OpenClaw, Claude Code, and the Future of Software | Peter Yang on The a16z Show

Transcript

And since that software will eat the world, I feel like coding will eat all knowledge work, right? And we're going that direction already. The whole agent stack is emerging. Yeah. Identity, payments, marketing, even CLI versus MCP. All of these are really new things. And I think a lot of the old playbook goes away. Yeah. It's a whole new world. I hope more companies will stay small. And I think the founders of this generation realize that. They want to stay as small as possible. Yeah. And instead of having a 10% product team, you have a two or three person product team. And you just have a bunch of agents to help you. Yeah. Someone tweeted that the job market is so bad that I can only pursue my dreams now or something like that. So it's, yeah, it's like, yeah. So maybe you lost your job, but now you can actually do your own thing. A hundred percent. [SPEAKER_01] And have a shot at actually achieving it. [SPEAKER_01] Yeah. [SPEAKER_00] Yeah. All right. [SPEAKER_01] Welcome everyone. [SPEAKER_01] I've got my friend Peter Yang here. Peter, welcome. [SPEAKER_00] Yeah. [SPEAKER_00] Great to be here. [SPEAKER_00] It's good to see you again. Yeah. It's great to see you. Peter and I worked together at Credit Karma for a brief stint. And then we went our separate ways. And I rediscovered Peter from his prolific posts on X and your YouTube. And you've got a little bit of a Clark Kent Superman thing going because you've still got a day job, right? [SPEAKER_00] That's right. [SPEAKER_00] I still have a day job. [SPEAKER_00] Yes. Can you share where? [SPEAKER_00] Yeah. [SPEAKER_00] I work at Roblox as a PM. Amazing. Roblox. [SPEAKER_01] And Dresden Portfolio Company. [SPEAKER_00] Yes. One of my favorites. Well, incredible, man. Let's get right into it. Maybe I'll start with a softball fun question. And then we're going to talk about everything in the Claude ecosystem. [SPEAKER_01] We're going to talk about coding agents. We're talking about maybe what students should study advice and some of the things that you've talked about online. [SPEAKER_00] Yeah. [SPEAKER_00] Sure. Maybe to start, what is the name of your Claude, how many do you have? And tell me their names. [SPEAKER_00] I only have one. [SPEAKER_00] I call her Zoe. Zoe. But I have multiple conversations going with her. [SPEAKER_01] Okay. Yeah. [SPEAKER_01] And why Zoe? I have two girls and I was going to call my younger one Zoe and I did not. So I call my Open Claude Zoe instead. [SPEAKER_01] I see. [SPEAKER_01] I see. [SPEAKER_01] Yeah. [SPEAKER_01] This is your fallback plan. Yeah. [SPEAKER_01] Peter, tell me a little bit about Open Claude, how you discovered it, how you're using it today and what you think the implications are. Yeah. I was lucky to interview Peter Steinberger before he became super famous and the whole thing blew up. And right after I interviewed him, I set up the thing. It took forever to set up. It was super janky. And yeah, it does a lot of things for me. It pulls analytics for me across YouTube and my Mercury bank account. [SPEAKER_00] It can update Google documents for me. [SPEAKER_00] It can build websites for me. But if I was honest with you, dude, I mostly just talk to it through voice and get voice replies. And every other day I ask it to give me a pep talk or look through all your memory and give me some deep insights that I don't know about. [SPEAKER_01] Okay. And it gave me a three minute pep talk that was really amazing. Really amazing. It was something about talking to me about your creator business and blah, blah, blah. And your job. But just remember that your kids, seven out of four are going to grow up very soon and they're going to want to spend time with you. [SPEAKER_01] Wow. [SPEAKER_01] So you should re-optimize for them instead. [SPEAKER_01] Yeah. [SPEAKER_01] Yeah. [SPEAKER_01] That's really cool. [SPEAKER_01] And I mean, very cool, but also something that all the language models could have done prior. Yeah. [SPEAKER_01] So what's the difference between this and in a use case like that? Yeah, that's a very good question. So I don't know, because I've installed it on Telegram, it just feels more personal than using Claude or ChatGPT. And it just feels something I can text in bed. It's probably not very healthy, but I text to it in bed. I've talked to it during my commute and it feels more personal, an actual human. [SPEAKER_01] Yeah. [SPEAKER_01] Yeah. [SPEAKER_01] So how much for you is Open Claude about the interface, pushing it to messaging and maybe helping to trick our brain into feeling like this is a person or a person-esque thing versus all the other components of the stack, the self-modification, the skills directory. [SPEAKER_01] All the rest. I think it's probably 70, 80% just the personal part of it, because I mostly just talk to it through voice. [SPEAKER_00] But I also think it's something, first of all, it is pretty janky. [SPEAKER_00] It tends to forget things a lot. [SPEAKER_00] Yeah. [SPEAKER_00] I have to keep reminding it. [SPEAKER_00] But any zany idea that I have, I just have to talk to it and it can probably [SPEAKER_01] a person-esque thing versus all the other components of the stack, the self-modification, the skills directory. All the rest. I think it's probably like 70, 80% just the personal part of it, because I mostly just talk to it through voice. But I also think it's something. First of all, it is pretty janky. It tends to forget things a lot. Yeah, I have to keep reminding it. But any kind of zany idea that I have, I just have to talk to it and it can probably just do it. Like, the other day I was doing voice replies with it. I was like, hey, can we just have a live phone call instead? And then it's like, okay, you got to connect to it. You got to do all this stuff. And then, okay, fine. I went off and did it. [SPEAKER_01] Yeah. [SPEAKER_00] And then we had a phone call. It called my phone. Oh, really? You have that set up? [SPEAKER_00] Yeah. I'd be dying to set that up. Okay. [SPEAKER_00] It's not very good though. The latency is bad. But the fact that I was able to get it going is pretty impressive. [SPEAKER_00] So any kind of crazy idea I have, you can do. And then in practice, how are you doing that? Are you asking it to write a skill on the fly? Are you discovering a skill? How much of the code gen are you actually using? I mean, I talk to it in a super casual way, just like a friend. So I'm like, hey, you know, hey Zoe, can you have a phone call? [SPEAKER_00] Yeah. It's like, okay, you got to do that. I was like, okay, fine. I'll open my computer. I'll do all this stuff. And then it's like, give me a call. And it will troubleshoot a little bit. And then it works. So with Claude, I have very fancy prompts, very long prompts. But with Open Claude, I just text it. [SPEAKER_01] It is really interesting. So we sort of touched on a couple of things actually. So one, there's mobile messaging. [SPEAKER_01] Yeah. [SPEAKER_01] There's the memory system. There's the code generation component. How much do you think the memory system is innovative because it's file-based? [SPEAKER_01] You said that it forgets things, but language models do too. Do you think the memory system is well done? Does it hold it back or does it enable it? I think the default memory system is actually not that great. Okay. The way I understand it works is just a memory.md text file. [SPEAKER_01] Yes. And then every day. [SPEAKER_01] Per day. Right. And every day it updates. And it tends to forget things a lot. Yeah. So I actually installed this three-layer memory system that, to be honest, I don't fully understand. But it has like... [SPEAKER_01] Sounds fancy. [SPEAKER_00] It has Toby's QMD search tool. [SPEAKER_01] Okay. So I installed that. And then I installed a two gigabyte thing. And then it got a little bit better. Okay. [SPEAKER_00] But I just have to remind it. I have to put it into the agents.md. Like, hey, before you answer any question from me, go through all your memory and check everything. [SPEAKER_01] Yeah. [SPEAKER_00] And it also tends to forget that it can do stuff. Like, you know, can you update my Google Doc? [SPEAKER_00] It's like, oh, I can't do that. [SPEAKER_01] Yeah. Yes, you can. It's in your file. [SPEAKER_01] Yes. Yeah. I have to remind it. [SPEAKER_00] Yeah. Yeah. Really interesting. Well, maybe let's get into a little bit of the controversy. [SPEAKER_00] Sure. [SPEAKER_01] You know, you'd said that apps will die. Claude is going to be everything and everywhere. Talk us through that point of view. Yeah. Well, first of all, I tweet all kinds of random crap that's not super well thought out. [SPEAKER_01] We take it all as fact. Yeah. [SPEAKER_01] Yes. But I do think ever since I set up all these apps like Mercury, MCP, and all this kind of stuff on my open Claude, I don't actually open those apps much anymore, you know? But I do agree with you. I think the ones that are going to die first or maybe get less usage first are apps that you're just opening to try to complete a task. Like you're actually trying to do something, you know? Like apps that you're opening to get entertainment can probably survive a little bit longer. But apps that I'm opening to complete a task, it's just way better to text my agent to do it for me. [SPEAKER_01] Yeah. It's like you have a really good admin just to do stuff for you. [SPEAKER_01] Yeah. [SPEAKER_01] Yeah. [SPEAKER_01] And so how much are you finding? Has this reduced your smartphone usage outside of modular open Claude? [SPEAKER_00] Yeah. No, because I'm a Twitter addict. I still use my smartphone way too much. But yeah, in terms of using those apps, it's definitely reduced it. Yeah. Yeah. Because you're not going to ask Zoe like, hey, read my apps for me and tell me what's interesting. [SPEAKER_00] I mean, it sends me a morning briefing of the top two tweets and stuff that trends. But yeah, I still open apps. I look through it. Yeah. [SPEAKER_00] Yeah. You know, it's interesting because I've always had this theory that people open apps on their phone because they want to feel a feeling. [SPEAKER_00] Yeah. [SPEAKER_00] I still use my smartphone way too much. [SPEAKER_00] But yeah, in terms of using those apps, it's definitely reduced it. [SPEAKER_01] Yeah. [SPEAKER_01] Yeah. [SPEAKER_01] Because you're not going to ask Zoe, hey, read my apps for me and tell me what's interesting. [SPEAKER_00] It sends me a morning briefing of the top two tweets and stuff that trends. But yeah, I still open apps. I look through it. [SPEAKER_01] Yeah. Yeah. [SPEAKER_01] It's interesting because I've always had this theory that people open apps on their phone because they want to feel a feeling. Yeah. [SPEAKER_01] And I think, of course, there's some functional set of needs, which is why you open calendar or something. [SPEAKER_01] But I also think that WhatsApp is you want to feel connected and Slack is you want to feel productive. [SPEAKER_01] And of course, TikTok is you want to feel entertained. [SPEAKER_01] So I do wonder with just one agent, how do you do the context switching of when are you flirting, when are you getting stuff done? In a sense, app gives you a nice division of the intent. Yeah. That's a good point. Which you don't get with Zoe. [SPEAKER_00] That's a good point. But I do have multiple channels set up with Zoe in Telegram. Like one is just for random voice replies. And the other one is we're actually working on our project together. Oh. And the other one, I have a public channel where I'm giving demos. I don't want to reveal private information. [SPEAKER_01] Yes. So I have multiple channels. And is that implemented as subagents or? No, it's just some janky setup I found online. You can set up multiple Telegram channels. And then I'm not sure if it actually remembers across contexts across the channels, but you can have separate conversations at least. Got it. Yeah. [SPEAKER_01] And how transparent are you with your agent? Do they see your personal email? [SPEAKER_00] I'm super transparent. Well, I did buy the Mac Mania and set up its own email. Okay. [SPEAKER_00] But I gave it read access to my email and calendar. [SPEAKER_00] And I also gave it write access to some docs. [SPEAKER_01] Yeah. But it can screw my entire drive or something. [SPEAKER_00] Yeah. So. How do you imagine OpenClaw, which is sort of an architecture and a primitive, how does it get productized, packaged for the world? I mean, I think that's what Peter Steinberg is working on at OpenAI, right? Yeah. It's probably going to build something to ChatGPT, which everybody uses. So that ChatGPT can actually get stuff done for you and maybe feels more human. [SPEAKER_01] Yeah. Dude, let me rant about ChatGPT. [SPEAKER_01] Please. Yeah. For some reason they trained the model so that at the end of every conversation it's always, if you want, I can also do X and Y. Yeah. [SPEAKER_00] And dude, I got so annoyed about it that I kind of churned from ChatGPT. Oh, really? [SPEAKER_00] Yeah. So it probably increases their metrics, but it's just super annoying. It's, why don't you just do it in the first place? Are you a Cloud guy now? Yeah, I'm a Cloud guy now. But I do use Codex to code. Yeah. Yeah. [SPEAKER_01] You like Codex, you prefer it to Cloud Code or you use both? [SPEAKER_00] Codex when I want to try to do something real and Cloud Code is when I'm just vibing. Yeah. [SPEAKER_01] It's very interesting. [SPEAKER_01] I think they live at different points and there's a sort of space of trade-offs. [SPEAKER_00] Yeah. [SPEAKER_01] Whereas I find Cloud Code and Opus 4.6, it's a little more chatty. It makes more assumptions, but it can be more pleasant for a synchronous experience. Yeah. Whereas Codex, it really thinks hard and it's more often accurate. But sometimes it's being in a conversation where the other person pauses for three minutes to think. [SPEAKER_00] Yeah. [SPEAKER_00] You don't have flow state, right? [SPEAKER_00] It's hard to get a flow. [SPEAKER_00] Cloud Code, dude, I tweeted the other day that Cloud Code is almost a slot machine. It has different things each time. It's just. Oh, 100%. It's a slot machine. [SPEAKER_01] Look, I do think that if you think, remember what we were talking about in the old social networking era? [SPEAKER_01] It was variable scheduled rewards, right? [SPEAKER_01] That was the whole magic of it. You open your Facebook feed and, you know, once in a while it's boring, boring. [SPEAKER_01] Oh, my God. [SPEAKER_01] This is so exciting. [SPEAKER_00] Yeah. [SPEAKER_01] And the coding agents have the exact same property. [SPEAKER_01] Also, the time is variable. [SPEAKER_01] So sometimes you get something in a second. [SPEAKER_01] Sometimes it takes five minutes. [SPEAKER_01] So up to a certain point, I actually think that both of those things give it that casino-like feeling. Yeah. And the other thing that's very different about the product strategy or maybe just the way it works is coding is self-explanatory. [SPEAKER_00] And Cloud Code, you have all this crazy stuff. You have hooks and skills. You have to. Plugins. [SPEAKER_00] If you're not following. Yeah. If you're not following X, you have no idea how to customize this thing. Yeah. But once you customize it, you feel it's part of you. So it's hard to turn. [SPEAKER_01] It's interesting with. So I've customized mine because also I read the long thing that Boris put up. Yeah. But I will say that I think Cloud Code, a lot of the reasons that I enjoy it are just inherent features. You have to use plugins. If you're not following X, you have no idea how to customize this thing. [SPEAKER_01] Yeah. But once you customize it, you feel like it's part of you. [SPEAKER_00] So it's hard to turn. [SPEAKER_01] It's interesting with. So I've customized mine because I read the long thing that Boris put up. [SPEAKER_00] Yeah. [SPEAKER_01] But I will say that I think Cloud Code has a lot of reasons that I enjoy it because of harness features. [SPEAKER_01] For example, if you cut an image, you have to paste it into a file and then paste that file into Codex. Okay. You can't just take a subset of the screen, screenshot it, and then paste it directly into Codex the same way you can with Cloud Code. Oh, really? [SPEAKER_00] Okay. Okay. [SPEAKER_01] So just little things like that. Cloud Code added voice. [SPEAKER_01] It's a little bit janky right now, but it's going in the right direction. [SPEAKER_01] So they've got a bunch of quality of life things. [SPEAKER_01] Yeah. [SPEAKER_01] Cloud Code speaks to Cloud in Chrome. Okay. And Codex doesn't speak to Atlas. [SPEAKER_00] Got it. [SPEAKER_01] So I think these are all things that OpenAI will fix. [SPEAKER_01] Yeah. I think Codex is actually a much better model, but they don't exist today. [SPEAKER_00] Yeah. [SPEAKER_00] Yeah. They need to fix it. [SPEAKER_00] They're going to go all in on Codex, I'm sure. [SPEAKER_00] Yeah. Talk to me about coding agents. What's your general view? Do you think it's the end of SaaS? [SPEAKER_01] Do you think these are just a toy? Well, first of all, I'm not an engineer, so I'm a novice. But I do hear that I was talking to some folks the other day at an AI native startup. And they're basically trying to build things—they have a bunch of vibe coders. And all the vibe coders are just trying to build internal tools that replace the SaaS that they're paying for. Really? Yeah. So it's an actual company that's doing this? [SPEAKER_00] It's an actual company. [SPEAKER_00] Yeah. [SPEAKER_01] It's an AI native company. [SPEAKER_00] It's one of the vibe coding companies, one of the more popular ones. [SPEAKER_00] Yeah. [SPEAKER_00] Interesting. [SPEAKER_00] Yeah. It's interesting. [SPEAKER_01] Oh, I see. So they're actually an AppGen company. [SPEAKER_00] They're an AppGen company and they're paying for a bunch of SaaS. They want to get rid of the payment. [SPEAKER_00] They want to just build AppCode internal tools. By using, okay. So in that case, they might be the most extreme form of adopter because their own product is AppGen. [SPEAKER_01] So they should use AppGen for everything. [SPEAKER_01] Yeah. [SPEAKER_01] I guess, is your prediction though that the average company will churn off of Slack or Deal or, you know? I don't think Slack has a lot of legs because Slack can also be the place where you talk to the agents themselves. But some of the other ones, they are pretty complicated. It's hard to vibe code that kind of stuff. But if you have an app like Calendly or something more simple, then why should I pay for it? [SPEAKER_01] The point is that it's not that expensive. Yeah. [SPEAKER_01] And do you really want to maintain your own Calendly thing? [SPEAKER_00] Yeah. [SPEAKER_01] You know, versus pay twenty bucks a month. [SPEAKER_01] It always gets updated. It's always up. [SPEAKER_00] Yeah. [SPEAKER_01] Because there's just a fixed amount of capacity that anyone in the organization is going to have for all this stuff. [SPEAKER_00] Yeah. [SPEAKER_00] That's true. Unless you hire dedicated vibe coders like the startup does. [SPEAKER_00] It's just one of the vibe cool stuff. [SPEAKER_01] But then it's the cost benefit versus just paying for Calendly. Yeah. Yeah. [SPEAKER_00] Yeah. [SPEAKER_00] It's interesting, for example, a lot of people are tweeting about Figma recently. Yeah. [SPEAKER_00] Like the stock is down. [SPEAKER_00] Are you going to survive? [SPEAKER_00] Yeah. [SPEAKER_00] And I feel the jury's out there. It's hard to say. [SPEAKER_00] Yeah. [SPEAKER_00] I feel all the designers are still on Figma, but as a designer, you need to learn how to vibe code. [SPEAKER_00] Otherwise, you're going to be out of date in a couple of years. Yeah. [SPEAKER_01] My counterpoint to that is that I've thought a lot about the sort of thinking tools versus making tools. Yeah. [SPEAKER_01] The IDE was historically a making tool. [SPEAKER_01] It's a place for execution. I think it's migrating away from that. And now with execution going to zero, I think these multi-agent, next gen IDEs are about trying things and using trial and error as a way to inform your thinking. Yeah. A lot of times I'll just build a feature in a really naive way and I'll hammer the coding agent until it works. [SPEAKER_01] Then I'll say, hey, write all the things that you would have done differently. [SPEAKER_01] And I'll go back to the initial point and redo it. [SPEAKER_01] So I wonder if, and I think Figma actually does both. [SPEAKER_01] I think it's a place for design execution, but it's also an important place for design thinking. Yeah. [SPEAKER_01] And I think that's their opportunity to be highly relevant in the new stack. Yeah. I totally agree. [SPEAKER_01] Yeah. [SPEAKER_01] A lot of times I'll just build a feature in a really naive way and I'll hammer the coding agent until it works. [SPEAKER_01] Then I'll say, hey, write all the things that you would have done differently. And I'll go back to the initial point and redo it. [SPEAKER_01] So I wonder if, and I think Figma actually does both. [SPEAKER_01] I think it's a place for design execution, but it's also an important place for design thinking. Yeah. [SPEAKER_01] And I think that's their opportunity to be highly relevant in the new stack. [SPEAKER_00] Yeah. [SPEAKER_00] I totally agree. [SPEAKER_00] I totally agree. But I think A16Z has, you guys invest in Pencil or something? Pencil.dev? Speedrun did. [SPEAKER_00] Yeah. [SPEAKER_00] Yeah. [SPEAKER_00] Speedrun. [SPEAKER_00] And yeah, Figma needs to level up its AI tooling because watching these agents collaborate with you and do stuff is very, very interesting. I know it's top of mind for them. Yeah. What do you think are the most under-discussed capabilities of coding agents? [SPEAKER_01] What's under-hyped and maybe what's over-hyped as well? This is probably not under-hyped, but I feel like software will eat the world. I feel like coding will eat all knowledge work, right? And we're going that direction already. I think Lovable recently launched today. Yeah. But they can support everything and can make decks. [SPEAKER_01] Yeah. So, yeah. [SPEAKER_00] So everyone's chasing this. And Anthropic is probably in the lead. [SPEAKER_00] Yeah. But I don't want to use PowerPoint anymore. [SPEAKER_00] I don't want to write a Google. [SPEAKER_00] Yeah. [SPEAKER_00] I hate writing Google Docs, dude. I've lost my entire life. [SPEAKER_00] Yeah, yeah, yeah. [SPEAKER_00] So the other day I was writing my blog posts and instead of just typing it out, I was like, hey, let me just use Claude and give you a bunch of feedback and you write it for me. Yeah. And then you just keep doing that. It did the first 80%, the last 20% I started to manually go in there, tweak, tweak, tweak stuff. [SPEAKER_01] Yeah. But that's the way I work now. [SPEAKER_00] I never start from zero. I always get the first 80% from AI. [SPEAKER_01] Right. Yeah. Yeah. [SPEAKER_01] It's interesting. [SPEAKER_01] You know, if you look at there, there are also historical analogs of this. I think Satya said this, which is that Excel is the most powerful or most popular programming language in the world. Yeah. [SPEAKER_01] And it's a programming language that millions and millions, I mean, a hundred million plus people must know, maybe even more. [SPEAKER_01] And yet we don't think of it that way. [SPEAKER_01] It's a way to describe and solve problems. Yeah. [SPEAKER_01] And I think coding agents are going to be that, of course, times a thousand. Yeah. [SPEAKER_01] Or even things that feel subjective, like writing Google Docs can be represented in the coding domain in such a way that it's more satisfying, more productive, more high leverage to use agents to do it. Yeah. Because Excel was popular because it's super approachable, right? [SPEAKER_00] Yes. And coding agents, the code is basically gone. You're just talking to some agent and getting them to do stuff. [SPEAKER_00] So yeah, exactly. [SPEAKER_00] It's going to be huge. Yeah. What do you think the future company looks like? Is it just a bunch of agents with a CEO? Is the CEO an agent? What is the role for people in a company in the future? [SPEAKER_00] Okay. [SPEAKER_00] Well, I have some hot takes. [SPEAKER_00] So we both worked at some companies together and let me give you a hot take, man. [SPEAKER_00] Maybe we cut this out. [SPEAKER_00] But I feel like as a company gets bigger, it tends to become a shitty place to work, dude. [SPEAKER_00] Yeah. [SPEAKER_00] Because there's a lot of people you have to align. I think that's axiomatic. Yeah. [SPEAKER_00] Right. And I remember, you know, maybe I should mention this comment, but I remember our company, we used to have all these OKR meetings and I remember sitting in a room for three hours talking about OKRs. [SPEAKER_00] I'm just like, dude, this is wasting my life. [SPEAKER_00] Yeah. So where I'm going with this is I hope more companies will stay small. And I think the founders of this generation realize that they want to stay as small as possible. [SPEAKER_01] Yeah. Instead of having a 10% product team, you have a two or three person product team and you just have a bunch of agents to help you. [SPEAKER_01] Yeah. Because I think it's way easier to cross-function a line with agents than with humans. [SPEAKER_01] Yeah. [SPEAKER_01] Yeah. [SPEAKER_01] Well, actually in a sense, the agents actually take the emotion out of it too. [SPEAKER_01] Yeah. [SPEAKER_01] Like you can imagine if I sent my agent, you sent your agent to go negotiate something. Yeah. And they came out with some conclusion. It's not emotional. It's not. [SPEAKER_01] For either of us. It's very objective. [SPEAKER_01] Yeah. [SPEAKER_01] Yeah. [SPEAKER_01] Yeah. [SPEAKER_01] Exactly. [SPEAKER_01] Yeah. It's funny. [SPEAKER_01] You know, one of the things that we've been talking about is what is the pro case for AI at work in terms of employee experience? [SPEAKER_01] Yeah. [SPEAKER_01] You can imagine if I sent my agent, you sent your agent to go negotiate something. [SPEAKER_01] Yeah. [SPEAKER_01] And they came out with some conclusion. [SPEAKER_01] It's not emotional. [SPEAKER_01] It's not. [SPEAKER_01] For either of us. It's very objective. [SPEAKER_01] Yeah. [SPEAKER_01] Yeah. [SPEAKER_01] Yeah. [SPEAKER_01] Exactly. Yeah. [SPEAKER_01] It's funny. [SPEAKER_01] One of the things that we've been talking a bunch about is what is the pro case for AI at work in terms of employee experience? [SPEAKER_01] And I think it's what you're describing, right? How do you increase the NPS of work? Yeah. [SPEAKER_01] So if we go all the way back or even broadly the NPS of the human experience, right? [SPEAKER_01] Think of the NPS of the day-to-day human in 10,000 BC when it's just don't get eaten by the lion. Yeah. [SPEAKER_01] And that's a good day, right? [SPEAKER_01] Or maybe a hundred years ago, it's okay, don't get killed at the factory, crushed by the steam press or whatever else. And now a lot of it is, oh, just don't get sucked into some high emotion negotiation with another VP's subordinate. [SPEAKER_00] Yeah. [SPEAKER_00] A 50-message Slack thread going back and forth. Yeah. [SPEAKER_01] Exactly. And then eventually everyone's like, I don't want to tell the CEO. And eventually it goes there and it's just terrible. Yeah. So maybe the future of this is that a lot of that emotional subjective work gets handled. Yeah. And we're sort of guiding the process, but not in the middle of it in a way that just doesn't suit us as humans. Yeah. [SPEAKER_00] I live that double life as a PM creator and I feel like all the PMs actually just want to create products. [SPEAKER_00] They just want to create products. Well, that's why we all got into it. [SPEAKER_01] You know, it's so interesting. [SPEAKER_01] I mean, Nikhil talks about this all the time, but every PM's sort of view of the ideal PM is the innovator, you know? [SPEAKER_01] I came up with the new thing and I sort of had the big insight and it unlocked the product. Yeah. [SPEAKER_01] I think the black pill is, I don't think most PMs know how to do that. [SPEAKER_01] In fact, many companies have zero people that know how to do that at all in any function. [SPEAKER_01] Yeah. [SPEAKER_01] So nonetheless, I think PMs aspire to be able to do that and they should either do it and either be successful or maybe not successful and move to a different function. I also feel my hot take is all the PMs I know are trying to code at home on weekends and I feel like if you're actually unemployed, you probably have more time to be a builder and to be innovative. [SPEAKER_00] Yeah. [SPEAKER_00] Because you can actually play with all this stuff and learn all this stuff. [SPEAKER_00] A lot of PMs are trying to- Or maybe being an engineer on the team, you know? Yeah. [SPEAKER_01] I used to be an engineer and I got sort of, I don't know if I got forced to be a PM. [SPEAKER_01] Maybe I also perceived PM as being a little more high status when I joined Google, but then eventually you come around the other side, you're like, this is terrible. Yeah. Yeah. [SPEAKER_01] You never really get the satisfaction of actually shipping other than once a quarter when you ship. I mean, the PM skills of talking to users and trying to figure out what to do, what's the problem to solve. Those are very important still. Yeah. [SPEAKER_00] And you got to wear multiple hats too. You got to go do that thing yourself, go prototype it and get some feedback and then maybe bring an engineer along. [SPEAKER_01] How much do you think that everyone has to go as fast as, I mean, Gary was talking about stimulants and skipping sleep and Gary 10, G-Stack. [SPEAKER_01] Is that the default way that we all need to work or do you think there's a trade-off for thoughtfulness? I think it's very easy now with all these AI tools to go 10 different directions at once. [SPEAKER_01] Yeah. So sometimes you do have to slow down and try to figure out where you want to go. [SPEAKER_01] Yeah. But I also believe that the traditional process where you do annual planning and all this, I just feel that doesn't really work anymore. Yeah. Yeah. [SPEAKER_01] And for the record, I love Gary and I think the world of him. [SPEAKER_01] Yeah. So here's my view on that because I thought a bunch about that. [SPEAKER_01] You know who asked me this the other day? [SPEAKER_01] Hithin, who's really, really impressive. Hithin Chow. Yeah. We're talking about this sort of productivity porn and everybody's got 20 agents running and 20 monitors and blah, blah, blah. And I do think when it comes to fully realizing a local maxima, a sort of local maxima, you should go very fast, right? So let's say you kind of hill climb, you get to the bottom of the new local maxima. I think with agents, you should be able to get to the top of that hill extremely fast, right? You have a new insight, build everything around the insight so it's fully expressed. But then I think to get to the next hill, you've got to probably slow down and almost stop and go touch grass and do whatever. [SPEAKER_00] Yeah. So I think there's this combination of fast and slow. [SPEAKER_01] That's probably the future way. [SPEAKER_00] Yeah, I think so. [SPEAKER_00] And you've got to go that random walk, trying to find market fit, which takes a while, right? [SPEAKER_00] Yes. [SPEAKER_00] So it's not like, yeah. So we were talking before we started recording about some of the business in a box platforms. Have you looked at them? [SPEAKER_00] Do you have a view? Slow down and almost stop and go touch grass and do whatever. [SPEAKER_00] Yeah. So I think there's this combination of fast and slow. That's probably the future way. Yeah, I think so. And you've got to go that random walk, trying to find a market fit, which takes a while, right? [SPEAKER_00] Yes. So it's not like, yeah. So we were talking before we started recording about some of the business in a box platforms. Have you looked at them? [SPEAKER_00] Do you have a view? I've looked at posts here that we talked about. I don't know if the guy intentionally made it the opposite of AI Slab or is it kind of a... [SPEAKER_00] Yes. I think so. Yes. Yes. Yes. Yes. Yes. [SPEAKER_00] That's funny. Well, I mean, I have a pretty big public presence, right? So I connect all my shit to it. And then, I mean, it definitely gives a good peek into what's possible. But right now, it's probably still pretty early stage. It's time to run Facebook ads. Yeah. Why am I running Facebook ads? Yeah. [SPEAKER_00] You know? So... I don't know. Yeah, yeah, yeah. I mean, I'm very excited about it because it does feel like it's a path for more people to build companies. Yeah. [SPEAKER_01] Even if they're single one-person companies. If you think about how competitive it is to build a billion-dollar business, the markets that support it, the number of people trying versus 100 million versus 10 million versus $100,000 TAM. [SPEAKER_00] Yeah. Maybe there are these pockets all over the country, all over the world where there are opportunities for $100,000 TAM products. Yeah. And that would change somebody's life. Now, that's not an enterprise venture-backed company, but that's okay. [SPEAKER_00] Yeah. So I hope that whole thesis works because I do think it's a way to get more people to participate. [SPEAKER_00] Yeah, that's my plan for my kids, dude. I want them to just build bootstrap businesses in high school. Yeah. And they can skip the whole college and corporate life. Yeah. Well, dude, I think for 10 years, there's this moral panic about the kids wanting to be YouTubers. [SPEAKER_00] Yeah. [SPEAKER_01] You're a YouTuber. You know, in the vein of Mr. Beast. And I think the pro case for that actually is that the kids wanted to be entrepreneurs or have agency. And the only channel for people if they weren't programmers was creating YouTube videos, at least online. Yeah. So if you're an online native generation, you want to create something, you're not a programmer, you make a YouTube show. And now you can make a lot more than that. [SPEAKER_01] Yeah. You can build whatever you want. Yeah. [SPEAKER_00] Exactly. So. Exactly. It'd be very exciting. Yeah. Any other hot takes for us? I'm curious about your thoughts about this actually. So I feel a lot of people are saying agents will interact with your product first, right? And then you see all these great companies building APIs and MCPs. But how do you think about, you know, you've been a consumer for a while. So the consumer is, you gotta get the user to come back and use your product, right? [SPEAKER_00] Yeah. [SPEAKER_00] But now the user is, hey, go send the agent to use it. So how do you think about retention and all this basic stuff? How do you, or even brand equity? Because the agent is just putting some API. Yeah. Yeah. Yeah. How do you? Yeah. I don't. Okay. So I think one of—I don't know is the truth. But there's, I have a few thoughts. So one, I think that a lot of the sophistication that happened in consumer happened because we had to have indirect monetization. Okay. We're never charging consumers directly for these products, which is why you got ads and stuff. Ads and just large scale networks. And we all obsessed with retention and engagement and whales and all of these things really mattered because we didn't simply charge people for products. So I think one big thing that's actually really helped in the AI era with that is that consumers are now excited to try new things. They're willing to pay. They're willing to pay a really high price point. There's also consumption revenue and consumer for the first time. Token and stuff. Token. You have your subscription plus your token. So, and then the actual, the blessing in disguise is that there are real costs as well. You have these inference costs. So you're, wow, we have to charge our customer on day one. So I think one thing is that the business model simplification, I think will really help with a lot of what you're describing. Two, I think that a lot of the products will have a sort of, it'll have an API interface for your agents to interact with or for, transactional sort of rote things. [SPEAKER_00] Yeah. And then it'll have a consumption based interface as well. So you can also imagine a mobile app or there's the feed, but then you can turn it over to where the wires are and you can just ask for things to get done or you can just see the log of the things that got done. [SPEAKER_00] Yeah. I mean, people will do both. Right. [SPEAKER_01] Two, I think that a lot of the products will have an API interface for your agents to interact with or for transactional rote things. [SPEAKER_01] Yeah. And then it'll have a consumption based interface as well. [SPEAKER_01] So you can also imagine a mobile app or there's the feed, but then you can turn it over to where the wires are and you can just ask for things to get done or you can just see the log of the things that got done. [SPEAKER_01] Yeah. I mean, people will do both. [SPEAKER_00] Right. I mean, you can imagine Credit Karma where we worked, you know, once in a while, you want to just take a look at your score history and a few other maybe credit card offers. I don't know. Yeah, yeah. If I get my score with all kinds of credit card offers, I'll definitely do that. Yeah. [SPEAKER_01] A hundred percent. [SPEAKER_01] Exactly. Yeah. On the other hand, sometimes you want to just be like, yo, can you just fix all my stuff or what stuff did you fix this week? How much money did I save? [SPEAKER_01] Got it. [SPEAKER_01] Got it. Yeah. It's definitely interesting. Yeah. But look, I also just think the whole agent stack is emerging. [SPEAKER_01] Yeah. [SPEAKER_00] Identity, payments, marketing. We don't even know, even CLI versus MCP. Like all of these are really new things. And I think a lot of the old playbook goes away. Yeah. It's a whole new world. [SPEAKER_00] And in 2025, I thought agents was overhyped, but now I think it's really coming. Me too. [SPEAKER_00] I know. [SPEAKER_01] It's just the word is frustrating because it gets so overloaded. [SPEAKER_01] Yeah. There's workflows, all this stuff. [SPEAKER_00] Totally. [SPEAKER_00] I've been trying to just say model in a loop. Yeah, exactly. Model that use tools in a loop. [SPEAKER_00] That's the best definition. [SPEAKER_00] Yeah. [SPEAKER_01] Yeah. But nobody likes to hear that. Agents is much flashier. Yeah. [SPEAKER_00] It's flashier. [SPEAKER_00] Yeah. My hope is that all this stuff makes human work more fun. [SPEAKER_00] Like our job is more fun. [SPEAKER_00] Yeah. Dude, I don't think we're all going to lose our jobs. Like I really think, and we see this in a lot of companies, so we look at a ton of companies and we've seen two different buckets. One bucket is, hey, we dramatically increase productivity for a person or a team. We see this in recruiting, but we couldn't do 100% of the job. [SPEAKER_01] So we could do the phone screen, but we couldn't obviously show the candidate around the office. Or we could do the phone screen and we could answer all the questions about the company and we could even do the comp negotiation, but we couldn't do the onboarding. Yeah. The other style of company, which we see, which is maybe a Decagon, right? [SPEAKER_01] Or a happy robot is, hey, we did 100% of a job like customer support. [SPEAKER_01] You know, the customer called in, they had a question. [SPEAKER_01] We hopefully resolve their query and then that's it. [SPEAKER_01] And that is 100% automated. I'd say that that second group where you have 100% automation of a job function is really rare. Almost every AI product, AI native X or Y we see is able to provide dramatic lift, but it's not able to do 100%. [SPEAKER_01] So the last 10% is still these humans. Yeah, it's still, today anyway, it's still humans that do that stuff. [SPEAKER_01] And it's interesting too, because the buyer looks at that as software, as expensive software. [SPEAKER_01] Whereas in the case of something like a happy robot, Ducky Gun Sierra, they look at it as cheap labor. [SPEAKER_01] So I do think there's a different buyer mindset, but because there's been this difficulty of getting to 100% automation, I think a lot of the efficiency gain shows up in just a different way. Probably not less jobs. Maybe we get the European style four day work week. Maybe companies get twice as productive. I have no idea. Yeah, but you don't think that I feel like there's going to be a transition from these 10,000 plus people companies laying a lot of people off to hopefully more smaller companies like solopreneurs and stuff like that. [SPEAKER_00] I think, yes, I think the shape of the economy is going to change, the amount of concentration. [SPEAKER_01] But I just don't think there's going to be less jobs. [SPEAKER_01] I think human ambition has no ceiling. [SPEAKER_01] That's true. [SPEAKER_01] Human desire has no ceiling. [SPEAKER_01] And just read any mildly interesting science fiction book. [SPEAKER_01] There's no way this is the peak expression of all the stuff that we want and we need and we're going to convince ourselves. [SPEAKER_01] And all the new things that you read about every day, these luxuries, peptides. [SPEAKER_01] And everybody is going to have all of that stuff and want even more. Dude, I saw a really good tweet about this. Someone tweeted that the job market is so bad that I can only pursue my dreams now or something like that. So it's that, yeah. Maybe you lost your job, but now you can actually do your own thing. [SPEAKER_00] Yeah, 100% and have a shot at actually achieving it. Yeah. [SPEAKER_00] Yeah. Yeah. Cool. [SPEAKER_00] Well, awesome, man. Maybe that's a good positive note to end on. [SPEAKER_01] Yeah. That's a good note. [SPEAKER_00] Yeah. [SPEAKER_00] Cool. [SPEAKER_00] Good thing you do. [SPEAKER_00] Thanks, Peter. Yeah. [SPEAKER_00] Thanks for having me. Bye. Yeah, that's my plan for my kids, dude. Like I want them to just build like bootstrap businesses in high school. Yeah. And they can skip the whole college and like corporate life. Yeah. Well, dude, I think this is like, you know, for 10 years, there's this moral panic about the kids want to be YouTubers. Yeah. You know, you're a YouTuber. Yeah. Yeah. You know, in the vein of Mr. Beast. And I think the like pro case for that actually is that the kids wanted to be entrepreneurs or have agency. And the only channel for people if they weren't programmers was creating YouTube videos, at least online. Yeah. So if you're like an online native generation, you want to create something, you're not a programmer, you make a YouTube show. And now you can make a lot more than that. Yeah. You can build whatever you want. Yeah. Exactly. So. Exactly. It'd be very exciting. Yeah. Any other hot takes for us? I'm curious about your thoughts about this actually. Like, so I feel like a lot of people are saying like agents will interact with your product first, right? And then you see all these great companies like building like APIs and MCPs. But like, how do you think about like, you know, you've been a consumer for a while. So like the consumer is like, you gotta get the user to come back and use your product, right? Yeah. But now like the user is like, hey, go send the agent to use it. So how do you think about retention and all this like basic stuff? Like how do you, you know, or even like brand equity? Because the agent is just like putting some API. Yeah. Yeah. Yeah. How do you? Yeah. I don't. Okay. So I think one of, I don't know is the, is the truth. But there's, I have a few thoughts. So one, I think that a lot of the sophistication, sophistication that happened in consumer happened because we had to have indirect monetization. Okay. Like we just, we're never charging consumers directly for these products, which is why you got ads and stuff. Ads and just large scale networks. And we all obsessed with retention and engagement and whales and all of these things really, really mattered because we didn't simply charge people for products. So I think one big thing that's actually really helped in the AI era with that is that consumers are now excited to try new things. They're willing to pay. They're willing to pay a really high price point. There's also consumption revenue and consumer for the first time. Like token and stuff. Like token. You have your subscription plus your token. So, and then the actual, like the sort of blessing in disguise is that there are real costs as well. You have these inference costs. So you're like, wow, we have to charge our customer on day one. So I think one thing is that like the business model simplification, I think will really help with a lot of what you're describing. Two, I think that a lot of the products will have a sort of, you know, it'll have an API interface for your agents to interact with or for, you know, for transactional sort of rote things. Yeah. And then it'll have like a consumption based interface as well. So you can also imagine like a mobile app or there's like the feed, but then you can kind of turn it over to where the wires are and you can just ask for things to get done or you can just see the log of the things that got done. Yeah. I mean, people will do both. Right. I mean, you can imagine Credit Karma where we worked, you know, like once in a while, you want to just take a look at your score history and a few other maybe credit card offers. I don't know. I mean, yeah, yeah. If I get my score with all kinds of credit card offers, I'll definitely do that. Yeah. A hundred percent. Exactly. Yeah. On the other hand, like sometimes you want to just be like, yo, like, can you just fix all my stuff or like, what stuff did you fix this week? How much money did I save? You know? Got it. Got it. Yeah. It's definitely interesting. Yeah. But look, I also just think the whole agent stack is emerging. Yeah. Identity, payments, marketing. We don't even, even CLI versus MCP. Like all of these are really new things. And I think a lot of the old playbook goes away. Yeah. It's a whole new world. And like in 2025, I thought agents was overhyped, but now I think it's really kind of coming. Me too. I know. It's just the word is frustrating because it gets so overloaded. Yeah. There's like workflows, like all this kind of shit. Totally. I've been trying to just say like, can we just say like model in a loop? Yeah, exactly. Model that use tools in a loop. That's the best definition. Yeah. Yeah. But nobody likes to hear that. Agents is much flashier. Yeah. It's flashier. Yeah. My hope is that all this stuff, like a lot of people think like we're going to lose our jobs. It was probably what would happen at some point. But like, I hope all this stuff just makes like human work more fun. Like our job is more fun. Yeah. Dude, I don't think we're all going to lose our jobs. Like I really think, and we see this in a lot of companies, you know, so we look at a ton of companies and we've seen two different buckets. So one bucket is, hey, we dramatically increase productivity for a person or a team. We see this in like recruiting, but we couldn't do a hundred percent of the job. So we could do the phone screen, but we couldn't obviously, you know, show the candidate around the office. Or we could do the phone screen and we could answer all the questions about the company and we could even do the like comp negotiation, but we couldn't do the onboarding. Yeah. The other style of company, which we see, which is maybe a Decagon, right? Or a happy robot is, hey, we did a hundred percent of a job like customer support. You know, the customer called in, they had a question. We hopefully resolve their query and then that's it. And that is 100% automated. I'd say that that second group where you have 100% automation of a job function is really rare. Almost every AI product, AI native X or Y we see is able to provide dramatic lift, but it's not able to do a hundred percent. So the last 10% is still these humans too. Yeah, it's still, today anyway, it's still humans that do that stuff. And it's interesting too, because the buyer looks at that as software, as expensive software. Whereas in the case of something like a happy robot, Ducky Gun Sierra, they look at it as like cheap labor. So I do think there's a different buyer mindset, but because there's been this difficulty of getting to a hundred percent automation, I think a lot of the efficiency gain shows up in just a different way. Probably not less jobs. Maybe we get like the European style four day work week. Maybe companies get like twice as productive. I have no idea. Yeah, but you don't think that like, I feel like there's going to be a transition from like these like 10,000 plus people companies, laying a lot of people off to hopefully like more smaller companies like solopreneurs and stuff like that. I think, yes, I think that the sort of shape of the economy is going to change, like the amount of concentration. But I just don't think there's going to be less jobs. I think human ambition has no ceiling. That's true. You know, human desire has no ceiling. And just read any mildly interesting science fiction book. Like there's no way this is the peak expression of all the stuff that we want and we need and we're going to convince ourselves. And, you know, all the new things that you read about every day is these luxuries, peptides. And, you know, everybody is going to have all of that stuff and want even more. You know, dude, I saw a really good tweet about this. Like someone tweeted that like the job market is so bad that I can only pursue my dreams now or something like that. So like, it's like, you know, it's like, yeah. So maybe you lost your job, but like now you can actually do your own thing. Yeah, 100% and have a shot at actually achieving it. Yeah. Yeah. Yeah. Cool. Well, awesome, man. Maybe that's a good positive note to end on. Yeah. That's a good note. Yeah. Cool. Good thing you do. Thanks, Peter. Yeah. Thanks for having me. Bye.