20VC: Anthropic CPO Mike Krieger: Where Will Value Be Created in a World of AI | Have Foundation Models Commoditized | When Do Model Providers Become Application Providers | What Anthropic Learned from Deepseek

24:02is it, you know, you ship as often as you can, but there's an admin opt in on each kind of thing. That adds complexity as well. And so it's a great question. We're, I would say like it's an active topic of conversation. How raw or how quickly we can ship knowing that we wanna bring things out to the world. And you don't know they're gonna be received and you wanna learn, but as you accumulate sort of notoriety or you know, the pet, people start depending on your for work for as you can't treat that completely sort of want to make. Are we in a product marketing nightmare? And what I mean by that, we have deep seat release something this week.

24:34We have open AI release something this week. We have anthropic release something this week. We have a mystery already something you attend days ago, where bunny, every single day, there's a new release that the world maybe gets apathetic. How do you think about that? And how does that inform how you think about product launches messaging? Yeah. I mean, it is much more robust in Instagram. The things that you had to watch out for the big rocks were very known in advance. It's like don't launch anything during WWE, that's going to be a flurry of a loud announcements with the September I -Guy OS event.

25:05You know, there might be some other big rock like holiday is so much easier from a product marketing perspective. We're here that reminds you a little bit of crossy road where you're like, okay, the car's going by. All right, there's a gap in the car like launch tomorrow or like, now it's good. But oh, now we hear this a rumor. It's so much hard enough heard from folks that other labs as well that everybody's kind of trying to read the tea leaves and be like, all right, is anybody? Is it quiet? Is it okay to launch now? Or like I think we're going to be doing it next Tuesday. So it's much harder.

25:32Oh, it's shipping. You know, it requires a completely different approach. And you know, I give credits our literally a product marketing team because they've had to worry it from a point where, you know, we were called 37 Sun every launch on Monday. And we locked the blog post for that Sunday night at 9 p .m. which is not best practice from a marketing perspective. You know, we were briefing press that day on Sunday. Thank you to folks that out on the phone with the sun Sunday, but that's the point where everything is done and ready and locked and we can like we can go. And so it doesn't involve that sort of ability to react quickly and be nimble.

26:09I mean, even things like the, you know, when we release a model, there's a model card and there's evaluations in a comparison table. There are things in that comparison table that were released the week before, right? Like Grock three was, you know, just a week prior. So it involves what happens when those are released? When Grock three releases there's like Joe, it's a side like, does everyone an anthropic and open AI get by? Oh, shit, they beat us again. All like, oh, shit, we won. Yeah. One of the things I try to do, you know, to support the team there is remind like, you know, it's the model releases are going to happen.

26:42And at any given point, you are going to be, you know, it's the, you know, it's so over and we're so back like that's like, oh, it's like that and you have to live that in AI and you can't get to down about one release because yeah, for sure, it is inevitable and sometimes you're lucky and there's like two or three months where the model that you launched or the product that you launched just still say to the art across all the things you really care about. Sometimes you'd last a week and you can't over -rotate on either of those. You can't rest on your laurels, you can't go that. Like I think the thing that's used really useful to is it's a chart I showed like almost every and every sales call, which is just mapping from like anthropics founding to where we are today in the milestones.

27:20And at any given point, you could say, wow, like, cloud two, that's like pretty far behind. Oh, cloud three, sit at the art. And then no, it's not. And it's, you gotta look at the trajectory and like trust that you were gonna continue to make improvements is number one. And then number two, remind yourself that, you know, if everybody switched every single day purely due to like eval being, you know, changed, one that would be like an insane thing to do to your user base says like a provider of software. But too, that would make for an even crazier industry. Over time, you start learning that people don't just deploy models.

27:51They're doing fine tunes, or they're deploying models. Plus, they've done a lot of really bespoke work to make that model be great for that use case. It's not a thing that's going to switch overnight. Or you're one of three or four options within a model selector, which, for example, in a coding environment. So you're still in the mix and you still have a chance. But I'm not sure if it's like like finding the meditative zoom out angle of it or just like get used to the bumpy ride or some combination of the two, but it is for sure a thing that like every time there's a model launch, I assume every one of those labs is like either watching the launch stream looking at the wheels and me either, hoo, all right, there we go, work to do.

28:27I would argue that brand is the most important thing. To your point, people aren't switching every day. They're kind of like, oh, I'm a Claude person or I'm a chat GPT person. And they kind of identify already with their models. Do you agree with that statement? Or do you think that's too glib? I don't, I think that is right. I think especially on the consumer front, you know, I was just reading Ben Thompson, you know, he has Nat Fudeman and Daniel Gross on there pretty often. And they're talking about some people being clad people and some people with chat GPT. I think that definitely happens where they, you like the personality, you like the interface design, you like the vibe again.

29:01It actually reminds me a lot, you know, we had this interesting back and forth with Snapchat over the years of Instagram. And then even before that, people would launch a new product that's like Instagram, but just for super high -end photographers, or with this like additional twist, or just one photo a day, you know, it's be real. And I had this like fake formula. I'm not the mathematician clearly, I'mthropic, but it was, you know, social networks are made up format or formats that you have in your product, audience and vibes. And format, you know, for Instagram, we got stories, we had feed, and then eventually we had a video.

29:32Audience, you know, initially was sort of hipster -y photographers. Eventually grew to be, anybody is really interested in sort of visual story teller, visual media. But the vibes of Instagram, even when we had more product similarities to a Snapchat, even to a Facebook, the vibes are very different. And I don't know what that fake formula is for AI products, yeah, but I think it's some version of that where there's like model, model personality is probably one of them. There's likely something around the scaffolding prescriptiveness of the product that you're working around it and then there's vibes and like I get harder measure but absolutely there.

30:10Can I ask you a hard one when we have so many different models and so many different providers? Open source is a very viable possible route and distillation is looked to have in a shady way. Is distillation really wrong if it ultimately propels spaces forward? Well, even like it's the sick within the labs. I assume every single one of the labs is using like we even within themselves. It is very valuable to be able to take the knowledge of your highest end model than be able to make it, you know, lower latency, more affordable, etc. So there's that loop, etc. overall. I think the places where this gets interesting are one, do we want any nation to be able to be able to distill models from any other ones?

30:51Like, personal answer is no. I think that there's value in like even like as AI gains and capabilities being really thoughtful about that from a national security perspective. And then the other piece is to have the advancements happen at the rate that they're happening, be sustainable long -term. Like, you do need the labs to be able to be able to commercialize all of that training and innovation, et cetera. And I think finding the right models for that long -term is important. I think the open source models take Lama, for example, that they've been able to do that from their own research of perspective and data and justion and trading.

31:23And so I guess I would say, distillation does not feel essential in order to unlock those things and poses other issues even just you know from a terms of service perspective. Does Lama show that there is no value in the model and all the value is in the data? If Facebook are willing to give it away for free because they know that no one can copy the data that they have, is that what that shows? I think it's a good interesting question is like whether Lama is the quality of Lama due to the fact that they can, I don't know if they've except that they do, but they clearly can train on Instagram and Facebook and etc.

31:59It was Gem and I better for being able to train on YouTube. They actually clear to me that Gem and I benefits from that, whenever they have a good video understanding demo, for example, somebody has probably the largest repository of video in the world and can likely train on a lot of those pieces. Let's clear on the Facebook front. I've never heard from people, gosh, you know what? Lama does extremely well, is generate good content that would work well on social media. Yeah, it just seems like a good general purpose model. So it actually go back to the values all in how good is your team? Do you have the underlying data that you need to do it?

32:33But then also, how useful is your model in actual use cases? And that is like the highest order bit. I almost wish I'd started with that because e -vails are really useful for hill climbing and for internal research, but they don't tell the story of like, is the model going to be excellent of what it needs to be excellent or deployed for or even if it is excellent at that thing, is it only excellent at that thing in very narrow situations or is it something that as a as an entrepreneur, you know, outside the good, the labs, you can rely on the model to be like your representative, I guess, in that product.

33:05So yeah, I think the values and for the labs, the values and the team, it's in like the the models of ability to actually perform the right actions in the real world without so much non -determinism that it becomes sort of unreliable. I mean, know I was one question on this, it's not a trap to go down, but I'm I've spoke to Alex Wang about it on the show and I saw a poolside on the show and they said we deeply underestimate China's ability in AI. Do you agree that we underestimate it? Yeah, I think that deep sick piece that people seemed surprised that there were sort of cutting edge research teams there and if you were paying attention, that part should not have been the surprising piece.

33:46You know, it's been we saw in Instagram was blocked in China fairly early and then we saw the sort of emergence of a parallel world of startups when if you take up Facebook and Instagram what happens and what emerges in those products were often like very high quality they like demonstrate a lot of creative thinking and and we're built at scale too like they were solving problems, you know people love talking about the like the super app and and we chat and there was some technical challenges solved by those at scale that were of the the same scale of challenges that Facebook was doing. So it would absolutely be a mistake to have underestimated or continue to underestimate China's ability to both train at the frontier, especially if they get access to compute and then continue to innovate there too.

34:31So I think it's a pretty western -centric view that I've definitely seen happen in more traditional software. I'm like, well, maybe it's caught in this 90s or early 2000s view of like, oh, all they're doing is like replicating what's already been working elsewhere and doing that, there's been products that I think take a differentiated view and grow, you know, internal to the Chinese market. And then sometimes make that very extraordinary. I mean, TikTok being an interesting example if that, on the other side. Final one before we move into like verdict, Clies, products. Just did DeepSeek cause you to rethink anything or change anything about the way that you progress?

35:05There's some architectural pieces that I want to speak for the research team because they're, you know, they're definitely the DeepExpress, was there, like, oh, interesting, like that, that's worth us considering or some ideas that have been considered and maybe were worth re -evaluating. So I think there's that piece there as well. It's interesting. Our plan was already to show the chain of thought when we launched our reasoning model. So that was not a reconsideration, but maybe it was interesting to see somebody else do that. And there's some user interface kind of details in there. And I think Grock does as well now.

35:35And there's, so it would be curious to see how that evolves to your distillation question. that might be a reason why more labs either choose to not show or otherwise obscure the chain of thought down the line. The other piece that like from a product perspective, there were two. I think that's like the under talked about piece of deep seek. It's, I think they were able to go from nobody knowing about them to dumb being like frankly in many circles better known than Claude, right? Like I like, great. I was calling me about deep seek. I'm going to leave a joke. Like it was like cliche was actually happening.

36:05Like I got a thing like, what do you think about deep to come like, great, like it's broken through. And that to me, what do you think they did to break through that maybe Claude had? I think there is a lot of interest, of course, in like world politics right now and like have the narrative be, you know, this was much cheaper, and rather that was exactly true or like what, you know, like, oh, they were able to figure something out. Like that was, you know, it's the story. And like, frankly, and I've had this conversation with our marketing team as well. I don't think we tell the Quad story well enough externally yet around what is different or what is notable about the fact that you know, the Quad three were training a model at the frontier that was state of the art with a team that was much, much smaller than any other lab, right?

36:49And I think we're we've been always very like efficient with our with our compute as we train. So I know I think that was whether that was a story that they told or was just told for them by the media because it was like was it really compelling story, The sort of uniqueness of the moment was a big piece there. And I think especially like it's January, you know, new presidency, China relations, like fed into the moment really, really, really well. So I think that worked well. And the second part of the product, like they went from not having a product to having like an iOS, but actually had a lot of like good details.

37:20And for me it was like a good like, I was just saying dodge, but it was like stronger than that. Like a shove around like we need to be getting some of the ideas out to market quicker. without your early question, focusing as much on exactly the polish that it needs to happen in every situation. And instead, be willing to put it out there and learn because sometimes the novelty of experience is itself valuable, right? It was the first time most people experienced the live chain of thought. That's interesting. And like, I wish we had done that sooner because it would have been novel for people to experience that.

37:49When you look at usage, you see emerging markets usage retains and you see Western markets, not really a tool. How do you think about them as a staining, credible threat. I think that the they already have this sort of like they're known at a level where that has some like ability to to generate the ongoing like staying powder etc. on a retention front. I think if all we're doing in these AI first sort of like lab generated products even six months from now or you from now. I was like asking questions maybe sometimes having like slight productivity. I don't think that's differentiated or interesting in the long and bread, it should be, wow, I can now do something uniquely because I am using deep sick or any one of these products and it unlocked hours of work for me and it made me smarter and it made me like a better partner to where the important people in my life is like, it has to transcend beyond surface level utility.

38:44Some people find the deeper level, don't get me wrong and those are the people that like are your DAUs right now. But for a lot of people, they'll try it, generate a poem with it, they write a letter to their son, There's all the stuff that they can do that provides some value in the moment, but I still think we are in day one around is a eye at indispensable part of most people's work. And I think the answer is no for us. And so I think deep seek and all of our honest products, staying power will come from who can get there and do that sustainably over time and have the right product design, the right integrations and the right deployment of that to actually succeed.

39:18And who can build those products? My is an investor's big question often, which is when as a model provider, move into an application provider. And I'm just fascinated to hear your thoughts around what is attractive enough where you dedicate resources to become an application provider, not just a model provider enabling. I think the two main criteria that I look at is because our team for all of them topic being big, I mean, I think we're across the thousand people our product team is, you know, maybe a tenth of that. Like by Instagram, year two standards very large, but by large SaaS company very small for summer in between all of those.

39:54It was the end resorting like, you know, you have cloud code now, we have the API, we have Claudio, a cloud for work, so it is across a lot of different surfaces. So I think generalizability is really important, even if we pick up persona or a vertical to go after, we are going to be building things that are general purpose as a rule with maybe some specialization at the user level, but not at the, I don't anticipate us building a lot of verticalized experiences that are like fairly bespoke to a given workflow use case. So I think that's one big highlight translation, transcription, customer service, quite horizontal, kind of hermortemist things.

40:32That seems like right in the pathway. I think it does, except for the fact that I think that there's a lot of valuable workflow knowledge that means that you can retain a differentiated product over time. Like, if you're a power user, yes, perhaps, yeah, if you're not a translator and you're your mom, you maybe use it once a month for that old thing that she needs. Yes. Yeah, I think the role of the great, we can help you translate this and from like, individual user will get you to pay, you know, $10 a month, this is description. That feels is iffy because I think that the models are quite good at that already, right?

41:11And maybe they're, you're right, there's not the like, if you play with the Leven's like console and workbench, a lot of the features that they've built are very clearly for people that are translating hours or like, uh, voicing hours of content with a reliable voice across the whole work stream, uh, descript, I need to script some of the best product design in AI. And like, they've clearly put so much time into the workflow. I've been I had to use it once for a personal podcast. I was like, Oh, this has clearly been built by people where day and day out sitting in this workflow and understanding it.

41:41So yeah, I think that maybe we've come to some sentences of our views, which is there's value in the more professional use cases, and the workflows that are unlocked by that. And I think on the consumer and maybe even for a consumer side, it gets good enough from a basic AI product perspective. You know, when you look at what your brilliant apps day, you do so well as we said on the code front. Is there a roadmap here to put your own ID in code agent? And how do you think about that? You know, again, with the product -focused lens, I think we have to pick our bets carefully. And even building, we built Cloud Code, which we just released as a sort of command line agentic coding tool internally first, because we just wanted to accelerate our own team.

42:22And after seeing your plan for a couple of months, we're like, this is good. Like, it's not as solution to all coding problems and doesn't object the IDE. But it's useful enough to us in enough case so that we want to see people use it out in the real world. And so shipping is never free, right? There's like, you got to name it something externally. We got to find the right packaging around it. There's a good market piece. We do it carefully. I think my view of where the models are today is you still need hands on keyboard and you still need that exchange of, hey, I did this, is this right? Well, let's pursue this direction down.

42:56And yes, this is great. Let's put up a lower cluster. No, we went down kind of like a false trail. let's like unwind the stack metaphorically and maybe an actual usage and then keep going. That's why I think that there is a role for this sort of in between IDE and the full -on, like cognition, devon, like full -on delegation of tasks, so that it can be used for a certain category of tasks. Our product engineers love Cloud Code because a lot of product engineering is, all right, we got to update the backend, we got to create the front end, we got to like submit these things for translation, we're going to like, you know, oh, the stillism work, let me do this, and it's that sort of build the product and to end workflow.

43:33That does well with the thing that can work energetically across a lot of different things. I did two full requests last week. I had encoded since joining in topic, which made me sad and so I got to finally use Cloud Code. I have not opened our code base before. So I don't really know like how it's even structured, but Cloud Code is very good at finding the file that has the right piece and then going on and and making edits and obviously not everybody's in the say a situation. I am in but it is really valuable for those use cases. So when I think about the coding space and where we can play and add value, it really is on the agenda side.

44:03It's not any idea you said, there are other companies that are spending like they wake up and go to bed every night thinking about how do we make a great idea and then involves things like low latency, auto -complete and involves like the right integrations, figure out how you play with the VS code plug -in ecosystem and all of that complexity. Right, there's a bunch of work there that is valuable and different than what we're doing. I think we can really play in let's be talking to these models and be doing real work with them in that Adjunct loop but recognize that they're not yet at the place where for many use cases you can let them kind of run free for hours.

44:35You need that more human in the loop piece. You power and you work with cursor, code, and stat blitz. My question to you is when you look at bluntly as you said that the first time you've coded since joining Anthropic and the changes that we see in developer behavior. What were the role of a software developer be in three to five years time, do you think? Yeah, I mean, I think it already looks, starts to look different already. I was a huge, early proponent of GitHub Copilot, I think, to like, my quote was on the homepage for a while, and I don't know if it still is, because I saw the potential, and then even GPT -4 came out, before it, if they had multimodal, and I was trying to do Swift with it, I would draw ASCII art of the screens I was trying to build for Artifact, and then go make coffee, because it was that time, right slow.

45:19You'd come back and it had like an 80 % version. Obviously now it would be a 95 % to 99 % version of a 3 .7 on it. I think the skills of becoming important. One, I think it becomes multi, what am I looking for? Like multi -disciplinary where it's knowing what to build as much as it is knowing what the exact implementation that you want. I love that about our engineers. Like, many, maybe even most of our good product that you guys come from our engineers and comes from them prototyping. And I think that's like what the role ends up looking like full out of one. The second piece is co -review really changes when all of a sudden you're mostly evaluating AI -generated code.

45:54I even experienced this. I put up a pull request and some of the comments that came back were, yeah, plot code does this sometimes. We don't actually do use default arguments in this case. So I was like, oh, I'm tired. So it was sheepish. If I was coding it, I would have probably noticed those patterns a little bit better. And so there's There's kind of two sides that need to happen. One, models and just the infrastructure on models need to learn from code bases and code reviews better so that they can like produce code that feels idiomatic to that company. But then also, how do we evolve from being mostly code writers to mostly delegators to the models and code reviewers?

46:31That's why I think the work looks like three years from now. It's coming up with the right ideas, doing the right user interaction design, figure out how to delegate work correctly, and then figuring out how to review things at scale. That's probably some combination of maybe a comeback of some static analysis or maybe AI driven analysis tools Of what was actually produced like is there security vulnerability? Is there some other flaws? There are bug computer use plays apart so you can tell you very excited about the space Automated testing of UI so that what would be great is you delegate the task, you know a year from now three years is crazy Like to take a year from now you delegate a task to it when you come back to it It says I've added these three approaches.

47:08I tested them all out But I had a different agent actually tried them out in a browser. This one is the one that worked best. I've run it through this initial agent that did a vulnerability test. It all looks good. All we need to do is help you resolve this one question. Let's review this particular critical section of code to make sure it's what you really wanted. That feels like you're suddenly empowered to be more of a manager and delegator to these things rather than just a partner in the loop. You said three sounds ridiculous. A year would be much more realistic. I agree and I get you when we look at the speed of scaling.

47:40Do we think that we hit a plateau or an asymptote in product releases the speed of development? Because it feels so fast now to our point earlier. Do we hit that plateau or do we continue in this exponential progression movement? There's a question I think a lot about. I started the year by looking at our product development process and looking at where we are plodified, like where we're using cloud and where we're not. And you look at him say, okay, you know, cloud can be useful and sort of taking an initial idea and creating a PRD out of it. And cloud can be useful. Obviously in the coding side, cloud can be useful and synthesizing.

48:14A lot of conversations that people are having about a product and kind of like finding like the kind of thorny issues of disagreement. Driving alignment and actually figuring out what to build is still the hardest part, right? Like that is actually like the only thing that is still best resolved by just getting together in a room and talking through the pros and cons or going off and exploring it and figment coming back. And so like any dynamic system, if you optimize one piece, all of a sudden something else becomes the the blocker or the or the critical path and Alignment deciding what to build solving real user problems and like figure out a piece of product strategy still very hard and probably like the malls are more than a year away from solving that that is the constraint.

48:51It's why I'm really bullish on at least stars to be able to explore the space because I remember this from my both Instagram and artifact days like when it's just a couple of you like like alignment is a coffee conversation in an afternoon, rather than steering the ship of a large company that has commitments to customers and all of those things. That's still a very human problem that I think we're at least three years away from the models being solving at that level of abstraction. Final one, I just have to ask before we do a quick five, but we mentioned kind of some M products there and building them.

49:21When you think about building M products for consumers versus building the API division of the company, which is very significant, how do you think about the balance and the trade -offs that between building an API business and building an end user consumer business. There's what we get out of beach, and I think about that trade -offs. So I think we learn a lot more quickly with first -party products. So as a specifically example with Cloud Code, within a week of it being deployed internally, we had found a way in which one of the tools that it has access to, the model wasn't using as well as it could have, and that made its way directly into 3 .7 Sonnet.

49:56Like that's a way in which internal dog fooding the first party tool directly led to a model improvement in the next generation. There's like a few other places where we've hit that even build a first party products much harder with the third party product. Like they'll tell you if something's wrong, but it's it's a bit more. I was like, then even though we work really closely, including with some of those coding startups that you mentioned, it's still not the same. So there's a lot of value in what we learn there. Then there's the sort of stickiness and sort of you've talked about brand and loyalty.

50:22I think it's easier from a consumer if you can build a brand around a product than just an API. The fact that we power a lot of these coding products is visible to people. It's often the default in the drop down selector. If you're in the know, you know, but not everybody does. And it's still not the thing that they downloaded, not the thing that they installed that they're going to tell their folks about. But yeah, it's also a place where we've gotten tremendous distribution. And we're not going to invent every company. And this way we can play this sort of, it reminds me of my investing days where we get to see a lot more.

50:53And there's more than one shout out on goal. And it's not all of those things. And so it's been actually a fairly, from like a resource allocation perspective, fairly even split. I think we've, if anything, underinvested a bit in two things. One is just having a faster duration spin on first party products, it's like my current obsession. And then on the second part on the API side, how do we build abstractions beyond, like, you know, tokens in, tokens out, you know. And every time we do that, we get great feedback from people. So whether that's helping the model like plan and work agentically, whether it's having the model build more knowledge graphs and repositories of like how companies operate internally if you're using the API to build more of like an internal knowledge product, whether it's perfecting tool use, whether it's understanding very large bits of context and having memory that transcends conversations like those are problems that I think are worth us solving on the API because there are things that we can take what learned what we learned on the training site and directly map it to the API and build good products around it.

51:53So that's why I think about those two, but it's a new problem. In history of disease, it was like 95 % product, 5 % API, it was, you know, that's all you really needed to do. What can and will you do to increase product speed on the first person consumer side? I think there's two things. One is recognizing that we were running, I think, a larger company playbook for what is actually like, we're still on like start, our products are even if the company he has good traction and like the API business is doing well while people are using Clod AI and upgrading Clod AI Pro. It's still early days and it's still like do or die or like make it or break it.

52:25So we need to operate in that way. And so getting the right people together sooner, faster, and ignoring organizational boundaries, we got to calcified, I think, and like, oh, well, this is on this team's plate versus this team's plate. Oh, you can't get this done this quarter because it's not on this team's, I mean, I get my organizations evolve and some of that, But it is natural, but we can't afford that right now. So it's been a lot more. We're the right people. Let's get them together. Let's clear all the other distractions. And then like, let's clear out my calendar so that I spend more of my time in product review and design review than I do in administration.

53:00Deep C showed the benefits of constraints. The Western companies, respectively, you and OpenAI have too much money. The way I would put it is the adoption that we've gotten of our products is ahead of their actual true product market fit because they are still the best ways of getting the models. And I don't think that's durable over time. So I think that's not a thing that to rest on. And two, I just think we're underserving people because I don't think we've gotten the right products yet. So it's what I wake up stressed out about every morning or inspired by defending on the day. It's like, I think we've got so much work to do on that side.

53:32Listen, I want to do a quick fire on. So I say a short statement, you give me your immediate thoughts. Because that sound okay. That sounds great. What's OpenAI done better than you on? They've moved faster at shipping V1s, even in ahead of where the model is sometimes. What have they done worse than you on? Probably personality and having the features they've built be cohesive. Which alternate model provided you most respect? OpenAI. I think they've balanced first -party product development and an API that people use at scale as well. But we had an Instagram principle that was do the simple thing first and I think they often do the simple thing first.

54:10If you could rebuild the anthropic product and stack from scratch, what would you do differently? I love this question. I do too, it's a good one, isn't it? Yeah, it's a really, really good one. I think the things that we built that were actually very valuable last year are now feeling like they're having, this is a long etsy, rather than quick fire, I'm sorry. Have some cost to the information architecture, which in this sense, like a very nerdy way of describing it, but basically like, people should not have to think about like projects versus artifacts versus chats and how they all relate. And I think Terry and all down and being like, what actually matters is do you have the right context into the right conversations?

54:46Do you feel like you can always know where to go next in the product and is anthropic and plot itself being a helpful sort of guide to what work is most important to do next? Is a different paradigm and like I know to create a project and then like if you get good that it's an amazing product. There's a lot of steps along the way. So that was that on the on the product side. I think that's the fundamental thing. On the stack, I mean, Claudet and probably chatgbt .com were very much initially just built to be showcases of the models and not really built in a lot of ways to be the foundational for a much more complex multi -product sort of thing.

55:24I think we have an active effort right now around tearing down some of that and rebuilding the Core UX to just feel good. It doesn't feel great right now. feels a little bit like it's been an evolution of a product that served a purpose at the time but now is being asked to do way more things such that the incremental thing is now both harder to add and getting slow. What have you changed your mind on in the last 12 months? How much first -party stuff is important? I think I saw the growth in the APN I was like this is what we should just invest a lot more of our time in and I think that there's you'll you'll miss out and not have enough of a durable mode if you're not equally investing maybe even investing even more on the first -party side of things.

56:03How much did it hurt you being late to that? I think significantly if you take a deep seek moment, like ideally the story of, oh, there's more than one, right, they're leading AI product to be used, is some narrative that we should have captured. I think it hurt us there. What's a major technical or product challenge on the horizon in AI that no one's talking about that you think is critical? The models as they get more, or maybe the headline, which is basically like discernment and privacy. So as the models get more capable, they'll also become more knowledgeable, right? They'll have you being conversations with them about everything from funding that might be quite intimate or something that's quite sensitive from a company perspective, or they'll have access to all of your, you know, particular companies things.

56:46And then everybody loves to talk about agent agent interaction, right? The intersection of those two non -know -people I think or talk about, I think, which is, Do you trust your Mike agent or your hairy agent to be out in the world and then not be jailbreakable or reveal something that it knows that is like quite personal or sensitive right? I think my metaphor is my five year old. It's great watching her with You know something that she's just met because she doesn't quite different You feel like stuff that's secret and private to our family and stuff that is like things that are okay to talk about with a new friend Or you know somebody at the checkout aisle so that That discernment is something you acquire over time for people.

57:24And I think models, this is very underappreciated and probably under research as well from like a model capabilities perspective because models fundamentally want to be helpful. And that is not always what you want them to be. And there's a safety case for that. But then they think it also a privacy and data security case for it too. Q. Why are you about your five year old becoming more comfortable talking to models and agents than they are humans? I've had so many conferences with Alex Mang about this because he has this whole thing of All in the Future, most friends will be AI friends. And you know, I don't think he's wrong.

57:55And I think that there is, there's ways in which that's already starting to be the case with, you know, people, you know, having lots of online game experiences and some of those NPCs that you might just have like more like a comfortable sort of existence in there as well, even if you know, breaking through. So I do, I worry, she is so gregarious that like, I'm not actually worried in her particular case, but like, let's abstract the broader sense, there is a lot you can learn from what it feels like. Like, here's the bull case. I was a fairly awkward teenager and I probably could have been benefited for some practice mode, like AI interactions around some of these things to build up.

58:32And at the same time, that's like not the real. It doesn't feel like it's totally closing the loop around like the consequences of real interaction. Like, it's a difference between reading about what it's like to have your first, like, really a hard argument with your high school girlfriend, and then actually having it. And like, when you're in that moment, you know, This is like now the classic like the Chinese room experiment where it's like you're it's not the Chinese room It's a different thought experiment where like somebody's in you know a black and bright room only reading about red and they gone to roll and they see red And like is there something qualitative differently about that?

59:01Absolutely and is there something different between talking to a model and engaging a model even an emotional roleplay and having that same interaction with a real human like Absolutely, and so it is a probably a helpful piece of feature of even interaction and absolutely insufficient as like the whole Does Europe become more or less relevant in an AI driven decade? Europe? I want them to do well because I love a lot of Europe and I have, you know, I've lived in Portugal growing up as well. I saw a funny, maybe, someone defeat his argument where if real world experiences and human interaction become more valued, Europe becomes more valuable itself as like the perhaps world capital of sensory and, you know, experiences, that feels weird as the, if that's all you're resting on that that feels a little limited in there as well.

59:48But I think it'll be really interesting from a Europe perspective or European perspective is what are the things like a thing I really respect about Europe is there's often in the case that there are things about the lifestyle or the society that they hold very, very strongly that then they not always elegantly, but at least attempt to enshrine in either like best practices or even laws. And so even as we think about doing our product design and data privacy and selling to German users or German companies. There's a different set of questions that get asked that are often very helpful questions.

1:00:19And so maybe the bull case there is that those are actually questions that are relevant to everybody and there will just be at the leading edge of asking some of those questions. I think from a lab's perspective, it's a lot harder question to answer. I think there's maybe some combination of access to compute. Maybe they move further up the value chain. And if it is the case that building applications above these malls becomes, it is a lot easier and you can go from zero to one and you can be more nibble than even these labs that are going to all have like tens or hundreds of millions of users and you have to move slowly at that pace.

1:00:49Can innovation happen there? Probably, but it probably involves a different both regulatory and startup ecosystem environment to really make that actually the case. Final one. Dario has said that this will be the generation that could live to 150. I'm slightly about butchering summarizing his quote, obviously. But like this could be the generation. I'm very optimistic. My mother has multiple cirrhosis that will point years for diseases like MS with AI. Do you agree with his optimism? And how do you think about AI increasing longevity and human lifespan? Yeah, I think the potential is huge. I think there's everything from today where AI is helping, which is in closing the loop on drug discovery and closing the loop on clinical trials where a nubanordist used to take, clinical trial reports, now they use cloud and get it done in 20 minutes.

1:01:40And like, that's a step change out. There is years of research I've preceded at. So I'm not saying that we've cut years to weeks, you know, or years to minutes, but that's a point, you know, of the process that we can make fast. And that's like with the models today. And you see ARC, which is this science and research is to do the Patrick calls and some others have started and funded. They're working on foundational models for cells, right? Where you have all of a sudden, a real cell model that you can run experience on in that kind of thing should also accelerate drug discovery and experimentation.

1:02:08They're tremendous things. All of a sudden you're you're cutting the loop there. So I'm very optimistic. There's a lot of places where AI is I think underutilized relative to its potential. And I think some of the smartest people in the field in the cloud like smartest minds of my generation were working on like serving more targeted ads. Maybe that was true at one point. I think a lot of them today are working on how do you make models that are tremendously useful and valuable and intelligent across a lot of domains. Mike, you've been fantastic. Thank you so much for letting me just completely unpack all of my questions on you without warning, but you've been amazing.

1:02:41The McPasher, really fun to do this. I just love doing that show with Mike, and if you wanted to see more from the episode you can find it on YouTube by searching for 20VC, that's 2 -0VC on YouTube. But before we leave you today, turning your back of a napkin idea into a billion dollar startup requires countless hours of collaboration and teamwork. It can be really difficult to build a team that's aligned on everything from values to workflow, but that's exactly what Coda was made to do. Coda is an all -in -one collaborative workspace that started as a napkin sketch. Now, just five years since launching in beta, Coda has helped 50 ,000 teams all over the world get on the same page.

1:03:21Now, at 20VC, we've used Coda to bring structure to our content planning and episode prep, and it's made a huge difference. Instead of bouncing between different tools, we can keep everything from guest research to scheduling and notes all in one place, which saves us so much time. With Cody you get the flexibility of docs, the structure of spreadsheets and the power of applications, all built for enterprise, and has got the intelligence of AI which makes it even more awesome. If you're a startup team looking to increase alignment and agility, Cody can help you move from planning to execution in record time.

1:03:54To try it for yourself, go to coder .io slash 2 -0 -VC today and get 6 free months of the team plan for startups. That's coder .io slash 2 -0 -VC to get started for free and get 6 free months of the team plan. Now that your team is aligned in collaborating, let's tackle those messy expense reports. You know, those receipts that seem to multiply like rabbits in your wallet, the endless email chains asking, can you approve this? Don't even get me started on a month and planet when you realise you have to reconcile it all. Or Plio offers smart company cards, physical, virtual and vendor specific, so teams can buy what they need while finance stays in control.

1:04:33Automate your expense reports, process invoices seamlessly and manage reimbursements effortlessly, all in one platform. With integrations to tools like Zero, QuickBooks and NetSuite, Plio fits right into your workflow, saving time and giving you full visibility over every entity, payment and subscription. Join over 37 ,000 companies already using PLEO to streamline their finances. Try PLEO today. It's like magic, but with fewer rabbits, find out more at pleo .io -4 -20vc. Don't forget to secure trust with your customers. Trust isn't just earned though, it's demanded. That's why over 9 ,000 companies, including Atlassian, Cora and Factory, rely on Vanta, to automate their security compliance.

1:05:17So Vanta helps businesses achieve certifications like SOC 2 and ISO 27001, turning months of tedious work into this beautifully fast and straightforward process. Their platform automates compliance across over 35 frameworks. Its centralizes workflows and it proactively manages risk, all while saving new time with automation and AI. So whether you're just starting or scaling your security program, Vanta connects you with auditors and experts to get audit ready quickly and build trust with your customers. Get $1 ,000 off your first year by visiting vanter .com -2 -0VC.

1:05:56As always I so appreciate all your support and stay tuned for an incredible episode this coming Wednesday with Anton at Lovable, the fastest growing company in Europe.