to be

accessible. We want the product to be good, right? Part of why I think we're

good, right? Part of why I think we're leaned in so much is we feel very invested in ensuring that people that are using Claude are getting out of it everything that they can. And so I think

our um you know our bothering you is really I think our way of you know feeling like to the degree we can we're standing up for our customers. We're

standing up for our users who are building sometimes their whole business around the premise of what these AI models are capable of doing. And I think the other thing maybe the thing that

makes product at Anthropic unusual or different than how I've seen it done um you know at other companies I've worked at is like product is is sort of one input and research is another. Like I'm

sure you've felt this in your role. Like

sometimes we're like man this is a clear, you know, this is a this is a great place where we should just be able to build something better that's easier to use or here's a product idea. We

really want to enable people to be able to access it right out of the gate. But

a lot of the time product innovation is driven by what new capabilities emerge in the model. And I think coding is actually a great example of this, right?

We didn't just sit out and say like from day one we're going to build a coding product. It was like once we saw that

product. It was like once we saw that the models were able to write code at a reasonably um accurate level, not perfect, we were like, huh, this is interesting. It seems like a lot of

interesting. It seems like a lot of people that are kind of claude files are are developers, right? And they're using it to write code. um this has always been a community that we've worked well with. We've wanted to lean in with and

with. We've wanted to lean in with and and engage and support. Should we build a tool for them, right? Should we build something that actually is going to enable people to be better at doing their day-to-day work in this way? But I

think that's I think that's just an interesting dance inside the company where there's a component that is sort of I don't want to say traditional product because nothing about anthropic is traditional but um but I think there's a component that that looks like

a normal product organization and then there's part of the organization that's like what is what is new from the models what's happening on the research side and how do we kind of marry those two things together

please? Yeah, I I I would maybe uh take

please? Yeah, I I I would maybe uh take this from two lenses, which is um uh uh building products for AI and building products with AI and and I think uh uh

the internal experience at Enthropic month by month and week by week has given us lessons about both. Um so I think over the last few years learned a lot of lessons about you know building

building products with AI and in some ways it's been an advantage that you know I was I was a researcher I was never in the era of you know building products without AI. So it's like you can end up in a situation where there are things you don't have to unlearn.

You can just you can just you can just learn the new world from scratch. Um uh

I I I I think you got the essential difference which is that if you go back to the product era in the 2010s, you had a slowly changing technological background, right? You were trying to do

background, right? You were trying to do new things with, you know, the the the kind of technology that was present and of course every once in a while you'd have a new framework or new way of doing but relatively slow. AI is moving

lightning fast and so there are a few consequences of that. One is that uh there are new products that are not possible with a given capability of model but then when you take the next

step when you go further far enough along the exponential it then suddenly they light up. Suddenly they become possible and so it puts a premium on internal experimentation because you

always want to be trying something. Even

if you tried something that did didn't work, you want to revisit it a few months later because you know it might it might work then when it didn't work before which sounds a little bit crazy

but you know if we had tried to do quad code in in like 2022 it wouldn't have worked because the models wouldn't have been strong enough. It was a frustrating experience. We did have some early

experience. We did have some early things that were a little like claude code in 2022 and it was like ah this is intra but you couldn't you couldn't actually derive value from it because

the models were dumb. Uh, so that's they were um I've been training these models since 2015. They they were really they

since 2015. They they were really they were really dumb. They were really dumb back then.

Um uh the the second thing is that products reach their uh uh uh products reach their um saturation when models start to get too good. So I think this

has happened with chat bots, right? Like

you know it's it's it's a big market, lots of people use it. it's gonna it's going to stay around. But like you know the ways in which we're making models smarter today are much more evident in

uh you know today's claude code fa form factor and in more generally in the augentic form factor than they are on the chatbot code factor. So it's kind that's kind of the other side of the coin which is that you always have to be

thinking about what the new thing is right the way this business work isn't isn't you make a product it becomes very big and then all the kind of stability sets in and you have to ask you're

you're constantly not only are you able to make something new you're constantly needing to make something new or at least update the things that you've made so it's it's kind of constantly an

innovation laboratory um and I think the other thing it it means of of particular relevance for developers and software engineers is API never really goes away

as a market. um uh because it's it be the fact that it's always possible to build new um products that's true inside anthropic it's also true outside

anthropic and and and so and so I I I I actually think you know both both within code as an application but just coders writing writing code with claude for you

know medical applications for you know law for finance or or or you know there are going to keep me new applications because the models get smarter and and

they enable them. Um so that's that's um uh uh building kind of in the age of AI.

There's a newer thing we've seen maybe over the last year or six months which is building with AI itself like using AI to kind of you know to to to enable the

the process of of doing uh product development uh faster. Um that that one is interesting and again we would go back to our old friend AMD Doll's law which is you know we've we've found with

the internal model acceleration you can write two times as many four times as many five times many you know we you just you you you see this within the company but then you see what breaks.

We've been able to ship a lot more products, you know, than we could a year ago, and they're of pretty high quality or or or I hope you think that. Um uh uh uh uh uh but but but it it is possible

to accumulate an extraordinary amount of internal technical debt when you ship that fast. And so then you have to say,

that fast. And so then you have to say, well, can we also use the AI models to like undo that technical debt or keep track of what it is that we're doing and and then you learn ah the team has to

work together in a totally different way. And these these these revelations

way. And these these these revelations come month month by month and you kind of learn how to do things um uh uh in in kind of a totally new way. So somehow it

somehow it increases the tempo not just of building but in which the way you have to change the way you build as a team.

I I really experienced that. I think the hardest part of it is you can get really familiar with the problems because the problems don't change that fast. the

problems are about humans, you know, like we always we're going to have similar problems. But a thing that's hard is to learn to be fresheyed all the time about what the technology can do

and like constantly scan for that. And

then I also feel it on a personal basis where your job just changes because you hit a new bottleneck, you know, and like the the way you spend your day is just a

little different. Um maybe I'll just ask

little different. Um maybe I'll just ask a couple more questions. Um maybe Daario very quickly you talked a lot about model capabilities and how they change

so often is there one thing that makes you most excited in the models that are coming in the next call it 6 months.

Uh you know I I would say this idea of um you know thinking thinking at the level of uh the organization rather than just uh rather than just one person. Um

and and you know again it ties to this um uh you know oneperson billion-dollar business which you know maybe that will turn out to be to be to be an under to be an underestimate um uh but but but

you know just just just kind of the idea that you know you know you can both have AI do the work of many people but that when you have a team of humans that AI is not just doing the work of many

people working for one person but that it it does the work of many people many times over by operating within an organization of humans.

Awesome. And maybe I'll just ask to close like Danielle, one thing we talked about is um some of the the kind of developers or use cases we see that are

just so inspiring. I talked a bit in in the keynote about some of my favorites, whether that's um people using it for efficiency at places like Stripe doing major infra upgrades, but also people

using it for these very um personal and specific ways like connecting kids to foster families faster. And I would just love to hear, you know, what are some of your favorite examples of how you see people use the tools? Yeah, I mean I

think even going back to the days when I was at Stripe, some of the the um something that I've just loved about the developer community in general is like for every interesting challenge or

problem there is in the world, there's like an incredible dev somewhere who's like trying to make that thing better.

And I think getting to watch people build things with Claude that are just bringing so much utility and value and meaning to people around the world is

really inspiring. I think you know for

really inspiring. I think you know for me personally um you know a pilot that I've seen with some developers who are building these like interfaces basically for mobile doctors in the global south.

So places where it's really hard to reach an actual doctor because of access issues right you're down a dirt road somewhere you know 50 miles from the nearest city but people there still encounter health challenges. How do you

actually like work with this really smart technology to build these like interfaces where people can just ask it a question and can give you some medical advice, right, in a way that's like

sanctioned. Um, I'm I'm also just like

sanctioned. Um, I'm I'm also just like in general blown away by the medical research field and what is coming out of Claude being able to accelerate biomedical research across many of the

developers that work with us that build on us. There also just some really

on us. There also just some really heartwarming like individual examples.

We have this user happiness channel at work. Some of my favorite ones, there's

work. Some of my favorite ones, there's a developer who used um Claude to help retrieve their wedding photos that were corrupt, like on a corrupted hard drive.

I thought that was so sweet. Um and just like such a like personal use case, but so like so meaningful. And then my last one, which is so random, is someone is using Claude to chart the growth of

their tomato plants in their garden. And

I was like, never in a million years would I have thought of it, but it's like it's so I like was like, do you have a live cam and can I subscribe to this? um because I would really love to

this? um because I would really love to see it. But I think just the range of

see it. But I think just the range of things that people are able to do um is is just astonishing.

Well, Dario, Daniela, thank you both so much for spending time with us here.

It's a pleasure to see you. Thank you to all of you for joining us. We've got a great rest of the day coming, but please join me in giving a round of applause to these guys.

Thank you.