0:00 Okay, this is a pretty special episode because we're celebrating one million subscribers on the AIU Code, uh, AI, AI Engineer YouTube channel. So go over there and subscribe if you haven't heard about it already. 0:09 The guest is Mihail Eric, who is not only teaching the Modern Software Developer at Stanford, which is going viral right now for basically covering all the AI-native development pract- best practices, uh, but also he is head of AI at Monaco. 0:22 So he's not only a teacher, but he's a practitioner. Um, and those of you coming to AIU Code Summit next week, um, it's gonna be the celebration of all the extreme progress that you've seen in coding in the last year. 0:33 I'm excited about all the themes of fighting slop, of system one models, and of increasing your ambition for major rewrites of basically everything that you've seen in code. 0:43 Um, there's never been a more exciting time to be an AI engineer, and with one million AI engineers, I think it's just getting started. So let's get into the interview. 0:51 Okay, we're in the studio with Mihail Eric, Adjunct Professor at Stanford and Head of AI at Monaco. Welcome. Thank you. I'm glad to be here. 0:58 You caught our attention because of your course that you are refreshing for the third time, third year? Second time. Second time? Yep. The CS 146, or 146S, the Modern Software Developer at Stanford. 1:09 This particular generation of junior developers, of junior engineers, the people that are now entering the workforce, will, I think, be the first kind of generation of that new shift. 1:18 And I looked at the syllabus, and I was like, "This is everything we cover." So every now and then, I like to invite people like yourself who, like, you cover a broad range of topics- Mm-hmm... 1:26 and then we can sort of go topic by topic and give people the state of the art- Yeah... on each of these things, which I, I assume you're doing with all your guests anyway. For sure. Tell us the history of the course. 1:36 What made you wanna spring out CS 146? Yeah. So this is the second time it's being taught, and I taught it for the first time last fall, so fall 2025. This was pre, like, Opus 4.5, 4.6 days. 1:49 But the-- Even before fall 2025, the course had been about, like, a year in the making. 1:52 The genesis of the class was like November of 2024, and I had sort of had this idea that we're seeing all this stuff happening in coding agents, and yes, they were an earlier version of the coding agents, but already it felt like there was something drastically changing in terms of how developers were doing their work, how quickly we were getting work done, uh, all that fun stuff. 2:08 And so at the time, I just-- It was all just scattered blog posts, scattered, like, things you could find online if you were trying to really get into the, the methodology of how to do these things. 2:18 And so I, I had this idea like, "Hey, well, is there a way to kind of structure a curriculum around this that's a little bit more formal?" Had the idea. 2:24 And you know, I, my alma mater is Stanford, and I always had this dream to go back and teach something and, and kind of pay it forward. Uh, I've always been very deep into education. 2:31 I talked to Chip, who also taught a class, uh, one of my good friends. Chip Hieatt. Yeah. Yeah, Chip Hieatt, exactly. Uh, spoke to her. 2:37 Who-- She had taught her, I guess, designing machine learning systems, and so, like, asked her what that process was like, and sort of, you know, took a lot of inspiration for how she-- from how she designed the curriculum and, and thought about, you know, getting something through the, through the school. 2:48 Put it together, spent about nine to ten months. This was now fall of 2024 through to, like, you know, mid to late of 2025. Got it through the committee, uh, made a case for the class to exist. 2:59 You know, I had to put together a proposal and everything. Uh, and that took me to fall 2025, when I taught it for the first time, which was very well-received. You know, students were really interested in it. 3:08 Uh, there was even, I guess, like, a Business Insider article that was written about the class last year. And then, yeah, this year, uh, a lot of things changed. 3:17 I would say really coding agents, autonomous coding agents really have picked up- [laughs]... in a way that, that we had not seen before. 3:23 Um, so it felt like it was time to do this again, uh, refresh things, uh, and here we are. M-Meanwhile, you're also head of AI at Monaco- Yeah... which is also a very buzzy startup. Yes, yes. 3:35 We are, we are, we are, we're in the airwaves right now and, and, uh, having a fun time doing it. Yeah. Yeah. Uh, so, like, I, I do like that you're also, I guess, a practitioner. 3:42 You guys are obviously implementing all this stuff at Monaco. Yeah. Any, I guess, principles or observations about just what you have changed from last year to this year at a high level? Yeah, yeah. I mean, a ton. 3:55 I think that when I reflect on what last year's version of the class was, uh, I mean, coding agents had been around, but we were still in kind of that phase where companies were emerging that were tackling every single part of, like, the ISDLC. 4:08 There was, like, a coding agent for code review. There was a coding agent for, you know, spec design. There was a coding agent for, uh, SRE. And, you know, some of those-- Of course, some of those companies still exist. 4:16 But it felt like last year it was a lot more about a broad tour of here's everywhere you can use coding agents. 4:23 And in a way, a lot of the curriculum last year was, was geared toward exposure and this, this kind of survey, this kind of like odyssey of sorts through, hey, everything that you knew a-about the ISDLC is now changing because of AI. 4:35 Um, this year I think there's some principles that are solidifying, right? The idea of, like, very careful context engineering. 4:40 The idea of, like, really thinking through sort of like design and architecture as a way of handling really complex tasks. And so the nature of my class is it's not intended for the, like, beginner programmer. 4:51 It is not a, it is not a replacement for, like, traditional computer science in the sense of, like, algorithms and if-- In fact, a lot of those classes are prereqs to what I'm, what I'm doing here. 4:59 But I do know that you-- if you have a good foundation, going now toward applying that and, and, and sort of using those principles in, in production environments in, like, real world, that's something that I think is, is now at a place where... 5:10 You know, at Monaco, we deal with ma-massive code bases. You know, I help a lot of companies also with thinking through their massive code base, uh, designs and things like that. 5:17 And so really, it's more of a practitioner's course for once the students have graduated or are now entering the workforce, how to be the most productive members of, of the workforce. Yeah. 5:26 You said you're passionate about education. I feel like students in that phase of being in college are in a pretty interesting spot. Any advice and any takes? Should people study computer science right now? 5:40 Is cloud code and Codex all we need? Like, I, I wanna, you know, hear your thoughts. Yeah. I, uh, you know, there's a lot of anxiety from students, as you can probably imagine. 5:50 And, uh, and I, I think a lot of what I'm trying to communicate this time around is that the anxiety, a lot of it still is here, and there was anxiety last year as well. 5:58 But I wanna make the case that A well-trained AI native engineer, and I know that's like a buzzy word- [laughs]... 6:05 but let's say that we'll, we'll, we'll, we'll, we'll dive into maybe what aspects of that, that, that are most interesting. 6:10 A, like, well-trained and native engineer can be some of the most, you know, the most useful assets like any company around. And so, 6:17 and, and, you know, I-- this is the first week of class actually, and like week one, and you know, first lecture I'm like: Hey, I really believe that every company needs you, but the floor is much higher, right? 6:27 So you're entering as like your first job out of school, but it's not sufficient to just be like a, you know, an L3 in like a Google term if you're gonna use like that, that kind of leveling system there. 6:35 It's like you have to generally be better than, you know, almost at the senior level going into this. So like what, what can we-- how do we get there more quickly, right? 6:42 And so, uh, I'm trying to say that there is still a core bit of like foundational software engineering that is important to have some of these core values of, of sort of how software systems are built and the trade-offs of those systems. 6:54 But then beyond that, we go and, uh, you know, that's where I think you as the humans should still be valuable rather than... And the rest of the implementation details you can kind of defer more often. Hmm. Yeah. 7:04 Uh, this is something I've said before, which, uh, is, is almost like the controvers-- I, I-- it's, it's like the kind of thing you say to be spicy, but I almost, like I, I do genuinely kind of believe it, which is most students are cooked, some of them are [chuckles] gonna be super valuable. 7:17 I, I don't think it's a popular thing. This is a delicate topic. Stanford's, Stanford s- Stanford's kids are gonna be fine, right? They're from Stanford. Sure. But like what happens to like the rest of America? 7:26 [laughs] Right. Right. That, that, that is an open question. It's a weird situation- Yeah... 'cause I feel like a lot of education is slow to move. Like- Mm-hmm... I'm even curious in how, 7:35 how do you determine where do you put the end date to curriculum in a course on AI? Like- Oh my God... you're teaching a Stanford modern software- Ah... like that's changing every year. You cheat. 7:44 You just get the people from the co- uh, the coding agents to come and speak, right? [laughs] Oh, yeah. Like yeah, have them be the... Yeah. I-- that, that's a funny- The old guest lecturers. It's a funny... Well, yeah. 7:52 I mean, you know, that was, uh... Okay, so, so on, on the guest lecturers point, I think that was, um... I mean, that was also last year, um, that was also the case. 7:59 So, so that was by design because I-- look, I, I came in, right, and I'm saying, "I'm a practitioner," a- as, as you pointed out, Swyx. Like I, I'm a practitioner, I do this thing on my, on my, you know, my day-to-day. 8:09 I design these systems. You know, we build, we build these things at Monaco and have seen that many times before. And so 8:15 that, that being said, I don't wanna claim that I have all the answers today, and that would be, that would be like very, uh, you know, be a lot of hubris to, to make that claim. So 8:24 the structure of the class was always, here's one practitioner's sort of aggregated set of insights from working in the field, from like talking to a lot of people in the field. 8:33 But then there's people that are like literally on the frontier, you know, seeing things even beyond that, and invite them to sort of put their p- their, their proposals forward, their view of the wor- with the, the world forward, and that's why the guest lectures are there. 8:45 No, I think it's really impactful for students to see big names, uh, people actually doing it, like hear from them directly, be able to like ask them. Of course, you know, they wanna intern, they wanna get jobs. 8:56 It's, it's good to have guest lecturers. Yeah. No, I think what he's saying... So okay, I'll, I'll sort of rephrase it. Okay. Uh-huh. Uh, no. So I-- so okay, there, you know, we've also interviewed Ange- Mm-hmm... 9:05 uh, who teaches the Coachella course, right? [laughs] Okay. Right? Which is like, here's all the names. Like it's only the names, right? 9:11 Whereas like he's actually also producing, you know, here's cloud code and 200 lines of code, right? And which is- Yes... no one's telling you to do that. Right. No one's, no one's gonna come to you to do, do that. 9:19 But like, yeah, that when you wanna teach fundamentals in a framework that you can use and working understanding- That's right... this is the kind of thing you do. That's right. That's right. Yeah, yeah, exactly. 9:27 And, and I-- but I do think that there's a balance for, for this kind of material, to, to your point, is that there is a... You can't-- it's not 100% solidified. I mean, by, by design. 9:36 Uh, and by design, it's a feature and maybe a bug of some, some kind, but a, a big portion of the curriculum did change from last year, right? 9:42 And so the, the- there's like a dynamism to the whole development of the curriculum. 9:45 But I'm trying to at least put forward that there are some things that I believe are kind of sticking around, and yes, prompting coding agents, they've gotten smarter. 9:51 What does that mean in terms of the new ways of dealing with them? But then still, Boris is also gonna know something that I don't know, right? 9:57 [laughs] So let's, let's have, let's have him also come to the table and, and share his learnings. Um, one of the fun things I noticed this time around is a good bit of the portion is on open source contributions. Yeah. 10:09 Uh, what's the thinking behind that? You have people- Yeah. Yeah. This is, um, this is one of the things that I'm honestly kind of most proud of and, and most excited about. 10:18 Uh, I mean, this is about to be week two of the class. Yeah. 10:21 So this is, this is the very gracious open source partners that, uh, I spent the greater part of this summer just kind of talking to and, and getting them kind of excited about and, and, and interested in, in sort of this offering. 10:32 Basically, the idea was... So last year when I taught the class, we had like nine assignments, and then we had like a capstone open-ended project, right? 10:39 The nine assignments were basically me sort of like coming up with these pedagogical exercises, like, "Oh, here's a repo, like use an AI code review tool to, to, you know, post a code review and then see what-- why is it good, why is it not good." 10:49 So relatively pedagogical. Honestly, pretty easy. I think for a lot of folks, they could've just like one-shot it, and I thought, "Okay, this is not that helpful." At this point, agents are around. 10:58 I mean, they, they have every student that I surveyed, you know, on Tuesday, not a single person really types by, by hand anymore. They're like- [laughs]... everyone's using coding agents. There's no... 11:07 There's-- and this is, you know, like all sophomores, juniors at, at Stanford. Do you, do you survey them what's the market share? Was that like what, what tools they use? What tools, yeah. Okay. 11:15 I haven't yet, but that's a good call. You should, you should do like a- It's fun. Yeah, yeah. Yeah. I, I will do that. You could be spicy. Stanford prefers [laughs] No, no. 11:21 Well, what I think through like the, the university, I think they have some... The edu-- like most of the, the, the providers have like an educational program, right? So- I mean, yeah. They're, they're very generous on 11:31 plans for students. Right. Right. So I think by default, even before the class, there's like, you know, a Claude license. I, I imagine Cursor license. There's like probably a Cognition license as well. 11:41 I don't know, you might know them better than I do. 11:42 But what I was able to also just accrue on top of that was, hey, like if I don't want people to be bottlenecked by like the $20 a month subscription, like I want you guys to go, if you wanna like token max something just to see what's possible, then, then you should. 11:54 And so, and I'll come back to the, the partners, but the, the additional thing that I was able to, to accrue through start talking to folks was like additional credits for students. Nice. 12:00 So like Factory is also getting involved. The, the Cognition folks are getting involved. OpenAI as well. 12:04 But anyway, so, so, so for the, the open source partners, I, I sort of like was like, okay, one of the things I'm most concerned about when it comes to educate the next kind of generation of, of students is that if everyone is saying, kind of to Swyx's point, like they're cooked, like, you know, what's gonna happen to them? 12:19 There, there, there's almost like this sense of like them being left behind, and I think that mentorship is like a really important kind of piece of just upleveling everyone and making sure that there isn't this like massive gap that folds, that, that occurs between, you know, the senior people that have already been in the industry before coding agents and know what's up, and they adopt coding agents, and they become amazing. 12:37 And then the juniors that are just like, "Okay, now what do we do?" So In a sense, one of the, the kind of goals was, "Hey, coding agents are really good. 12:44 I wanna put you guys in situations where you're dealing with complex code bases, large code bases at scale. 12:49 We're gonna work with someone like what I think are the, among the best like AI-native repos, you know, that all do AI-focused things, or either coding agents themselves, or building tools in the service of agents, you know, in the modern world. 13:00 I want you guys to get involved with these projects. I want part of your assignments to be actually contributing to these projects. 13:05 In the course of being in all those projects, you'll actually be talking to the maintainers, they will be reviewing your code, they'll be helping you build mental models of, like, good software." 13:13 In that way, there's effectively this, like, passing on of, like, the guard of true good software design in a way that, you know, I don't think most students get the opportunity to do. 13:21 And then there's the benefit, of course, that, like, I'm sure some of them will get, you know, poached and picked up as part of, uh, doing well in those assignments. Yeah. 13:28 One of the interesting things, you know, uh, for the, for what it's worth, I'm, I'm gonna get a bunch of hate if I don't say this, is I think you say cooked and versus cracked for dramatic effect, but, you know, like, it's just nice to, to state extremes. 13:42 Um- Sure. Sure. Yeah. And so, so maybe we should go into the sort of detail of the curriculum- Yeah... or is there anything else you want to explore at a top level? I mean, like that, that's, uh, very high... Yeah. 13:51 All the, all that's there. Okay. I, I have high-level questions about students using tools, best practices. Are you pro it, like individual tutors? 14:02 Um, I'm assuming this will also be picked up by students, so, you know, before we dive straight into the course, just anything you have guidance on there for- Like, in terms of what, uh- Yeah. 14:12 You said you're passionate about education, right? Like- Yeah, yeah... there's also a stigma of one side is how, how do teachers, professors use it- Yeah, yeah... and then how can students best use these tools- Mm... 14:21 that's not just throw your assignments in, it can kind of one-shot it. How do you still learn, like, if you have anything there? Yeah, yeah. 14:29 And I think that there's-- I still have some assignments thrown in there, which I think are not intended to be the hardest things in the world, but are meant to be more about, like, concept, concepts, like more conceptual things. 14:38 Like, you know, what does a skill mean? It's sort of like the, maybe more of the, the, the tactical of some of these tools, skills like agent, you know, agent-ready code bases like AgentsMD, things like this. 14:47 So there's a, a certain amount of just, "Hey, if you haven't heard these terms, kind of get up to speed on them." 14:52 I do think most of the learning, though, in terms of the kind of putting the tools to the test, will happen both in, in the OSS and then also in the, like the final project, where it's, like, very open-ended and they'll have to just kind of explore something interesting. 15:03 Cool. Okay. Yeah. So, so maybe, uh, uh, yeah, I just wanna top level, let's go through one, two, three, four, five, uh, to ten. Yeah. Uh, and then we'll dive in deeper. So like breadth-first, basically. Sure, sure. Yeah. 15:15 So we start with internals of coding agents. That's right. You, you teach people how to write their own Claude clone. It's very helpful that it's now open source. [laughs] And then? Yeah, yeah. 15:24 So, uh, week one, it's all, uh, internals of harnesses, you know, kind of build a coding agent from scratch. 15:30 Uh, week two is more, okay, now assume that there's, uh, how do we start thinking about, like, context engineering at scale? Uh, so beyond just vanilla prompting. 15:38 Uh, and then I think coupled with that is just, um, in, in doing that, thinking about, like, tool design, MCP, you know, people... [chuckles] I think for a while MCP was getting a lot of hate. I'm still pro MCP. 15:47 I see it all the time, and so I think that there's, at this point, it's a standard that's existed, and I think it will continue to exist. So, you know, talking about that, effective tool design, uh, that's week two. 15:56 Uh, week three I would say is more about just like, uh, yeah, skills and, like, what does it mean to, like, extend beyond simple, single-purpose tools to more, like, procedures, and that's where we'll talk about skills and, you know, that kind of starts leading into, like, the automation of agents. 16:09 Um, week four is, is a lot of just more of, like, the foundational pieces of how can you customize an agent, and this is things like agent config files, a la AgentsDB, Claude MD, hooks, um, you know, kind of sub-agents. 16:23 Like, there's a lot of mechanisms that exist in modern harnesses, uh, which you can use for different purposes. So I wanna expose students to that. Agent-ready code bases, this is, uh, probably one of my favorite topics. 16:33 It's like-- A- and this is especially relevant for folks that are going into, like, industry and thinking about code at scale is, you know, essentially what does it mean to have a repo that's ready to be ingested, consumed effectively by an agent? 16:45 And so thinking a lot about, you know, the trade-offs of designing of kind of surface level boundaries, but also docs and, and verifiability and things of that nature. Um, so that's week five. 16:55 Uh, yeah, week six, uh, sort of agentic code review. 16:58 Um, and here I think what is interesting is thinking also kind of extending from week five, week five around code base, like agent-ready code bases, going to week six is more about like, okay, in order to... 17:09 Like, code review is now becoming a bottleneck, which I don't know that students, unless they've, you know, have done a lot of... 17:14 Some of them are master's students, so they've been in the industry, but a lot of them have not seen this. 17:18 But for most people that are using code agents today, we're all experiencing something of a new bottleneck in code review. So how do we unblock that or sort of unbottle the, the bottleneck a little bit? 17:27 Um, what mechanisms are there there? And so verifiability and, you know, things of that nature and how we can extend that into the code review process. Week seven, uh, security. 17:35 I mean, obviously, I think there was a point-- So this was in last year's class as well, and this, this I wanted to reinforce. 17:41 I mean, with any autonomous system taking a lot of actions, we're seeing a lot of, this year especially, a lot of things getting hacked. 17:48 Um, you know, not least of which are, are just some of the, the core pieces of software from, from Vercel to, you know, uh, gateways, LM gateways entirely. 17:56 So I wanna spend some time just, like, talking about what new attack vectors exist, exist in security and, and, you know, how we can mitigate some of them. [lip smack] Uh, week eight is background agents. 18:08 So this is, uh, what I wanna start touching on here is, you know, how do you, like, 18:14 basically run agents in many different, like, product service areas, uh, not just in what was traditionally an IDE, but also having things that are, like, based on triggers in your code base, in, in your repo at, you know, during CI/CD, uh, things that get triggered even when, you know, based on events, things like, you know, SRE-style bots. 18:31 Um, how do you even build a system like a... Uh, I think this past year was when a lot of companies started releasing their posts like Ramp Inspect. 18:39 Um, there was Spotify has their own version of this, you know, I think Stripe Minions. So all these folks were doing really cool write-ups of, like, their internal infra they were building to do this. 18:46 I wanna touch on that, uh, get people exposed to what that is. 18:51 Yeah, this is, um, yeah, this one is, uh, more on, like, what happens when you're in, in a team in, in the modern world, and sort of like the core pieces of, of team-level infrastructure. 19:01 So everything from, uh, LM gateways to, you know, MCP portals, which is like MCP at scale, and then just also how to think about the evolving roles, I would say. So this is maybe a little bit like a soft thing. 19:12 It's like what, what, what is now... 19:14 You know, and the, and the v- the vanilla maybe, like, not super nuanced thing to say is, like, everyone's a builder, but I wanna, like, go a little bit deeper beyond that and, like, sort of like what thing is actually getting subsumed? 19:22 What, what roles are actually, um, either evolving or merging that didn't once exist? Um, what should students think about in terms of defensible skill set that's maybe wasn't true before? 19:31 Um, and so a lot more, like, team dynamic type, type considerations. 19:35 And then, yeah, this is, uh, it all kind of builds up toward the, toward the factory, which I think is top of mind for now, um, for a lot of folks today. 19:43 Once we're, we're building factories, where do humans still exist and all that? What is the, you know, where are the new choke points that are, that are intentionally there around, around human involvement? 19:52 Things like intent, things like, uh, setting up a verifiable systems, and what happens in the limit when maybe 89% of code is fully driven by just agents, uh, where we g- we kind of filled into all that. So, yeah. 20:05 Nice solid 10 weeks. [laughs] Uh, I was gonna say, like, you know, some of these are more established patterns than others, and I, I think people need to know that, like, not all of this is settled science. Sure. Yeah. 20:19 Uh, you know, but it is... Everything-- I'll endorse everything as, like, things that people talk about and you should know. You just-- we know the open problems. We may not know the answers. That's absolutely right. 20:28 Yeah. Yeah. Yeah. I mean, think even, um, I mean, you probably have your own version of this. I'd be curious to hear what you think is, is still not as defensible, but, or not defensible, but not as, uh, set in stone. 20:36 But, you know, even, like, the, the software factory. Are we talking about, like, dark factories? Are we talking about these sort of half lights on factories? Like, I think that even there's a spectrum there. 20:44 Yeah, and there, you know, there's the people that are on the, on the very frontier, like, like the warps and the factories [laughs] that are, that are, like, building factories and selling them as services and, you know, even they will only claim, like, 30 some percent, if even that, of, of, like, fully, you know, agentically driven contributions. 20:58 So- Yeah... what does that mean in the limit? It's, it is hard to say. A- and there's people who went very, very far dark factory and then came out saying, "I was wrong, guys." Yeah. Right? 21:06 So, like, uh, we covered that at, on AIE. Mm-hmm. Um, just be aware that, you know, not everyone... They, they probably believe the things they're saying at the time- Yeah... 21:16 but they have the right to change their mind, and they do. And even Boris recently tweeted that, like, "Yeah, I am often wrong." Yeah. 21:22 Like, and, like, everyone is like, because he created Cloud Code, is, like, treating his word as gospel. Totally. And I'm like, "No, well, sometimes he disagrees with himself." Yeah, yeah. [laughs] And, and that's okay. 21:31 I mean, that, like, this isn't politics, right? This is... We're, we're still scientists. We're engineers. We're, we're subject to... I, I don't claim to be dogmatic, but, you know, I just want people to experiment. 21:40 I think that's, like, a principle as well. You asked what principles earlier. Just, like, experimentation. I mean, like, you just have to be ready to... 21:45 A- and oftentimes when I, when I sort of, like, give sort of these lectures or, or talk to people about this stuff, they're like, "How do I... There's so much going on. How do I stay on top of this?" 21:52 And it's like, you know, there's something as simple as just, like, experimenting for 10 minutes a day. Like, just spending some 10 minutes doing something that you didn't do before. 21:57 If that's a traditional way that you were solving a problem, there's now a new tool that you wanna try or a new, like, framework or something, and just spending, maybe even doing the exact same problem just with that new kind of thing is, is, is, alone will cover most of the gap. 22:09 Yeah. Uh, I will call out, so some of your, uh, for example, you have cognition on the week that you're talking about code review. Yeah. Uh, I actually did work with Silas on Devin review. Yeah, yeah. 22:19 Pretty good, I would say. Like, you know, it's hard to benchmark these things. Sure. There is a review bench. No one uses it. Um, but, like, qualitatively, Devin review is pretty good. Well, we use you guys. 22:29 That's also- Oh, really? I mean, like, yeah. So, so Monica does use Devin. Oh. Uh, and that's, that's, uh... I mean, and Si- Silas actually spoke last year, and so S- I, I mean, Silas is great. Just super, super smart. 22:38 But, um, yeah, we, we use you guys, and so I was like, okay, among, among the providers of, of kind of code review tools that we... And we sampled a lot of them. Um, we actually did kind of converge on Devin for- Yeah... 22:46 a lot of our use cases. Silas is actually more on the model side these days, and then actually it's Walden that you want. Yeah, yeah, yeah. Um- I was talking to, yeah, I was talking to Walden as well. Yeah, yeah. 22:53 I will also shout out, you know, from Factory because- Yeah, yeah... uh, you know, talking about agent readiness, they've been all about that. They build signals, tooling on that. You, do you guys use that? Yeah, yeah. 23:02 Any opinions? Yeah. So, so we do use agent readiness of our own, of our own kind. I mean, Eno is, is awesome as well. 23:08 Like, I, I think they were super early on this, on this train and I think when he codified that in his talk, I guess at AIE, what was it like last... I forget when it was. But one of his talk- Europe. Yeah. 23:16 Yeah, Europe or, um, just fantastic talk, and I think it was kind of, in a way, ahead of its time, and so I was like, there's only one person that I could see being, like, the most- [laughs]... 23:23 the most appropriate person to talk about this, where I, like, really wanna give him the spotlight, and they're super excited. Oh, yeah. It was funny. 23:28 I, I was at a call, we were talking about this stuff and, like, we were like, "Okay, we need to form some opinions about it." And then, like, Factory, like, shipped- Yeah... 23:34 a full product on it that week that we were talking about it. Yeah. And I was like, "Okay, you, you won." Yeah. Like, obviously. Yeah. I mean, like, everyone kind of agrees on the same things. Sure, sure, sure. 23:41 And it's just a question of, like, do you want to turn it into a dashboard? Yeah, yeah, yeah. Yeah, yeah. Because it's a dashboard. Right, right, right, right, right- [laughs]... right, right. 23:48 But you know, there's, there's value and, like, I think they did well from a marketing perspective jumping on, like, the, the readiness report, and I've definitely spoken to... 23:54 You know, I, I will regularly kind of pitch that to other people and, and say, like, "Yeah, you should run readiness reports of your own kind." 23:58 And, um, you know, of course, now people have open source versions of it, but, but I think the, the principles are super important. Yeah. Uh, do you wanna go piece by piece or, or do you have any other things? 24:08 I, I can go piece by piece. Mm-hmm. Yeah. Yeah. Okay. So the first one, right, uh, you focus on Cloud Code, but then you also have Py as an open source sponsor. Yeah. What's up? Okay. So, so, so, okay. 24:17 This one is maybe a little bit more marketing-y, but, but okay. Um, the one thing I do wanna highlight a- and I've made that, I also made that clear last year, but I'm not a, I'm not endorsing any tool, okay? 24:26 [laughs] Like, like I, I don't, I, I, I think that is, uh, the wrong way to approach something like this. 24:30 Um, there are tools that come and go and when, you know, transparently, when, when the c- class was taught last year- Claude Code was not as popular. It wasn't as like readily accepted. 24:40 Boris was only at four thousand Twitter followers, and now he's at, you know, whatever million. [chuckles] I don't even know if he is, but you know, like he's at a lot of followers. 24:47 Um, so, so Claude Code has really like e-emerged on the scene. Um, I think what- I think, yeah, I think we, we also underestimated Claude Code when we had Boris and Cat on the podcast. Right. Right. 24:54 It was like two months old. I hadn't really-- I, I used it once. Uh, a query cost me fifty cents. I was like, "Who's gonna use this shit?" Totally. Totally. It was still in beta for a while- Yeah. 25:02 But I think it's good on them for how they built it before the model was ready. Yes. Like you hear so much build for where the model will go. Totally. Super ambitious. Like models were not that good. 25:11 You couldn't feel it, and then, ah, we see, we see that it's where it came from. 25:14 That-- it's, yeah, it's funny that even-- 'cause when he gave his talk last year, I remember, um, you know, it was like a very high-level sort of set of slides. 25:20 But I remember one of the slides was literally like built for where the model is gonna be six months from now. And this was like October of last year, and 25:27 I can see why he would've made that point if, if he knew wh- in six months where like Opus was gonna be, 'cause that was also what I think was a huge unlock where I would argue was one of the reasons why the curriculum changed as much as it did, right? 25:37 Where there was suddenly this huge arms race around true like autonomy in these agents, like long-running agentic coding agent tasks. 25:43 But yeah, so, so, you know, back to your point, so it's like less on-- I'm not endorsing Claude. 25:47 I mean, you know, Claude Code, I, of course, I use it and, and use other things as well, but, but here I was more about like the what does a modern coding harness look like? 25:54 I wrote a blog post that actually, uh, yeah, this one, where this was, yeah, earlier this year, where I sort of broke down some of, some similar-ish principles. 26:03 The TLDR was that, you know, you could sort of through like pretty much introductory algorithms level understanding, you could just define like essentially an agentic loop that does the bulk of what you would call a coding agent. 26:15 Um, and of course, there's like bells and whistles. There's no sub-agents. There's no MCP, none of that stuff. 26:19 But if you wanna see something that you talk to, and then it literally will edit your file, and you can do s- pretty sophisticated things with it, you c-- it doesn't require more than two hundred lines. 26:27 So I, you know, a lot of what I try to do in education is to try to like peel away the stuff that's like the secret sauce and show that like really at its core, I mean, these are still students, and you want to teach conceptual principles. 26:37 There's nothing more to this than in Stanford, there's this class called 106B. It's like the introductory to intro to algorithms. You don't need more than 106B to, to build something like this. And so there you go. 26:45 [laughs] Um, there it is. Given that you already taught this class, any pushback, any questions, any weirdness that you observed from the students? On- What do they not get? What do they ask- Hmm. 26:56 Seem to cluster questions around? Hmm. You mean like now that this, like this particular concept or just in general? Uh, this concept. 27:04 Of, of like- Of like, of like building it from scratch and then like doing, doing what you did because this is basically what you needed for Pi. Right. Right. Right. Exactly right. Yeah. Yeah. Yeah. Right. 27:12 So yeah, it's, uh, it, it-- not really. I think they're-- I haven't got- received any like pushback on, on like this idea at all. I think they're actually more just most of them are consumers, right? 27:21 Most of them, most of the students, uh, do just like end up using the tools. I don't think that they get a lot of exposure, um, if any, certainly in the coursework around peeling that onion back, right? 27:29 And I think w- um, so this is the week of the harness. So, so, so Tuesday, first lecture, you know, logistics, blah, blah, blah, all that stuff. 27:37 But then first half is fir- you know, the first true like bit of curriculum is the two hundred lines of, of, of like a, of a harness. 27:44 And then yesterday, when I had sort of the follow-up to that is like we actually took Claude Code and then, you know, you have these like proxy things. 27:51 You, you-- these like pr- proxy tools where you can intercept the request. MITMProxy? Yes, exactly. Exactly right. So use that. Uh, use, use that tool. Intercept the requests and responses. Nice. 27:59 Look under the hood, see what's happening there, and then, then it's like, okay, I did say you have Claude Code in two hundred lines, but really, okay here's like actual Claude Code. 28:07 And so what I literally did is I took Opus 55, so this is like, you know, model two days old, and was like, this is-- I intercepted those requests. Here's exactly what's going on. 28:14 Here's the prompt, here's the system prompt, here's the plan prompt, here's the compaction prompt. Mm-hmm. And then we just like went through all of that and sort of showed it to students. 28:20 So now they really, really have like a s- a state-of-the-art harness that they can understand. Yeah. It's nice to get those three levels. Like students are very much consumers. Here's the baseline agent- Yeah... 28:30 in a loop, and then the real one. I like how you have a video walkthrough of your blog post as well. Do you see students are on like extremes of some people are power users, they consume a lot. 28:44 For some people, you could assume that this is kind of very straightforward if you know how agents work. Right. Uh, how do you kinda balance that in the classroom setting? Like do you tailor towards the average? Yeah. 28:57 I think it's a bit self-selecting- Mm-hmm... because the curriculum is all here, right? Yeah. 29:01 So the curriculum has been available since, I mean, you know, definitely it's like a few weeks and, and, uh, I, I guess I announced it a few weeks ago, but it had kind of been in the works for like probably at least one to two months. 29:10 So if- if you looked for it, you could probably find it. 29:11 Um, a-and I feel like the reality is if students are not interested in this, then they won't, like, they'll, they just won't show up and, [laughs] you know, they, they will, they will choose not to show up. 29:20 What I'm finding that is interesting is there are still definitely students that show up, but then there's also just a lot of folks from, you know, broader Stanford post-docs at other departments. 29:27 There's a lot of like interdisciplinary- Mm-hmm... 29:29 sort of interests where folks that are in, you know, somewhere in pathology or like, you know, like medicinal sort of fields or, or even like random other engineering fields that aren't pure CS that are coming in and are also interested in, in kind of getting exposed to this. 29:41 A lo- also a lot of, uh, GSB students, MBAs, um, that also, uh, want to [chuckles] I guess, learn how to not have a technical co-founder or whatever the- [laughs]... motivation there is, but... Yeah. 29:52 So I, I think, by the way, if anyone hasn't done the MITMProxy exercise- Yeah... you should do it. Yes. You'll be shocked. 29:59 Uh, Claude Code has the longest system prompt of all the coding agents, something like twenty thousand tokens. Yeah, yeah. Although, uh, yeah, that, that is true. That is true. Yeah. Yeah. Yeah. 30:08 Although they've also reduced the size. They reduced the size. They changed it. They cut it by eighty percent. Yeah. Yeah, I know, but then they grew again, and then they cut and grew again. [laughs] It's, it's a seesaw. 30:15 Yeah, yeah. No, it, it, I think it's interesting. Yeah. You'll be surprised like what goes in there- Yeah... that actually sometimes you don't use. Mm-hmm. Sometimes it actually hints at the internal w-workings- Yeah... 30:23 of some of the tools. Um, anything that you found there that could be cool to talk about? Yeah. No, I mean, in fact, yesterday when I was talking about this, it was a little bit of this like 30:33 one of the things that I thought was most interesting about Claude Code back in the day, and I think part of why it really blew up in terms of popularity was their focus on simplicity, right? 30:41 So in the early days, they were among the first to really lean into this, let's fewer tools do more, right? 30:48 Lean into SQL, lean into Bash, lean into like the things that are, that are powerful, like Swiss Army knife tools, and have that be a lot of the driver of capability. 30:55 Um, I think they did a fantastic job leaning into that. I think that's why they, they were as popular as they did. 31:00 I think that when I presented something like the system prompt, it was like, well, there was a different system prompt about two months ago. 31:05 It is now-- They have reduced in complexity, and it turns out that now what was once part of the system prompt, things like how to use even Git, right? 31:13 Or how to like set up tasks, or things of that nature, are no longer part of the system prompt. It's actually expected to be more kinda captured in the models, like the model weights itself. 31:20 Um, so they've kind of baked in a lot of those learnings more at the model level. And that just raises interesting questions around, you know, the model versus the harness, right? 31:26 And how do you kind of think through that trade-off? What do companies invest in? What should you invest in to really push the, the, the frontier of your own capabilities? 31:34 And so I, I think we still, even as a, as a community, we seesaw between the two a lot. Um- I think you don't wanna fault them for growing a system prompt and then cutting it. Yeah. 31:43 At some level, it's also how the model was, right? Uh, weaker, older models, you need to give them examples of this is good, this is bad, and then obviously you cut that away at some point, right? 31:54 And that's just part of learning what has been done, right? Yeah. You can apply the same towards if you're using a local coding agent with not the strongest model, you might need to add these examples, right? 32:05 That's right. That's exactly right. Yeah. Yeah. 32:07 And then, and then of course, there's, uh, interesting AI learnings to follow out of that, which is why, you know, as the models get better, when are you over-constraining it by adding those examples, right? Yeah. 32:16 [laughs] And then how do you pull away the over-constraining? Uh, and that's, I think, part of the learning as well. Okay. We too, you're going to talk about context engineering. 32:24 Repit, I haven't seen, uh, but there's many versions- Mm-hmm...of spectrum and development. Totally. 32:29 Um, what would you, would you highlight to a broader audience about, uh, what you're putting in here that people should know? Yeah. This is like the, you know, one-hour version of the course. [laughing] Yeah, exactly. 32:40 Like please condense everything. I think after this is going this- [laughing] That's, that's fair. Yeah. That's, uh- Well, no, it's, it's also a trailer, right? It's like, well, if you want more, go to this source. 32:49 If you're signing up next, next year or next fall for courses, you know, this is your one hour. This is like, this is like almost like the Quibi of the class, you know what I mean? 32:56 Where, where it's, it's, it's too long to be a trailer, it's too short to be the class. Is this like, what, what, I guess people will double click- And you know the best part? 33:02 If it's a trailer, by the next time you teach it, it'll change again. [laughs] That's true. That's true. That's true. That's true. Yeah. Well, yeah, so, so, okay. 33:08 So a good call out on that, Andrew, because it's, it's funny. 33:10 Um, the-- I think that the, the main thing that I just wanna reinforce here is, uh, and I see this even when I'm like interviewing folks, uh, and the very few people in industry even progress beyond like just vanilla prompting, right? 33:21 There's like a vanda- form, like, "Hey, do me the thing," right? Like do, do the thing. And, and that certainly as the models get good, that can carry you a certain distance. 33:30 But inevitably, you know, and any folk- folks I speak to that are come out of like a Google or some place where there's a massive like code base, the reality is you have to be a lot more thoughtful about how you're like ingesting context, right? 33:41 And so thinking through what are some of those principles and those techniques around how to like point, you know, k- and I sort of like, you know, the Repit framework that I'm sort of describing here is, is functionally similar to like Dex's RPI, right? 33:52 Or something of that nature, where you have this like re- research kind of proposal and sort of s- a spec, and then, and then you do like, uh, your implementation, and then you kind of tie it together with some testing and whatnot. 34:01 So what I just wanted to reinforce here is that like you should have this step, especially when you're entering unfamiliar code bases, when you're trying to, uh, sort of infuse into the context something that an agent needs to understand about a very large code base. 34:14 There's value in something that is a research phase. 34:16 And then in the case of this particular sort of framework that I'm putting forward, it's like also then suggests some proposal where there's a few alternative ways of solving a problem once you've done the research. 34:25 And so this is where I sort of also invite the human in the loop to the sort of be the decision maker on one appr- one approach versus the other. And then you enter the plan, which is kind of the spec. 34:35 All to say that the, the TLDR is like there is a way to do this that actually does scale out to code bases that are, you know, millions of lines long. Mm-hmm. Um, and it's not just the vanilla prompt. 34:45 Uh, and I think that's what I wanna highlight. If you're gonna go work in these big code bases, you should use something a little bit more, a little bit more involved. Anything to add on context engineering? 34:55 I think that probably the weirdest thing about how I think about context right now is when you split out to a sub-agent. Mm-hmm. Because sub-agents can benefit from inheriting your entire context of the main thread or not. 35:11 Uh, sometimes, uh, there's a regular disagreement on, on that, and then sometimes you want to condense what the sub-agent does and researchers put it back onto the main thread. And it's very hacky. 35:23 I don't like anything I'm doing. So I don't have any answers. All I know is like, uh, when you want to fan out work- Mm-hmm...uh, it is not just about you have a big spec and you, you, you know, you do RPI, you do R- Repit, whatever. It is not one agent anymore. 35:39 It is about fanning out work and making sure they all have the right context, and they all have the right guidance. 35:43 If you have a review agent reviewing all the other works, you have to make that guy have, have everything. And you always go through this phase of expand and contracts. 35:51 Like- Yeah...oh my god, these guys don't have enough context. I'm gonna give them more context. Okay, great. 35:55 And then they're gonna overtune to some random one-off comment that you made some t- a while ago, and then make some really stupid mistakes 'cause you, you didn't mean it that way. 36:03 But like [laughs] they took it as the law. Right. Something like that. And so I don't know what, what the word is I'm saying, right? It's just like multi-agent management for context. Yeah. Yeah, yeah. 36:13 Yeah, that, and that, that's absolutely right. There, there's like some aspect of, uh, this is maybe the part that's more maybe among the most art than science, right? It's like some combination, I think, to your point. 36:23 Yes, the fan-out for the exploration, the fan-out for like the subtasks, and then, and then the consolidation, compaction, whatever, you know, you might wanna call it. 36:29 I think being thoughtful about refreshing of the context, understanding when things are getting a little bit too bloated. 36:34 I think, you know, Dex had this concept of like the dumb zone and sort of like being a little bit- Mm-hmm...more thoughtful about the dumb zone. He's good at these equating things. Oh, yeah, yeah. 36:40 I think he's a- The little diagrams. The little di- yeah. Oh, yeah, I mean, he's- Why didn't we have him on? Uh, yeah. Well, I think he's-- I mean, he hel- definitely informed a lot of the, the curriculum here. Yeah. 36:47 And so he's just fantastic, I think, thinker in all this. Um, no, I think the diagrams are huge. That's useful. Yeah. I would add, um, breadcrumbs for people that are trying to look into this. 36:57 It's good to see best context engineering preferences, like on your sub-agent inheriting. Cloud-managed agents has a version where sessions are separate and it's not inherited. 37:07 This used to be a big talk- topic before reasoning models- Mm-hmm... when agents didn't run as long. Um, we had shorter context windows. Everyone was very, very managing that precious context window, right? 37:21 Uh, Link Chain used to write a lot of blogs about this and good context engineering. All this stuff is good for people to know. I think in the student context as well, 37:30 you can expect the average student to take your class 'cause it's a great class, and they'll know all these things, so you should know them, know them better as well. Yeah. Um, this is a fun space for research as well. 37:41 Mm. One of the papers, research papers presented at AIE, they looked at the delta in the Claude Code system prompt, what got changed, and they like, um, they, they plotted a whole bunch of changes in performance. 37:55 They tried to reverse engineer it. It's just good research that you can do. There was another one that also looked at context window management and how it affects performance downstream. 38:04 They weren't the best papers, but like, it's quite accessible, and it puts out good blog posts, so people that are trying to work on it, you know, I just- Yeah... recommend it. 38:12 It's a, it's a good area you can easily poke into. Yeah. It's a good shout-out. Mm. Agent skills, CLI. Yeah, yeah. I think all this is, uh... 38:20 So a little bit of an extension from a single, single kind of quote "function tool" calling, uh, like MCP, right? Using... 38:27 I- interacting with like third-party external services and then extending into like what you would call like procedures, right? 38:31 Like how do you automate some set of actions over and over again that you then formalize in the form of like a Skill MD or something like that. 38:37 So I think just making that a little bit more, uh, giving people exercise in, in codifying some of those procedures. 38:44 I think what I'll also sort of like include here is, is a few, few kind of fun skills that, that I think are, are good to use. Like, I think Matt, Matt Peacock's like Grill Me- Yes... is like a good one. 38:53 Uh, and then I-- there's ones that I also like from more from the educational perspective, like Jeffrey Blitz. 38:58 Uh, he has this like Explain Diff, which I think becomes especially interesting once you start talking about code review because, you know, the whole- Understand the code. Understand the code, exactly. 39:05 Yeah, you gotta understand the code and, and, uh, I don't know if you guys have used that, but effectively, like you run... 39:10 You know, you have your, your, your kind of the, the, the diff you've put in your code, and then you run the skill, and at the end of it, you have this not only like kind of a visual representation of what's happened, so you can, you know, kind of skim it and understand what's, what's actually at the core of it. 39:22 But then what I really liked was this almost like quiz at the end of, uh, at the end of it, where you go through and sort of like really reinforce- Yeah... whether you understand what's happening. 39:30 And I think that, that level of interactive sort of back and forth, I think is, is quite nice. Yeah. There's, there's still a couple blog posts in this domain that I really like to bring up. 39:38 Uh, this one, uh, I think recently, which is like kind of, there's no situation in which turning your brain off is a good idea. 39:44 Like, you know, and, and I think these skills actually force you to turn your brain on, and when everyone wants to use coding, you just turn your brain off. They're just like, "Make me an app and get... 39:53 Read my mind and guess-- get everything correct without me telling you about it." That's right. But like, a lot of these skills are mostly in the genre of like, "No, no, like, ask me more questions, please." 40:03 Like [laughs] Yeah. A hundred percent. I think there's a lot more output than people are used to. 40:07 Like, I can type in a one-line prompt and get tens of thousands of lines of code and a whole structure, and- It's like, "Look at my productivity, guys." [laughs] Um, but you know, 40:18 more effectively, you should know what you're working on. Yeah. You can really be put out of touch, out of the loop. You should spend more time thinking about what you're actually doing. Mm. 40:29 Uh, doing the work and getting the code written is much easier than it was. Yeah, yeah. The understanding there's no, no excuse for. For sure. Sorry, what is a web skill? Yeah. 40:38 So I mean, effectively, like skills that interact with- With the web... uh, with like the web itself. Yeah. Computer use. Computer use, yeah, things like this. Anything else? 40:44 W- Uh, I mean, mainly things that involve like pull and pulling information from the web. 40:48 I think it's, it's- Are there best practices on skills that will interact with the web versus- I, I think this is more of like a, a bit of a, a survey or kind of a demo of stuff that I found that was kind of interesting. 40:58 I think it was... It might have been out of... Was it like a Vercel? I'm trying to remember now. 41:03 I'd have to look back at the, the full, like reading materials, but there was like one site that I found where I think they... Maybe it was... It might've been browser-based actually. They might've like done- Ah. 41:11 Do you remember this? Yeah. Uh, well, it would be them that been- Could be... coined this thing. It's just that I was like, "Did I miss something? This is like a, a big part of the- No... skills community." 41:19 You know, yeah [laughs]. This is good. He's being educated [laughs]. [laughs] No. Uh, yeah, I think they, they like made some sort of... 41:24 'Cause I guess they were, yeah, I mean, browser-based obviously like- They like browsers. They like browsers and um... 41:29 But yeah, they, they had like a good, I think, website where they showed some of these interesting capabilities. [clears throat] Seems legit. CLI is a fun one. 41:36 Do you have a lot of students, or this is their first time in a terminal in a CLI or is it kind of the norm now? Uh, no. I don't, I don't think so. 41:43 Uh, I think that given that most of them, again, uh, last year, I would say that some of them had probably still remember the days of, of finger on keyboard, uh, typing. Uh, the, this is, um... 41:54 Most, I think students have like interacted with Claude Code at this point and so- I mean, I, I'm curious the split of people that are now just using the apps, right? Mm-hmm. Claude Code also has a desktop app. Yeah. 42:05 So does Codex. Codex is primarily an app now. Yeah. You're kind of not in the CLI, not in the terminal. It's a thing that puts off a lot of average people, but- Mm-hmm... just- Yeah... random note. Yeah. 42:17 Uh, I would say, you know, I think there's also another thing about skill. There are skills that are bundles of skills- Mm-hmm... where they, they sort of cross-reference each other. Yeah, like the plugins, right? 42:26 Uh, like plugins. And then there's also, I guess, the art of writing a skill, which is a thing that I've also focused on for AIE. Mm-hmm. Mm-hmm. I, I think a lot of these, uh, are focusing on skill consuming. Yeah. 42:39 Um, and you have all these influencers like Matt Peacock and what have you, they're like, "This skill will save you- This, this-... will solve all your problems." But no, you should write your skills yourself. Right. 42:46 And actually, when you try to write them naively, they're usually bad. And actually, it takes a little bit of skill to write a skill [laughs]. That's absolutely right. That's absolutely right, yeah. 42:56 What are Swix's best tips on writing skills? I, I have a skill... I have a problem with skills, people. I went through definitely a phase of psychosis of, like skill everything- No, no, no... earlier. 43:04 Don't give us skills. How do you write good skills? Yeah, yeah, exactly. Oh. So I have a skill for cutting skills. Okay, okay. So I, I, I like there... 43:11 So every like, uh, Cloud Code and Codex, they all have skill creation skills, right? So then there's... I have a skill cutter skill that cuts it down because they're, they're all... 43:20 They- it just documents every way in which those default skills that are created are bad. Hmm. So you always run this after any skill- [laughs]... is created to cut it down. Nice. Right? Um, so- I kind of like this. 43:31 Uh, yeah, you... [laughs] It's really s- uh, annoying because sometimes there is slop and like... but there's good in the slop. Right. And then you, you have to like sort of whittle down what they're, what there is. 43:41 Right, right. Oh, interesting. As in, you're saying like even the slop is... it's still in the language of the agent, so the- [laughs]... the slop means something. It's like, it's like a, it's a... 43:48 Yeah, okay, okay, got it, got it. Yeah. It's fun. Everyone has their own version. I use... First, there's very easy to like vibe check a skill, right? You just call it, run it a few times- Yeah... 43:57 and then I distill these preferences into a skill that's like, "This is what I like, this is what he doesn't like." Mm-hmm. Mm-hmm. 44:03 And it ad- it adjusts over time and then, you know, it can automate a lot of the- Yeah, well, so the, the problem is that y- yeah, you're calling it, but the, the actual... 44:09 the worst part is when a skill triggers when you're not expecting it to. Like you wrote it one time to fix a thing- Yeah, yeah, yeah... and then like now you're working on a completely different thing and it triggers. 44:17 Yeah. Why? 'Cause it's badly written. Because the trigger says like, "Oh, you call me whenever blah, blah, blah happens." Totally. And you're like, "Okay, read's good." 44:25 And then you do a different project and it, and it triggers, and you're like, "Oh, what the hell?" Oh, what's happening? What the hell's happening? Yeah. 44:29 You know, there's a, there's like a few people that are just like, "Okay, I have a bunch of skills I keep adding on sometimes when I mess up." 44:35 And then I know some people, like, uh Eugene Yan, they are pretty particular about how many skills- Mm-hmm... they have. They prune them. If not used, they clean them every few weeks. Yeah. Yes. Yeah. 44:46 So, uh, so yeah, th- this is like a, a huge problem if you're, if you're sort of in that skills world. I think this is actually why I suspect, uh, there are good... 44:55 there are very smart people at OpenAI and, and Anthropic that want skills to die. Hmm. Uh, which is not a popular statement. No, not... They, they won't say this publicly. Hmm. But it just feels sloppy. Hmm. 45:06 This feels sloppy. And I know it's... I, I can see my own skills people being sloppy. It doesn't feel, uh, AI tilled. Right. Yeah. You know what I mean? Yeah, yeah. 45:15 It's, it's not, it's not what AGI was supposed to look like. Uh, it ju- it's like everyone having a giant bunch of Markdown files. I mean, I think part of it is like- [laughs] Yeah, yeah. 45:22 Well, but there's a gardening problem, yeah [laughs] in some sense. A lot of people actually don't look at their Markdown files, right? Like the- Right... 45:29 the UX of it, of just chatting what you want, creating a skill in chat, using it, seeing output. There's a lot you can... There's a lot of value you can get without ever looking at your Markdown. 45:41 Don't, don't overcomplicate it at some level for yourself. Yeah. Uh, you can get a lot of value. 45:46 If you do stuff repeatedly, just be like, "Look at the last five times we did this, make me a skill," and you can iterate it with chat. There's a good side of it as well, right? Th- there's another thing. 45:56 Um, there's like I guess another way to look at it as well 'cause there's a, there's the use of the skill as just procedure that the agent kind of runs whenever to, you know, do like a code review or whatever. 46:05 But then in the case of something like Devin, right, where we've codified skills as a way of, of sort of, of representing our best practices, right? And then making... 46:13 and exposing that to the agent or the code review system to then, you know, institute the APIs the way that we like to define them, institute kind of the service level boundaries the way we define them. 46:23 That's like an alternative way of sort of representing like org level knowledge, and maybe not so much this like arbitrary thing you want the agent to do so you don't have to write the prompt each time. 46:33 Uh, and so I like y- I like that. I feel like that positioning might... That, that purpose might exist longer than just, uh, you know, these kind of pro- random procedures. 46:42 I just Googled Monaco AI skills and it's giving me the, the, the country. [laughs] Hey, maybe the country is- One day, one day- We're gonna become- Well, we're gonna beat them on AEO, that's for sure. 46:52 Can you just buy the country somehow? [laughs] Well, here's the thing. Okay, on this- [laughs]... we, we, we, uh, we're about to buy... No, we're not about to buy the country. 46:59 But, but, but, um, you know, we, we did, uh, we did get into a little bit of flack when we, uh, bought the monaco.com domain. I will say that. Uh, we did get, uh, sued by them, so. Ew. What? 47:09 Yeah, there's, there's like a whole public, [laughs] there's a whole public thing about this where like... [laughs] Yeah, so they didn't like that, but, but here we are. And, uh, we won. We won, so. Yeah. 47:18 Okay, then we're keeping it in. I was about to say this is spicy, but you won. Yeah, yeah. No, no, no, it's, uh- [laughs] Is that why we, is that why we see the flags flying around in here? 47:26 [laughs] Because it's like a victory. Because you won. It's because it's a victory. It's our victory. That's funny. [laughs] For, for those that are in SF, just look up and you know- Yeah, yeah... 47:35 you'll be- [laughs] So, uh, for, for those who are on audio, we're, we're showing this on YouTube. Uh, it's... You fly a plane around. I think you're still one of the few that actually do this on a re- regular basis. 47:45 Sam Blan came up with it, $6,000 a day. Uh, I don't necessarily think it directly causes me to Google what Monaco is, but the... all of Go- Monaco's ads are just Monaco dollar sign. Right, right. Oh, yeah. 47:59 We, we stand for money. It's also like dollar sign and the next number or whatever it says. It, it starts the conversation. Right. All... The, the whole point of the ad is what the hell is Monaco? Yeah. 48:08 And, and, you know, you, you could, you could ar- Like that's a debate to be had, but, but in some sense, like what is the purpose of marketing if not to like get attention, right? 48:15 And, and yes, maybe there's a one-level beyond this, which is if somebody... So you know, I'll, I'll go to events and people are like, "Oh, you guys are the ones with the planes. 48:20 You guys are the ones with like the big number on the billboard," blah, blah, blah, 193, you know, million, whatever it is. Um, and, and that... But they're like, "But what do you do?" 48:27 And then- [laughs] But that's, that's, but that's... But maybe that's the whole point, right? Is that I just needed someone to come and use it as an icebreaker. Oh, God. 48:35 [laughs] I will say it's what I said as you walked in as well. I was like, "Ah, the planes." Yeah. And now, now like, "What do you guys do?" "Well, well, let me tell you about what we do." 48:43 And then- So, so, you know, as someone who's been in those meetings deciding what we do on the, the Devin billboards, we spend so much time on the words, and maybe we should just be like, "Devin." 48:52 [laughs] I think we do it. Which is so fucking dumb. [laughs] But like, okay[laughs] You know? Look, look, I, I think, uh, Sam had another post, uh, this past week where he was just talking about, like... 49:03 You notice, I mean, Sam is effectively, like, our, our, you know, marketing sort of mastermind here. 49:07 And, uh, yeah, I think there's a little bit of history behind Be-Behind the Planes as well, but, but the TLDR is that hi-his whole shtick is, like, "I don't have all the answers. You just gotta try things." 49:15 And I think a lot of this is- It's good, yeah... we just try things, and, and if it sticks, great. If it... Like, did we know this was gonna blow up as much as it did? 49:21 I mean, not really, but, but now that it did, it's, it's- You had an idea. Yeah, exactly. So- And, and minimal is start small, go high, instead of start big, go small. Right. 49:29 Uh, way easier to get start small, go, go big. Yeah, exactly. Do this post. That's right. Um- Also, you know, it's 6K a day. Like, it's- Oh, yeah. Yeah, yeah... it's very much just try things. Yeah. 49:37 I think, I think people- Cheaper than billboard. Cheaper than a billboard, for sure. Yeah. Um... And it's fun. Yeah. Not, not much harm done. But, you know, it does, it's, um... 49:44 But, you know, the kind of simplicity here that, that's worth mentioning is then when we're trying to, like, one-up the plane, we're like, "You know what's better than one plane? Two planes." Ooh. Two planes. What? 49:53 [laughs] Three... [laughs] My mind went to rocket. Well, okay. Like, I, I also had that- I was gonna say, I've seen the memes... I had that idea. 49:59 Uh, I would, I would like for that, for that to happen as well at some point, but, uh, we'll see. [laughs] This is the next SpaceX launch. We just see Monaco- Yeah. [laughs] And they- There you go... 50:07 for the, the ship to start. [laughs] There you go. There you go. Oh. Gotta, gotta warm up to our, our SpaceX friends. Okay. I, I'm hap- Uh- You don't wanna? [laughs] Well, I mean, we're, we're on Monaco. 50:16 Is it worth talking a little bit about Monaco? Do we leave it there? Yeah, yeah, yeah. Uh- Because this is like an intermission between the, the topics. You decide. Okay. We're, we, we mentioned Monaco, right? Yeah. 50:26 Like, let's, let's give... I, I, I think it's- More... I think it's a little bit annoying to, like, mention what the hell is Monaco and then not answer it. Sure. Can you answer it? [laughs] Sure. Well, I, I hope so. 50:35 Yeah. So, uh- Whoa. [laughs] This is the country. So it's a small little, uh, I guess- Oh no, dog poop. Oh, no. There you go. There we go. There you go. Okay. That, this is Monaco. 50:45 This is the Monaco that, that I know and love. [laughs] Uh, yeah, so we are, uh, you know, we're, we're a Series B company, uh, here based in SF. We are, are building effectively, like, an all-in-one sales platform. 50:55 Uh, fully AI native, agents doing all the fun stuff, uh, or at least the stuff that humans don't like to do, but, but that agents are-- do like to do 'cause they [chuckles] don't really have opinions, but do in a less error-prone fashion, so the salespeople and sales teams can focus on doing what humans are good at, which is, you know, closing deals. 51:12 So presumably, Marc Benioff doesn't like you guys. [laughs] Probably not. But he's, like, also honestly a very AI native person. Mm-hmm. Very aggressive in agent force and everything. Yeah. 51:24 What are you doing that he do- that he doesn't do or he cannot do? Well, okay. So, so right now we are, uh... We started the, the company in a sort of like we're... 51:31 The, the market that we went after was not enterprise from the get-go, right? We sorta said, like, we're gonna start by... 51:37 Like, if our goal is to build the all-in-one, then we need to kind of build this in an iterative fashion with, with, as a group of customers, and we need to, like, grow with them into this. 51:44 It's much harder to pull someone away from, from Salesforce than to say, like, "Hey, if we meet you at the point when you are still just a Google spreadsheet tracking pipeline, and we sort of show you a, you know, more AI-pilled sort of answer to the problem, then we can grow with you." 51:59 And in a way, it kind of follows the same, the same, uh, model of like, you know, Brex or all the, like, YC companies, where you kinda sell to your batch mates, you kinda grow with them, and, and you- then you go up, up, up market. 52:07 And so we are, I would argue, are the, you know, the, the... I would, I would claim that the- among the best, if not the best sorta like, uh, startup-focused, uh, sales platform. 52:16 And so all of our customers are, like, early-stage startups. At this point, probably through to, like, Series A is, is kind of where, where we've primarily been focused, kinda seed to Series A. 52:24 And then now we're going more up market, and we've kind of learned a lot about the all-in-one in the last year and a half of doing that. And you are the head of AI? Yeah, that's right. 52:33 What's some cool AI stuff you guys have built? Yeah. I mean, uh, we have built our own agentic harness for, for sales use cases. 52:40 So, uh, all this, I mean, to, to call back to the, to the coding harnesses, uh, taking a lot of inspiration from, from the systems like the Codexes and the, and the Clauts of the world, and sort of building bespoke tools that are very sales focused, bespoke workflows that are all kind of geared toward GTM use cases. 52:56 Um, so we built our own version of that, that powers pretty much every single thing and that, that powers the Monaco platform. 53:01 So from, you know, uh, scoring of, scoring of TAMs, uh, scoring of ICPs and, and develop- dev- defining of ICPs- Mm-hmm... 53:08 uh, through to pipeline evolution, through to, like, all the mundane kinda things that maybe a salesperson doesn't wanna do, which is like, oh, update my meeting, you know, pre-meeting notes, update my summaries, like all these things. 53:19 Um, you know, we have agents powered, powering all the, those workflows. And so we, we've done that. It, it's, it kind of comes, comes out of the box. Um, and, uh, yeah, so long, long agents, long-horizon agents for... 53:31 long-horizon agents and short-horizon agents in some cases for, for, for GTM use cases. And it's pretty domain specific, built on foundation models. Would you do any of your own models for this? 53:42 Yeah, it's a good call-out, uh, because so far... 53:44 I mean, I think this is generally true of a lot of, like, vertical AI companies, is like you kinda start where you sort of use the frontier models, like as third-party services, to see what's possible, 'cause they are, like, the very frontier. 53:55 And then I think we've seen this, I guess, in the past week with the whole, like, Harvey sort of where they- Right... like the fif- plus 50% and minus 50% margin kind of question. 54:03 I mean, you know, I think ir- irrespective of the numbers, I think what is interesting is, like, at a certain point, things do become interesting from a cost perspective. 54:10 You, you start to wonder, can you do it faster on o- you know, open weight models? And then that's a lot of that... 54:15 Then a lot of what I think the, the purpose of, of the AI pieces are is just to post-train those models, build environments for, you know, kind of to post-train using RL and things of that nature. 54:23 So, um, we're kind of in that phase right now. Ooh, fun. 54:27 Yeah, I've been calling this, uh, playbook to the, the agent lab playbook, but that is kinda my coining of, like, you know, like agents start from the user first, and then they, uh, they, they use whatever models you need to, to build products, but then you can backfill the m- the model side Once you have the user base. 54:43 And then the, the model labs go, kinda go it the other way. They, they start- Yeah... with the big pre-trains, and then they try to find users. Um, and it's, like, very, you know, it's hopefully working- Yeah... 54:51 as a, as a, as a split. And Monaco would be on the sort of revenue GTM side. That's right, yeah. Um, and I think it's interesting, uh, they would want to serve you while also trying to compete with you- Mm-hmm... 55:03 uh, as, as model labs. That is correct. One of the things that I've been... Obviously, I've been, I, I dabble in, in, in some of this. I think one of the hard problems is actually enrichment. [chuckles] Mm-hmm. 55:13 [chuckles] Uh, I would say I spend a lot of my time doing that for, for basically SEO, and like, uh, research and what have you. And I imagine for you it's actually, like, a core part of that. That's right. 55:26 Do you find that models are good at it? Uh, do you have any vendors that, like, you like working with and you wanna give them a shout-out, whatever? Yeah, yeah, absolutely, yeah. 55:33 So we, we have used-- we were early adopters of Parallel, and then I think they're very strong. Okay, so just ge-general search. Je- yeah. I mean, they're- Not like ZoomInfo or some- No... one of those guys. No, no, no. 55:43 Yeah. Um, but yeah, we use Parallel, we also use Exa, um, so we'll give a, give a shout to both of those folks. Can you compare them? Uh [laughs]. No chance, brother. But they're, but they're both friends. 55:52 They're both friends. Okay. They're both friends. They're both folks that, uh, that we like to work with. They'd be very interested if you like one over the other. [laughs] Yeah. Look like lion. 55:58 I'll keep it, I'll keep it, uh, neutral. I'll stay, I'll stay Switzerland in this today. But yeah, no. So, so and then, of course, we do our own version of enrichment as well. So we... It's, it's like a... 56:06 Most of our data systems, I mean, part of like, one of, one I think of the most valuable pieces of software that we built or most valuable systems is kinda like our world database, like the database of companies, enriched companies, enriched accounts, uh, enriched contacts. 56:18 So, like, that whenever we are constructing ICPs for folks, we have the most up-to-date information about who you should reach out to at which company at this point in time. Like, I would say that one of our... 56:28 Like, I think where we are, like, world-class, hands down, is just demand. Like, we are a very good demand company. Like, we help companies increase their meetings booked in a way that actually- I see. 56:36 So find buyer at [indistinct]. Find buyer ri- at the right time, right? And it's kind of like the, a combination of those two pieces. What do you mean at the right time? Like, there's, there's times when... 56:42 Like, here's an example. Like, if you are, um, if you are targeting someone that just, you are selling to, um, VPs of sales, let's say, okay? 56:50 Like, the, the, the right time to, to reach out to them is not when they are, like, sort of in the middle of some, like, they've been at the same place for the last, like, five years. 56:59 If they are now in the middle of, like, a job... 57:00 Like a, a good signal is, like, once if they've gone from one job to a new job, they're now, like, entering a new place where they're more likely to be, to b- purchase or be open to a purchase of a new sales platform because they're in the middle of a job transition. 57:11 So going from company A to company B for, like, the VP of sales, a good signal of, like, hey, should you reach out to them at this point in time is did you detect that their LinkedIn changed or something like that. 57:19 How much of this is your expertise put into a model and harness versus models are good, there's generic best practices? Like, is there... There's obviously some value in both. Yeah. 57:31 Um, is it just the glue, getting it reliable, doing it all in one platform, doing it well? Uh, it seems like there's also- Yeah... stuff that we wouldn't know that you guys are also domain experts in that's baked in. 57:43 Yeah, I think without the former, it would, it would be like a fully commoditized. Yeah. I mean, not fully com- I mean, 'cause there's still a lot of value in the all-in-one. 57:50 I think, like, that is something that we definitely pioneered. 57:52 Like, most companies in the sales space started as just, like, point solutions, and I think they're, they tried more of, like, the YC approach of you kinda, like, land and expand. 57:59 You kinda pick a niche and then you expand from there. 58:01 Uh, and we, from the get-go, I mean, that's why we stayed in stealth for the greater part of a year and some months, was just building, 'cause we knew we wanted to be the all-in-one from the start. 58:08 But you know, that alone is, is powerful, it's meaningful. 58:11 But I think the real power is the fact that today, Monaco as Monaco has, like, a world-class go-to-market team, like a truly, like, world-class go-to-market team selling Monaco, of course. 58:19 But we also have the same go-to-market people that sell Monaco are the ones that are, like, what we call four deploy account executives. They're FDAEs. 58:26 They're effectively, like, well, I say human, human salespeople that are, like, the consultants of sorts that help onboard all of our customers, get them, like, building agentic workflows, and are r- effectively there to, like, help guide them through every part of their own sales go-to-market life cycle. 58:40 And so to your question, it's the all-in-one, yes. 58:42 But it's also the fact that we have these, like, you know, Sam and, and you know, we've, we've have brought on [chuckles] a lot of, like, former Sam, uh, Sam's team members and stuff like that who were, like, the best people in, in GTM at Brex and other places. 58:53 So, um, we're codifying all of that in the platform. Yeah. Interesting to have. You talk very differently than a normal head of AI I, that I might also talk to, because you're also a former founder, right? Mm-hmm. 59:04 And, like, you're, you're also think- used to thinking about these things in this way, and it's, like, interesting to- Right... it's important to have that- Right... uh, in your role. Yeah, cool. 59:12 Uh, I, I, I think you've, you've, like, introduced Monaco enough there and, and, and obviously, I think as the Monaco, the story develops, we can revisit. Sure, sure, sure. 59:20 Um, maybe the bridge back is also what is coding, uh, or agent native coding or modern software development like inside of Monaco? 59:31 Um, you know, I think you're, when you're teaching, you're maybe trying to be, "Here's the lay of the land," but maybe inside of Monaco, you guys are more opinionated on one thing versus the other. Yeah, yeah. 59:41 No, I mean, we have... So we, we made investments in coding agents from basically, like, day one, and we've, like, learned a lot. We have our own, like, scars that we've sort of, like, picked up and, and learnings. 59:50 And, and I think we've now been getting much deeper into sort, sort of sharing some of our learnings. 59:55 So, like, in our blog, which might be worth sharing here, we've been increasingly posting some of what, uh, maybe not the... So that one is more of, like, a product, the f- the one at the very top. 60:03 But the stuff I would say, like, yeah, when AI gets desperate and before, like, scaling contributions, build or buy, SRE. 60:09 So you know, we spent a lot of time looking at different providers, looking at vendors, and we ended up building a lot of things in-house. 60:15 So for us, something like a background agent platform, we have our own version of, like, a Ramp Inspect, um, that we call Monocoder. 60:20 [chuckles] Um, which is, is built on the Coder platform, um, but it is what we use to essentially scale up our engineering productivity. Coder, like coder.com? Coder.com, yeah. Oh. Yeah. Okay. Yeah. 60:29 We actually, yeah, have sort of adapted it to some of our use cases. And yeah, it's where a lot of our agent development happens. 60:35 It's just this really well-designed piece of software for essentially isolating of, of sort of, like, um, dominant environments, scaling out to multi-agents, um, you know, being able to run these, like, long, long horizon agentic coding agent tasks. 60:47 So Monocoder, shout out to John Tracy and, and, and co for, for building that out. And, and then, yeah, like, from there we've also built our own versions of, like, a, you know, like, like, a Slack bot which is... 60:58 We call Hal. 60:59 But, um, you know, effectively Hal is, like, this part AISRE, part, like, customer success, part, like, uh, kind of QA engineer because it, it, it lives in our Slack, but it, it serves all the purposes of helping us debug things and put up PRs when things are broken, but also used by our own salespeople to find out how the product works, find out, like, how to get to the bottom of something. 61:21 And so there's, like, interesting questions we've had to answer around the different personalities that Hal has, uh, depending on whether it's talking to, like, a salesperson versus a, versus, like, a, a develop- developer. 61:31 And so systems like this that we've built out over time, a lot in the last, like, year or so. Uh, shout out to Yunja for, for Hal, uh, and, and John Tracy as well. So, uh, you're building a lot in-house. Yeah. 61:43 Uh, you also use other tools? We do use other tools, yeah. So what is the line? Yeah. So- Builders buy... was that... Yeah, the builders buy. It, it's a good question. Uh, and we... 61:53 Yeah, I think there's this, um, [lips smack] uh, like, a thing, a general principle that I, that I generally... Which is like you build to learn and you buy to keep. Or sorry, you buy to learn and you build to keep. 62:03 [laughs] Yeah, the other way around there. 62:04 [laughs] So, so you, you buy to learn and you build to keep, and I think that's generally true here, where in the early days when we were still trying to figure out, like, you know, get our feet under us a little bit, we were... 62:14 You know, CodeReview as an example. One of them where we spent a lot of time looking at different code review platforms. 62:18 It turns out today we still use Devin 'cause we've actually still find a lot of value in how the system's been built, so, you know, shout out to the cognition team. 62:24 Um, but then along the way h- we did try out some, like, SRE tools and, you know, we tried some sort of, like, off the shelf, maybe more like background agent platforms. 62:31 Decided that it was just, given our infrastructure, given how we had set up our own cloud platform, it just made more sense to kind of build this out internally. Of course, we're not dogmatic. 62:39 We, we are open to revisiting them if, if something comes out. But here I felt... I think usually we ended up building in-house when we felt that there was no, like, mature offering out there. 62:49 So yeah, and CodeReview's maybe an interesting example because we did do a whole survey of code review tools, like, maybe over a year ago, and at the time- And I'll shout out, like, CodeRabbit, right? 62:59 Like- Yeah, CodeRabbit and, and, and them and, you know, that... There was also, like, Coderate Graphite. Yeah. Graphite, yeah, Graphite for a while before they got acquired. 63:05 Um, like Ellipsis and, uh, which, like, a YC company before the fact. So but there's a bunch, right? 63:09 I mean, so, so there were a lot of folks that we did try, and they, they were okay, but w- I, I don't think they, like, crossed the threshold of, like, they're immediately useful yet. 63:16 And so then we kind of, like, we sort of, like, put it aside a little bit. This was, you know, I'm saying, like, well over a year ago. Then we kind of went back to this and sort of organically... 63:23 I mean, again, again shout out to, to the, to the Devin team and, and cognition folks, where we tried it again. I think, um, I had sort of, like, gotten a, the... 63:30 Somebody had told me like, "Hey, you should try this and, and, you know, it seems like it's gotten a lot better." And I've been actually during Silas's talk during the class last year, 63:36 and it was just good, and it, it was picked organically. [laughs] Like, like the- So purely vibes, right? Like this is... There's no way that you know this. There was... Yeah. This, this was like... 63:44 I, I mean, okay, okay. Here, here's what I'll say. 63:46 When we did that first evaluation the first time around where we didn't really pick any vendor, um, and we were like, "There's no vendor that we're super happy with," I did do, like, an analysis where I looked at our PRs. 63:55 Basically did sort of like a, you know, set of pie diagrams, like here's what this is solving, are we, are we happy with the kind of the trade-off of verbosity and how often it's posting- Mm... with utility. Mm. 64:05 Um, and so there was... It was more statistically driven than just vibes. The second time around I wasn't gonna push hard on it. The first time I was kind of like, "Hey, we should adopt an AI code tool." 64:13 Second time I was like, "No, I'm not gonna do that again." Um, but then organically people were just like, "Hey, Devin is, like, good. Like, can, can you, like, add me to this?" 64:18 And so today people are sort of just, like, getting organically added. So this is a vendor where, you know, one of the cases where we're still leaning to the vendor. 64:25 You know, we'll see in the long term what that looks like, but, um, everything outside of that we just didn't feel like the maturity was there. Yeah. And it's not painful enough that you would, like- Yeah... 64:33 we've, we've involved Devin right here on. That's right. That's interesting. Yeah. Yeah. It's, it's, uh, yeah. So good, good talk for Devin who doesn't focus on code review. 64:39 [laughs] I, yeah, I, I didn't, I didn't, I didn't even know. Uh, you know, for the record, I'm only an advisor. I'm not that involved anymore, but, like, it's interesting to, to hear about it externally- Mm-hmm... 64:48 because internally, uh, there is a sh- certain amount of, let's call it, um, the, the, the bubble where like- Yeah... on Slack you're, you're like, you only hear positive things- Right... 64:59 'cause people only wanna circulate positive things. Uh, but, like, you, you have no dog in the fight, like you, you just choose the best tool. I mean, if you wanna get us like free credit- Yeah... 65:05 that's totally, that'd be cool too. [laughs] Like that wouldn't... I would say no one say no to that. [laughs] I think, I think Silas, I think Silas, uh, will hook you up. [laughs] [laughs] Okay. 65:12 So, so, so there's that, right? Um, and then, uh, and that informs all your teaching work. Yes, that's right. That's right. And, you know, I think that's, that's just really good. That's right. 65:19 Uh, and we can bring in more of this as, as you like- Sure... uh, because I think this is, this lived experience is very good. Yeah. Um, uh, I'm happy to go back to agent-ready code bases or more- For sure, yeah. 65:28 Like, so, so week four and five kind of, uh- Yes... you know, matching- Yeah. They, they blend a little bit, yeah. Blend a little bit. Yeah. Yeah. That's right. 65:36 Whatever you want to, whatever strong beliefs you have introduced in Monaco- Yeah... uh, and recommend people to start with. Yeah. No, for sure. Yeah, week four and five, I would say they, they are a little bit, um... 65:46 Four is maybe a little bit more like, uh, conceptual foundations and ways to do this, 'cause there are a lot of ways to customize agents. 65:53 And then I would say week five is where I really kind of like lean into, "All right, let's think about this and apply this in practice on, like, a true code base." 66:00 And so here I will definitely, you know, give a shout-out to, to the Factory team and, and, uh, and wanna just highlight the value of something like a readiness report. 66:08 I, I think that the main piece that I want to, uh, reinforce, something like agent-ready code- codebases is to me like a precursor and a bit of a building block toward everything to come afterwards. 66:18 When I say everything to come afterwards, I mean, like, people always got, you know, for a while, and I think people still are, like, super hyped on, like, multi-agents, like parallelizing agents, like tens of agents doing all kinds of things. 66:27 It's like you actually need to set a foundation in order to do that well, and the foundation is how verifiable is your code base, how, like, well-structured is your code base, how understandable is it by an agent. 66:36 And so I think, like, agent readiness to me is like almost, almost like piece one of the next four to five weeks where... 66:42 Maybe until, like, week eight or nine where, where I spend more time talking about, like, broadly, uh, you know, AI software teams and the kind of org level structure. 66:49 But- Definitely building up the, the building blocks for, for, uh, background agents, uh, the software factory. I think this is where it all kind of gets interesting from, from that perspective. Yeah. 67:00 One thing that I observed you talking about for Monaco, but is not here- Yeah... is the SRE side. When you close the loop, um, you actually want to have the agent read your logs for builds and then put... 67:12 I- ideally read your analytics for live production. That's right, that's right, yeah. Um, not as ex- not explicitly, but it's a really good shout-out, so I think that the... 67:20 I'm gonna weave that in, in, in, like, two places. One is in the background agents, uh, week, and then also in the software factory week. Mm. So, absolutely right. Um, I do... It's a really good call-out. 67:27 This idea of, like, trigger/event based closing of the loop, um, which, you know, now is one of the, one of the probably two or three entry points when people talk about agents kicking off tasks in the software factory. 67:38 Um, that's absolutely right. Uh, that'll come in there. Yeah. It's interesting that people call it trigger or event based, which I guess that is the actual technical implementation of it. Yeah. 67:48 It just seems too low level almost. Mm-hmm. Uh, but, uh, so, so the trigger is usually a cron job, right? Right. Like, let's, let's be real. Sure, sure, sure. 67:56 Uh, event based is usually a cron job that, like, surfaces an event and then it's event based. But it's, it's all crons. [laughs] Yeah, yeah. That, that's right, that's right. Yeah, yeah, yeah. 68:04 [laughs] Or, yeah, like web host or something like that, but yeah, that's absolutely right. 68:07 Um, yes, and, and you know, the technical term of course because [laughs] at the end of the day, yeah, uh, being sort of like a, a CS person by, by, by trade and, and yeah, we, we, we called it those things and so I guess that's, uh, I fall back on that terminology. 68:19 Yeah. And have any AI SRE companies made an impression on you? So, okay, I mean, like, I think the Resolve folks are doing well there. The Resolve is doing well. 68:30 I would say Night Shift is also v- doing quite a, quite a good job there in terms of broadly, like, background agents. I am a big fan of kind of wh- like, they're Greylock backed. 68:38 I think they're doing, doing real well in this area as well. Yeah, I, I would say that where... 68:43 I think SRE is an interesting use case, and this might, back to the build versus buy question, I think what's interesting about it is, like, where do... When does it become... 68:52 Like, where does, where does a, an SRE, an AI SRE, like, really see the most value? 68:55 It's typically when you have, like, a tremendous number of, like, service dependencies and these, like, massive sort of, like, systems and tons of Kubernetes and Helm files, like s- you know, uh, just, like, tremendously large Kubernetes infrastructure Helm that's, like, s- files kind of strewn ab- everywhere around. 69:09 I think that's why, um, Resolve has done really well in, like, the enterprise, right? Where you go and you sell to Ford, you go and sell to, like, a Walmart, and you kind of say, "Hey, you have... 69:17 Your service level dependencies are crazy." 69:19 Um, and you see these, these, like, you know, these diagrams of how things are set up, and then they kind of go in and they're like, "Well, we'll, we'll try and make sense of all of this." 69:26 Think that, like, at, at smaller scales it's usually still manageable, that you can kind of roll your own. But you know, I will say that for the enterprise, I think they, they've done a good job. Yeah. 69:33 For, for people who have never worked at this scale, 'cause it, it took me a while to, uh, to, to just, to see this, it is on the order of let's, let's call it tens of thousands of repos. More repos than developers. 69:44 These are the situations where, like, you don't know what code is running- That's exactly right... any- anywhere in the company- That's right... 'cause probably the guy who wrote it is no longer here, right? 69:52 That's absolutely right. And, and, like, ag- by, by the way, it's still load bearing. [laughs] That's absolutely... [laughs] That's, yeah, that's, that's the scariest thing. [laughs] That's, that's absolutely right. 69:59 It's, it's like, it's... 70:00 And, and then you're, you're the person, you know, you're the person who's on call and you're interacting with a system that goes down where n- that person still isn't there and there's, like, old out-of-date runbooks and you're like, "What the hell is going on here?" 70:10 [laughs] It's a scary, scary thing. And yes, you're... At the, at the enterprise level it's, like, it's a whole, whole different beast. 70:15 [laughs] Yeah, so, so, like, the, the, the, the kind of, like, college hacker, they, you know, one person per repo, they've never- Totally... no appreciation. 70:23 But, like, I think, I think if you just give that idea, like, you drop them in, like, a simulation of, like, "Okay, like, day one on your job this happens. What now?" Yeah. [laughs] Yeah. 70:33 So, so w- another shout-out I'll give, I'll give here. So the Resolve folks, actually last year they, they made an assignment for the class where they effectively created, like, a sandbox for students to go and... 70:45 Then the sandbox was, like, stre- I mean, the, within the sandbox there was, like, many, many services that were doing their various things and there was even, like, bugs in the services, uh, within that sandbox, and then as part of the assignment the students had to, like, go and actually, uh, like, use the Resolve agent and kind of, like, probe and play the, the, the, the human on-call person probing the system, discovering what's happening. 71:06 Um, wow, that's, uh- They've got an interesting- Yeah, yeah. No, no. It's an interesting website. I do not... None of the last... It's been a while since I've seen this. Yeah, this is very interactive. 71:14 Okay, they're, they're being a PagerDuty now. Yeah. Yeah, yeah, yeah. Yeah, that checks out. For sure. Yeah, but also this, like, this, this, uh, video. I was, uh, it's- Well, you know, you were shouting Adam. 71:22 [laughs] Interesting. Interesting. Sure. But okay. Yeah. It's, it... So okay, you know, we never get a chance to talk about- Yes... this side of the world except for conversations like this. Any other segments? 71:32 Because I, I think AI SRE is maybe too broad or even too narrow- Mm... uh, depending on how you look at it, right? Like, too broad meaning, well, there's several subdivisions of AI SRE that's interesting. Mm-hmm. 71:42 Um, or too narrow being like, well, actually that's not even the biggest problem that we have. Mm. So take it however you like. Yeah. 71:48 Uh, you mean, like, specifically for, like, Monaco or just, like, broadly for, like, organizations thinking about this? Broadly because that is our audience. Mm-hmm, mm-hmm. Take it however you like. Yeah. 71:57 I, I think that it, it... Again, it's a, it's a function of how we just scale, right? I think that for, for the enterprise, they, there are a number of issues that aren't always... 72:07 I, I almost feel like the answer is not even always just more agents. 72:11 Uh, and, and since we're still actually on the week five agent ready code bases, like, when I talk to companies of, like, a certain scale, the, the things that are holding them back are really more just, like, low level, like, DevOps, CICD stuff, right? 72:21 They're, they're, like, their, their merge queues are getting backed up. They... It takes too long to- Okay... stand up an app. Okay. 72:26 There's just, like, actually more foundational things that have almost nothing to do with... 72:29 They're, like, throwing in one more SRE bot or a- a la, you know, a bot wouldn't be the answer to the problem because there's, like, more systemic things that they need to figure out. 72:36 And so, uh- So that w- brings us to week six. This... Yeah, exactly right. [laughs] So that's, that's where we get into, yeah, easy. [laughs] Exactly right. Yeah. There we go. To the merge queue and beyond. 72:45 Um, yeah, this is a, a, a love letter of sorts to, to modern agentic code review. Um, yeah, I think the, the, the, the biggest thing that maybe happened between last year and this year was just- 72:56 It suddenly became very clear that code review is a bottleneck, especially once you-people got really AI pilled and, and peop-there's the velocity on generation became so high that, you know, you could in theory like put up 30 plus PRs a day, uh, and then what? 73:09 Like, what's... At what point do you decide that 73:13 you either have to institute yourself and, and put yourself in the middle as some way to, to gate and review things in the traditional way, but then you're, you're literally the constraint, you're the choke point of the velocity, and that's not good, right? 73:24 Then, then what was the point of generating at scale? 73:26 Or you say, "Well, how do we maybe get the best of both worlds where we're instituting ourselves enough that we have trust in what's getting merged, but we're also, uh, able to kinda like merge at scale, but we're also not kinda like free-freewheeling it here and anything goes?" 73:40 So a lot of, I think, what I wanna touch on here is just that, that middle ground, right? 73:44 Which is like there are, there is a class of problems that if you've put in thoughtful design into your agent readiness, you've put in thoughtful design around your verifiability, there is a class of problems that you can feel comfortable, you know, some percentage of the time essentially like auto-approving. 73:58 And what has to be true from a kind of a design perspective, from like an architecture perspective to just auto-approve, and when do you have to insert yourself and say, "Actually, here I still wanna review this carefully"? 74:08 And even if you do review it carefully, are you still doing line-by-line review or are you doing something else, right? So yeah, some of this stuff. Yeah. 74:14 We, one of our hottest topics of this year, uh, we have a guest writer program for Latent Space. Uh, this is from Ankit from Aviator. Nice, nice. 74:20 And he has a whole thing on killing the code review because I, you know, that, that, that is like s-maybe the final boss- Yeah... uh, of, of the, that's, that's a bottleneck. 74:28 He has a few, uh, gates by which he does this. I, I don't know if you, you've talked to Ankit or- Yeah, no, I've, I've read this post. Yeah, yeah, yeah. Okay, yeah. 74:36 I've, I haven't spoken to him personally, but I've, I've definitely read this. Was he right? Yeah. Any criticisms? No, I think that was right. Okay. I mean, like, like broadly right. 74:42 Um, I guess you mentioned one thing like the last boss. Um, maybe not f- I think there's two choke points. There's the review and there's probably like- Yes... the intent or like kind of the thing at the top, right? 74:49 Like this is the- Yeah. And there's some, there's some stats as well that, that I like to pull up. Uh, you know, everyone's stats is different. For sure, for sure. But like f-clearly more people are merging. 74:58 Clearly the number of reviews are not increasing in line with- Yeah... uh, re-with merge PRs. Yeah, that's right. That's right, that's right. 75:03 And if you look at any modern, uh, sort of company that's, that's, that's really invested in this and thought about this carefully, uh, and some people have already done posts about this and, and you, I'm sure will do posts about this as well. 75:13 Um, but I think recently like the Amplitudes and, um, I think Cloudflare has, has done a good one. I think Augment has also done a good one where they talk about the design of their AI code review system. Mm. 75:23 Um, these are a bunch of really good... If you, if you wanna share some of those links, I think they're fantastically written posts. Um, yeah, so shout out to... 75:29 I actually, um, shout out to, to Cloudflare because I think like, um, very like kind of like, you know, in a sense like legacy company or like, you know, bigger company, but I think they've done a fantastic job in terms of just their, their content around the AI, the AI SDLC. 75:41 Um, so I think they have a good post there. But TLDR, um, yes, I think this is true. I think that there's... We're still fi-finding that sweet spot. What is auto-approvable? What is not really auto-approvable? 75:53 And if it is not auto-approvable, is even in that world of something that has to be somewhat human reviewed, what do we review? Are we still doing the line by line or are we doing some, some combination of the spec? 76:04 And, you know, and I think this is... Yeah, again, [chuckles] this a whole separate question there as well, but... Are, are people keeping, uh, are you guys keeping the spec up to date at all times? 76:14 Uh- The spec is in the way, right? Yeah, no. Well, I mean, okay, in a few different places, right? 76:19 So like there's a spec that as it exists in like a linear issue, and there's a spec as it exists in like the context of a PR description, right? 76:25 So you can like, you can have your peer description be the place where you like describe more or less the entirety of the implementation in a more condensed format. 76:32 And as long as that's the most rep- that's the most up-to-date thing representing this point in time of the, of the code base, I think that's sufficient actually. So we're thoughtful about like our peer descriptions. 76:40 We make them be not just like, you know, a one paragraph. 76:43 It's more like it is effectively like a spec of what has happened in the PR, and there's even like kind of line, um, file level pointers to things to where things have changed, so you not only get the high level of what's happened, but even the kind of the, the construction of the files and sort of the surface level boundaries that you're touching. 76:58 So yeah, I think that's where we, we've invested quite a bit. So I push back on that because to me, the, uh, y- of course it's, uh, you know, bare minimum it should be internally consistent with the PR spec. Sure, yeah. 77:09 The... No, what I meant is the spec for the feature, the code base, because that is the summation of all PRs. 77:14 That, that is like y- if your, if your PR spec got it wrong or missed something, it's gonna be the delta between what the whole code base or the whole history of the code base is supposed to do versus what your PR is trying to do. 77:24 Mm-hmm. Yeah. So, so I'd like, I mean, I, I don't expect you to have an answer. Yeah, yeah. Because like I, I really stress about this, uh, myself. I try to instruct models to do it. They're bad at it. That's it. 77:34 [laughs] So the, so the humans are the s- the spec gods of like, well, f- five weeks ago I did this other thing and now you're reversing it. Have you, have you seen the, um, the post from Basis? 77:44 Like they have a, they had a tweet, tweet little thing, a tweet article. 77:48 I thought they did a great job talking about, um, [lip smack] like different sort of file artifacts they, they, uh, they track via Git, and some of them are like the specs, the intent of what is to be, and the other is like what they call like the canonical state of the code, which is like a little bit more maybe what you're describing, which is implicit knowledge, implicit intent that also has to be codified somewhere, right? 78:07 So that part, that's all version controlled, and I think they found that to work for them, so that might be another thing to share as well. Yeah. V-Vibu was trying to pull up that. Uh, this is Basis- The, the-... 78:15 the accounting company... the accounting company. Exactly right. Yeah, yeah. Yeah, yeah. We're, um, we're having Mitchell at, uh, my New York conference, uh- Amazing... soon. Yeah, yeah. 78:20 Talk, talk to, talk to him about that. I like that they're c-coming up with this. 78:24 My solution for this is, um, [lip smack] I don't know if like how ready I am to talk about it 'cause I may never ship it, uh, is something that is replacing Git in order to- Mm... capture all of this. 78:37 Uh, so I call it a program database, uh, where it captures all traces and all... And, and, and reifies specs to something that, that we c- that we can ignore and, and keep. 78:45 Uh, and it also con-controls the skills that are attached to it, uh, as well. That's cool. 78:49 Uh, because I think like basically what Git, Git is only focused on code, and then y-you just need a source of truth that is, uh, not only, uh, everything else outside the code, including the, the implicit... 78:59 What would, what would you call it? Yeah, like implicit assumptions or like the constraints- Uh-... the intent, the... Yeah... other stuff. Yeah. Uh, and, but also multi, uh, like highly concurrent. Right. 79:07 Uh, 'cause Git is, Git by intr-by design introduces mutexes everywhere. Mm. Which sucks. Uh- Yeah... so, so you need like super concurrency. [chuckles] Have you seen, um, uh, the GitButler is it? There's like- Yes. 79:19 Yes... yeah, like, um, the guy who used to be CTO at GitHub. Ian, I think is his name I forget his name, but yeah Scott, Scott. Um, so yeah, the, so the... I think there's a, there's a lot of people going after this. 79:30 Um, and I think there's the, there's the question of like really what is the ideal representational tasks. Yeah And I think the winner, probably not me, Linear. Really? 79:40 [laughs] 'Cause like that's where the, the tasks are Yeah, where we're going. Yeah, yeah. Uh, and like clearly they have decided to become a, a, a coding agent company. 79:48 Right And well, you know, that they have all the source of truth there as far as you, as, as long as you buy into Linear. Yeah. Yeah, that, that is true. That is true. 79:57 I, I wonder, I, I have like this somewhat like kind of held belief that I think a lot of companies are gonna evolve into some version of a coding agent company, which is any company that has like a, a, interacts with B2B software where you're collecting something about the user or some like product insight, I think you can also close the loop. 80:12 'Cause ultimately like what a coding agent is, the purpose of a coding agent is to act on some event or thing that's happening where a u- user's involved or customer's involved, and they use that to like fix ideally some aspect of the product, right? 80:22 I think what Linear is doing, I imagine the same thing like an Amplitude could do as well, right? Because they're also getting insights about user analytics, user problems. 80:30 I imagine even like you wanna really stretch it, you could even have like a Decagon or CRM do something similar where, you know, "Here's what's broken. Go close the loop." Mm. Mm-hmm Um... Yep, all right. 80:38 So set all those things up for agent-ready code base and set up code review. Security is week seven. Yeah It's not week one. [laughs] Students What do you, what do you mean security's not important? 80:47 [laughs] No, I make fun it 'cause I, I always say like, uh, you know, at Amazon, job zero is security, and they always say this. Uh... Yes, yes Anyway, uh, but yeah I, I- Security review at the end. 80:57 Okay I was at Amazon, and I don't remember that, but I, I believe you This is a thing Andy Jesse says. Okay, okay, okay. I see, I see. Yeah And that was, uh, I guess he... 81:03 I, I was there partly during Jeff's time and then partly dur- when this transition to Andy, so... But yes, I, I think, um, this is a thing, a topic last year as well, and it's, uh, it's also one this year, and I think that there I, um, Isaac also spoke last year. 81:18 I find that he's an incredibly thoughtful thinker, uh, thoughtful thinker as [laughs] You know, a thoughtful, thoughtful leader in sort of this space. 81:25 I think Semgrep is an interesting company 'cause they've been around even before sort of agents, right? 81:29 I mean, they're the OSS repo was, w- was, it was well before that kind of s- traditional SaaS kind of, uh, defense mechanisms. 81:36 Um, but I think they've done a good job of pivoting and, or not pivoting, or like augmenting their offering to, into this space. Um, you know, I think it's, it's, um, it's an interesting time, right? 81:45 Uh, there's just a lot going on where these, these attack vectors are now kinda coming to the surface. There's a lot more slop and AI-generated code, and I think 81:53 maybe one of the core insights that I wanna drive here is just the idea of like dependency, kinda dependency tracking and supply management, basically. Like supply chain, right? 82:03 Um, you have to be careful about like your dependencies and, and pin things appropriately, and don't just arbitrarily hit NPM install and expect the chain to go. Yeah. 82:10 So potentially, the most secure code base is one with no dependencies. Yes. Right? That's correct. 82:15 You, you look at the dependency, you're like, "Cool depend- uh, cool implementation, bro, but I'm just gonna make my own." That's absolutely right. 82:22 That's absolutely right [laughs] A- a- and that's an interesting insight because it sort of to the limit, and you know, I think it's, it's probably easier said than done, but if you believe that with code now being commoditized, code being easy cheap, why not just build everything in-house, right? 82:36 [laughs] Like that, like of course that's... And that would be the safest thing 'cause you've seen everything, you control everything, instead of, you know, depending on Axios or whatever. 82:43 But you know, we'll see how many people actually act on that, but that is the safest thing. I agree with you. No, I... 82:47 So that statement is dangerous because then you would just make your mistakes instead of making, uh, m- uh, relying on other people to discover mistakes. 82:54 Right The whole point of open source is that, you know, what, what is it? Like, uh, sunlight is the best disinfectant. Hmm. It's like Linus Torvalds thing. Mm. 83:03 And it's, it's like, well, yeah, you will be very slow to discover your own custom software dependencies 'cause you only have your users, and they won't report it to you, but if this, this thing has 100,000 users worldwide, someone will find something. 83:14 Hmm. And then you can depend on it more, right? Hmm. Yeah, this is, it- Yeah... this is, there's no, basically I, I- Yeah... I don't wanna be, I'm trying to be a voice of, uh- Yeah... reason. 83:24 Right, right And you cannot be absolutist in both, and uh, you, you, I, I kinda trapped you there. I talked, I talked to you into the talk track of- Yeah... vendor everything. Right, right And there is a bug there. 83:34 Yeah, yeah. Yeah, it's an interesting one. I, I wonder how much that, the, how much does that... I, I buy some aspects of that, but how much of that is impacted by the fact that everything... 83:43 I, I always believe that like most OSS repos, if not all of them, will just become software factories that are effectively agent run in the future. Hmm. 83:48 What happens then where like, I mean, even, um, like even if you look at Warp, right? 83:54 Like Warp, which is a monocoding agent, of course they're also, you know, selling a software factory product, but, but they are running their Warp, open source Warp terminal as a software factory. 84:03 And that almost suggests that there is a part where, yes, humans are still involved, but increasingly less so. 84:08 Certainly not every single person is reviewing every single thing that comes in to the, maybe the same exact level that we did in the past. Does that make this code base more or less vulnerable to security? 84:17 It's a bit hard to say, right? Um, I think that's just where there's a little bit of nuance that I think we're still... 84:22 And even the folks that I know that is, that work in OSS are trying to figure out exactly what that stable point is. Yeah. There, there's a couple of related arguments, and then I, I'm, I'm happy to- Yeah... 84:30 to keep going, which is like security is like, it's like a nonstop, like huge- Yeah... huge, huge, huge surface area thing. 84:35 There's one argument from Malta of Vercel who's like, "Okay, guys, let's take every single like, uh, high depend, highly dependent on software, uh, that's written in C, C++, port it to Rust," and it's like, "We'll be faster, more secure." 84:50 Uh, and like it, it won't, it'll probably take like $6 million and, but you'll j- you'll, you'll get it done. And, and then there's, there's another argument for... 84:58 This, this also links to Pacing the Frontier from the, the open source, uh, from the Model Labs- Mm-hmm... 85:02 who are like, "Okay, let's slow down so that we," uh, or like, "We'll do Daybreak, we'll do what's, uh, Glasswing. We'll make sure everyone's patched, and then we can let loose on-... rampant acceleration. 85:14 Uh, so I think, like, there, there's, there's, there's an inherent belief that the models will fix all security. Mm-hmm. 85:19 And I think it's just unproven to me, because what if they introduce new bugs, and they're just different bugs? [laughs] That's fair. That's fair. [laughs] That's fair. 85:28 So, so I, I mean, well, I'm, I'm looking for a pushback. I'm looking for a, like, "Well, no, you're, you're, you're an idiot," like, obviously models are way better at this. [laughs] Yeah. 85:38 Uh, I, I guess, like, that's almost, like, in the limit. It- there- there's this... the question's like is, is a model... Do we believe a model will create perfect software, right? If a model's creating- Yeah... 85:46 perfect software, then, then you're done. Is there such a thing? Then, then that's done, right? Then, then we have... Then we're done here at all. 85:51 [laughs] Like, then, then, then we're just gonna go home and, and, and chill on a beach for the rest of our lives. 85:55 I, I think there's also the question of is it good enough to first give it for security, so let people patch then just delay the release and do it after? Hmm. 86:04 Or is there a problem of, oh, no, it'll be much better at finding vulnerabilities. You can full send test it for longer even if you've patched or, like, how long, right? Yeah. You can give it for three months. 86:15 Someone can try to exploit it for six months. The problem's not solved, right? Right. Right. Yeah. Yeah. And, and my whole problem with, like... And, and again, like, this is a whole rant that we don't need to go into. 86:24 [laughs] But like, it's, like, you know, you name 100 less organizations that, like, okay, these are, like, priority. 86:30 They, they have access to a s- to a cyber model that, uh, you've just created 100 attack vectors for me to go compromise. Mm. And, like, they're... by the way, they're pretty easy to compromise. [laughs] Mm-hmm. Mm-hmm. 86:40 Mm-hmm. Yeah. Yeah. So I don't know. Anyway, um, the- above our pay grade. [laughs] Uh, prompt- It's always fun to get juicy in, you know, security. Like- Yeah, I mean, security is right now very juicy, right? Yeah. 86:52 Yeah. Like, and, like, I think actually the security folks love it because they, like, they usually don't get that much attention, and now they have, like, a ton of attention. Is that true? Like, I- What do you mean? 86:59 I, I mean, in that they don't get attention. I find that, um, I, I find that at least from, like, a VC perspective, people love... VCs love investing in security companies, and have, like, pretty good... 87:07 I mean, like a Wiz or something, like, has done very, very well. Like, like Wiz- Wi- Wiz is, Wiz is argu- Wiz and Socket prot... Uh, I, I literally, I had this conversation with a VC- Yeah... two days ago. Yeah. 87:15 Apparently Wiz and Socket are the only security companies in the last five years that have worked. What about, like, Abnormal? I don't know. What... It, it worked for a VC point of view of, like- I see. 87:24 From, like, from, like, an economics perspective. Yeah. Hmm. Yeah. Interesting. Yeah. Interesting. Not, not, not, not for me to say. Right. 87:30 Obviously many, like, great players in, in, in that field, but it's, it's a tough, tough industry, I think. And so basically, like, the, the best time... But I think it's not controversial. 87:39 Best time for any security company is when an, a high-profile incident just happened. Yeah. Yeah. Right? Obviously. Right. Like, then- Yeah... obviously you're gonna sell a bunch of security. Totally. Um, yeah. 87:49 So, so I, I think, I think that's a, a, a fair thing. Like, people, you people should be aware. People should know the, the base- basic migra- mitigations. 87:56 Uh, I do think, like, s- code scanning is important, but, like, I, I just wanna give people the, the idea that, like, oh, well, models introduce bugs that you wouldn't make as a human, and, like, if you're, you're not checking those bugs, then, like- It's, it's not even that they make bugs. 88:10 It's that you have less awareness of what's being written, right? Yeah. Like, when you- Yeah, 'cause you're not even reviewing everything. Yeah, yeah. Forget bugs. Models are pretty good. 88:18 They'll have, on average, I would say, less bugs. The problem is also that you don't even know what you're putting out. [laughs] Like... [laughs] Yeah. There's a lot of that. Yeah. 88:29 And that's, yeah, I think we have to do top of mitigations and, and, like, the... And there's sort of a combination of security and, and security as part of code review as well, right? 88:35 I mean, I think this is also, like, doing those checks I think is important. Yeah. Yeah. 88:39 So, so I think, um, maybe I would say one more thing here, uh, or, or just check your, check, check my intuition on, on this thing- Yeah... is that, uh, not all parts of the code base are created equal. 88:49 The parts of the code base where security is very important is data storage- Hmm. Yep... um, and, uh, billing. 88:56 So let's call it, like, data, billing, uh, auth, and auth is obviously related to all those things, and those are basically don't write code, and then everything else go- [laughs]... go nuts. Yeah. Right? Yeah. 89:08 I think that's absolutely right. Uh- That's absolutely right... and, and, and it's really funny how, like, okay, well, if you don't know what the difference between auth and, and auth Z is, please learn it. Mm-hmm. 89:17 And then, you know, if you, if you don't know... If you have never looked at AWS IAM, which you have, please learn it. Like, it, this, like, these are basic, like, conceptual things to understand that you can prompt for. 89:28 Now, once you have the language, you can prompt for it. You don't have the language, you're fucked. [laughs] Totally. Totally. Yeah. 89:33 Yeah, to an extent, I, I would even extend that to, like, just general, like, certain, certain, like, cloud pla- platform of things, right? I mean, there was the, uh... 89:39 Forget which company got, had that huge incident, huge incident where, like, it was just, like, a DNS issue that brought down... Was it... Do you remember this? Was it... Yeah, it was DNS issues- Many, many DNS issues... 89:47 well, you're gonna need to narrow it down maybe. [laughs] Yeah, I know. I know. There, there was like a particularly prominent one. Was it- There have also been many prominent DNS issues. [laughs] I don't remember now. 89:58 [laughs] List them off. [laughs] Let's go, let's go down. It's always DNS. It's always DNS. Yeah. Yeah, so DNS is more of, like, a reliability uptime. Sure. Sure. 90:04 Uh, you know, security, I think people are like, "Yeah, did you lose money? Did you leak customer information?" Yeah, yeah, yeah. Yeah. Uh... Yeah. Yeah. That- That, I would, yeah, probably put that above that for sure. 90:11 Only DNS 'cause of uptime. Yeah, general uptime. Yeah. SLAs and stuff like that. Uh, background agents, uh, now you got your buddy from Cloudflare here. 90:18 But, I, I mean, your guests aren't necessarily correlated with that week, right? Yeah, no. Yeah. 90:21 I mean, like, uh, the, the 9/1 was a little bit of a, a scheduling thing where I've thrown in, you know, Elad and, and Amjad. Yeah. [laughs] I guess Amjad is, is with definitely week nine. 90:30 Elad, I was trying to get him at the end, but I think it was just timing wise couldn't work it out. Timing, yeah. Um, so yeah, it's, uh, it's a bit of a weird one where I don't... 90:35 I guess I'm not lecturing at all week nine, but, [laughs] but we're coming back to those topics in week 10, so it's a little bit staggered in a, in a bit of a bizarre way there. 90:41 But yeah, week eight, um, Rajesh, who has been leading a lot of, like, the infra stuff, uh, at, uh, kind of cloud infra, uh, DevX, it's, and, and beyond at Cloudflare, which as I earlier said was I think they're doing a great job here, and definitely their adoption for s- a company of their size and a public company of their size is, is fantastic, and then their content I think is also really good. 91:00 I mean, this is fundamentally about all the ways you can use, like, asynchronous background agents. Um, I think I'm also gonna spend quite a bit of time talking about, you know, the Stripe Minions, the Spotify... 91:10 I forget the name of their version of this, um, the Ramp Inspects, the, you know, things of that nature that I think are all kinda showing what does it look like when you scale this out. Cloudtag. Instinct Grokbot? Yeah. 91:22 Yes. Muse? Yeah. Muse? Yeah. I mean, I guess less on consumer, more on like, you know, software teams using them at scale. They're all coding agents under hood. It's true. Everything is a coding agent. Yeah. 91:34 So like Grokbot is one of those interesting things where like very, very much presented as consumer-y. Uh, it's b- it's Grok under... It's, you know, it's, uh, it's Cursor under hood. Yeah, yeah, yeah. 91:44 It's the Cursor team. It's- No, no, for sure. Like, so I- [laughs]... I, I, I've seen the like, you know, Cursor working in, in the, like the data dots, but, but, but they are not... 91:50 They don't have like a code-based integration, right? They do. Oh, okay. I haven't s- used that. Uh- I've only been using it for like OpenTable reservations and- Exactly... okay. 91:57 [laughs] Uh, which, uh, I think that's the point of what they're trying to do. I see. Yeah. 92:02 As in like- They're, they're trying to be like, uh, well, you can use it for consumer stuff, but here's the extra power if you know how to use it well, uh, because, because it's actually Cursor under hood. 92:12 I think this goes to all coding agents, right? Mm-hmm. Like Claude Code, Claude Design. Mm. Uh, people using Codex not knowing they're using Codex, right? 92:20 A random code interpreter window pops up and you're just asking ChatGPT a question. At, at some level, there's coding agent under the hood, right? Yeah. Yeah. 92:29 So, so a- and I'll give you another sort of practical example. Uh, so I, I run all my SEO and AEO inside of Codex, and it has a code base- Mm... 92:37 even though like a traditional marketer would just check Ahrefs or something. Yeah. Uh, because I write all these scripts for doing all the, all sorts of crazy stuff. Sure. 92:44 And, and also I need to store data, and also I need to visu- visualize data, and like pretty soon you go f- like, as long as you're doing anything on a ongoing basis that has like maybe a lot of steps or a lot of data, you need a code base. 92:59 Hmm. That is an interesting take. Uh- I think models also like building code bases. [laughs] Not one of the... But that could be a temporary thing, right? I, I guess maybe my, my point was more on the interface, right? 93:09 Like, will... Are you going so f- like, are you gonna go so far to say that a Grokbot will have a place where like the power user will be able to see the code under the hood? Is that... 93:17 Or is it- Well, there is a computer attached, right? Maybe they'll, they'll just jump to Cursor I s- I guess. Is, is that what you're saying? 93:21 Well, exactly, 'cause I'm like, otherwise then, then you basically cannibalize Cursor. 93:24 Like if, if in the limit, what you're saying is Grokbot is the Cursor, then you don't even need like a Cursor if all under the hood is just becoming- It's the same team, so they- Maybe... right now they don't care. 93:33 But yeah, it's- Yeah, it's all Grokbot. Yeah, yeah. Yeah, yeah. I, I, I buy that... 93:37 I mean, I do accept that even things that seem outwardly just like non-technical use cases under the hood are technical, can be like written and, and designed and implemented in a code. 93:45 Like that, that much is absolutely true, right? The question is just more like, is a developer, is that their single point of entry? Yeah. I think it's not even developer, right? 93:53 Like, you think of a regular person doing marketing- Mm-hmm... that is using Claude Code or just Claude in the app now is working in a code base that they would never know, never touch, but 94:04 there's a code base under the hood. Mm-hmm. Mm-hmm. Yeah. I think Claude Tag is a good one for, uh, background agents. It's a big way Anthropic works, and, you know- Mm... they make it. You should- Yeah... 94:14 see how they work. Yeah, yeah. Check it out. Uh, okay, AI Native team, uh, we talked a lot about these topics of like gateway, uh, we just put a routing episode out today, and MCP portal. 94:24 Uh, I would say MCP portal is like the least, uh, established- You said it's bullish on MCPs. Uh, I'm bullish on MCP. The, it's, it's not bullish or bearish, it just is. Mm-hmm. 94:33 It just is a part of the landscape now- Mm-hmm... right? Yeah. Um, MCP- That's, uh, that's a statement. It is. There was a whole MCP is so back, MCP is dead. People, people like extreme [laughs]- MCP- But yeah... 94:43 portals is a- Yeah... unproven business. Uh, I st- I thought about starting one. Uh, I actually pitched the guide Anthropic on, on potentially starting one- Hmm... and then I lost interest. 94:51 [laughs] I lose interest in a lot of, a lot of things. But like, uh, what, what's, what's in there? 94:55 Uh, so okay, so what I'm thinking here is if this is, um, if this is the week desi- uh, more kind of geared toward scaling this out to like what happens when you take some principles, and a lot of this was informed by like the conversation I had with folks that were, you know, leading AI Native teams, AI Native teams at, uh, at respective companies. 95:12 And so here, uh, I think that something like MCP portal really becomes interesting once you're at a certain scale where you effectively have to have a single point of entry to do things like governance, security, auditing, cost optimization. 95:25 Uh, and so, you know, I, I'd be curious to hear what you see as the future or maybe lack thereof of something like this. But it to me is... 95:32 And it's if you believe in like an open router or, you know, some place, like a single point of entry for, for many things to happen and connect to and be effectively, you know, governed, I feel like there's a world where this also can easily exist. 95:43 Yeah. My, my only tweak on this is that this folds into the LLM gateway. So they- That I will give you... they are the exact same platform. That I will give you. That I will give you. You know? I can, I can accept that. 95:52 [laughs] That actually can, can agree with, yeah. [laughs] Uh, and, and then, then by the way, it also folds into the other thing of, um, what did I say it was? It was, uh, off Z, you know- Mm-hmm... 96:02 permissioning and all these things- Mm... and billing. So like e- like all, to the limit, all infra converges to the same exact problems where they have the box in the middle, and then you have the box on two sides. 96:11 [laughs] Uh, the, you know, God box will solve everything, right? For the enterprise. And it's true only if you take, uh, if you understand the job of the God box. 96:18 And so separating MCP portals and LLM gateways means you don't fully accept that they are, uh, you know, the, the, the same layer problem. 96:26 Uh, in, in a very similar way, I also find that people don't understand database querying, uh, to have a lot of the similar things where they will separate filtering from ranking- Mm... 96:37 from, uh, uh, or, or just like general recommendation. Mm-hmm. They're the same problem. Mm-hmm. Mm-hmm. Mm-hmm. 96:42 Uh, and, and so like then you just comb- uh, because like when you take all these things, you combine them in the same layer, then you can sort of cross-depend on, on things in useful ways. 96:50 If they're different layers, then they cannot. Uh- Yeah. Yeah. I, I, I actually will not disagree with this. Uh, I think, um, maybe the... 96:58 Yeah, so as a slightly pedagogical, I think that the, this breaking up of like separate terms versus one, I think I... It is absolutely correct that the underlying architecture is the, the God box- [laughs]... 97:08 where one thing goes in, and there's a bunch of other stuff that is some third party, some internal, some first party that all kind of exists, uh, on the other side of the God box. So yes, I agree with this actually. 97:18 I, I have a, I have a, I have a spicy, uh, spicy tweet I, uh... Oh, God, where, where is the God box tweet? Oh, yeah. So this is the M-M-M actually- Sure. Sure... N to M plus N- Let me bring it up... architecture. 97:29 Uh, so anyone working infra needs to understand, uh, okay, congratulations, you now have a job in infrastructure or you want to be an infra engineer, you need to understand that your entire job is making God boxes and then taking them apart. 97:41 [laughs] So either, yeah, your problem is either you don't have enough... You don't have a good God box, or your problem is someone else's God box you don't like, so now you have to take it apart and make your own. 97:51 That's funny. [laughs] That's very funny. I like that. That's very funny. [laughs] Into the heat death of the universe. Wow. [laughs] That's funny. Which, which is a... 98:00 This is a, a, a restatement of, like, there's only two ways to make software, which is- Mm-hmm... you bundle them, you unbundle them. Right, right, right. But, like, what are you bundling? 98:07 It is the things that go together, hopefully. Right. And then, like, maybe o-over time, people's opinions shift as what, what goes together, what should not go together. Right. Right. 98:16 So, so, like, there, you know, there is that, that whole thing of, you know, maybe, uh, Monaco should be a more of a point solution, maybe Monaco should own everything, right? 98:25 And, like, there's the question of, like, in some ways, you're bundling for, uh, the, the kind of startup segment that you serve. 98:32 In some ways, you're unbundling from Salesforce because Salesforce that, and then that's how you win market share from Salesforce. You unbundle from Salesforce. Hmm. Hmm. That's interesting. Yeah. 98:40 That's, that's an even broader statement, I mean, connecting it to- [laughs]... broadly all of vertical AI or, uh... And now, like bundling, unbundling, I mean, that's like the whole, like, Craigslist and Airbnb- Yes. 98:49 Yeah... you know, the, the, uh, Tinder and all those are separate. [laughs] Literally, literally separate. Tinder. Well, you know, like a personals basically. Yeah. That- Yeah, yeah... that was the- It is. Yeah. 98:58 Never thought, not thought, thought about it that way. But yes, uh, you know, one, one of those things, so, so we just shipped the Open Router pod today, right? And, um, Open Router, huge success. Mm-hmm. 99:07 Lots of people tried to build gateways. None of them as successful as Open Router. Open Router only builds gateways. So there's an example of, like, this... 99:14 They disagree with me that the, the o- the router layer should also do ops- Hmm... because Arise is doing ops, um, and observability and, and what have you. Hmm. 99:24 And, and they, and like the MCP gateway, they don't have that. Hmm. Uh, and they, they just, they only handle model routing. And so, like, I, I, like, I am... 99:32 I love being wrong like that because they're like, "We are the point solution, and we only do this one thing really, really well." 99:37 And, uh, that's our, that's our en-engineering focus, which is good, but it's also the marketing focus, which is nice. Mm-hmm. I, I think the, um, this becomes... 99:46 I think the, the, what you've described of like the one, one-stop shop for all the things is ideologically what makes sense, but I think what ends up happening in practice, and, you know, part of it is probably also marketing, is that especially when you look at the enterprise, is like this is broadly under infra, like broadly under, under platform, but then there's separate teams owning different parts of the stack. 100:04 And I think part of that consolidation at the, at the scale of an enterprise, if we're going to hit MCP portal and gateway, someone has to come together and say like, "Now we're... 100:12 There's a single endpoint that you're hitting- Yeah... for both of these things." And that fundamentally is like a hard thing to, you know, get- Yeah... everyone on board with. Yeah, yeah. 100:18 Um, so oftentimes it's like more like political and, and even like people problem rather than like a, from a software, pure software design perspective, this makes a ton of sense. Yeah, yeah. But that's not it, right? 100:26 They, like software systems sometimes are bottlenecked by people problems more than, than pure software- Yeah... right? The way, uh, my, my framing of this is typically, you know how you ship your org chart? 100:34 This is, uh, well, I forget what the, the name of this principle is. Uh, you need Conway's Law, you ship your org chart. Right. Uh, you buy your org chart. Like, like who... 100:42 Like when you start a startup or you, or you make a tool, who in the company is your, are your best friends for? Hmm. And if it's for everyone, it's for no one. Hmm. [laughs] That's interesting. That is interesting. 100:54 So, so like- Hmm... a, a conscious part of, uh, latent space in A- in AIE is like, okay, there's this category of software engineer that really, really, really cares a lot about AI. Is it gonna make it their day job? 101:06 Let's only make it for them. Right. [laughs] Right. Right, right, right. Not a, not a bundling of engineering. Never gonna be the dev everyone. Yeah, uh, uh, look, we talk about databases- Yeah, sure... 101:15 talk about security, but like there's an, there's an AI angle to it. Mm-hmm. Mm-hmm. Mm-hmm. Um- Yeah... we'll see. We'll see. That makes sense. Uh, cool. Software factory. 101:23 One thing I'll just like shout out as well related to the, the point is, uh, I think role of the future that is increasingly important, um, platform engineer. I think this is like a huge one. 101:33 Um- You know, okay, more lore. Uh, Temporal, pre- my job before all this, I would have... Uh, yeah, the, so we wanted to own platform engineer. If I hadn't started AIE, I would've done platform engineer. Mm-hmm. Mm-hmm. 101:43 Uh, because that is the role that deci- that decides all of this mission-critical stuff, and then the kids over in applications can do whatever they want on top of our platform. 101:51 But w- but, you know, the big boys do platform. [laughs] Yeah, and I think in the modern day, they own, they own cloud, they own DevEx- [laughs]... they own security, SecOps. So I think it's, it's all the things. 102:02 Like everything where you can actually like lose the company or like, or make or break, like yeah, it's platform guys. And then product guys can do, you know, adjust the CSS. [laughs] They're not gonna... 102:12 [laughs] That's saying... I mean, that's how you're starting a fight here. [laughs] I mean, I'm not the CSS adjuster, but, but I, I, I, uh, I, I appreciate everyone. [laughs] I mean, it's a fun time. 102:23 What's, what's the future? I feel like all the guys can adjust all the things, um- Yeah. Yeah, yeah. So, so there's, there's a few aspects here. 102:29 Okay, so, so there's a little bit of spillover from, from the previous week. So yeah, coding agents and big teams, some of the concepts. But then, you know, we'll touch on the software factory. 102:36 Uh, I want people to, to really start thinking about this, this world of, of this machine that builds the machine and where, if at all, are people still in the loop there. 102:44 You know, today, I think the answer is some at, some combination of intent and kind of review or maybe the places, like, so intent where you're defining the specs and what needs to be built and, and review where you're, you know, doing some amount of your own checking of the system. 102:58 And of course, this also, this is... I think that bottleneck is gonna be loosened up before even the intent piece does. But for now, those are the two choke points, I think, in, in a fully automated system. 103:07 Um, so yeah, I, I just wanna give people exposure to that. 103:10 For one of the OSS partners, actually, uh, Warp, they're doing this thing as part of their contribution to the OSS program where they are gonna essentially make a version of Warp that is running through the factory, and students will just be like building the factory, contributing to the factory- Yeah... 103:23 and keep it running. Like, like work in our factory. Work in our factory, exactly. Uh-huh. And so that's gonna be one of the ways. 103:27 And then they're gonna, yeah, find things that they find that are interesting there and pull it back in. 103:31 This is one of those things where, like, I'm not sure, like, obviously factory itself is, are very, very keen on this. Yeah. I feel like everyone wants to build their own- 103:38 Yeah, that's-- Yeah, I, I think that's, that is... It, it is a good question, and I think they're, you know, of course, entitled to kind of r- try and sell the products they do. 103:48 I think it's a hard system to own its entirety. It's a, it's a lot. It's, it's a lot, right? Like, decide everything for me, please. Yeah. It is a lot. [laughs] And, and, and I'm trying to... 103:56 You know, I haven't, like, gone too deep in understanding, like, the, the, how they approach their GTM here, but I suspect you would have to do it piecemeal, right? 104:02 You would have to be like, "Hey, we first do your coding agent. Oh, then also..." Like, you basically land and expand, right? I mean, we talked about this bundling, unbundling piece. 104:09 You don't start by saying, "Here's your whole factory." There's no way you could get, like, a Ford on board with that. You know what I mean? Where you're gonna go on and just plop the whole thing on their, on their desk. 104:17 Um, that's at least how I would think about it, is you first be their, their coding agent of choice. You, you, you build up some background agent for them. 104:24 You, you own some of their SRE pieces, and then you're like, "Well, now we can just weave it all together if you define everything in code or something like that." Or like, you know, configs. 104:32 There was a fun blog post on this from OpenAI. We had- Our buddy Ryan... we had Ryan on the podcast. Basically, they spent five months making an internal tool with zero lines of manually written code. 104:45 They kinda walk through, one, it's called harness engineering, so you can learn harness engineering. 104:50 But you know, the learnings of there was kind of a slump where it's slower at first, and then you offload, and you had peer review. It's, it's at this point, it's from February, but 104:59 it's, you know, a vision of a software factory, right? Like, you are full sending, um, letting agents fully write everything and how to do that well. Yeah. Absolutely. Yeah. I mean, some version of this is gonna exist. 105:13 Uh, just to fix this point, I do wonder about the, the sales piece here, like the actual business of this. That's an interesting question. 105:21 From a principles perspective, I think this is the, the, this has been version of the future, 105:25 and then beyond that, I think there's gonna be some amount of dabbling into and speaking about, like, okay, what does it mean to... Yeah, what does it mean to be a developer in, in the future, right? Like, are we... 105:35 People talk about taste. People talk about, like, you know, understanding of systems and architecture and, yeah, like software taste. Mm. 105:41 Uh, I wanna, like, go a little bit deeper here on maybe starting to, like, at least put a some answers to the question around this. Yeah. Uh, so this is your answer to where, where AI software engineering goes next? 105:51 Yeah, so I mean, the factory is, is like the tactical, but I think beyond that, there's like a, there's a question around, like, the role, right? Like, what does the role evolve into? 105:57 Um, if, if the factory is, is running and, you know, full steam ahead doing its thing, the conveyor belt is, is churning out code, what, what do we do in all this, right? 106:05 And I think we-- I think you and I, we had this discussion with Jerry at the dinner once upon a time, a few weeks back with Dex and Jerry- Yeah, yeah. Yeah, yeah... um, around the, around the factory. 106:13 But yeah, I, I, I think that there was this part of at least the discussion points there, which I am pretty much on, on board with, is like increasing levels of abstraction and then somehow there's someone still, even a CEO of a company has to have enough oversight and understanding of like different constituent parts to be able to make a judgment on direction, right? 106:31 To be able to course-correct, to be able to set metrics and things of that nature. I think there's a version that translates of that. 106:36 There's a, there's a version that of that analogy that translates to like someone who oversees the technical. We call it a manager of agents. We call it a, you know, different terms. 106:43 They're all buzz, buzzwordy, but, um, in some, some respects, someone who understands product requirements, constraints, technical constraints, kind of loops through all of those and, and, and knows how to adjudicate. 106:53 Yeah. Um, I, I will say that, uh, in the chair before you, we just had, uh, Anthopic's, uh, head of products and head of engineering, uh, for the- Platform engineering... 107:02 platform, uh, and I was like, "Uh, when are, when are your jobs gonna combine?" They're like, "Not yet." You know? Uh, and so if Anthopic's not doing, not there yet, you should not feel like it's, it's imminent. 107:13 [laughs] Good. Good. Good. Good. Um- Did they say how long or are they just not yet? Not yet, but who knows? Uh, I, I definitely-- There, there was definitely a lot of like, "Okay, yeah, we're-- I'm doing her job. 107:24 She's doing my job." Like, yeah. Yeah, I think it's also interesting perspective for students, right? You said there's a lot of sophomore undergrads, like- Mm. Mm... 107:31 you got two, three years to go to plan out what you wanna be doing. Um- Yeah... there's different ways to look at it from that, right? Yeah. You're not straight in industry. 107:39 The, the thing that I care, uh, I, I'm thinking a lot about is... Well, so one part of this discussion that resonates more with me is the loop, uh, stuff- Mm-hmm... where like, um, okay, you, you started writing prompts. 107:51 Now you no longer write prompts. Right. You write prompts that write prompts. That's right. Right? That's right. And then, then you just keep going up. 107:55 And I, I think to me, the sort of recursive concentric circles nature going outward but also going inward- Yeah... is something that builds the factory. I think you build up to the factory- That's right... 108:07 not build down. That's right. But I think a lot of people will eventually start building down. Like, you start from the top level, like what is my goal, and then you figure out the systems. 108:16 Um, that currently creates a ton of slop, and you should not do that. But someday you will. [laughs] Mm-hmm. Mm-hmm. Mm-hmm. Yeah. Yeah. So, yeah. So, so in, so instead of... 108:26 So, I mean, so, you know, that would be my question to, to prompt you, um, is, uh, uh, do we, do we, uh... What is the sort of prime question that people exist to solve? Like, that is your, that is your thing, right? 108:39 So in, in- That's right... let's, let's say an engineer today, um, you are measured by your PR merge rate. Right. PR contribution rate. Well, two years from now, that's gonna be ridiculous. 108:51 Okay, well, what's your main thing? Right. I don't know. Yeah. Money? Yeah. I think- People always go to money too quickly. It, it's something in between money and- Right... PRs. Customer satisfaction- [laughs]... 109:00 CSAT scores from your features. That, even that's like a product decision, not really a code decision, right? Mm-hmm. Like, I, like I don't know. 109:07 Like I- But, but if the, but if the engineering is evolving into like more of a PM engineer, like product engineers, you know, hybrid, then they... 109:13 Then I think that's, you know, that's my answer to the question of like what does the role evolve into. I think it goes without say that, 109:19 um, I would today recommend to students spend more time thinking about holistic systems, not just your like single part of the, the engineering puzzle, but how that ties in with- And you know, and this is like, like in practice ends up happening, like students for their capstone projects, I encourage them to go out and like, in fact, what I give them as the seed of ideas for final projects is like the, the YC request for startups. 109:39 I'm almost like, "Go build a company." [laughs] Like, you know what I mean? Like, go, go build something interesting that- What, what do people build last year? Uh, I mean, a variety of things. 109:46 Like someone was like, um, uh, someone came in the E department, uh, uh, was in the E school, uh, and E- was studying EE, and they were like, they built effectively a translator for like parsing of, um, like manuals to... 109:57 that have like, um, like memory addresses in, in C and then converting that into like, uh, like code that they could then like interpret and even like visualize. 110:05 So, so stuff, stuff like that is specific to their, their use case. 110:07 Someone was like a designer and is now actually at Vercel, and she like had, you know, like take a picture of my like design notebook and generate me an app based on what I, I take a picture of, right? 110:15 So literally she was like, for her demo, like, "Here you go," snaps a picture of like some random like scribbles that they, she was doing, and she got a fully fleshed out Vercel app. 110:22 And from that, and I think because of that, probably got a job at Vercel doing that, doing, doing some version of this. Wow. Nice. Um- That's a high bar, actually. Vercel, uh, very known for their design engineers. 110:31 Yeah, yeah. And I think she's now working on Eve, uh, from what I remember. Ah. No Eve in your... [laughs] Eve? Eve is an OSS, uh, or like OSS partner. 110:40 I mean, like, like they-- One of the, the ways Vercel is partnering with us is that they are having- Ah... students build with Eve. Yeah. Yeah. So the, this is the concept of meta harnesses, right? Right, right. 110:48 Uh, harnesses of harnesses. Uh, you can always go one level deeper. 110:53 [laughs] Um, but I, I do think like it makes sense, uh, at least for me, uh, in the same way that model orders wrap model APIs, you, you need to wrap harnesses as well if you're gonna- Yeah. That makes sense... 111:01 swap between them. Agreed, agreed. I think at some level, students should also stay grounded and like there's a lot of stuff happening at the edge. 111:08 There's also a lot of companies that are still like single call ChatGPT in production, playing around, haven't tried it. Um, you know, there, there is a lot of that, so there's, there's still time for it to work. 111:21 But I think your point on taste is good. Have taste, try stuff. 111:26 Like, uh, whether each company has their own version of a software factory, uh, you're building tools to help companies do it, like staying on top of it, actually using the tools, playing with it, like you'll, you'll figure it out along the way, you know? 111:41 Yeah. And I think the, the... 100% agree with all that. 111:43 I, I mean, first to your point that are a lot of companies still in the prompt ChatGPT or some version of, you know, ChatGPT, maybe they just got a cursor license and are just now getting people onboarded. Hmm. 111:53 Um, that is way more true than we believe, certainly here in the Valley. Um, that is- [laughs]... I can, I can say with 100% confidence. Um, 112:01 but yes, I think for the students being exposed to this is just, at a certain point, this, the fact that they know this and are on the edge is what will eventually allow them to also be some of the more interesting and more valuable contributors when they go to maybe places where, that maybe aren't up to snuff in terms of being on the frontier. 112:16 Yeah. Uh, one, one of the, this, this kind of progression type talk, still very influential, about a year old, Steve Yegge. Mm-hmm. Yeah. Talking about the, the levels of psychosis. 112:25 Uh, he's all the way through level 10 now. Um- Where he's- He's, he's, uh, he's given up and he's moved to a, a farm. But, uh, yeah. Like you want, you want... [laughs] Sorry. I'm, I'm making fun of him. 112:34 Uh, he built Gastown. Yeah, I remember. Yeah, yeah. All these things, moved to Gas City. Now he's burned down the city and said, "I'm building my game. I'm, I'm no more Gastown." I thought, I thought that... 112:44 I mean, I, I remember talking to him at, at, at the engineer fair, uh, earlier this year, and, uh, wasn't he, he, his whole like thing, like the only useful thing that was built at Gastown was Gastown or- Yeah. Okay. 112:53 [laughs] Right. With, with that, that's a horrible failure. Horrible. Totally, totally. 112:57 Um, and, and so like people should be aware that like psychosis is real and y- if you're not holding yourself accountable to delivering real products, you're just screwing around with LLMs. Yeah. I- 113:09 This is like a, maybe, I mean, relevant to it, but I wonder how much of... I mean, I played a little with Gastown, very involved system, maybe a little ahead of its time. 113:17 Um, but I'd also wonder if they were like bottlenecked a little bit by the, by the marketing in that it was a very difficult thing to grok from a terminology perspective. Like there's just a lot going on there. 113:24 It makes sense to him. I think the Gastown marketing also is one of the things that propelled it, right? Like what was right before Gastown? Hmm. There were like- Beats. The, like the, like his, his, yeah, Beats. 113:36 Yeah, yeah. He did- Unless you're talking about someone else's stuff. No, no. There were like, there was the whole Open Claw, there was something else, and there was Gastown. Mm-hmm. 113:42 And the whole manifesto also really helped spur it into the conceptual. The people that would never use Gastown but learned about Gastown 'cause you can- Yeah... read the manifesto. 113:53 Like there, there's a marketing angle to it. Yeah. I mean, so, so, I, I, but I do th- so Steve is n-not necessarily someone to take literally, but direction, like s- like, you know, directionally correct. Mm-hmm. 114:03 You know? Yeah. And, and so I think he's directionally correct on this one. And so if, if you're at the idea stage, it's time to move into the power user stage. Yeah. 114:10 If you're a single agent, you probably time to move to multi-agent and blah, blah, blah. Uh, and you probably should try to build your own orchestrator, right? And, and that's, that's all that. Understood. 114:17 The, the arguably, uh, this is not even, this is like kind of out of date now. Yeah. There, there is more to the software factory beyond this, uh, but this, this article predates the software factory. 114:26 I, I do think, you know, o- one of those things where like, uh, I'm open to... 114:30 As an employer, I have this problem in my company today, which is I do not want to hire anyone who hasn't been through psychosis and come out the other end. 114:38 Because I am currently paying for somebody to go through psychosis. Mm. And it's very expensive- Mm... from his salary, uh, very expensive in tokens, and I'm getting nothing out of it, and he doesn't get it. 114:48 Like my, my feedback doesn't get through. Mm-hmm. Mm-hmm. And so like, uh, it's almost like I would encourage your students to have psychosis and then come out and go like, "Okay, what was wrong with me?" Right? 114:58 'Cause like that's what S-Steve just went through, like a very public two-year journey of this. Yeah. It's weird on the student side. 115:04 There's like a lot of, uh, industry, you think you're being productive, you're not, is probably the student version, right? You're- Mm-hmm... 115:10 Notion charting and making all these optimizations, personal ads, and it's like, dang, I went through the psychosis but nothing has changed, right? Right, right. So go through it, go through it. Right. Yeah. 115:20 Uh, I haven't seen this. I've seen a version of this from Boris, actually. 115:24 Obviously more geared toward cloud use cases, but I remember he made like a cloud artifact which showed, uh, I think more at the team level, but I think he had four stages, not eight stages, but it kind of rough- roughly followed the same progression of like no adoption to through to 115:37 You know, I guess- Yeah, only four is good. This is a-- this is too many levels. Yeah, well, see-- we'll see if, uh, people can pull it up. 115:44 But other than that, you know, zooming out, have we, you know, discuss, discussed all the things- That's the one. That's the one. Oh, that's the one? Okay. Yeah. Here we go. It's the Boris- Yeah, it is an artifact. Okay. 115:52 So it's gated, assisted, parallel, and what's the last one? Supervised autonomy and AI native. Okay. So it's zero index, so it's five levels, right? All right. Things start at zero. Yeah. Wow. Actually, it's pretty good. 116:03 I need to, uh, read through this in more, more depth. And use AI native. Yeah, yeah. And do the best one. I, I would give a shout-out to dynamic workflows. This is one of those things that is a sub-agent pattern- Yeah... 116:15 that the OpenAI guys are ex- also exploring, um, that is clearly fitting. And, uh, and so basically, you're writing a harness on demand for the thing that you're trying to do. Yeah. 116:26 And so, uh, this is maybe-- like a lot of sub-agent conversation is just you're copying the, the parent agent- Yeah... 116:34 but you actually should be writing way more code and then, like, running that as a, as a, as a sub-agent, and that is what the Anthropic people have chosen to call dynamic workflows. Okay. Uh, but [chuckles] you know. 116:45 Yeah. We, we built a version of this as well. Yeah. It's a, it's a good pattern. For Monaco? Mm-hmm. Yeah. What does it do? Yeah, I mean, on the fly, sort of like harness sort of use cases for GTM. Ah. Yeah. 116:54 Ah, that's great. Yeah. Um- I will note that when Boris wrote this, it's for team adoption, for people that are listening, and also he says, you know, "Anthropic is on stage three. 117:06 We're still trying to get to stage four." Wow. He's like, "I am at stage four. I'm the guy." But you know, it's, it's not as simple as Steve's. Like Steve is a very good personal do this, do this, do this. 117:17 Uh, this is a much more if you're running a team and you wanna AI pill your team and not just- Yeah... be a user of one. Yeah. You know, Anthropic is still at stage three. Don't, don't- Yeah... be like... Yeah, yeah. 117:30 Okay, I mean, you should teach this. This is, this is great. Yeah. This is-- this has got all this stuff that- [chuckles]... this is everything that we talked about. There you go. Yeah. 117:35 I, I can't believe I haven't come across this. Yeah. Um- Good resource, yeah. Very, very good. Okay. [book drops] That was a wild tour. One semester of Stanford in one hour. [laughs] Or two, two hours. 117:46 Thank you for doing this. Any parting thoughts? Any call to action? No, this is, uh, thanks for-- I mean, ton of fun. It's, it's been a, a, a, you know, nine months in the making. I'm excited. 117:54 This is, uh, end of week one. A lot to do, a lot to learn. 117:58 I'm excited to see what happens with the, especially the OSS partners and just getting students like deep in, in the code of, of some complex code bases, um, having them build some stuff. So you know, it'll be fun. 118:07 [chuckles] Keep you guys posted. You're doing another version next year? You, you need credits for students if OSS people wanna get involved, they can reach out? Sure. Sure, yeah. Do, do, does credits matter? 118:17 Do they want it? What? Uh- Do, do they actually want the credits? Like the students or- Yeah. They want credits. You, you do have credit card. I don't think a lot of students wanna spend five bucks. 118:24 Um, I, I work at AI credit card, but- Yeah, no, I, I, I think. 118:27 Look, I will start like, uh, I remember, you know, it was like, you know, there's a certain number of credits for sure, but then on top of that, he's like, "Look, if they wanna go hog wild with this, like, let me know, and we'll do-- have these, like, one-off experiments where they can just, like, scale up to, like, a thousand plus agents." 118:38 And so- Great. This is email from Vector. Yeah, email from Vector. Yeah, yeah. Yeah. No, those kind of people are val-valuable. So this is Wilson from Cursor- Mm-hmm... right? Who is like, "I'm gonna write a browser." 118:48 Right? Yeah. That's right. Um, those people who are ambitious and, like, can actually execute- Yeah... those are very valuable, and that's basically a- That's right... a recruiting call. That's right. 118:56 Um, but- Yeah, that's a good point... 118:57 yeah, I, no, but I, I do think, like, students should, should, like, understand how to be ambitious because snapping a photo, I mean, respectfully, snapping a photo and turning it into an app is not that ambitious in, in th-this day and age. 119:09 Uh, writing a browser is. Yeah. And there-- and, like, there actually is a lot more projects like that. Yeah. Yeah, I think that's right. Yeah. It's all about raising the floor. 119:15 I mean, I think, yeah, last year it was, uh, it would, you know, you know, credit's good to the student of, of last year, but, uh, yeah, this year I obviously- [laughs]... 119:22 we're, we're gonna expect, we're gonna expect more. Yeah, that, that, that's a fail now. We're gonna expect more. [laughs] I mean, yeah. But that's, but that's, that's the whole point, right? 119:27 Like we're leveling up, right? Last year it was like, "Yeah. Hey, have you heard of prompting? Like, here's what it is." 119:31 I mean, I spent some time talking about like, you know, K-shot prompting as one of the lessons, and like, I don't even think you guys even think about that now. 119:37 So, um- I guess there's a, there's a version of thinking you stick out in your cohort of classes, and then there's also a version of you're also in early engineering roles with a bunch of people in industry, people still trying to get jobs, and it's not, it's not a, you know, not to call out one demo, but you can, you can do a lot to stick out. 119:56 For sure. I'll, I'll give, uh, students a freebie, uh, of a week 11. Uh, m-make benchmarks. Make benchmarks and evals. Uh, this will get you, uh, a, a good one. 120:05 It's gonna get you in the door of every frontier lab and every front, every top agent lab. Good, good engineering blog posts with some rigor on anything interesting. 120:14 There's a lot of benchmarks, a lot of hanging fruit there, but the engineering blog posts do really well, right? You, you went deep on one thing. Yeah. 120:21 Um, and probably also there's like big-- a lot of money to be made in that as well. So not just, uh, in the doors, but if you make that into businesses, you, you can, you can really, really cash in on that. 120:32 Sell some- Yeah, people don't know, like C- Sweet Bench was like undergrad project. Yes, yes, yes. That's right. From, from Princeton, right? Yes. Yeah, yeah, yeah. [laughs] Um, great. Okay. Well- Awesome. Fantastic... 120:43 that's it. Thank you so much. Thanks a lot. Yeah, thank you, guys
Community discussion