34: Ryo Lu - It's All the Same Thing

[00:00] Before we get into the episode, I have an announcement. [00:03] In case you missed it, I'm going full-time on Dialectic, thanks to the support of my new presenting partner, Notion. I guess first and foremost, I'm just excited and grateful. I'm about a year into this. I crossed my year anniversary of starting at the end of November, and it feels fitting to be able to... [00:20] fully lean in and consolidate and focus on something that has just felt like being in my lane, getting to amplify people I'm excited about. And I've been reflecting on this and I think this ties to notion too. Like I've been reflecting on like, what is, what is the show? What makes it good? [00:37] what am I trying to do here? And, [00:39] There's been a handful of patterns that have become more obvious over time, things that have become more legible. I think it's definitely a show about ideas. Particularly, I think I love to talk to people who make stuff about the ideas and philosophies that underpin them. But I was also reflecting on, like, what are the patterns that stand out most? And I think they... [00:56] tie into why notion is such an ideal partner for me. The first is [01:01] I think it's a show about where ideas meet action. I love introspection and reflection and thoughtfulness and philosophy, but I think I also love people who are able to take those things and use it [01:11] to make contact with reality. This combination of introspection and agency and action, ideas are powerful but we got to put them to work. The second pattern is craft. [01:21] craft is aspirational. Craft is when we deploy our taste. Craft is a human touch. Craft is saying, I'm just going to

[01:30] push things a little bit more to make them a little bit better. And whether my guests are [01:36] People who design things or write or invest or whatever else they might create, I think there is a deep amount of craft inside of how they approach what they make and inside the things that they make. [01:48] And the third pattern is soul. [01:50] or soulfulness. [01:52] This war is... [01:54] Obviously a little bit hard to pin down and you might, [01:57] Instead, say authenticity or originality or even aliveness. [02:02] But soul is about when somebody line is lined up, I think, like in who they are with. [02:09] the way they're showing up in the world, and maybe even more than that, a willingness to reach deep. And so I think when I think about what I'm drawn to and all of the people I admire, and certainly the people I talk to for this show— [02:19] It is... [02:20] soul at its core. One of the things I'm most proud of for this show is the audience. It feels like it's my kind of people. Some of my guests are listeners. Some of the people I've met through the show have been incredible. [02:32] And [02:33] Akshay Kothari, co-founder of Notion, is a listener. And so we've gotten to know each other over the last few months. And when I started to think about what it would look like to go full time on Dialectic and bring on a partner, it was ultimately a pretty easy choice. I think it was clear to me that he really got the maybe even intangible elements that made the show special to me and to the people who were listening. But also, I think those patterns I mentioned earlier really do embody Notion too. And that's why it made it [02:57] Such a right fit. Notion makes beautiful tools for your life's work.

[03:01] I think I'm someone who's certainly interested in tools. I've talked to a bunch of toolmakers on the show, including Notion's own Jeffrey Litt. He wasn't at Notion when we spoke, and he is now. But also on those... [03:12] Themes from earlier, I mean... [03:14] Notion is a tool for taking your ideas and turning them into action, whether that be tinkering with them or expanding them or sharing them. It starts with ideas with Notion. It's a brand and a tool that, despite a long road, a tremendous scale, and a great deal of complexity, has embodied craft, I think, at every step of the way, both as a brand and as a product. [03:44] the product they use. And I think Notion's community and templates and remixing and creative expression are all evidence of just that, a product that [03:53] is full of aliveness. So it ultimately wasn't a very hard decision to partner with the ocean. And I feel so grateful to them for helping me embark on this journey. As for what's to come, I mean, I think a lot more of the same. Hopefully people who are inspiring to you, people you're really excited about and people who surprise you. I would like to keep you guessing. I think too, a lot more video for those of you who are listening or haven't tried. Video is coming. And more than anything, I hope to amplify people who can or have the ability to shine. Last but [04:23] notion. I'm even more grateful to those of you who have listened, watched, read, whatever, found a way to support me. I feel so lucky. I hope I am doing you a service when you spend your time here listening to these conversations. I hope you go

[04:35] take your ideas and turn them into things. I hope you do it with craft. I hope you do it with soul with that. I will, I will turn it over the episode, but thank you so much. And, and I'm so excited to continue to share dialectic with you. [04:45] Welcome to Dialectic with Rio Liu. Rio is the head of design at Cursor. Prior, he was a designer at Notion, working across so many different projects and features, including Notion AI, for about five years. And he was a designer at Stripe and Asana. He grew up between China and Montreal, and now lives in San Francisco, where he's focused on building Cursor and helping anyone create software. [05:09] We talked extensively about his design philosophy and how he is constantly moving between simplicity and complexity, bare material and abstraction, and why, in his words, so many of these ideas and these patterns are all the same thing. [05:26] We also talk about how design is changing, where in the past, using tools like Figma, it felt more like painting or drawing. Now much of Rio's design feels more like sculpting clay or finding David in the marble. [05:39] So much of his philosophy is about getting closer to the material. And in the case of digital things of software, that is working with code. And that's why I think why he's so excited about cursor. [05:49] The line between vibe coding and real engineering is also, I think everyone's feeling that it's flattening. And there's no better example of that than Rio's personal project, Rio OS, which you can find on his website, which is essentially a nearly a full on operating system of apps and games and simulations. You can talk to Rio's agent. I've watched him literally make games and new apps for Rio OS in Rio OS. And in some sense, it's entirely vibe coded. He's built it using cursor. And what's

[06:16] I think so outstanding about it is that it's quite literally the opposite of AI slop. It is so deeply personalized and has so much soul. It feels so much like Rio. [06:25] So we talk about how he is iteratively designing both his personal projects as well as all of the design decisions he's making at Cursor and helping more and more people across the team work with him in a range of different ways. This is definitely a philosophical discussion. Much of it is about designing things that feel true or even inevitable. [06:45] But in many ways, I think Rio is also an amazing example of somebody who is doing a lot more doing than thinking. And so I think that marriage together makes him so effective. And I hope and I think we we really dove into that today. If you already make things, especially software, I hope you are inspired to be all the more. [07:05] willing to try things, to be more flexible, be more dynamic, and expand the boundaries of what you can personally do. And if you feel like you could be making more things, [07:13] I hope you are inspired not only to try tools like cursor and make software, but to apply some of this philosophy to making any range of things. Um, I just so love, um, [07:25] the way Rio thinks about getting up close with material and how, [07:29] learning with material, getting feedback from it is how we design anything. It's addictive. It pulls us in. And... [07:37] in The Limit, we end up making things that [07:39] other people get to enjoy. I hope you enjoyed the conversation as much as I did. With that, here's Rio Liu. [07:46] Rio Liu. Okay, let's go. We're here. Thank you for being here. I'm really excited about this. Yes. We're going to start with a...

[07:54] I guess what you could call a catchphrase of yours, which is you love to say it's all the same thing. Yes. What does that mean and what does it tell us about design? [08:04] Hmm. [08:06] Thank you. [08:08] It's like when you look at all the apps you use, [08:11] where even like everything around you... [08:14] even looking at ourselves as like humans, as like life forms, we are always... [08:20] Tot de volgende. [08:21] It's almost like with the same parts... [08:23] Thank you. [08:24] that are really simple. But when you merge them or combine them [08:29] Reconcline them. [08:32] they give a rise to complexity. [08:35] Um... [08:36] Like the most fundamental elements... [08:39] are the same. [08:40] Thank you. [08:41] Like a lot of the concepts that we use... [08:45] You know, regardless if you call it like, this is a task management thing or like a document thing. [08:52] They're all just... [08:53] Like information organized in databases. Yeah. [08:59] that much difference. [09:01] And there's always... [09:04] Like something at the core that is like the simplest form of the thing itself. [09:09] Thank you. [09:10] And this most likely thinks that... [09:13] you've seen before, or... [09:15] There's like analogs in nature where like patterns... [09:19] When you talk about those simple things, [09:21] Are they abstract things? Like, are they, as you say, are they patterns or like metaphors or sort of like ideas? Or are they, can they be...

[09:30] Also very concrete. Oh yeah. [09:33] I think they can be very concrete. [09:36] And it's like the same thing. [09:38] manifested at different levels, different levels of abstraction. [09:43] Okay. [09:44] So you can think of maybe like, ah, these are my core ideas, but then how do I say... [09:51] visually represented in like this constraint 2d space which is like a screen [09:56] like a phone, or like you stretch it to like a window, then you have more space. Then what are the things that should be shown? Like what are the relationships between them? What are the more important bits that... [10:09] that you want people to get in. Like, it's almost like... [10:13] It's like a multi-floor apartment. Yeah, yeah, yeah. And then you want people to go to the lobby on the top floor with the best view. They can kind of see everything. Ah, this is cool. Now let me go to the place I want. [10:29] That's more like for the users, but the same thing applies to say, like, you're designing UI, you're designing some flows, you're designing how the data model works, you're, like, conceptualizing how do I, you know... [10:41] make this into like a big scalable distributed system, [10:46] And when you're operating on all these layers, they're still like just... [10:51] manifestations of those core concepts or ideas. Then you keep everything [10:59] Together. [11:00] And they feel cohesive.

[11:02] when a lot of people... [11:05] Maybe they think of these things as separate things. And then they treat them as like, I need to do this box first, and then do that box first. And then each people doing the boxes don't talk to each other. Then they build something that's kind of [11:20] It's like... [11:21] Google. [11:21] Wiggles. [11:23] Yeah, it doesn't have the connectedness. You have a little essay you wrote about complexity coming before simplicity. One part you say, it's like a swan serene on the surface, but paddling like hell beneath. Yes. Which is an amazing metaphor. Why does complexity actually have to come before simplicity? Yeah. [11:44] I do think, say, conceptually, it is possible to say... [11:50] These are the core building blocks of my world. [11:54] And that's it. Let's just go. [11:56] Yeah. [11:58] But... [11:59] It needs to survive in the real world that we live in. There's people who don't come here to look at your essay or look at your academic idea of like, ah, these are the ways we need to connect these computer ideas. They're here to do something. So they come here. [12:19] They should ideally do the thing [12:22] they want to do first. Without thinking too much. Yes. Without thinking too much, they can do it. They can actually like... [12:29] slowly master it, configure the thing, customize it, then they kind of know what is in there.

[12:37] You can do it from both ends, and they kind of are... [12:42] It's like two sides of the same coin almost. But a lot of people, they only see one side. [12:48] Say like we do a lot of like [12:52] user-centered design or like [12:54] You know. [12:55] Let's start with a user problem and then decompose it or do some research, look at some numbers. [13:02] figure out if solution A, B for this problem 1 [13:07] Which one is the best? Ah, A is the best. Oh, let's just do A. And then you keep doing this A, A, A, A, A, A, B, B, A, B, B, B. And then now you have a platter of random choices. [13:19] And then they don't connect. And then they're all like discrete buttons on your UI. Yeah, yeah, yeah, yeah. And that's kind of crazy. [13:26] When fundamentally maybe all of these ideas are the same ideas or maybe they are like... [13:32] better versions of your original ideas, like a remix version of that, or like a reconfiguration of the thing. [13:39] - Yeah, you're sort of seeing both the swan, both aspects of the swan at the same time. You're seeing the elegance in the kind of-- - It's like you need to test your model [13:47] With the real world examples and people. And then as you do that, you figure out, hmm, this part of the system is a little weak. I need to make it better. Yeah. We're like, ah, maybe we really need to add this new thing. [14:02] then we should probably do it. Because a lot of people need it. [14:06] But if you're just conceptualizing yourself and you're kind of in your own world thinking and you're just like,

[14:13] ideating [14:15] then you're not really doing anything. [14:17] Like you're not helping anyone. You're just... You're untethered. Yeah, you're just like... [14:22] I don't know. [14:23] Thank you. [14:24] Having fun yourself, I guess. Another line from you. You say, the universe is fundamentally modular. Simple rules, endlessly recombining, creating emergent complexity. Design is the human practice of participating in that process consciously. We look at the world, identify the patterns, extract the rules, and use them to build new realities. [14:46] much of this sort of, it's all the same thing inside of that. I'm curious, maybe at the most zoomed out level, like what... [14:55] What initially drew you to what you describe as design there and what kind of keeps you coming back? Like, what is it about this? [15:02] almost like philosophical approach to the world that's so compelling to you. [15:07] Hmm. [15:09] Okay. [15:11] I did not come here like... [15:13] you know, [15:14] When I started, I did not know... [15:17] the difference between even like engineering or design or products or anything. Yeah. [15:24] I just saw this [15:27] things that were made [15:28] by people [15:30] Like, I started... [15:32] playing with software when I was a kid, I would get these pirated CDs... [15:38] And then they're almost like... [15:41] Software subscription packs monthly. Like they get...

[15:45] You just load them on your PC, and then you play with all the new apps. And then I started playing with, like... [15:53] or the office tools... [15:55] that [15:55] like all the fonts, Excel, PowerPoints. [15:59] Photoshop, video editing things, 3D making things. [16:05] programming tools, [16:07] starting making websites and stuff [16:09] And as you do these things, as you make things, you start to realize... [16:14] Like... [16:17] The end output of what we do is just code. [16:20] Okay. [16:21] But there's like a lot of different depth... [16:24] in all the layers. [16:26] Um... [16:28] And if you're curious enough, you can go to every layer. [16:32] really deeply. [16:34] Um... [16:36] But the more you do these things, like make more websites for different kinds of people or make... [16:42] different apps for things, you realize [16:47] Like a lot of it is just the same ideas. [16:51] And then you also can trace it back to history. [16:54] Thank you. [16:55] like when you look at people when they started this or when they were just [16:59] Again, like... [17:00] ideating, things were not real because things weren't ready. [17:04] but the ideas were there [17:07] And then all you're doing is like remixing the idea, repackaging it a little bit. [17:13] And then you want to find out.

[17:15] What is the core essence? [17:18] things that you cannot remove, that will always be there, and then you keep making those better. [17:25] You used the phrase, things weren't ready. [17:28] Yes. Obviously... [17:30] technology [17:31] design applies across disciplines. Technology is an area where design, you actually are dealing with that sort of the rate of progress, right? [17:39] I'm curious, especially maybe now since what it – [17:41] you have this great future site you made for Cursor, where you're listing the kind of arc and the lineage of computing. We're in the middle of-- [17:49] an immense amount of readiness, you could say. But I'm curious what your relationship has been like to [17:55] things being ready [17:57] or maybe not ready, even let's say the last two years with AI models and cursor. Yeah. Yeah. [18:03] Thank you. [18:04] Yeah, there's like the technological level of whether it's ready. Right. But there's also the conceptual level of whether it's ready. It's like... [18:17] For example, a notion... [18:20] even though [18:22] technologically as like, [18:24] Everything is kind of fully ready. [18:26] Like Notion itself is almost like just databases in the cloud. [18:29] And then you can do live editing with people. You're just manipulating, like, blogs and databases. Like, the ideas have existed for a long time. Right. But then people have not caught up or people are not familiar with these ideas. Yeah. Then it's, like, still, like, kind of foreign to people. And then, boom, AI happened. Then it's almost like using this new primitive new technology, we can actually, like, help people understand better.

[18:57] or make translations of... [19:00] ideas. Yeah, it's bridging the conceptual gap. Right, right, right. You can use that to bridge the gap and [19:06] Basically, instead of people making databases manually, they have to learn about... [19:14] coding is like [19:16] There's so many layers and there's so many dependencies in order for you to do a running program. You need to know so many things. You can actually reduce that to like... [19:28] Nothing. [19:29] Thank you. [19:30] But then it's like people kind of start from... [19:33] the other end. They get some output, they play, they... [19:37] They tweak. And as they do that, they learn. [19:40] Instead of like... Backing into it. Right. Instead of doing it in the reverse. That's like... [19:46] We are... [19:47] fundamentally the limiting factor. [19:50] like us humans, like, like, [19:52] Our brains can't process too much information. We can't hold too many concepts in our heads. Yeah. [19:59] then [20:01] Like what we're doing is... [20:05] You're simplifying the amount of information or ideas that you're giving to people. [20:11] It used to be like designers have to do it. [20:14] The thinkers have to do it. The inventors have to do it. They're thinking about what is the simplest configuration of the thing, what are the parts... [20:21] But now it's almost like... [20:23] A lot of it can be... [20:26] handled by the AI, then you can reach to like

[20:29] Lower level primitives were even connected. [20:32] more things. You can pull in more complexity. Yeah, yeah, yeah. But then the presentation layer can still be simple, and the simplicity can be more subjective. It is not designed by the designer. It is actually to you, the person using the thing, or you're doing this thing. [20:51] the ideal configuration for that thing. AI can kind of... [20:56] Do the translation. Yeah. [20:58] there's [20:59] We're talking about simplicity. There was another comment you made that [21:02] um, [21:04] is very similar to something you wrote about making things true. Um, and I think truth and simplicity next to each other seem interesting. You say design is the practice of seeing through the surface of things to understand their underlying structure and then rearranging those elements into new forms that didn't exist. Design is philosophy because it forces you to ask, what is this thing really? What are his essential properties? You talked about that. What can I remove before it stops being itself? And once I understand that, what new things can I build? [21:34] Thank you. [21:35] I think I have a sense, and the listener probably does too, but what is, maybe not what is the difference between truth and simplicity, but what is it, maybe even what does it feel like? [21:45] when you're designing and you're approaching trueness or truth. Oh, yeah. Yeah. [21:52] Yeah, it's like you... [21:55] Yeah, the thing is, I think... [21:58] I believe...

[22:00] there is actually like an ultimate... [22:02] solution, given the amount of [22:06] this space and the constraints. [22:09] and the things you know. Yeah. But the problem is you never know everything. And the things always change. So it's like, maybe it is the ultimate solution for this point in time, for this condition, but then... [22:22] Maybe tomorrow it's not true anymore. [22:28] Thank you. [22:30] There are always, like say when you're doing a product or making software, like a set of things that don't really change. [22:38] And... [22:40] It is so important to figure out what those things are. [22:43] those are almost like your fundamental building blocks or ideas of the... [22:49] the software [22:51] deep. It's like I see software as just like a tree of concepts. [22:56] And you package it up, give it a name And then give it a UI, put it out [23:02] Are those concepts changing a lot or are they changing very little? [23:04] Like... [23:06] Most likely they don't change. Okay. Or... [23:09] It is really hard to change them, especially the ones that are core to the thing. [23:14] For example, I worked at Asana. [23:17] Asana is basically projects and tasks. And everything revolves around it. Every data model is kind of locked in there. [23:26] And then, for example, it will be hard for Asana to expand into, like, whatever.

[23:31] But then it is easy for Notion to do that because Notion is building blocks on the... [23:37] And the underlying abstractions are more flexible. Yeah. And then they actually don't change that much. Yeah. [23:44] All you're doing is... [23:46] Like you're fixing some problems with how they connect to each other. Or now there's like a different kind of data that we can present better. What are the better views for that? How do people like, you know, combine these things so that they can... [23:59] do a lot more crazy things. How do you help people? [24:05] instead of them building this thing, maybe the AI agent does this thing. Um, [24:10] And say for cursor, it's like [24:12] that common layer is even lower, which is code. And it's so generic. It means you can actually do anything. [24:20] Is truth universality? [24:22] Is it the same thing? Kind of. Or like, it's like... [24:26] Given this constraint... [24:28] what is that ultimate answer? Or what is that simplest configuration of our system that does everything? [24:35] the most beautiful... [24:38] state. [24:39] You have another idea about inevitability. You said the best future solutions seem almost retroactively inevitable. The philosopher who said that the truth is what never had to be said might as well have been talking about a product so perfectly aligned with its context that no competitor can propose a simpler alternative. [25:00] is...

[25:01] - Thank you. [25:02] That, I mean, it obviously connects to the truth and the universality of [25:08] Maybe really what you're pointing to there is what you said earlier, which is there actually is some objective... [25:14] Final... [25:15] at least final for right now form. Mm-hmm. [25:19] How do you design towards inevitability? Yeah. [25:25] Yeah, you kind of project... [25:27] It's like you always design... [25:30] There's a set of fundamentals that don't change. [25:34] And then there's like an ideal future that you want to go to. [25:38] Then you figure out... [25:40] What are the... [25:41] deltas between that. That future, sorry to interrupt you. [25:46] You could certainly think, take Notion example. Yes. We are going to take a really, really simple set of very flexible building blocks. Some of that... [25:57] You, when you were working on it five years ago, or Ivan, when he was working on it 10 years ago, may have had some sort of future conception. I've seen some of the early decks Ivan had. There's crazy stuff in it. It's amazing. But on some level, of course, he didn't fully know. And so I'm curious how important it is for the specificity of that inevitable future outcome. Right. It's more like... [26:18] It looks retroactively [26:20] inevitable. But when you get there, it's very ambiguous. Like you actually don't know. Like you start with you actually don't know. And then you're, you're looking at what do I have?

[26:33] Hmm. [26:34] What do I want to do? Or like, you know, my future state, my ideal. You can just imagine it. [26:41] Like, [26:41] Don't limit yourself. And then you start thinking, hmm, [26:46] Maybe there are [26:48] These kind of big changes I need to do, these are the little steps that I need to take. The closer you are to the present, the clearer the step is, the further out, the muddier it is. But then the only way you can start doing or start... [27:04] going towards it as you [27:06] Do things. You build, you know... [27:09] Stups or I kind of like [27:13] Thank you. [27:14] say like prototypes were like pieces of it. And then as they... [27:21] Get built, get used, get feedback. [27:24] you kind of clarify the thing and you move forward. [27:28] obviously a lot of this is philosophical. Um, [27:32] Someone might listen to this, and this combination of complexity and simplicity, it's really appealing. Most designers, most people making things along a long road... [27:43] are forced to compromise somewhere along the line. And so it almost feels like maybe one of the things getting in the way of getting to trueness or inevitability is practical compromise. You're also very practical. You're sort of just pulling this thread in many ways. Yeah. How do you sort of fend? I'm sure there are a million compromises Notion could have made along the way. I'm sure there will be many compromises Cursor is faced with. Yes. How do you relate to that?

[28:06] Yeah, it's like... [28:09] I don't want every single thing to be perfect. [28:12] I [28:12] Or like there are certain things that are like say... [28:16] they're actually okay to be... [28:18] a little divergent or [28:21] Like you kind of let it go a little bit. [28:24] Let it roam a little bit. [28:26] And then... [28:28] see what people feel, see how the thing does. [28:32] And then... [28:35] You're on this constant loop of... [28:39] re-examining what you have in your system. [28:42] um [28:43] all the things you add. [28:46] See how they're perceived. [28:48] Thank you. [28:48] And then you're trying to [28:51] Maybe now we need to unify these things together. Maybe now we need to clean this part up. And then once you do that, then you maybe open up... [28:59] Boom. [29:00] I don't know. [29:01] this amount of like people can use it now or you make this [29:05] Thank you. [29:05] part of the experience better. [29:10] And it could... [29:12] It's not a future-level thing anymore. It's more like... [29:17] all these things together... [29:19] Because they make a better system [29:22] because the system is more flexible or extensible. [29:26] And you also like... [29:28] increase its capabilities, then that can do a lot more for a lot more people. [29:35] And it's not just about...

[29:37] Yeah. [29:38] Uh, let's make this feature A and then see how it does and then run some numbers on the, I don't know, like adoption, retention, whatever. It's, it's, I, it really kind of feels like it goes back to the swan. It's like, um... [29:52] or maybe use another metaphor, it's like, [29:55] you seem to be constantly taking stock of both, like, what is this pixel and also what is the picture of the clock. You need to, like... [30:03] Go... [30:04] around these layers of abstraction. Yeah. If you really want to make something truly simple. It's like a lot of people also think, ah, simplicity is about removing things, or let's just get rid of all the... [30:18] I don't know, any feature that gets used less than 5% of users. And then you're like... [30:26] Removing something that maybe the 0.1% power user really loves and depends on. [30:33] Maybe the better way is to just like... [30:36] You just... [30:38] Marie Kondo it. It just cleaned up a little bit or reorganized it so that most people get the... [30:46] Thank you. [30:48] the most easy path. But there are still, like, little pathways for others. You don't have to take things away. You just... [30:55] Tuck them away maybe Or like you build like elevators [31:00] What do you say to, it's funny you bring up movie condo. [31:04] I think for many people that's very aspirational. For other people they're like, "How unrealistic. She doesn't live in the real world. She spends all her day cleaning."

[31:14] You've written and talked about minimalism, which maybe is a little bit [31:17] I think minimalism may be... [31:19] People take it too far, it gets a bad rap. [31:22] How do you relate? [31:24] It doesn't seem to – you present – you're very refined. You clearly care about aesthetics. And yet, Rio OS, like, it has, like, a little lived-in messiness almost. Uh-huh. [31:38] I don't know what my question is there, but like, do you... [31:42] and [31:42] How do you have that sort of tidy, thoughtful, careful, and also like alive – [31:49] in a system. [31:51] I think it's like a lot of people think these attributes is like... [31:57] you have to have this or this, when you can actually have both. So like... [32:04] Should it be simple or should it be complex? Should it be flexible or should it be rigid? [32:11] Um... [32:12] Thank you. [32:14] To me, that's almost like... [32:16] Because software is... It's almost like a life form. It's like it runs. It can... [32:22] mutated, it changes itself. [32:25] You don't have to be... [32:28] So, opinionated. [32:29] Like your opinion is actually taking the stance of I don't have... [32:34] Too much opinion. [32:35] I [32:36] But you always make things start really simple. [32:40] Um... [32:42] And then you let people play with it. You let people discover what they want or

[32:48] the way to do things. [32:50] what is their way to do things. [32:53] That's not my way. [32:56] I don't want to force my... [32:59] Like my way of thinking, or this is how you do it, one, two, three. [33:03] unto you. [33:04] I [33:05] I just kind of give you [33:09] Like... [33:10] pathways and elevators and the tools to do the thing you want. [33:15] Yeah, you have a line somewhere you say, no point solutions, always spectrums. Yeah. Which I think captures that. Yeah, yeah, yeah. It's like... [33:22] Like fundamentally all these tools are the same things. [33:26] So, [33:27] Like if you're... [33:29] Okay with that. [33:31] then you don't have to really pick, like, ah, do I want to do this, like... [33:35] Cursor for salespeople. Or a cursor for... [33:41] Coding. [33:42] It might be the same thing. Mm-hmm. [33:44] Thank you. [33:45] I want to talk about that kind of process of making. And you started to get a little bit, you have this metaphor of sort of like sculpting or finding what's in the stone that is really powerful. That's not totally intuitive for how people think about creating. Yeah. You say there's a quiet, almost mystical art to starting with something so unrefined that you're unsure if it's mud or marble and patiently revealing its shape until others recognize its beauty. In the end, they'll say, of course, it's so obvious. Yes. Yes. [34:14] Why can't greatness... [34:17] Why must it be emergent?

[34:20] Thank you. [34:23] Because you haven't seen enough, you haven't tried enough. [34:26] you think this first idea [34:29] I have this perfect. [34:31] And then you throw it out there. [34:34] And they realize, hmm... [34:35] Maybe only I think like that. [34:40] Maybe people like it, but they don't really understand the words or the nuance in there. [34:48] then you need to keep tweaking and keep getting input. [34:51] and [34:51] It's like... [34:53] You never start with something that's... [34:57] like the [34:57] The ultimate answer. You always start with shit. [35:02] And then you make a better and better. Is that the case for every medium? I think so. [35:06] Like even when you're painting. Yeah. Yeah. [35:10] You start with like the pencil sketches. [35:13] And then you layer on top the paint. Or like you're sculpting. You start with just... [35:19] like a blob of clay. And you're like... [35:23] making the high-level shapes good enough, and then I start... [35:28] Working on the details. It's the same thing. [35:30] Like you never... [35:33] You never get... [35:34] The first shot rate. [35:36] Even more true with like AI. [35:39] Yeah. Um, [35:40] But with AI, it's like... [35:42] Or like, say, with Cursed Composer 1, because it's so fast, it's like... [35:49] It's a different way to do things now.

[35:52] Like you're building as you're seeing things, as you're thinking. Right. And as you're designing, and it's all... [35:59] together. Yeah. I wonder, like you, you referred to software earlier as almost like an organism. Um, [36:06] And maybe that's something that's true about software inherently, but it feels especially true with AI now. [36:14] And one of the things you said to me when we met, you talked about sort of how you used to work being much more like painting or drawing and now much more sculpting or finding something new as stone. Yeah. [36:25] That way of thinking is intuitive to people, even people who make software. [36:30] And so, [36:31] maybe one question I'd have would be like, [36:32] Have you started to think about it in a fundamentally different way with AI, or is this actually just a continuation? [36:39] I think it's almost like going backwards... [36:42] It's like I started building things myself and designing everything... [36:48] A lot of times I do not use like... [36:50] Pixel tools. I just coded it. [36:53] And then... [36:56] I became like a professional product designer. Yeah, capital D designer. Yeah, yeah, yeah. And then... [37:02] Oh, interesting. Now I just make mocks and fancy animated prototypes. And then I would drop that mock into my PMS PRD. And I'll wait for things to happen and things don't happen. [37:19] And then now it's like going backwards, meaning like,

[37:23] I have an idea. I'll just prototype it out. [37:27] Like a kid. Or like, uh... Like a kid with a piece of glass. Oh, yeah. Or... [37:30] Oh, there's a bug. [37:32] Okay, I'll just... [37:35] make a screenshot... [37:36] And then circle the thing up. [37:39] Add cursor, fix this. And then it'll get fixed. It's like... [37:44] Instead of waiting, instead of getting stuck in pictures or words. [37:48] You actually make the thing. [37:51] where you use software or use code as a tool. [37:55] to communicate your ideas. [37:57] better. [37:58] And because we're software makers... [38:01] The best tool is code. [38:04] I interviewed early on, I interviewed a couple of designers, like industrial designers, physical designers, Sele and Taylor. And one of the things that they feel really strongly about is they hate renders. It's like make the prototype. Oh, yeah. And I almost feel like this is the digital version of that. It's like get it down in the metal code. Exactly. You have to play with the material. [38:24] Like our material as software makers is never the pixels. [38:29] It is the code itself that renders the pixels. [38:33] Thank you. [38:33] Yeah. [38:35] Yeah, you have a line I love. You say, but it existed, and because it existed, it could be improved. [38:42] which so captures the like power of working with actual material. Uh, uh, [38:48] I do wonder, like... [38:52] You...

[38:53] When we were first talking, you said ... [38:56] I use Figma when I want to go into my old way of thinking, which obviously relates to what you just said. [39:02] . [39:03] Thank you. [39:03] I'm curious today, and maybe part of it is that you're designing [39:07] cursor, which is [39:09] especially conducive to it's less about the pixels already. Mm-hmm. [39:14] But [39:14] When do you find yourself sort of like, [39:17] Tempted towards the old way of thinking and like is it a yo-yo is it a? I [39:21] will you be using Figma at all in a year? Oh, yeah. [39:25] Thank you. [39:26] It's like they're just tools... [39:29] And... [39:30] Like sometimes we think in words. Sometimes we think in pictures. On podcasts, we definitely think in words. Yeah. [39:38] Making videos too. [39:40] Some people do that. Yeah. [39:43] Or like slyes or whatever. [39:46] Like those are just, you know... [39:50] Different artifacts were like forms. [39:52] to help us think. [39:54] And I think... [39:57] Like I don't want to take them away. [40:00] Like, [40:01] Different people have their preferred form. [40:05] To think. [40:07] Maybe some people are more like linear. They just write text. Yeah, yeah, yeah. [40:12] I like Bullitt's [40:14] I think I got the disease from Notion. It's like... [40:19] All I do now is I go out [40:22] And the Night Walk.

[40:24] I have ideas out [40:26] Open Notion Dock and the [40:29] But in a list. [40:31] and then once I'm done with my walk, I'll go back. Ha, maybe now draw some pictures. [40:39] Then maybe I'll do Figma. [40:41] Because it's so... [40:44] Because I've been doing those for so long. It's like water to me. Like I don't think. Yeah. When I make more artboards or when I do the Figma like shortcuts. So when they change shortcuts or like they move around my things, I get mad. They keep doing it. I saw you were really mad that they had changed the. [41:03] checkbox. Oh, for notion? To-do box, yeah. Oh, yeah. That's for another thing. [41:09] That's more for like... [41:11] It's like I feel like... [41:12] Thank you. [41:14] Like every piece of software is almost like a person. It has a vibe. It has like a... [41:19] history has some [41:22] Character essence like you don't want to lose that yeah, you don't want to water everything down to like a [41:28] border radius for pixels. [41:31] Sometimes it's good to keep that. Keep a lineage and [41:36] Keep a thing that's maybe a little weird. [41:40] But it's so characteristic. Yeah. [41:43] On the note of sort of your thinking time, [41:46] And you talked about thinking and... [41:48] using different tools, you're thinking using Figma, [41:52] you've talked about your walking and the value of this sort of idle time, the space between...

[42:00] thinking time isn't wasted time. Hmm. [42:02] Are you... [42:04] and maybe this is running against what you just said about it feeling like water, but are those like different modes? Like watching you use cursor, at least on your phone when you're hanging out, [42:15] it didn't seem like you were doing very much thinking. You were just like, you were just throwing like paint at the canvas. Right. And then when you write about your walks or like that, that feels like a very structured. Is that maybe a template for how? That's more for the longer term things. Ah. Yeah. Or like vague ideas, ambiguous. Dreaming. Or like, ah. Ah. [42:34] Maybe we should do this. I'm not sure. Maybe we should do it this way. What are the components in there? How do I break it down? [42:42] What are the things people care about [42:45] da da da da da da da da [42:46] Whereas when you're using Figma, you're using Cursor, you're... Those are more for maybe like Figma... [42:52] Just like... [42:53] Thank you. [42:55] There are still some, say, like difficulty where... [42:59] Let us just take [43:00] It just takes more time to say, build a really crazy prototype in code. [43:07] If you want to just communicate ideas into the space really quickly, draw some pictures. That's fine. [43:12] And then when the thing gets... [43:15] to the state where... [43:17] I think I know what it is. [43:19] I want to figure out how they fit together, how they work together, what are the... [43:26] you know [43:28] especially with building AI stuff, there's so many...

[43:32] Like... [43:33] both like procedural and like non-deterministic things that you need to think about. It is really like really hard to simulate in Figma. [43:42] Or like in... [43:44] Static pictures. Yeah, and you're not with the material. You're not up close to the material. Like you actually need to [43:50] glue it up, and then see how they fit together, see how the states transition. If I get this, like, error, what happens? Or if... [44:00] The return gets too long, what happens? [44:03] Like you'd never get that in Figma. [44:06] I want to talk a little bit about real OS, both because I know you're very obsessed with it. And it does feel like the perfect embodiment of this sort of working with clay. [44:20] And I think it's, I would strongly encourage people listening or watching to go. [44:24] to poke around with it, um, [44:26] As I understand it, RyoA started as a soundboard app you made for your friends when you were leaving Notion. And it sort of feels like... [44:34] it's this just infinite thread you keep pulling or this piece of clay. You just kind of keep turning over in your hand. Yeah. Um, [44:41] For people's contact, when we first met, you had your phone out, and we were just talking, and you were literally making apps as we sat there and talked. [44:51] What have you learned about [44:55] making things and maybe even about yourself from this crazy project. Uh-huh. [45:01] I learned that

[45:03] Oh shit, I can't do all of this? [45:06] I think that's the biggest thing. [45:08] And it's like... [45:10] Thank you. [45:14] It's all like little ideas piling up on each other. [45:17] Um... [45:19] You start with like... [45:21] something simple small [45:23] Um... [45:26] And you just keep building and building and building and building and see it grow. [45:31] And then when it grows to like a size where it's like... [45:35] Thank you. [45:36] There's some constraints. I actually started the thing in V0. [45:40] Not cursor. Like the soundboard thing. Yeah, yeah, yeah. [45:44] Um... [45:45] Like I ran into some errors. Then I'm like, hmm, I need to do it in cursor. [45:50] Thank you. [45:51] And you used cursor much prior to that? [45:53] Not really. I tried three times, I churned three times. [45:57] Oh, interesting. Why? It's like the first time... [46:01] It was like... [46:02] Oh, cool. New code editor. Let me try it out. [46:06] I type some lines. It completes like five lines of code instead of one line of code versus like GitHub Copilot. Then I turn. [46:15] Because you felt like it was trying to do too much. No, it's like, it's just completing code. [46:20] With more likes. Yeah. And the second time it was the chat. [46:25] Yeah. [46:26] It's like chat GPT next to your code. It can read the code. It can answer some questions.

[46:34] I can't do much. [46:35] So I tripped. And then the third time was like... [46:40] discovering the agent. [46:42] Is this post using V0 or pre-used? After. Okay. [46:46] So [46:47] I needed some tool that can let me do anything. [46:52] Then I found cursor. [46:54] - [46:55] and I'm like hooked. [46:57] Yeah. [46:58] Thank you. [46:59] And you start from simple things and then... [47:04] You just ask them maybe a little crazier idea. [47:07] And then... [47:09] you see it getting built. [47:12] And... [47:14] So now with PlanMo, you actually see how the models think. [47:19] and you can change. [47:21] You can be part of every step. But it's still your clay. [47:25] Yeah. [47:26] But it's like the model now handles all the parts that I don't really care about. [47:31] I actually studied computer science because I love... [47:36] computers and software. [47:39] I think that's a good thing. [47:40] But I hate it. [47:41] writing code or like all the algorithms and stuff we learned that's like kind of useless and yeah [47:49] What I care more about is like... [47:52] Like what are the ideas... [47:54] How do people, you know, [47:56] Feel. [47:57] How quickly can I make this thing I thought of? [48:01] Exactly. It's like the thing, the idea, the concepts, I want to play with the concepts. Mm-hmm.

[48:07] You mentioned it. Like, real OS... [48:09] It doesn't really seem like something like that could be... [48:14] should be able to be built by just [48:16] throwing more paint at the canvas. Like it feels like the type of thing that should have needed to be more planned. There is a lot of things that say like it's not just throwing. [48:25] So it's almost like it's a constant throwing things and cleaning up shit. [48:31] Okay, same one. It also happens there in ReOS. What is the cleaning up? That's what we're not seeing, I think. Yeah, you don't see that, but you can see it in my commit logs. The maintenance. Yeah, it's like... [48:42] The more things you add, the more things you realize, hmm... [48:45] That's the same thing that I just talked earlier. It's like... [48:49] Haa. [48:50] All these apps need, say, [48:53] Some AI endpoint and some auth. And like they need to store their states. They need to write or read into the file system. Da-da-da-da-da-da. Like... [49:07] Maybe I started, you know... [49:09] Doing the file system part... [49:11] from the TexEle lab. [49:13] but then now I want all the other ones that [49:18] that can use the same ideas. [49:24] to use the thing [49:25] then I need to kind of abstract the system. [49:29] Thank you. [49:29] Like, put that part out or... [49:32] unify some... [49:34] you know, state management things, [49:36] Um...

[49:38] And then you need to kind of refactor your original... [49:42] Even though maybe to the user it looks exactly the same. [49:46] That part of it, though, I think is where, like... [49:50] for lack of more precise language, people get stuck. Again, I watched you use cursor. It's like, you're literally, it's like you're just nudging the model. And your prompts are not, yeah, it's that demeanor for the listeners. You're just poking it. It's not these long specs. I'm watching you just be like, can you come up with an app idea? Like your language is really casual. And so... [50:13] I think to the average person using V0, um, uh, [50:17] or the person who tries cursor and is churning. I think we'll talk about it later. You're very clearly focused with cursor on building for the hardcore user. [50:26] for someone who has somewhat of a computer science background, hadn't written a lot of code, maybe what I wonder about is like, [50:32] in the poking process, you're getting more invested that you care enough to do the hard maintenance part? Oh, yeah, I learned a lot. [50:41] But... [50:42] building real us [50:44] Like the four... [50:47] Even since I became a professional product designer, I would have little projects I do on the side. [50:55] Like... [50:56] Thank you. [50:57] The first few years, I kept doing those. [51:00] And then I got busier or something, and then I stopped. And then every time I tried to go back... [51:05] Oh, shit. I need to learn, like, React 18, Intel, WinCSS, whatever. All of those, like, new things. And then...

[51:13] It takes a long time. I have to read all the docs. I need to understand how people do things now. [51:19] um... [51:20] But it's like... [51:23] Now with the agent, you don't have to do that, but you're still doing that. It's like the agent maybe helps you do the research. [51:32] it comes up with some ha. [51:34] Here are how people do it. [51:36] Now. And then maybe it gives you some alternative options. Maybe you know certain things. You also don't know certain things. [51:45] but the agent can kind of help you find your way. [51:49] And then you can say, ah, now just do this. It will write the code. You can look at the code still. [51:55] You can learn from its output. [51:58] how things work. [52:00] You're getting deeper into the complexity. Intentionally or otherwise. It's almost like [52:08] Just by... [52:09] reading... [52:11] Like a lot of users say this too. It's like they love reading how the models think. [52:17] They actually want to expand everything and then they want to look at all the output because it helps them understand... [52:24] Thank you. [52:25] what the model is doing, gain trust from it, and learn. [52:29] you know? [52:30] Especially when they're starting. [52:32] Yeah, it might be a strange comparison, but somebody I interviewed, he was talking about reading with his seven or eight-year-old daughter and how reading with her, these books that were actually far beyond her ability level, it pulled her in. And now she's reading whatever. I don't know if she's reading Anna or Corona enough, but she's reading well beyond her level. And there is something about sort of like...

[52:56] Yes. Being exposed to someone else's thinking. Yeah. Yeah. [52:58] Even if it's GPT-5 codex or Composer or whatever. Yeah, it's like... [53:03] Thank you. [53:04] Most of the... As you said, most of the prompts that I did in RioOS is just... [53:09] really short, simple things. [53:13] It's like... [53:16] Theoretically we are ready. [53:19] Like, you can actually build a lot of things and you just vibe. But there is, like, you know... [53:25] I'm a little cheating too. [53:27] Because I know things before. Right. So I know, like, when the AI gets stuck, how to, like, get unstuck. [53:36] Thank you. [53:37] As I play more [53:40] So my full-time job is to... [53:42] play with all these models and use cursor. [53:46] So I kind of developed some intuition... [53:50] on how this... [53:52] say different models behave as I make it. [53:57] What are their limits? [53:59] Maybe this one's faster, this one's slower, this one's smarter, add certain things. Um... [54:05] that a lot of people... [54:07] Like... [54:10] They don't know. [54:12] They don't really know what to do yet. [54:15] Thank you. [54:16] So that helps me like put this back to the... [54:19] to the tool [54:21] On that last note... [54:23] When is it? [54:25] Your job.

[54:26] as the desire or maybe a better way of asking, when is it cursor's job to try to solve those? [54:31] things versus the model's improvements job to sell those things? I think it's both. Okay. The models can kind of raise in capabilities or like say – [54:42] Now the models are... [54:43] getting better at, say, using terminal commands, uh... [54:48] clicking around in the browser, [54:50] Stuff like that. [54:51] It's like as they get better... [54:54] Like, [54:55] you still need a way to unlock those capabilities. [54:59] So you need to fit them back to the twits. I'll package them up. [55:03] make them [55:07] just really obvious... [55:09] Um... [55:10] So people can just... [55:12] play with them. [55:15] They don't have to think too much like how do I... [55:19] Got a little... [55:20] trigger that we're [55:22] Thank you. [55:22] get it out or use this crazy like script or mcp thing to do something yeah um [55:29] Like you start simplifying... [55:33] making things that are possible more obvious. [55:37] Yeah, for more people. [55:39] Ah, that's an interesting way of thinking about it. Yeah. Making things more obvious, making the next step more obvious. [55:45] Yeah. [55:47] It's like you're constantly simplifying, unifying, [55:51] figuring out like now that I have... [55:53] This and this and this [55:55] No. No.

[55:56] How do I, like... [55:58] clean it up even better. It feels like it relates a little bit to the readiness thing we talked about. [56:03] I talked about earlier, which is like, [56:06] It feels like maybe the model's job is the technical readiness and your job at Cursor is the cognitive readiness. Yes, absolutely. [56:13] Thank you. [56:14] like [56:15] Again, humans are kind of [56:17] Thank you. [56:18] We're like single threaded. We've been trying a lot with multi-agent or parallelization of agents. [56:27] And nobody has really solved it yet. [56:30] Because most people are still thinking about, now, let's just give you... [56:36] Fifteen agents. [56:37] Here you go. 15 agents have done all these changes. [56:42] 2,000 lives of changes. It's like all horsepower, no steering wheel. [56:49] figure out [56:51] you know, these, like... [56:54] I'm not even sure if there will be new patterns, but... [56:58] Just like better framings or... [57:02] packaging [57:04] or interfaces... [57:07] For people to just get out... [57:09] get utility out of these things. [57:12] Um... [57:13] Thank you. [57:15] without... [57:18] breaking their minds or like [57:20] changing too much or feeling [57:23] Overwhelmed. Yeah. Yeah. Yeah, you...

[57:26] You've obviously worked on a lot of different types of systems, and you're sort of drawn to almost like this container... [57:33] Type. [57:34] tool or product or something. Um, [57:36] and certainly at least with cursor notion, you have, you have a line where you say systems thinking is essential because the only path to building products that scale, not just technically, but cognitively along the lines of what we were just saying. Yeah. Um, [57:49] Thank you. [57:50] What are the... [57:52] Is the goal when you're designing a tool like that, um, [57:56] to allow the user to stay as single-threaded as possible? And, like, is that essentially what you're designing for? No. No. Or it's, like... [58:06] It's up to you. It's like you need to design the zero state, the one state, and the end state for everything. And then see how they melt together. This is the simplicity of complexity. [58:20] Like when you have N times N times N, it will be kind of crazy. But if you really want to be there... [58:26] So be it. Yeah. You should meet the user where they're at. Yeah, yeah, yeah. Like, if you are actually, like, someone, I don't know, you have ADHD or something, like, you want, like, eight different windows, all, like, running... [58:38] so be it. Yeah. Yeah. Like the average person probably wants, right. Maybe average people just want one main thread. And then it's like how we're thinking right now is like, [58:48] Instead of having you, like... [58:50] you need to review changes from these 15 different agents. Maybe help you kind of cluster them a little bit, organize them semantically.

[58:59] Maybe instead of talking to each of them separately, you just talk to one person or one agent. And it's almost like your PM or your assistant. And then it's going to figure out, ah, these guys are blocked. Do you want to approve the terminal command? These changes, I think they're pretty good. This is bad. You should look at it. [59:19] A very small subset of users want StarCraft, and most people want Candy Crush. Right. I'm fine with both. We can actually do both like a... [59:32] I don't know, a TikTok and a... [59:35] StarCraft. [59:36] Because of AI. Yeah. Yeah. [59:38] Um, there's an idea that I think is really interesting that I think is connected here, which is about slack in systems. You say the best systems have slack in them. Redundancy isn't always waste. Yes. Optionality, multiple paths mean you can explore without breaking everything. The core remains simple. Uh, while layering itself into more complex permutations, controlled chaos means you're stable enough to not collapse, but loose enough to, to evolve. Mm hmm. [1:00:06] I think that's such a powerful metaphor. [1:00:08] And maybe... [1:00:10] Slack is that like willingness to go as complex as I want to. Yeah, I wonder about like... [1:00:16] You have somewhere else you talk about that sort of [1:00:18] chaos and order together. It's like you let diversions happen and you let things evolve. It's like evolution. It's like... [1:00:29] nature is constantly...

[1:00:32] Like making more... [1:00:34] you know. [1:00:35] permutations of the same thing, a little different. [1:00:39] See which one works better. [1:00:41] How do you give a tool more slack? [1:00:45] This... [1:00:47] What does it mean to add Slack to cursor? Right. [1:00:50] It's a little complicated, but also it's like... [1:00:54] Sometimes you just kind of... [1:00:57] All designers were... [1:00:59] We're like kind of perfectionists. [1:01:02] We want things to be exactly what we wanted. [1:01:06] Thank you. [1:01:07] But... [1:01:08] Sometimes you just allow this ugly thing to pop up, or this random button someone else added. And then... [1:01:17] I cannot keep a blind eye on it. [1:01:20] Um... [1:01:21] And... [1:01:22] You let it simmer a little bit. You let people play with it. [1:01:26] our internal group of people [1:01:29] Um... [1:01:30] And then... [1:01:32] As you do that, or maybe people through the first... [1:01:37] Bucket of paint. [1:01:39] And then... [1:01:41] Now that it's there. [1:01:42] You can see it. You can play with it. [1:01:45] you can... [1:01:46] Think about it more, understand it better. [1:01:49] Versus sort of roping off the canvas. Yeah, yeah. Then it's like, ah, now I know how this thing fits with the other things. [1:01:55] where like, ah, this thing is actually like a start of something much bigger. Then it's almost like this constant...

[1:02:05] you know, chaos convergence thing. Yeah. [1:02:10] And it gets into like an equilibrium. And then you want that thing to be like... [1:02:16] almost at the edge of like [1:02:18] the maximum chaos you can allow. [1:02:21] Thank you. [1:02:22] for the thing. Your job as a designer is almost... You're trying to help people like, here is the line, don't cross it. Yeah. And then you're also helping people like, bring this like... [1:02:33] It's like reducing like entropy. [1:02:36] Just... just... [1:02:39] tame it a little bit back. We're like, ah, these, you, you should talk together and then make those things actually the same thing. We're like, [1:02:48] You're making a new thing? Cool. Think about these four things that we have. And that's it. [1:02:54] I'll just let them think about how does this new thing relate to the four things. And then... [1:03:00] Ideally, they come back with a good answer. You're almost like... [1:03:04] You're like the game maker or you're like you're the agent of evolution, like setting the rules of a little bit of what is tolerated. But critically, you're not snuffing things out too early. Yeah. Yeah. [1:03:16] And it's like you're [1:03:17] You're mostly like an observer. [1:03:19] or like [1:03:23] Like I'm not dictating how things should happen. [1:03:28] I just... [1:03:29] tell you like given all the things I know [1:03:32] Here's probably how we do it. [1:03:34] Thank you.

[1:03:36] And this is also maybe why it seems like you're very attuned to not just the different ideas for cursor inside of the company, but all over Twitter, different stakeholders. [1:03:46] students, whatever. Yeah, yeah, yeah. Because you're almost trying to, like, broaden the aperture of what... [1:03:52] is allowed in. [1:03:53] Because fundamentally, it is the same thing. [1:03:57] all the vibe coding tools, [1:03:59] Cursor or the CLI agents... [1:04:02] It's all the same thing. [1:04:04] but only like cursor kind of [1:04:08] tries to bridge all of them. [1:04:11] And... [1:04:13] Like, I try to give people, like, their ideal form. [1:04:19] And I think like when they... [1:04:21] that [1:04:21] Like a big reason... [1:04:24] Cursor got popular as because it looks exactly like VS Code, at least before... [1:04:30] But as we kind of noticed, like people changed their patterns of usage, [1:04:34] people kind of moved from like manual coding at like reviewing [1:04:38] every line and, [1:04:40] To do more agents. Yeah. You have to move with them. Then we just flipped. Right. Right. [1:04:45] Like, our defaults changed as... [1:04:47] The world moves. [1:04:49] And as the product evolves. But fundamentally, it's still the same thing. [1:04:54] but [1:04:55] What is cursor? [1:04:57] Hmm? [1:05:00] Obviously, Cursor is... [1:05:01] a plugin or a skin of VS Code on some... No, not just that. Of course, of course not just that. And it's changing every day. Like...

[1:05:10] Again, at least when we spoke first, you talked about Cursor like it – at least the way you seem to relate to Cursor is almost like it's your little butler – [1:05:20] that just does things for you at your hand. [1:05:24] And [1:05:24] And we talked about code being the universal language. Like in many ways, it almost feels like [1:05:28] Cursor is just this medium to work with code with a computer. Right. Right. And so I'm kind of asking about what cursor will be when I ask what cursor is. Right. Like, do you have a conceptual do you have a metaphor? You like it is a tool, but it's sort of this more. Is it just the agent? [1:05:45] I see it as just like we... [1:05:48] We started from one slice of making software. [1:05:52] which is you're just actively coding... [1:05:56] when you're sitting on the computer. [1:05:58] We put an AI next to it. [1:06:00] so that I can help you write the code. [1:06:02] Thank you. [1:06:03] And now it's like [1:06:06] Like I want Krishna to be... [1:06:10] It's like one place where you can do everything about making software. [1:06:14] And that is not just... [1:06:17] Writing code. [1:06:18] And it's not just. [1:06:20] the developers. [1:06:22] There is like... [1:06:24] the PMs thinking about what to do [1:06:28] How to measure things. [1:06:29] you [1:06:30] aggregate all the data, synthesize it, figure out what are the problems to fix. [1:06:35] Um... [1:06:37] breaking it down into tasks. [1:06:40] There is the designer.

[1:06:41] Maybe they're [1:06:43] trying to kind of you know explore in in [1:06:46] in 2D space. [1:06:48] higher level abstractions [1:06:50] There's the... [1:06:52] Engineers writing the code, but also they need to review, they need to test whether it worked. [1:06:58] once you put it out [1:07:01] You need to like... [1:07:02] gather feedback and input from the market and people using it. [1:07:06] Thank you. [1:07:07] Like all of this is making software. [1:07:10] Especially in a team or a company. [1:07:14] Um... [1:07:15] And now people... [1:07:18] People's workflows and tools and the metaphors they use, the artifacts, are all scattered and disjoint. [1:07:24] Yes. [1:07:26] Whereas I think cursor can actually help everyone. [1:07:30] put everything together again. [1:07:32] and then using the agents. [1:07:34] It's the same agent to help you translate between, say, your form... [1:07:40] of thinking... [1:07:41] your preferred artifact into the code itself. [1:07:46] Then it's almost like anyone who wants to build software or any team, they can just... [1:07:53] Be closer together. Yeah. And then the agent kind of helps them... [1:07:57] Yeah. [1:07:58] They say solving a lot of the issues [1:08:00] that we have today, that were kind of created by all the tools that... [1:08:06] that we've made in the last... [1:08:07] Yeah, we just need one more tool. [1:08:10] You need a thing that kind of melds them fully. What about cursor shape, though?

[1:08:15] People have been trying to build the final tool forever. What about cursor shape? [1:08:21] makes it [1:08:22] what you're describing theoretically possible, acknowledging you're still currently serving mainly devs. Yeah, I think it's like... [1:08:29] Like people joke about like cursor is like a fork of VS code and it's just code editor. [1:08:34] and [1:08:35] But if you look at VS Code deeply, there's actually really good... [1:08:39] Low-level primitives. [1:08:41] For example, [1:08:44] Like in VS Code, there's a concept of editors. [1:08:47] Like you can open different files in different kinds of editors. [1:08:51] Some of them might be looking like the code editor. [1:08:55] Maybe there's like a diff viewer. [1:08:57] Maybe there's like a markdown preview. Maybe there's a browser. [1:09:03] Like just having this... [1:09:05] allows me to just present different things to people differently. [1:09:09] even though underneath it's still the same code. [1:09:12] Is that because it works with files? Yeah. Yeah. [1:09:16] That's another thing. It's like in VS Code, there's a concept of workspace, which is just like folders and files. Maybe they're tied to a repo. It's like a lot of these low-level ideas... [1:09:27] Again, it's like... [1:09:29] They don't have to change. [1:09:30] And I don't intend to change them. [1:09:33] So like... [1:09:35] I don't know if we will ever detach from VS Code at some points. Maybe once we kind of, you know... [1:09:41] Go fool the agent. Yeah. Or at least a lot of the people using Coz. Yeah, exactly. Right.

[1:09:46] but I think it's still like [1:09:50] The challenge for me is to come up with a way to... [1:09:55] Thank you. [1:09:55] So you're tying all of these different workflows and people's preferences together. [1:10:00] together into one thing. [1:10:03] And you're trying to come up with like... [1:10:05] different reconfigurations of that thing, how they transition between these states. [1:10:12] for these different people what do they each see by default [1:10:17] How do they like [1:10:18] customize it [1:10:20] How do they actually talk together? [1:10:22] That's a really complex problem. How do we move from, like, cursor from, like, a single-player thing to, like, a multiplayer thing? [1:10:29] Not sure. Your work cut out for you. Yeah. [1:10:33] On the note of literally using Cursor, we talked about the way you poke it, at least when you're using ReOS. Yeah. You had given me, your advice was treat it as someone who's a little dumb, composing things it's seen before. Yeah, yeah, yeah. Don't expect to come up with full components. You shared a list of 12 rules or tips for using Cursor back in April. Yeah. [1:10:53] So those are almost like two time stamps of advice around cursor. One of those I think that stood out to me is if the code is wrong, just write it yourself. Cursor learns faster from edits than explanations. Obviously, that's... [1:11:06] That works for someone with a coding ability, not without a coding ability. [1:11:10] How often is advice like this changing? Oh, yeah. It changed a lot. Okay. I would say a lot of the things I said in April don't apply. Okay.

[1:11:19] For example, like... [1:11:21] The agents now are so good at [1:11:24] Finding stuff. [1:11:26] that you don't have to say like [1:11:28] at the exact file anymore. Back then, it was like, if you don't include the right context, the agent will just come up with something random or it will make some mistake. [1:11:39] What is the... [1:11:42] Is there anything that stands out? [1:11:45] As long as you've been working on Cursor, that has been... [1:11:48] true consistently? Or even like the type of person who consistently remains good at... Like what is staying the same, I guess is what I'm asking. Not much. Not much. Yeah. So you got to be surfing the new wave. Yeah. Things are constantly changing. [1:12:02] Even the things that appear the same might be replaced under the hood. [1:12:07] That's both exciting, but also back to the, you were talking about, I don't know, it was Notion or something else, like you have a tool you're used to and they change a little. Jeffrey Litt has this metaphor. They change your chef knife. That's hard. Yeah, I guess there are things that don't change. [1:12:25] the agent [1:12:26] So [1:12:27] It used to be like, you know, before I joined Cursor... [1:12:32] There were like five things. [1:12:33] Like there was... [1:12:35] Comment, K, Tab, Chat, [1:12:37] Composer or composer agents. [1:12:40] The first thing I did was to merge the agent's [1:12:44] So Chad, composer, composer-agent, became agent with more specific modes if you want, more specific behaviors. And then the agent, the idea is they're all the same. They're just like...

[1:12:58] apply configurations on top of the agent. Maybe for this agent, it has some custom prompts. It has a specific model set to it. Maybe it has some tools that it can use or cannot. [1:13:10] That's it. [1:13:11] And you give it a name. [1:13:14] Um... [1:13:15] Yeah. [1:13:16] And then these agents all operate on different models... [1:13:19] Those don't change. [1:13:21] They need context. [1:13:23] that don't change. [1:13:24] And then you need to show something. [1:13:26] with the editors, that don't change. But all of these things are changing. Yeah, it's like all the things inside are changing. I guess your bet is how somebody's... How they know that are changing. [1:13:37] So if your bet, too, is that if somebody's playing with the clay, they're okay with change because they are living with the material in a way that... Oh, yeah. You have to. Or, like... [1:13:46] I think [1:13:48] Like, in my career as a professional product designer, the thing I hate the most is like, oh... [1:13:55] Like people want the design to be final. [1:13:58] Where's the final version of this mock? If you don't have it, I won't start building it. [1:14:04] Like, that doesn't make sense. [1:14:07] Because the first mock is never right. Yeah. Like, you have to keep building it. Yeah. Like, now it's almost like the reverse happens at Cursor, which is kind of chaotic, but I'm actually okay with it. It's like our engineers were like some of our like enterprise PMs. They start like vibe coding. [1:14:26] And then... [1:14:27] some weird patterns emerge or...

[1:14:31] It's saying you need to clean it up again. Not architecture. You need, like... [1:14:34] wrangle it back. And then now it's like... [1:14:39] Because AI is really good at composing parts... [1:14:43] Thank you. [1:14:44] I'm actually thinking we need to build bricks. Really good bricks. It's like... [1:14:50] from all the things that we have that kind of suck all the patterns, the core... [1:14:56] Bricks. This is something that seems like you guys did a really good job at Notion, which is like pretty principled about what the bricks were going to be. [1:15:03] Yeah, Notion did it more like on the conceptual level. Oh, you mean like... [1:15:08] tangible feature bricks almost. More like, I don't know. [1:15:12] It's like low-level components up to patterns that people can just reuse that are not just... [1:15:20] Dialogue is different or this view is different. [1:15:23] and [1:15:24] You, you, um... Like you start... [1:15:27] helping people create this [1:15:31] Patterns that just work and just fit together that both humans and agents can, you know, make things a little better. [1:15:40] by not reinventing the wheels every time. [1:15:44] Because the agents, when they're lacking guidance, they have a tendency to do that. [1:15:50] Thank you. [1:15:50] We talked a bit about like... [1:15:52] I think you're clearly designing for hardcore users, even if people are vibe coding with cursor, like maybe the lines are thinning. [1:16:00] There was, I think, a line from you somewhere that I found where you...

[1:16:03] or maybe I made this up, but... [1:16:05] I think you talked about designing for power, to give the user power. [1:16:10] What does that look like, maybe in the context of cursor more broadly? [1:16:15] Yeah, I think a lot of people... [1:16:19] So I don't see your users as like they're dumb. They're not. They can't figure things out. They don't have to be like babysitted. [1:16:28] They can... [1:16:31] It's like... [1:16:33] I want to make things... [1:16:35] The simplest that you can when you start. [1:16:39] But as you go, [1:16:41] Thank you. [1:16:43] you get all the depth that you want. [1:16:45] I [1:16:46] Like... [1:16:48] As a beginner, you get the same tools as what the pros use, just maybe packed a little [1:16:53] Differently. Yeah. [1:16:55] You don't have 18. You don't see everything yet. [1:16:58] But... [1:16:59] I'm not sure. [1:17:00] Maybe this thing that you get can do like... [1:17:04] 80, 90% of what you want it. [1:17:06] Maybe on the other side, like [1:17:08] Currently, I think most people's [1:17:11] My intuition would be that most engineers' relationship is like there's five coding and then there's real engineering. Obviously, that's... [1:17:17] That's the same thing. Yeah. [1:17:20] What does it look like designed for power and for serious hardcore users on the vibe coding dimension? [1:17:27] Ah. [1:17:29] And part of that is conceptual, right? Because they have to be willing to say, I'm going to give up the wheel. Or not the wheel, maybe, but I'm going to let the engine be. Yeah, yeah, yeah.

[1:17:40] We do like little nudges and we change our default sometimes. [1:17:43] And I think those are probably the most powerful tools that you can do as a... [1:17:49] as a product or like a piece of software. Yeah. [1:17:52] And then you want to introduce them in a way... [1:17:56] that people can still get out of it if they want, but you want to show them that, ah, now here's the new world. Here's how you do it. If you don't want it, you can get out. [1:18:06] But... [1:18:08] It's almost like, again... [1:18:09] Okay. [1:18:12] The same thing. [1:18:13] but reconfigured. [1:18:15] Or slightly more optimized for the new way of doing things. There's a little trust there too, right? It's like, actually, if you trust us for a minute, let us show you how much the agent can do. Yeah. Yeah. [1:18:26] Thank you. [1:18:27] Yeah, people like a lot of... [1:18:29] People haven't [1:18:31] felt it yet, or maybe they've tried it before. [1:18:35] But it didn't work. [1:18:36] And then they kind of lost their trust. [1:18:38] Right. And then they never come back. Yeah, they turn three times like you. Yeah, yeah, yeah. And it's like, so it's... [1:18:45] I would say for now... [1:18:49] You can probably do something pretty impressive even on the first shot. [1:18:53] But even say like four a month ago is not the case. [1:18:57] So maybe the first time you tried cursor... [1:19:00] It didn't work. Or it got blocked. Or it did something stupid. [1:19:05] And now you're like, I don't want it. [1:19:07] um

[1:19:10] It's like we need to figure out how to... [1:19:13] Like get the new people in [1:19:16] without too much thinking and setup, they can do stuff. [1:19:20] Okay. [1:19:21] Get the... [1:19:22] existing users... [1:19:25] onto better ways to do things that are more up-to-date. Without feeling like they're both apps. It's like you want to carry them over instead of teleporting them to the new world, and then they're like, ah, what the fuck is this? Yeah. And then there's getting the people who maybe tried Cursor before that thought, [1:19:45] It was not good. [1:19:47] to come back, because it's good now. [1:19:50] Um... [1:19:51] Yeah, there's like work for us to do there. Mm-hmm. [1:19:55] Solvable problems [1:19:57] Many, many problems to solve. Some questions about process and some other stuff. You have this amazing essay about creating something great. A few things in this process. [1:20:08] broader vein. First, like, [1:20:10] I guess we kind of talked about this, and maybe this is silly, but is design kind of just writing now? [1:20:16] Like, it seems like most of the design you're doing, you have your walks, you go on, and then you go to cursor and you write. Yeah, yeah, yeah. [1:20:24] Maybe you write a longer spec sheet. [1:20:26] As cursor improves. Yeah. Yeah. [1:20:28] I do write docs and stuff... [1:20:31] I think it is just about like... [1:20:37] communicating your idea and all the details that you can think of.

[1:20:41] in a way, digestible for your peers and the agent. And the agent, critically. Yeah. Yeah. [1:20:49] So depending on who I work with even, I will change the way I... [1:20:54] make these things [1:20:55] So like... [1:20:56] I work with an engineer. His name is Ian. He loves mocks. [1:21:01] He loves pictures. Like when I do like live code prototypes, he doesn't like it. He just wants Figma mocks with all the, like every detail in one picture. Yeah. [1:21:12] So I just do that with him. [1:21:16] talk about something more vague, people have also vague ideas. Then I keep it more... [1:21:22] Maybe they're just bullets. Maybe they're like... [1:21:26] Simple writing. [1:21:27] And maybe when we want to do something like [1:21:31] It's going to be like a multi-month stage thing that's a little bigger than a big RFC. [1:21:39] Yeah. [1:21:40] It's all inherited from the way we do it at Notion. [1:21:44] Yeah. [1:21:45] the writing part. But with Cursor, it's like, now there's also... [1:21:51] Thank you. [1:21:52] you just kind of [1:21:54] Ah, I have this idea I'll add it to my prototype, and then... [1:21:57] Ho ho, look at this. [1:22:00] Should we do it? Yeah. Let's do it. I suspect those two modes together are quite powerful. Mm-hmm. Yeah. Yeah. [1:22:05] Yeah. [1:22:07] You get from the most high-level abstract level to the most detailed.

[1:22:14] Writing, when you say the abstract level, you mean long writing. [1:22:17] Or like even just high-level bullets. What are the ideas and the constraint? Is a really detailed spec doc and an actual prototype two forms of... [1:22:28] Like two almost different trees of detail. It's like the same thing, but... Right, yeah. Visualized differently at different levels. Totally. Yeah. [1:22:38] On that note, maybe, like, what is a week? [1:22:40] what does your time look like? I think Cursor has like one meeting a week. You're going on walks and thinking, you're prodding VOS, whatever. You're... [1:22:49] in Figma sometimes. Like, what is that, like, pie chart of time... [1:22:53] Thank you. [1:22:54] That's kind of random. Every week's different. Yeah, very different. [1:22:58] Yeah, we also jam with people... [1:23:01] Yeah. [1:23:02] at the office. [1:23:03] People are always like there. [1:23:05] Um... [1:23:08] Not much meetings. [1:23:10] um [1:23:13] But a lot of talking, it sounds like. Not scheduled meetings, but a lot of... Yeah, chatting and talking and jamming and... [1:23:20] um... [1:23:22] Drawing pictures, finding people. [1:23:26] to help [1:23:29] Join us. [1:23:30] Yeah. [1:23:31] podcasting sometimes. Oh, yeah. [1:23:35] getting designers to turn into coders. [1:23:39] You're a big ringleader for that. Yeah, I want to make it happen. [1:23:42] What do you say to the average designer currently who's feeling stressed out? You're ready. You're ready. Yeah.

[1:23:47] It's time. Just start building. Just start pulling the thread. Get in there with the clay. And then send me all the feedback. And if you don't like what you're seeing, we'll fix it. [1:23:57] Maybe on that note... [1:23:59] Although this could buy to engineers or any maker, too. I think one intuition people have around AI, maybe the average creative or artist, non-technical person especially, is that... [1:24:08] Vibe coding or AI or whatever can make slop, but it can't make soulful things. You've certainly made the most soulful vibe-coded thing I've ever seen, if that's what it was. Right. You just need to put your soul in this. You need to care about every detail. You need to not accept whatever purple gradient the AI gave you. [1:24:29] It's the end. Like that is just the beginning. Ah, yes. You always start with shit. You always start with sloth, with AI. [1:24:37] and then you refine it that's the beginning not the end you just poke at it with little prompts yeah [1:24:45] And then it'll get better. Mm-hmm. [1:24:48] I'll take some turns. [1:24:50] You say in the age of AI, the question everyone's asking is, will I be replaced? The real question is, do you know yourself well enough to become irreplaceable? Mm-hmm. [1:25:01] I don't think we're through with technique and skill and craft and mastery. I am curious if there are any of those that you think are worth mastering now, but it seems to me that it's actually more about what you might call intuition or sensibility. Yeah, yeah, yeah. [1:25:16] Can you talk about that? Yeah. What goes into that? Because that feels like it's not the end, it's the beginning. Right. The beginning feels like I don't like the purple slot. Uh-huh. That's like I know what I like and I know incrementally what I like. Right. Right.

[1:25:31] True. [1:25:32] It's like the AI models are trained on all the public knowledge information and the code that it can see. [1:25:39] And you are trained on the same thing. [1:25:42] Like all the books you've read? [1:25:44] Other fonts that you know. [1:25:46] All the artists that you admire [1:25:49] the world around you. [1:25:51] And you build that. [1:25:52] intuition or taste or whatever... [1:25:55] And you start forming an opinion about how you want to shape the world. [1:25:59] and [1:26:01] And... [1:26:02] in [1:26:05] You express it by building. [1:26:10] Yeah. [1:26:11] That's what it is. [1:26:13] Like... [1:26:13] not by thinking by the way yeah not by thinking not just thinking [1:26:17] Yeah. [1:26:20] then it's like you have to [1:26:22] Keep making things and keep looking at things. [1:26:25] yeah one of the things that get missed in the when people talk about taste is taste is eating food yes it's not thinking about food yeah yeah yeah you need to keep uh eating and making shit yeah [1:26:37] And then make the shit better and better. Mm-hmm. [1:26:41] You critique design as aesthetics, I think, a lot, but you're also very attuned to aesthetics. [1:26:48] Real OS is, like, the most specific thing ever. Like, you've perfectly handcrafted, recreated Aqua, among many other things. Yeah. [1:26:57] What is the... [1:26:58] Maybe it's back to this taste thing, but what is your relationship to...

[1:27:04] sort of like not holding aesthetics too tightly, but also still clearly... [1:27:09] really putting a ton of time and effort and energy and thought. I think it's like, [1:27:15] So how you present things visually will always be there. [1:27:19] And... [1:27:24] Like I don't really think about it anymore. You just start noticing. [1:27:28] Like this feels off. This feels wrong. [1:27:32] Thank you. [1:27:33] Once you have almost like a... [1:27:36] the set of patterns [1:27:38] Then you don't really think about it anymore. [1:27:40] Thank you. [1:27:41] Unless it's like something new that you want to stress on or... [1:27:46] Thank you. [1:27:47] You want to like put a little bit more flare into it. [1:27:50] Um... [1:27:52] But it's like all the foundational bricks... [1:27:56] they need to fit perfectly. Even in the visual space. It's like the visual space, the bricks are... [1:28:03] It's like the color, the spacing, the layout. [1:28:06] The grid. [1:28:07] And [1:28:08] the different like type, type scale font sizes and [1:28:13] All of that. [1:28:15] It's sort of part of the big picture. Yeah, it's part of it. It's more like one layer of it. Yeah, yeah. But it's like ideally the thing... [1:28:24] is also constructed in a way [1:28:29] that is like... [1:28:32] It's almost like the simplest form...

[1:28:36] for the low-level ideas that you want to convey. [1:28:39] Yeah, I like that. [1:28:42] it's a, it's a, it's there compressed. Yeah. Yeah. It's like compressed into pixels. What are they? Yeah. Yeah. [1:28:48] Um, [1:28:49] So you still think about it, but you don't think about it too much. Yeah. Mm-hmm. [1:28:55] Once you're over... It has its hierarchy. It has its role in the hierarchy. [1:29:00] And I also dislike how people think of them separately sometimes. It's like... [1:29:06] So at Google, they have interaction designers and visual designers. They're split. [1:29:12] And that's bad. [1:29:14] Then you create a world where the visual designers only think about how the button looks. [1:29:20] Yeah. [1:29:21] And then they fight. Not what it looks like to press the button. Yeah. Or, um... Or feels like, I should say. Like, how should the buttons be fitted together? Why is there so many buttons? Mm-hmm. [1:29:32] Yeah, yeah. You're always backing into this. You need to have the cohesion in mind when you're in the micro. [1:29:40] It's like... [1:29:42] I know you in that in that greatness piece you wrote about focus and breath. Like we're taught to focus early, choose what's important, discard what's peripheral. The genesis of a thing that might be great. Strict focus is a ruse. The treasure lies in expansive searching and stitching together a tapestry of interrelated issues. Later, once you roam far enough, clarity will guide you toward the right edges until then, like curiosity roam. And it almost feels like that is going in two axes, which is the axes of like incremental new thing and the axes of like hierarchy and cohesion.

[1:30:12] Yeah. [1:30:12] You do that at the same time. And that's why it's chaotic. Yeah. And ambiguous. Yeah, yeah. And you have to rein it in with the order. Yeah, yeah. Like when people try to put this into like a linear process or order, they just fuck it up. [1:30:28] I [1:30:29] Yeah. [1:30:30] Because there is no more like emergence. [1:30:34] Do you think that [1:30:36] One view just says that Google doesn't have... [1:30:39] Real ill. [1:30:40] or whatever and pick your favorite designer. Sure. Another view that says the people at Google are talented and actually like they are. Oh, yeah. Of course. The system is failing them. Yes. It seems like you think the latter. I think the latter. Yeah. [1:30:51] And I think, say, a tool like cursor or its ideal form can help with this. Mm-hmm. [1:30:56] meaning like people with different roles or... [1:30:59] They're kind of stuck in boxes right now. Yeah. You've just break the box. [1:31:04] And let them build the thing they want. [1:31:07] Yeah. [1:31:08] Another part of that essay on greatness, pursue agility and quality in equal measure. The myth says you must choose, move quickly and break things, or move slowly and ensure elegance. But genuine excellence emerges from a dance between speed and depth, agility and quality. I love this. Like a skilled musician who can improvise yet still maintain impeccable technique, you must learn to adapt fluidly without compromising the integrity of the final piece. [1:31:36] I'm curious how this dance, it makes sense to me that it could happen working solo on a short-term project without that much of a plan, maybe reOS. How does that happen maybe at other modalities, either with wide collaboration or let's say you're working on Cursor 2.0 and it's this big long-term project? How do you embody that?

[1:31:59] in that type of context. [1:32:03] It's kind of like the... [1:32:07] Thank you. [1:32:07] You let chaos be and you wrangle it at the same time. [1:32:12] Or... [1:32:15] It's like you're... [1:32:19] You don't pick size [1:32:22] You find like an equilibrium. Yeah, between the complexity and the simplicity. [1:32:27] And same thing with how much fast you want to go versus how much thinking do you want to do. [1:32:36] And I think... [1:32:38] especially in this age. [1:32:41] It's actually so easy to just try things out. Maybe it starts with so much, in so many of your answers, it starts with just saying like, [1:32:50] it doesn't have to be a choice. You're allowed to do both. They're the same thing. [1:32:58] People get stuck thinking they need to pick sides or they need to make these hard trade-offs when... [1:33:05] All of these are just like... [1:33:07] variables and you can add a little bit here, lower a little bit here. Um, and, [1:33:13] It's all dynamic. You want to be... [1:33:18] more flexible, [1:33:19] to [1:33:21] the situation you're in and the change that's coming. [1:33:24] you don't want your system to be stale or stuck in lack of form...

[1:33:31] that you can't get out as the world is changing. [1:33:35] Um... [1:33:37] You want to keep the essence clean and simple? [1:33:41] You won't create like a space for people to play with ideas. [1:33:45] so they can ship really fast. But maybe it doesn't disrupt the rest of the system as much. And then once you have, say, more... [1:33:54] More room or even like... [1:33:56] you're constantly doing this, like, let's wrangle things back, let's unify things. [1:34:01] Then you keep the core system better. [1:34:04] As you add more things or as you experiment with more things [1:34:08] It's like a complex system can actually be... [1:34:12] Quiet. [1:34:13] high quality and fast if its parts are simple. [1:34:16] Yes. [1:34:18] Yet we build all this complexity and scaffolding and arbitrary bureaucracy, whatever. All these things, all these shoulds. Right. And ideally, you'd actually get rid of all that crap that is not even part of the system. [1:34:31] The software itself. Yeah, it's bloat. Yeah, it's like everything around it, the processes... [1:34:37] A lot of it just don't make sense. [1:34:40] where they slow things down. They slow this loop down. You have an idea to see it real, to test it out. [1:34:48] And a [1:34:49] reiterate on it. [1:34:51] How does this... How many people are cursor now? [1:34:54] 300. And you were obviously a notion for a long period of that growth. Yeah. [1:35:00] When Cursor's 3,000 people, how does this not happen? You guys don't really have that much of a roadmap. I hope the planner agent will be ready by then.

[1:35:11] And then multiplayer Cursor will be there. Sure enough. Then people can be still pretty... [1:35:18] Like... [1:35:18] I think Hal Cursor does it really fast. [1:35:21] And pretty good. [1:35:23] As a lot of people we hire, they're just really high agency people. They were founders before. They haven't made stuff before. They just want to build. They don't want to think too much. Sure, but that definitely works with $30,000. Maybe that works with $300,000. Yeah, yeah, yeah. All wisdom would say that doesn't work with $3,000. Even if you had $3,000 Steve Jobs, it would actually be a disaster. Yeah, yeah. [1:35:44] I'm not sure. [1:35:45] I think that is actually one part of the [1:35:48] It's like a part of the questions we need to answer, which is like... [1:35:54] In this new world of building with AI, how do teams work? [1:35:58] And I think it won't be that... [1:36:00] Like, it won't be... [1:36:03] too close to what we had before, like layers of management and linear processes. It's probably not going to be that. [1:36:10] So what is it? [1:36:11] Yeah. [1:36:12] How do you like... [1:36:15] both make sure... [1:36:17] Like people are kind of aligned on the general direction, but each person have agency. Each person can build whatever they want. [1:36:24] to an extent. [1:36:25] have systems to kind of manage this and help people control. [1:36:30] Hmm. [1:36:32] Making sure that these people are actually talking to each other and share the same information. [1:36:38] When they do stuff. [1:36:40] Like, that's the main problem we have now, I think. It's like people are so...

[1:36:45] So does Trayv. [1:36:47] They talk to their own teams that are created with row boundaries. [1:36:54] They work in their own files, own tools. [1:36:58] Thank you. [1:36:59] One thing that maybe helps that you also have in that essay is [1:37:03] about the quality of a team. You say, "The team that molds greatness is not a conscript army, but a band of pilgrims." [1:37:12] such people don't hide behind process or hire. Oh yeah. What does it feel like when you [1:37:18] meet a group of people, you're in a room or you're in a visiting office or when you first kind of met the cursor people or whatever, what is, how do you know, how do you, how can you tell that it's a band of pilgrims? [1:37:30] Just see what they're doing and what they care about. You ask them, why are they here? And then they tell you, because I love programming. [1:37:41] They just like doing this thing. [1:37:43] Like they're into it. They're passionate. [1:37:46] Thank you. [1:37:47] Take care. [1:37:48] deeply. [1:37:50] Thank you. [1:37:50] and they want to make the best thing. [1:37:52] And they want to put the work in it. [1:37:55] And you see it. [1:37:57] Like they don't talk about [1:38:00] I don't know, equity or whatever, you know, investment or... [1:38:05] I don't know. [1:38:07] They talk about... [1:38:11] Like the latest models, the new ideas...

[1:38:17] They exchange their ideas. [1:38:21] And they're there [1:38:23] for quite a long time every day. [1:38:25] And they're doing that like... [1:38:28] Thank you. [1:38:30] Not being forced. Yeah. Yeah. [1:38:33] Thank you. [1:38:33] uh, [1:38:34] on the note of the sort of essay about making something great, [1:38:38] Mm-hmm. [1:38:39] Do you aspire to greatness? [1:38:42] Oh, yeah. [1:38:43] What does that mean for you? [1:38:46] To me, it means like you make something... [1:38:50] that helps a lot of people that lasts. [1:38:56] And ideally it's pretty close to the ideal configuration of the thing. [1:39:04] Yeah, that truth, that trueness we talked about. Right. But sometimes you fake it. [1:39:10] It's like sometimes we make the upper layer really nice and pretty and cohesive, but under the hood is like chaos. But that's fine. You just... [1:39:19] You do that like slowly. [1:39:20] Yeah, it's like the picture of the SpaceX rocket, the first SpaceX rocket, the iPhone. The iPhone Air now is like… Oh, yeah. Even if you look at the inside, it's so pretty. Yeah. I want the clear iPhone Air, too. That would be amazing. [1:39:39] Mm-hmm. [1:39:40] I have... [1:39:42] We have a little time left, so I have a bunch of quick speed round questions. We don't have to take super long on each one.

[1:39:49] First off, maybe it relates to your last answer. What does it mean for technology to feel more human? [1:39:56] Not exactly the easiest speed round answer or question, but... I think I should, like... [1:40:02] Fit each human better. [1:40:06] And it's different for everyone. [1:40:08] Like some people prefer something really simple. Some people actually want to see every button. [1:40:13] some people like talking, some people like reading, some people like... [1:40:18] like watching YouTube tutorials. Some people like going to a course, buying a book. [1:40:25] It's fit. It's a personal connection. It's about like fitting the human... [1:40:30] In the way they do things, not in the way I do things. Yeah. Or like our engineers do things. Yeah. Those can be good examples. Mm-hmm. [1:40:41] Mm. [1:40:44] And as a, you know... [1:40:47] As it fits you better, it [1:40:49] inevitably needs to understand you better. [1:40:53] your preferences of [1:40:56] even like your way of thinking or how you talk and, [1:41:00] Thank you. [1:41:01] The things you care about. [1:41:03] It's like almost being seen by a... [1:41:05] a design or a product. [1:41:09] Thank you. [1:41:10] When you do it, [1:41:11] It just feels like... [1:41:13] Like you're in slow and you don't think... [1:41:15] Kind of like how I use Figma. Yeah. But that took like years of training. Yes. But now it's like maybe...

[1:41:22] A couple of tries, you were like there. Yeah. [1:41:26] You write a lot, and you clearly are really thoughtful about how [1:41:29] not only what you have to say about Cursor publicly, but the narrative and the conversation around Cursor. We spoke about this briefly, and you said, like, tools are all selling ideas. They're all attaching themselves to ideas. There's a lineage of ideas they're sort of pointing at. How you talk about tools matters tremendously. You have to plant seeds. Mm-hmm. [1:41:47] What do you mean by planting seeds and how do you think about shaping plants? [1:41:52] what people think and perceive about cursor. Right. [1:41:57] Yeah, I think like software... [1:42:02] To me... [1:42:04] Kind of like what we said, that's just like a tree of concepts packaged up in the Word. [1:42:09] Cursor or notion. [1:42:10] Notion is blocks, pages, databases, cursors, agents, models, contexts, and [1:42:15] Editors, maybe. [1:42:17] Um... [1:42:18] Thank you. [1:42:20] But... [1:42:24] You also want to create something like a brand that lasts... [1:42:30] That is not just... [1:42:31] Your present form... [1:42:34] That is a little bigger. [1:42:36] that ties... [1:42:38] With... [1:42:41] the past and the future. [1:42:45] And that is definitely not, say... [1:42:47] Cursor is the AI coded. [1:42:50] It is not even like, say, cursor makes you...

[1:42:55] Extraordinary productive. [1:42:57] It is bigger. Yeah. [1:43:00] and then you want to tell the bigger story, [1:43:03] And then you want to also tell... [1:43:05] smaller stories to like... [1:43:07] different groups of people. But tie them all together. Yeah, it's almost like the tool itself, the product, is like the ship. And the story is like, "We're going to the Americas." or something. [1:43:19] Having that broader context is important. People attach a lot of identity to the things they use to make things. Yep. [1:43:26] Like, I think it's actually... [1:43:28] A service. [1:43:31] Like, we need to do more of this. [1:43:35] to kind of paint... [1:43:37] a picture for people to see how we came here. Yeah. Yeah. [1:43:41] and how these things are actually the same things, same ideas, how the ideas originated. [1:43:47] how they kind of interweaved. Yes. Well, that's so important to the AI, especially. AI is so alienating to people. A lot of people, like, now when they start, they actually just start from, like, now, now. They don't see the past. They don't know how we came here. Or they're living in the past, and they're like, I don't like this future. They're stuck in the past, and they don't know how this future can... [1:44:07] Take them. Yeah. [1:44:10] You said there was a tweet where you said you're talking about a bunch of things. You said don't build slot machines. And a few people accused Cursor of being a slot machine. Right. What do you say to that? [1:44:20] I don't think cursor is a slot machine because slot machines, they don't let you open it up.

[1:44:26] It's closed, black box. But Cursor is like... [1:44:31] I actually don't want your primary way to interface with cursor to be like [1:44:36] Kind of like, say, Cloud Code or Codex CLI is like... [1:44:40] You're in the terminal, you're in this little box, and then you're kind of constrained and just like that input. And you're just typing the thing in a little box and then enter and see what happens. Wait for a little bit. See what happens. [1:44:52] Versus like incursor [1:44:54] Like that is say... [1:44:57] Is this possible? [1:44:58] and you can do it like that, but it [1:45:01] That is just like... [1:45:02] One form of it. It's the beginning. Yeah. [1:45:05] Or like... [1:45:06] you will just naturally hit these, ah, I see a code block. Maybe I want to click and see what's in it. Ah, I'm, like, done with this chat. I hit this review button, and then now I see all the things. [1:45:19] Um... [1:45:19] And it slowly teaches you, say, like, now we're doing code reviews. We're going to stitch the agent reviews with the code review with Git and, you know, all the other stuff. [1:45:29] Then that's a new [1:45:31] newcomer even like you come in and then you started with a simple thing and [1:45:35] you slowly get to the [1:45:38] Like if you want. [1:45:39] I don't force you either. It's like, if you don't want to open the code, you don't have to. Yes, keep hitting the slot machine if you want. If you want that, it's fine. And I don't think that's a slot machine either. Again, it's like... [1:45:51] customizable, [1:45:53] You can open it up, you can do whatever to it. Even in the simple form, you can still customize the thing.

[1:46:00] Um, [1:46:01] I [1:46:02] And you have full control. [1:46:04] And you have the whole spectrum of control from the most manual coding, which is you just type... [1:46:11] And it's still your thing. I don't do anything to like you type and then [1:46:17] Our tab model is still the world's best thing. You type and then boom. [1:46:21] It kind of completes your thought. It jumps you to the next place. You keep going. So if you prefer that and you're in your flow state there, you should keep doing that. [1:46:32] Um... [1:46:34] And then for say like [1:46:36] There's now like... [1:46:39] a small chunk of professional developers who have become very age-encoded. [1:46:45] Thank you. [1:46:46] Like they don't do manual coding as much anymore. Then for them, we have tools for them to [1:46:53] Focus on one agent, spend multiple agents... [1:46:57] manage them at a higher level, see, and... Then you get the whole spectrum, and for these people, again, it's like... [1:47:05] they can find their preferred spot. [1:47:07] And then... [1:47:08] They can open it up and... [1:47:11] Do more if they want. [1:47:12] but I don't force them to be like, ah, you're always in this little box and all you can do is... [1:47:19] Put the problem in a little box, see what happens. [1:47:22] Is there a pattern from stripe to notion to cursor? [1:47:25] as you've spent most of the last decade. Yeah. [1:47:28] I don't see them as too different either.

[1:47:31] Or like... [1:47:33] Thank you. [1:47:34] They're actually very similar. [1:47:36] Like Stripe to me is just passing messages around the internet but the messages are transactions or money related. [1:47:42] Yeah. [1:47:43] Notion is just like... [1:47:45] basically like [1:47:46] the MetaSAS tool kind of [1:47:49] databases and [1:47:52] all the archetypes of views and patterns. [1:47:57] Cursor brings it more low level, but it's also more flexible. [1:48:01] Like you actually break all of these patterns and parts. [1:48:04] completely [1:48:06] and [1:48:07] At some point, like you will... [1:48:09] Okay. [1:48:09] get it composed by the AI or... [1:48:13] With our, like, precess or something. So you get the toy you want. [1:48:17] Yeah, you have a line somewhere where you say building stuff that frees up people's minds. And it felt like that's kind of true for all of those three things. [1:48:25] helping people make the thing they want. [1:48:28] Hmm. [1:48:29] What did you learn from, what did you and what have you learned from the Notion founders and the Cursor founders, respectively? [1:48:36] or maybe even Stripe. [1:48:37] Hmm. [1:48:39] From Ivan's, like... [1:48:46] I think he kind of showed, like, system thinking and aesthetics. [1:48:49] can be melded together. Like you don't have to pick. [1:48:52] Wow. [1:48:54] And then from cursor people... [1:48:58] It's just like, you should just YOLO and do stuff and don't think too much. Yeah. Keep doing it. Ambitious naivete. Yeah. Exactly. It's like, that is actually so, so good in this age. Because actually, nobody knows what they're doing. Right?

[1:49:16] Yeah. [1:49:17] Like all the old ways of... [1:49:20] doing things don't really apply anymore. [1:49:23] What do you love about Steve Jobs? [1:49:27] Hmm. [1:49:28] I love him. It's almost like... [1:49:33] It's like a spiritual figure [1:49:35] Kind of. [1:49:37] Like I don't [1:49:38] I'm not religious. [1:49:40] Thank you. [1:49:40] But I feel like sometimes people need like a... [1:49:44] So it goes thing there. [1:49:47] And I kind of put... [1:49:49] as a symbol there. [1:49:52] That helps me a lot. [1:49:54] What does that symbol represent? [1:49:57] It's like forcing you to be... [1:50:01] thinking about everything, all the details and... [1:50:03] Coming up with the simplest thing. [1:50:07] Yeah, and he kind of helped me start all of this. [1:50:12] Like he got me into design. [1:50:14] Or like, you know, the old apple. Yeah. [1:50:19] Like, they showed how... [1:50:22] like computers can be beautiful [1:50:25] Maybe on that note, what is the difference to you between liquid glass and audio? [1:50:32] I mean... [1:50:37] Like aqua. [1:50:40] It's more like... [1:50:42] What they were trying to do was they bring a lot of the physical metaphors into the computer. Yeah.

[1:50:49] so that people feel more familiar with things. [1:50:52] Like, if you look at all the icons, they almost look like the emojis we used today. Yeah, yeah, yeah. Like, they're super detailed. Yeah. Like, with real-world, like, reflections and material. Yeah. Yeah. [1:51:04] And it's like back in the days, it looks completely different from, say, the gray boxes people used. [1:51:11] Like the double 3D buttons and stuff. [1:51:15] So that was like... [1:51:17] Pretty game-changing. [1:51:19] they also master how to render fonts. [1:51:22] Like back then, how Aqua was made, it's like all just kind of PDF rendered. Yeah. [1:51:28] on your screen. [1:51:30] You can stretch the UI freely. The text was not... [1:51:36] He liked the map. [1:51:38] little pixels, but... [1:51:41] It's all anti-aliased, perfect. [1:51:46] Liquid glass almost feels like... [1:51:49] Thank you. [1:51:50] It's almost like a flex. [1:51:52] on what Apple can do now. And it's kind of weird. [1:52:00] I get the point. It's like they're trying to unify the design language across all of their platforms, coming up with one thing. [1:52:08] But it's like how you use a phone versus how you use the Vision Pro when you stare at things and then, you know, they need to track your eye, your finger and your little pointer on the mouse button. They're all different.

[1:52:19] and [1:52:20] So your interface probably can't be the same thing. [1:52:23] But they tried to make it the same. And this material, even though it's inspired by glass, is purely digital. [1:52:31] Thank you. [1:52:32] they're just flexing that they can build system-level shaders and make them perform across every single UI. And then my menus can morph into a button and out from the button. [1:52:45] But then to the users, like, what's the point? Yeah. [1:52:48] It's just... [1:52:49] It actually makes a lot of the UI like [1:52:55] Like you can't see much anymore or like the tabs take so much space. Like you need to keep clearance for, for the tabs. Yeah. [1:53:04] their shadows, the little blur under it. So you actually, like when you compare the old iOS and the new one, [1:53:11] you actually see less text or like there's like less stuff you can do. [1:53:16] So maybe like the priorities have changed. [1:53:19] Like instead of being truthful to the platform themselves and the way you interact with it, either it's a finger or your eye or your local pointer that have different precision, [1:53:29] Let's just make everything the same. [1:53:32] I have to stop you because I know you can rant about this all day. I'm really good at finding things to get my guests to rant about in the last few minutes. Just a couple more questions. Yes. I know I had to get this one in. Okay. What makes new genes stand out in a world of factory farmed caves? Uh-huh. [1:53:48] I think it's the same idea. Like, I think all of the things that we make, the new things, are just kind of remixes of the old things.

[1:53:57] And what Neugeens did was they just mix things really well. And then they give these girls like a space to just... [1:54:04] be themselves and have fun. [1:54:07] And that's why... [1:54:09] It feels so different from all these scripted, manufactured K-pop songs. [1:54:15] Um, that were... [1:54:18] It's almost like [1:54:21] Most people, they're just kind of mixing all of crazy things together now. [1:54:27] Whereas, like, new genes are more like... [1:54:30] Southfall. [1:54:31] And... [1:54:33] So, again, it's like about taste and... [1:54:37] Like the constraint. [1:54:39] Yeah, K-pop, in some ways, K-pop can feel like it's just like, what does the algorithm want? Yeah, yeah, yeah. [1:54:45] Yeah, like you find a concept and then you kind of... [1:54:48] Like what they do is they get a lot of sound writers and they buy a lot of songs and they're just like... [1:54:54] Let's mix these parts or mix these genres. Boom. Put the English, Korean, Japanese lyric together. Boom. [1:55:03] What can you say something about Zhuangzi's butterfly dream? [1:55:08] ah [1:55:11] Butterfly dream. [1:55:15] Thank you. [1:55:17] It's like life... [1:55:19] In a sense, it's like reality is not that real. [1:55:23] And a lot of it is just in your head. [1:55:26] All right.

[1:55:27] So sometimes you feel like... [1:55:31] So almost like you're... [1:55:34] Thank you. [1:55:35] You're living in a dream where you can actually mold anything. [1:55:39] It's an old Steve Jobs video. It's like when you figure out that the world is moldable and plastic, you can poke it and you get feedback back. Yeah. [1:55:49] And it's like the butterfly. [1:55:53] And sometimes you just let things go and see how it works. [1:55:57] how it happens. [1:55:59] And sometimes you go back and you take control. [1:56:03] Like you wake up from the dream. Right. Or sometimes you're in between dream and reality. [1:56:10] We're always all doing that, by the way. We're on autopilot and we're not. You know. [1:56:16] I was talking to Rio OS and... [1:56:19] I was talking to Steve Jobs, Pope Francis, and Rio. And the Pope said something about a revolution of tenderness. And Steve was skeptical. So I asked Rio what tenderness means to him. He said, tenderness to me is when a system or tool feels intuitive, almost invisible, making things smooth and delightful. It's the empathy baked into the design. Right. We didn't talk a lot about empathy today, although I think it's kind of running in the background of our conversation. [1:56:49] you are deeply empathetic to the people you care about, which is, I think, people who make things. What does IRL Rio think about tenderness?

[1:57:00] Thank you. [1:57:01] tenderness. [1:57:04] I [1:57:06] It's just like [1:57:07] putting the care into things and people you meet and... [1:57:11] the [1:57:12] People we serve... [1:57:14] Um... [1:57:19] Being truthful. [1:57:22] that like [1:57:24] you know, the ideas that we work with, or the technology even, is like universal. [1:57:29] It's general, it's like generalizable. [1:57:32] It's not exclusive. [1:57:35] to like a group of people. [1:57:37] Thank you. [1:57:37] And you can always start [1:57:40] by like... [1:57:41] Like you understand what you need, what you are frustrated with. [1:57:45] And then you find a group of people who are maybe similar to you. [1:57:49] So like the people working at Cursor [1:57:53] And they all shared similar problems. [1:57:55] And they, you know, [1:57:57] make stuff for themselves and [1:58:01] Make this tool. [1:58:03] And then it's about like... [1:58:06] How do we bring it out to more people like us? [1:58:10] beyond people like us. [1:58:13] And that's maybe like the next breakthrough will be... [1:58:16] like the vibe coding tools and the pro coding tools today. [1:58:20] are still very split. [1:58:23] Thank you. [1:58:24] Like it's really hard for... [1:58:27] So the non-technical people?

[1:58:31] to come into cursor today [1:58:33] But also very hard for them to progress from a vibe coding thing to a real thing. [1:58:38] Yeah. [1:58:39] So maybe we can help with that. [1:58:41] We can help with turning the designers into coders, the PMs into coders. [1:58:47] the coders into designers. It's all the same thing. It's all the same thing. And we start realizing, oh, we can actually like, [1:58:55] We don't have to like, [1:58:57] put boxes around our heads or our eyes. We can actually do things. We can... [1:59:03] do things better with other people who have, say... [1:59:06] different areas of specialization. [1:59:08] But we're all thinking about the same thing. [1:59:12] People don't have to fight. [1:59:14] Thank you. [1:59:15] Like instead of fighting about, I don't know, bureaucracy... [1:59:19] You fight about the truth. [1:59:21] Like what is the best thing to do? [1:59:23] What is the ideal configuration of the thing we are doing together? [1:59:29] and you're helping people you raise. [1:59:33] all the parts in their job that they don't really like doing. [1:59:38] You help people like amplify their strength. [1:59:41] Like what they care about. [1:59:43] what they're really good at. [1:59:46] And you help meld these people's drinks together. [1:59:50] And then the agent covers the rest. Yeah. [1:59:55] Really? Really? [1:59:56] Thank you very much. Thank you.