Web3 Galaxy Brain 🌌🧠

Nicholas: Welcome to Web3 Galaxy Brain. My name is Nicholas. Each week, I sit down with some of the brightest people building Web3 to talk about what they're working on right now. My guest today is Vidy Thatte, co-founder of Trace. Trace is a website that lets you build SwiftUI components with LLM text prompts, screenshots, and drawings on a canvas. In short, Trace lets you build iPhone apps with AI. On this episode, Vidy and I take a walk down memory lane, chronicling his illustrious history of app development projects and startups that led him to where he is today. We discuss the technical details of how the current version of Trace uses GPT-4 and a code database to create smooth-feeling UIs, and how powerful Trace may become in the future. It was excellent getting a chance to meet Vidy, whose journey I've been following online for years. I hope you enjoy the show. As always, this show is provided as entertainment and does not constitute legal, financial, or tax advice or any form of endorsement or suggestion. Crypto has risks, and you alone are responsible for doing your research and making your own decisions. Vidy, welcome to Web3 Galaxy. I'm really excited to talk to you about Trace. Thanks for coming on the show.

Vidy Thatte: Yeah, thanks for having me. This was spontaneous. Did not know it would be streaming.

Nicholas: We're going to YOLO it. But I think everybody's going to be very excited about Trace if they haven't already heard about it. So we're here primarily to talk about Trace, but I'd also like to talk to you a little bit later about kind of the subculture that we both belong to and that you've been grinding in for so many years trying to make a piece of software that people love. So first up, let's just talk about Trace. What is Trace?

Vidy Thatte: Yeah, so Trace helps you build. So if you want to prototype something really quick, if you want to build a screen for your app, you can just type it in, describe it, and it writes the code for you, and it gives you a file so that you can drop it into your Xcode project really fast.

Nicholas: Awesome. So you introduced it in November 2023, right?

Vidy Thatte: Yep, end of November.

Nicholas: So you've been explosive with the product development lately. For folks who've never built an iOS app, what are the main components of an app, and how does a SwiftUI fit into it?

Vidy Thatte: So yeah, if you want to build an iPhone app, I think any software product, you have the front end, and then you have all the logic and the database. So SwiftUI is mostly the front end part of it, which involves the buttons and text fields and images and all that. It really makes it so much more intuitive than all the previous things that Apple had, had given us to make apps. So it's really great. I love using it. Very simple. So Trace makes it really even more easier.

Nicholas: For people who haven't done it themselves, what kind of like level of, like what kind of prompt might I type in and what kind of code might I get out of it, or at least simulated experience? What part of the app can I, how much detail can I get into? basically?

Vidy Thatte: Yeah, I mean, honestly, you can push it as much. A lot of creation in and of itself is programming. And I mean, so yeah, you can do something very simple. You can just make, you know, something like a settings page or something, or you can just take it further and make all kinds of interesting, you know, UI stuff that, you know, kind of pushes the limits of what an app UI does. You can do all of it. The only limitation probably is sometimes when it gets too complex, you might run into a lot of bugs because like AI isn't like at a point It writes flawless code. But yeah, it's kind of a skill, learning how to prompt and getting what you want.

Nicholas: And it's like distinct from like ChatGPT because I guess you have this alternate affordance. For people who haven't seen the app, basically you have like a text box. You can have a canvas if you want to paste in an image or draw something and you sort of type it in, hit go. It generates the code in the background. You don't even see the code. And then it launches a simulator and you're dropped right into the app, which I think is just amazing. It's really, really cool. Congratulations on that.

Vidy Thatte: Thank you.

Nicholas: I'm curious, like, I guess, you know, so one thing this might not be good for is if you need any kind of backend logic, like a database, if you want to like, you might be able to build the whole UI of some part of Facebook, for example, but you couldn't have it currently do the whole backend for it.

Vidy Thatte: Yes. Yeah. I mean, you can if you wanted to, like you can call an API and stuff, but it's not ideal. It's better to, it's mostly, you know, built for it. To an Xcode project and wire it up with your backend.

Nicholas: So it's kind of pitched at someone who wants to like prototype a frontend, but then maybe has the chops or is learning how to do the whole Xcode piece as well, right?

Vidy Thatte: Exactly. Yeah.

Nicholas: Is it like a perfect tool for you, given your background, which we'll get into later?

Vidy Thatte: Yeah, totally. It's been in apps for like, I don't know, 10 years. And I think in the last year, since GPT came out, I found myself using GPT a lot, like to write a lot of my frontend code, especially. Like things like, like stacks and like lists and like all these like grids, especially like I did, I would have to like put a lot of energy thinking about and plan it out, but I just like delegated all that thinking to GPT. And I was like, okay, this is kind of perfect. So in a way I ended up building this for myself because I wanted to get, you know, I wanted something to generate the UI code and then see it, tweak it, and then, you know, drop it into, into my project and continue building.

Nicholas: Because the process, I mean, it's, it's kind of similar to building an Xcode, I suppose, because you have to kind of build and then run. So it has this compilation loop feeling to it, but you don't have to do any of the code editing directly. Yeah.

Vidy Thatte: Yeah, yeah, exactly.

Nicholas: Is it, I'm curious how it got so good. or actually, I mean, let's talk about a little bit about how it works. Like what is, is it using any proprietary data or are you fine tuning? An LLM to be able to do this, or is it just the raw capability of something like GPT-4?

Vidy Thatte: I mean, I think it's less fine tuning. I mean, we, we tried fine tuning. It was actually worse when we tried to fine tune. So it's just raw capability of GPT-4, but we also have like a massive database that it gets, like gets inspiration from. But it's all based, you know, it's like a vector database or like you do this RAG. So whenever you, you're asking, we may already have in the database, it, it takes stuff from there. And repurpose it for purposes it for whatever, whatever you want to build.

Nicholas: - So that's why it's so good at like really knowing how to build a UI. It's not just guessing from nowhere. And it's not just based on stack exchange tokens.

Vidy Thatte: - Yeah, totally. Yeah. And it's also like constrained to like one single view or a screen. So it doesn't have to make a whole app. It can just make like little things, which is also something that I think when you're building with ChatGPT, you have to like take like a iterative program programming approach where you're not trying to build this whole complexity and you want to try like, but you know, block by block and add up to something bigger. So it's really good at that. So I think, I think we've kind of shaped it to work from UI point of view, as well as like how the prompt works. So yeah, I think that's probably why it's, it's working.

Nicholas: - Do you imagine it being useful for people who are not familiar with Xcode at all eventually, or is it right now really most useful for people who kind of know what they want and could maybe program it themselves, but this will make it a lot faster.

Vidy Thatte: - I mean, I think right now, because of how, like you, if you, if you don't know how to program, I mean, a lot of people have been using it who don't know how to, they have never built a SwiftUI app before, have never programmed, have used it. One day anyone should be able to just make an app for some niche thing that they want.

Nicholas: - What I really want to know is, so Trace is kind of becoming like a library of small components built with Trace. Is that kind of sharing of SwiftUI code already common in Apple development scene?

Vidy Thatte: - Yeah. Yeah. So I think if you go to GitHub, there, there are these like really popular lists called awesome SwiftUI. I don't know. I don't know. Do you like, have you, do you make apps or?

Nicholas: - No, I've never made an app. - Do you have any SwiftUI background? - No.

Vidy Thatte: - Okay. Okay. Yeah. So yeah, I mean, if you, so one of the things that a lot of the developers use is these lists called awesome, awesome Swift, which is like a massive library of like things that people have built. It's not just limited to UI stuff. It's like everything from like what, like networking libraries to all kinds of stuff. Like a big, big part of inspiration for Trace was actually that because we were like, one of the things that I use awesome Swift lists is for getting good UI that I can incorporate into my apps. But you couldn't, we couldn't like, I couldn't see the UI on GitHub. I had to like go to the thing, see some screenshot, download it to my laptop and then try it out and then decide if I want to integrate or not. And I was like, oh, like if we had a massive library of like components that, you know, you could preview and then decide to incorporate into my app or not. Like that would have, that'd be like a massive, massive, you know, friction removal, whatever. So, so yeah, that was definitely a big thing. And I think one of the things we did recently was like redo the website so that the first thing you see is like a library of components. - That people can browse and, and you know, kind of use and the reaction was crazy. Like all, all the devs that we showed it to were like, this is amazing. I need this. So yeah, that's, that was the whole thinking behind it.

Nicholas: - Yeah. So who, who's loving it the most right now?

Vidy Thatte: - It's, it's been developers. When we, I think one of the things that we, when we started out was like, okay, we want to, we can build. So I think we had three different target audiences that we were, you know, thinking about. one was like, like developers, people like me who want to make apps and need this to like, you know, may reduce like the time to build apps by like a big chunk. The second target audience was designers who, who have a lot like designers have amazing ideas. They, they're always prototyping, always like coming up with new concepts and being, giving them the ability to just like turn their ideas into actual apps that they can run on device. Would be really powerful. So that was our second target audience. And then the third would be like, you know, a random person has an idea for an app one day, they should be able to just come in and build it. So at this stage, the developers love it. They, they want to, they want to use it for, for, for their projects. And yeah, so we've kind of like, we're kind of working towards nailing that workflow really well. And then, but what I'm really excited about is figuring out like how to make this more useful to designers.

Nicholas: - Do you think those three target users will sort of, you'll be able to reach them sequentially as the software gets better?

Vidy Thatte: - Yeah, I think, yeah, that's exactly how I'm thinking about it. The first obviously is developers. The second, as we build more, more like experiment with more different kinds of interfaces for prompting the AI and also improving the AI itself for code generation, I think we'll be able to tackle, take on designers. And then one day everyone will be making their own apps.

Nicholas: - It's a beautiful thing. And I, you know, I think a lot of people have wanted that for a long time, honestly, looking at your site, it reminds me a lot of, I don't know if you've encountered the Alan Kay stuff around the original design for object oriented programming and like this idea that,

Vidy Thatte: you know,

Nicholas: functional pieces can talk to one another without needing to know each other's guts. That was kind of the original object oriented. And this feels, obviously it's not really there yet, but it does feel a little bit like if you can sort of skip some steps in terms of programming, with LLMs, then maybe people can be creating little applications or little pieces of software that can talk to each other. And in this case, really UI centric ones.

Vidy Thatte: A hundred percent. Yeah. I think that's the dream. I think I told, I've been saying this for a while now, I think like software is a new medium of self-expression just like music or photography or writing. - It's just not accessible as much to everyone. It's just limited to experts and you need like, all these credentials and you need all this experience. But I feel like in the next decade, we'll see more people creating software for self-expression and less like utility.

Nicholas: - And I mean, this kind of intersects, I don't want to lose the trail of talking about Trace, but this kind of intersects with like your history of making apps, cause I've been following you for several years. I don't know. You've been making apps for almost a decade and a lot of cool and interesting things, but they all seem kind of to come from some personal inspiration maybe. - Yeah. - They're a bit personal to you. So I feel you're kind of in this world of seeing software as craft, as art and something that you're inspired by. How did you come to that? And yeah, do you have any thoughts on why some people treat software that way? - Yeah.

Vidy Thatte: I mean, I think like, I think for me, like for my journey was I loved making music as a kid and I was like, I always wanted to become like a musician full time and all that. But then I discovered software and making products. And I was like, oh wow, like this is a really even more powerful medium of like, you know, like contributing to culture in somewhere because the scale at which like all these software products spread around the world and how like they kind of contribute to like people changing their habits is really, really powerful. So I think, so like my whole like perspective on that, comes from there. Yeah. And I think a lot of the stuff that I built is also like very much like, oh, I have this problem. I don't know if a million people have it, but if I can solve this for myself, I think there should be some, there should be at least like 10 people who would want this solved. Yeah. So I guess that's, that's like the more personal side.

Nicholas: Okay. For people who aren't so familiar with your background, maybe I'll say the name, you say roughly what it was. Graham. What was that?

Vidy Thatte: - Oh yeah. So Graham was, I, I don't know if you had, I don't know if you've seen this thing called use intro that I've been seeing a lot, use intro, but for IRL. So if you wanted to learn something new, I would pair you up with an expert who has done that before and you, you would work on a project together so that you would learn how to do this. So one of the things that I experimented with, so I started this in college when I was in senior year, I put out this website, on my Facebook. I was just like, I'm going to teach you how to make an app. I'm going to, I created like six modules or something. The first one's like how to make a to-do list app and then how to make a weather app, how to build Instagram. And yeah, I just shared it with a bunch of friends. They were all like, this is great. I want to learn how to do this. And then I just did like one-on-one sessions with them where within an hour I would show them how to build a weather app and then they would have a project on their GitHub and, and they paid me for it. I think I made it. I think I made like up to like 500 bucks within like the first few weeks of launching it, which is really cool. So that was a concept, but I think it was my first attempt at the startup. I continued working on it right after college, but I think within four or five months, I realized like I couldn't really scale it or turn it into anything much bigger. Maybe I just gave up too early, but I was also bootstrapping it and like, didn't have it, didn't know any investors or had any funding for it. So I was just like, I'll pause this and try something else. But yeah, that was my first.

Nicholas: Well, another one that's close to my heart because I kind of built a similar thing for the web, but Stream, what was Stream?

Vidy Thatte: - Yeah, yeah. Yeah, so Stream was like one of those niche things that only I wanted and surprisingly people used it. And a lot of, like somebody recently reached out. So we have this, the chat in Trace where you can like directly contact us and ask us questions or whatever for support. Somebody messaged through that and was like, "Hey, I still use Stream. If you have like this export feature, can you build it so that I can export all my notes?". And I was just like, this has been like five years. So what the hell? This is crazy. Yeah, so Stream was this, like. I think one of the things that I always did was text message myself whenever I had notes or like wanted, if I wanted to resolve something in my head, I would just like text it to myself and kind of have a conversation with myself, which would, it was pretty effective. It's still effective. I still do it. It's something about it, like something about like sending a message to yourself is very, it's really satisfying. It triggers something in your head. And every time I have like something I wanna like think through, I just start doing that. And I think the whole like thinking behind it was, like all the notes apps and like, you know, writing tools we have on our phones, they're just like, they've just been adapted from desktop. And, but like, you know, nothing's been built like ground up thinking about, you know, mobile first UI. So I was like, one of the most native things for your phone is text messaging. So there should be a notes app around text messaging. So yeah, that was my, that was something I built, like I think in a few days and put it out there, which was cool.

Nicholas: - Same, I did something very similar around the same time. I didn't share it with anyone. I just used it personally, but yeah, there's something fun about it. I stopped using it, but I could easily see if I, I think there were some technical problems and I just didn't give them the attention to fix them, but I really loved it.

Vidy Thatte: And

Nicholas: I think you both in stream, you had both the ideas which appealed to me too, which were texting and Twitter are both very kind of convivial composition frames for not just for mobile, but just honestly for thinking it's much nicer than a huge blank page in word or something ugly like that. And it's not as sort of, I don't know, Markdown. I write a lot of Markdown notes and I use this thing NV alt, which is kind of like a single browser for a folder, like where you can sort of edit really easily. But I agree with you and the theory behind stream anyway. TLDL, what was that about?

Vidy Thatte: - Yeah, so, so yeah, so stream happened. And then, so I was interning at this company when I was building stream and then I ended up, you know, finishing my internship and I was like, okay, I got a, you know, got a job or figure out what to do next. And a bunch of people, like I was in, I was in Silicon Valley, I was in San Francisco, a bunch of like VCs and everybody had heard about trace and there was not trace, sorry, stream. And they were like, you know, I was like, just jump jumping on calls or meeting them. And then I was like, oh, maybe I should just like turn this into a company or something. And one of the, like one of the things that I was really excited about was audio. And I was like, okay, like I, I just, I just gotten AirPods and I was just like using it all the time. And I was like, I think like audio is a really interesting, like could be the next big platform, whatever. I think a lot of people who are hacking on audio back then, especially on Twitter, people talking about audio, AirPods being the next platform clubhouse started around the same time as well. So, so, so I was really interested in that. So I, yeah, I, I was like, I wonder if there's some thing to do here with like, like stream, like making an audio stream. I was jamming around that concept and then met up a few investors, like raised our, like a little bit of pre-seed money, hit up, you know, a partner with one of my friends from college. And we just like started like, we're like, okay, let's, we just started a company like in a few months and then we, we started like hacking on like audio, audio stuff. and then one of the one of the things that we both wanted really was uh like we listened to a lot of podcasts but podcast discovery was it's still pretty bad like it was pretty bad because um there were a lot of these shows really good snippets but the shows were really long. they were like two hours four like like every episode is like really long um but and within the episode you speak you don't talk about one thing you talk about like a bunch of different things. so we were like hey what if you know each each like topic is a starting point for the podcast. so that um if i'm interested in like you know how to in a particular topic you don't have to listen to the whole podcast. you can just listen to that topic and then more to the next podcast listen to that topic um and it would kind of change the way people listen as well as discover stuff. uh so that was the whole insight. so we just built something uh around that and uh and launched it and honestly like you know pretty great uh pretty great we got a lot of really good um like super engaged users. uh people were using it people loved it. uh yeah so that was. that was tldl. it's basically an app where you can make um you can bookmark part bookmark a podcast uh or like drop bookmarks on a podcast episode and then it would turn into a feed um of like bookmarks so that you don't have to listen to a podcast from the beginning. you you can you can listen to the most interesting part um and it was social. so you could see other people's bookmarks as well. um and uh that would be like a much better listening experience than you know just going through episode by episode.

Nicholas: i remember it i tried it out at the time. is that michael you were working with?

Vidy Thatte: yeah yeah i'm still working with michael.

Nicholas: yeah so you guys went to school together.

Vidy Thatte: yeah we did. yeah we we met on day day one at nyu orientation. nice yeah yeah that's cool.

Nicholas: so you're close friends.

Vidy Thatte: i assume we weren't close friends. i mean we were friends like we knew we we we knew each other. and uh we like usually usually hung out and like similar like groups of people um but i think when i was working on stream uh i was talking to a bunch of vcs. they were like yeah you should are you? do? you have a co-founder? you need a co-founder. and i was like yeah i think i need a co-founder i can't do this alone. and then i was like talking to a bunch of people. and then michael was i think around the same time like looking for jobs or he was he wanted to be a pm um and i was like yo why not team up let's do this. and then he flew to sf um and we just started working together. awesome um so i was skipping some steps.

Nicholas: uh you then got to re-meme. slash collage i guess is the next sorry. did bounce ever release anything i should ask.

Vidy Thatte: no it was. it was mostly a test site. um it was never released. we never got it to like be production ready and also one of the things um with the whole goal like a a mobile music creation studio and um. but i think we were trying to go after like consumers and and we wanted to build something really simple but also useful. we never really tracked it um and the app itself was like technically super hard to build um and like it was like buggy and and yeah. so maybe i was never able to launch it but um. but yeah i mean the time for music is really small especially if you like. why would anyone have a reason to make a song like who cares? i mean people take pictures. uh people do all kinds of stuff but i didn't see like. yeah i mean i mean okay people do want to make songs. i think one of the things that we identified which was really cool was um people were making a lot of cover songs for tiktok and they were like making like slow slow down versions um sped up versions lo-fi versions of songs um and i think that would have been a really interesting um that interesting thing to like go after. but it didn't feel like it was going to be like a mainstream product. it would be like a niche tool. um i like something like cap cut which does yeah cap cut is amazing right but it's for video uh but cap cut for music would be like somewhat of a niche tool but could be could have been cool yeah

Nicholas: but still still kind of acquisition target for tiktok rather than probably its own network. yeah exactly yeah uh which leads us directly to re-meme aka collage. uh where did that idea come from? yeah i think it's a really cool idea.

Vidy Thatte: i think it's a really cool idea. so i i think so we were working on bounce. it had been a year. we were running out of money. uh i was in like hawaii. uh i'm staying at my parents place. and then uh i was like what am i doing in my life? this is not working. um. and then like yeah i think i think i was gonna like like lose my health insurance to return 25 and i was like oh god like um like this is this is bad like i thought i thought it would be like you know uh in a better place. so i was. i was freaking out and i was like i'm just gonna like try a bunch of experiments launch a bunch of apps and like see see if anything sticks. um so i had a bunch of ideas and then collage was basically like. it was basically like okay what if we take bounce? so bounce had this feature called remixing where if somebody made a soundtrack you could remix that right. um basically build on top of it. i was like what if you take this remixing and apply it to memes? um i wonder how that will work. and and i had this idea like before i told michael about it i think like a few months before i was like yo somebody should be like instagram but with remixing and focus on memes as a starting point um. but we never really took action because we were super busy like working on bounce um. so yeah i just like whipped it together and like in a day and and gave it to a bunch of people. um and i think like within two days it was like less than 50 people had already made like. it was a crazy number of people and i was like oh my god i'm gonna do this and i'm gonna do this. and i was like okay this is blowing up like nothing's. nothing like this has happened before and i think within a week there were like like set five to seven k memes or something posted on there by less than like 100 people.

Nicholas: wow can you describe the app a little bit?

Vidy Thatte: yeah yeah so so collage is basically a meme app where uh you can you can create memes on it and then other people can remix your memes. so if i make them you know like if i make them i'll put them on there and then upload my picture. somebody can take that and add text on it and draw stuff and um you know turn it into something else.

Nicholas: um so you can kind of build off of each other's posts but then in on the back end you retain the like raw source for the memes such that it's not just flattened into an image but you can then like move things around subsequently.

Vidy Thatte: it actually no yeah it was. it was much more simpler than that it would. you would we had a canvas you would just draw stuff and then it would just take a screenshot of it and then so if you were editing something you would be editing on the screenshot that would. this is how this is why i built it in a day because it was super simple and dumb um and also like one of the one of the things uh in the beginning was i was compressing all these images. so if you had like 10 it was like 5 to 10 remixes. oh god the first meme would be like super lo-fi um and people loved that. they were like we want this like we don't. we don't want hd images. um so yeah it was pretty fun. uh yeah. so so collage started there and then because of like crazy activity in the beginning um like we were able to raise like a big seed round and then all that happened. um hire team and then keep working on it. for so i think at its peak it was like seven people so not not too big still a lot of people to manage. so it is a lot of people to manage. yeah um yeah. so we just so we did had the team like kept working on it. uh we tried. so many experiments with it. we uh i think at first we tried we tried to like launch in colleges and then we um we tried. we we did like this competition to like uh get a lot of k-pop people. uh we experimented with groups. uh we experimented with meme pages. like we sponsored a lot of meme pages and like got a lot of them on board. um for some reason it was spreading through all the web three conferences like that. um yeah so that was also.

Nicholas: that's what that was because of your twitter connection and kind of web three's twitter presence probably because of that.

Vidy Thatte: yeah so that's probably. that's why? yeah because a lot of the twitter people were going to these conferences and like talking about it. um so that that's probably.

Nicholas: what were they using it for?

Vidy Thatte: just for making me they were making memes about tech like yeah whatever like whatever new controversy on twitter. they were just making memes about it on on collage and were they mostly being viewed?

Nicholas: not theirs particularly but just everyone in general. were they viewed mainly in app or elsewhere on the web? it's like you screenshot it mostly in app. really wow so it was really a network not just a network.

Vidy Thatte: that's how it was built from the beginning and people were sharing it on other platforms. but you can't really tell because it's a meme right. so like nobody knows like where it's from. but i think a lot of the people were like they would hear about this this meme app and they would come try it out and then they would be like oh this is cool.

Nicholas: so yeah very cool um but you decided that wasn't the way forward.

Vidy Thatte: i guess yeah yeah so we i mean we tried we tried a lot of things and i think early 2023 we like it took off in the beginning the first few months and then we were like oh we're incredible. and then um the engagement started going away. uh you know like it was really hard to like get more users and and grow. uh and i think yeah i think after trying a bunch of stuff after even trying like launching a side app on on the side like with the same similar pieces but like focus on friends um it was. uh yeah i would just realize like it wasn't going to grow um and um and like engagement was down. uh and we were just like we should like try try other um other ideas.

Nicholas: so one thing i want to ask you about here is there's this funny thing where it's like the app store has millions of apps in it and they're all getting made by somebody but there's also this kind of smaller culture of people who are in a kind of silicon valley or at least culturally silicon valley kind of yeah movement trying to make apps that are going to become maybe very popular or somehow like truly viable without just being like. i don't know the app for a bank that has a separate business that like. naturally they need an app but but there are also things like meme apps that exist out there. yeah so did you do any kind of like market research or did you get a sense of like? what makes those tick tick?

Vidy Thatte: yeah i think so. um our goal was not to build a meme app because i i think a lot of these memes are cool but they're not facebook right. i think what we wanted to build was like the next facebook or instagram or snapchat like the next big network that i'll be. you know people spend time on um and uh like we we i mean ream is like the like collage which changed the name to re. we changed the name to ream it's it's like it's still on the app store and we kind of stripped off stripped off the network. it's just a tool now so it's just a meme app um but that just by itself wasn't compelling at least to us to like keep working on um we wanted to build something more impactful and yeah we just didn't see

Nicholas: because the remix behavior itself wasn't distinct enough from other social networks.

Vidy Thatte: i guess yeah and i think the remix behavior was like really fun when you when you got remixed or when you remake something um but you had to. it's almost like you have to experience it to know what it is um and you wouldn't know unless you tried. and and it was really hard to like convince like random people to be like yo download this app create an account make a meme maybe yeah exactly yeah yeah so yeah too many steps there. so um yeah so we just didn't see how it was going to become like the next instagram. so we were like okay it's fine let's try it try their stuff.

Nicholas: and what is it that makes you want to make the next instagram?

Vidy Thatte: yeah uh i mean i think like i don't know just ambitious ambition. i just want to build something impactful that um reaches a lot of people. uh kind of goes back to like. uh contributing to culture and like you know um kind of pushing pushing humanity forward with like new tools. yeah i think that's that's what. that's probably what it is. yeah

Nicholas: definitely i mean i completely understand but it's kind of a part of the this culture that i guess people do talk about for you know making meaning in the world and contributing back to the world that i saw there's like a steve jobs quote or something to that effect on your blog.

Vidy Thatte: yeah oh was there oh yeah maybe something like that.

Nicholas: i was just reading your recent musings um but it's kind of not a part of the. it's sort of like either you're in the culture and you get why you should be trying to do the most like the sam altman quote that keeps going around or you know yeah uh it's just like you must be doing the most ambitious possible thing you could possibly do right now. uh versus like. yeah totally most of the apps on the app store are not built in that in that mode yeah.

Vidy Thatte: so i think like i think this year i've been like reflecting a lot about it and like like you know maybe maybe the when you're too ambitious you kind of miss um you you end up like focusing on the wrong things um and instead of that you should be more focused on just like stepping back trying to make the best thing trying to make trying to solve your own problem um and and like kind of taking it step by step uh and maybe one day it'll compound to be big enough that it everyone is on it. but yeah i don't know i'm still in the in that reflecting phase.

Nicholas: i'm curious about. so for collage aka rememe you said you built the original prototype in one day. surely that was possible because of the prior five or so years of experimentation.

Vidy Thatte: right yeah yeah totally i think a lot of i'd actually experimented with building um like a video editor app. uh so the canvas stuff were like. i just took cord from it and then dropped it into this and it was easy.

Nicholas: that's it. the sort of youtube thumbnails of overnight success often obscure this. five or ten years or more of learning something really well so that you can be really fast when when you have an idea and then you do enough of those little ideas and you get somewhere.

Vidy Thatte: yeah hopefully yeah hopefully hopefully.

Nicholas: i mean it seems to be going on for a while but i think it's very well. so i i guess we don't need to dwell on it. but there's wap also which is kind of be real for drinking water.

Vidy Thatte: yeah so after so after we decided to stop working on collage um i think i was. i just told everyone i was like okay we're not gonna we're not. we're not gonna obsess over like one product idea we're not gonna like you know like get attached to one thing we're gonna be like emotionally unavailable with apps. we're just gonna come up with the come with the randomest ideas. uh for the at least for the next two three months and we're gonna keep shipping we're gonna keep putting stuff out there.

Nicholas: that's a team of seven people i don't.

Vidy Thatte: uh. no so yeah. so one of the things that happened was uh i think we decided to like like kind of lay off like half the team. um mainly because um it was uh it was just like really hard to coordinate and and like you know like it was really hard it was getting harder to move fast. um and i think one of the biggest things was uh if everyone on the team is like a kind of a founder like it's easy to like change ideas and and you don't have to worry about things like morale uh things like career advancement things like you know um giving a raise like all that stuff that a big company you know or even a medium-sized company has to care about. um i just felt like we're responsible for like people's careers and all that which i i feel like is a very bad place to be in when you're pre-product market fit. um so we were like kind of came up with the decision to be like okay i think like three of us can actually do a lot more um than like having a bigger team. having a bigger team kind of slows us down. um and i think february was probably the worst month ever like it's in the last like few years. because yeah we we had to let go people um have those conversations and and like it it was. just like it was just terrible. yeah it's tough feb feb 2023. but i think right after we did that though i think me rafi and michael like met up in san francisco and we started building and it was kind of obvious that that was the right decision in hindsight because of how quickly we were able to just spin out spin up things again and uh and and just like ship build not not have like too many discussions and like trying to build get everyone to on the same page and a lot of the even like decisions we took in terms of like the kind of tools we were going to use to build the apps like all that like. it was a whole different game uh which made us move really fast because you're already all on the same page. yeah yeah it's yeah. so i think three people can do a lot. yeah yeah yeah.

Nicholas: and all three devs are not a designer.

Vidy Thatte: yeah um so so um. so basically the way it was. um i did product and design and then rafi was the engineer and then uh michael was doing helping out with gtm but recently we're all devs. now like michael became super technical has been propped engineering and and building and react like me and rafi build build stuff.

Nicholas: um so yeah so cool so cool. it reminds me of fluid. i often think about the early days of facebook and kind of yeah roping in his roommate and whatnot and i wonder how long that lasted. i suppose once they had product market fit and then teal money and etc. it maybe pretty quickly became like bring in someone who has many years of experience in some other place that can kind of build a team or knows how to coordinate a larger team.

Vidy Thatte: yeah i don't know. i i'm really curious to see how they transition. but from what i've heard i guess they stayed pretty flat for a long time

Nicholas: right but i yeah it's just interesting like it seems like maybe two different activities the kind of scrappy three people just grinding on something and then sort of okay we actually have like a million users. we need more than three people to oh yeah or maybe a million you can do. it's not such a problem. but although then again sam allman says one person billion dollar company so maybe yeah it's not even true.

Vidy Thatte: yeah it's definitely i think to the when they started in 2004 2005 like uh we definitely need less people uh to do similar things because there's just more sass and apis you can plug together and pretty much yeah yeah yeah um do you mind sharing your stack?

Nicholas: what? what are? what are some tools that you your team loves to use?

Vidy Thatte: yeah yeah so um i think i mean the main thing is if for database we use this thing called super base uh it's pretty great like post. it's built in postgres it has api calls it has like sdks for um any any anything you're using and then so for uh for web views we use next.js which is react and then we're cell to like host everything. and then for ios stuff we just use swift ui. that's it. that's pretty much it.

Nicholas: i don't know that one. what's the last one? 50 ui. oh swift ui okay yeah yeah yeah very cool. yeah i think that's kind of the stack most people are using these days or are excited about.

Vidy Thatte: um for like quickly building stuff yeah i think when during collage days like uh we we tried to spin up aws back end uh which took like weeks or even months like which took a while to get uh wire everything wired up. and yeah probably a bad decision. um but things that you do when you have a bigger team oh we have a bigger team.

Nicholas: okay okay because it's also something that just any any builder will run into they're like well i got to do this thing but i don't know how to do it so i guess i'm just gonna have to learn how to do the whole thing. yeah so do you think you're kind of progress through all these projects? i know we didn't quite get to the end of the list. uh there was uh what happened introspect also which is very cool. uh basically texting and ai comments uh sort of suggesting what you should text next as you kind of read a personal diary like a co-pilot personal diary. um yep so much would you say that the like among the things that you learned through all this trajectory was like write less code rely on apis. uh that seems like one lesson you learned. maybe you oh yeah work with less people too.

Vidy Thatte: yeah work with less people move fast um get. i think the biggest thing is just optimized to getting something out there. um doesn't matter how you do it just get it out there quickly so you get feedback and and you can learn from it. that's the best. uh that's the most valuable thing.

Nicholas: do you really have to jettison this idea of tech debt? because like it's wrong anyway and we're gonna have to rewrite the whole thing anyway?

Vidy Thatte: yeah but i mean i'd rather have that problem than you know.

Nicholas: not having teched up and not having users right having having a beautifully designed app that no one's using.

Vidy Thatte: yeah who cares?

Nicholas: do you feel like in your experience you've done that before.

Vidy Thatte: uh what do you mean like?

Nicholas: have a design like worked spent way too much time on some technical aspects and not really validated the interest in the product?

Vidy Thatte: i i don't know what it is. i have this like anxiety where if i'm not solving like the most like existential thing. next then i just freak out like i can't i just have to do it. um so so like? i think so for me it's like oh i can't spend like a like a month trying to build something or two months trying to build something uh without like knowing that you know there's gonna be like 10 people who are gonna pay for it or use it. um so i just have to. i just have to move fast and get it out there um which i think is also not a good thing because i don't think every product like especially now. um people like like really well made things and and it depends on what market you're going after um people like for certain products i feel like they have to be well built. uh they have to be fast and and well designed.

Nicholas: and uh is there an example of what kind of product people have high expectations for in your mind?

Vidy Thatte: yeah yeah i mean like. i think one of the things is like superhuman. i think they took like two or three years to get the first version out right um but when they got it out everybody was like this is great like we want to use it like. uh people started paying paying a lot of money for it. i think even figma they i heard they spent like 10 years trying to build their web gr canvas. no i think i heard three three years maybe okay okay i i don't know but the whole i guess yeah um three years to build a web geo canvas uh which is yeah which is a lot of time and i heard three years of experimentation.

Nicholas: to be fair i don't know if it was exactly on that one feature.

Vidy Thatte: yeah i don't i to be honest i don't know. uh this is just i heard this at a party yesterday. what was i saying? yeah so i think i think like yeah there are i mean. but then that became their biggest mode and they were able to like take on sketch and all these other competitors and it did feel like something nobody else could do. yeah exactly exactly um so i think it really depends on context. there's no one way of doing something but that's how i've been operating in the last few years.

Nicholas: that's interesting. uh you mentioned being at a party like it seems to me not being there myself that there are sort of a variety of kind of tribes in san francisco that are propelling. yeah not not only young but you know new entrepreneurs um into you know incubating their projects helping them and it's not just yc anymore. i know there's always been other things but it feels like there's a lot of different places. now um what's the what's the landscape like?

Vidy Thatte: honestly i i haven't been like too active i just because i've just been like heads down building stuff um but it's just these. i just know like i just know a bunch of people from twitter um and i just meet up with them and i just i just i just i just know a bunch of people from Um, and I have been seeing a lot of activity though. Like, I think, I think, uh, I, I see a lot of people from actually like F Inc. Um, I think the, the part, the podcast you were talking, I think you were talking about the podcast.

Nicholas: Um, yeah, I forget which podcast it was, but yes, they do a podcast and what is it? The build space that person was on also. Yeah.

Vidy Thatte: Build space is pretty cool. Yeah. So like a lot of these things are coming up, uh, which is really like exciting.

Nicholas: There's a place, um, where Avi is with the tab project. They're doing. Yeah.

Vidy Thatte: Yeah. Yeah. So yeah, that's pretty cool. Yeah. That's pretty cool as well. Haven't, haven't been there yet, but, uh, South park commons. Oh, I see South park. Yeah. Yeah. I mean, those, those guys were around. Yeah.

Nicholas: I don't know.

Vidy Thatte: They've been around, but that's cool.

Nicholas: Okay. So it's not, it's not really that relevant to you. You're, you're just like, you're locked in. You're, you already know what you're building. You don't need to socialize about it.

Vidy Thatte: Yeah. Yeah. I have like, yeah, I have a good group of people that I get inspiration from and I'm just building, but, but it's pretty cool though. I'm really glad all these, all these things exist.

Nicholas: Yeah. It is very cool. Um, okay. So I want to talk more about trace a little bit. So you said developers are the ones who are using sort of people in your own vein are kind of the ones using trace most right now.

Vidy Thatte: Yeah.

Nicholas: And what, what are like some of the most popular components that have been built so far? What, what do people love the most?

Vidy Thatte: Uh, I think people have been making like charts, people have been making like chat interfaces. People have been making, I don't know, like, uh, what else. I should go look at it. But like, I think loading screens are pretty, pretty popular. Okay.

Nicholas: I see like map things. There's like lists of sort of social posts.

Vidy Thatte: Yeah. Yeah. I think one of the things that I made, I was surprised that trace could do it was, uh, I built this thing where you could, you could draw on a, you could doodle and then drop the doodle on a map.

Nicholas: Maybe there's, there's a social app there, but drop the doodle on the map, like painting on a map or stickers on a map kind of. Yeah.

Vidy Thatte: Yeah. So, and then you, even if you zoomed in, zoomed out, moved up, moved around, you could see it dropped in that location. Wow.

Nicholas: Sounds like a, a one. Yeah. Like a Snapchat feature or something like that.

Vidy Thatte: Yeah. Yeah.

Nicholas: Yeah. But I guess what you're saying is, is it kind of like the speed of iteration on this does make it like, how long did it take you to make that? roughly?

Vidy Thatte: Uh, less than 30 minutes. Not even.

Nicholas: So you could imagine like, uh, because I think this is where we're headed, where we're, it's not so much about building like, you know, software as performance art has kind of been my premise for a long time. And with the AI stuff, it seems much more within reach.

Vidy Thatte: Absolutely. Absolutely. Yeah. Um, cause I think you don't have to learn a lot of this stuff you just do, and then you learn by doing.

Nicholas: It's very exciting. I guess it does kind of still require that, you know, the like tooling enough to be able to integrate these new features and, you know, that map, you want that sticker to not just live inside of the UI component, but also be like on everyone else's phone. Also, do you imagine that that's something trace will be able to do in the future or in the near future?

Vidy Thatte: Absolutely. Yeah. A hundred percent. I think we have to do it. Um, we definitely have to, uh, make, build the whole pipeline. Um, but I think, yeah, I think one of the things we're thinking about is, is that we're going should we be doing, doing it like on web? Should we experiment with like other paradigms? Like what happens if like people start making vision pro apps? Um, so a lot of, a lot of questions to answer.

Nicholas: I still feel like, you know, whatever, 2 billion plus install base. I know which one I'm going for in terms of impact.

Vidy Thatte: Yeah, absolutely. A hundred percent.

Nicholas: Now you raised from a 16 Z for a trace. Is that right? Yeah.

Vidy Thatte: Yeah. We haven't made an announcement. We have kept it low key, but yeah.

Nicholas: Uh, but it is out there. I found it. I found out somehow. Uh, I don't remember how, but it is out there a little bit, but, but I guess that gives you the runway to kind of decide what, you know, what the long-term play is that you're most interested in.

Vidy Thatte: Totally. A hundred percent. Yeah.

Nicholas: That's pretty cool. Um, but I, it's, it's funny because you got this round, but also like the site has a kind as it stands currently has this kind of homemade feel to it. It doesn't feel like, uh, um, you know, it's not, uh, what's the, no, no, it's not like the, uh, rip off of, Oh God. What's the bug tracking software. Everyone rips off their website.

Vidy Thatte: Linear. Yeah.

Nicholas: Linear. It's not a linear website. It's, it's not just a UI component library. It's like, if you, and maybe you are using UI components, but it does feel kind of like we rolled it ourselves. Yeah.

Vidy Thatte: Yeah. Yeah.

Nicholas: Is there any deliberate choice in that? Or is that just, this is what it is right now, but it'll get more professional feeling soon.

Vidy Thatte: Uh, it's a combination of both, I guess. Um, yeah. It just, whatever we could put together quickly.

Nicholas: Do you have any kind of, I have some kind of faith in this idea that like along, like, yeah. People have always wanted it, but now maybe with the tools, especially it's even enabled to have, like, to move away from apps that look like everybody else's app, you can do stuff that looks a little bit different. I guess that's a little bit contrary to the vector database of, uh, components.

Vidy Thatte: Yes.

Nicholas: How do you think about that? Yeah.

Vidy Thatte: I mean, I, the way I see it is like, you know, I think, I think there's this quote, like, if you want to be great, um, you have to learn the basics and the rules first, and then you can bend, bend the rules. Uh, so you, you kind of have to like acquire the skills and then you can really, really push the boundaries. Uh, and I think, uh, software X like AI is, is essentially accelerating that where you don't have to learn the skills, right. Um, you're getting a really customized template or like a route, like a starting point. Um, that's far more advanced than anything you would have, you know, building from scratch. And I think that you, you have to start from there, you get to start from there. Uh, and then you can actually, you know, do a lot more things with, uh, with whatever you have. So, so like, I think, you know, a lot of these AI tools are essentially like really great template generating, uh, products, even trace. And then, and then it's up to you to like tweak it and make it, uh, push it further. Uh, so I, I think this is gonna definitely help. people who care about this stuff, who have taste will, will take this and, and push it, uh, rather than, you know, being limited to whatever it generates.

Nicholas: Makes sense. When do you think someone will be able to make a little app? Yeah. Like a small app, one, you know, one button, click it on, click it off with trace.

Vidy Thatte: You can do it right now. Yeah. If th if you can export whatever you made to Xcode and, and push it to test flight or upload to the app store, you can do it right now. Wow.

Nicholas: Uh, has anybody done that yet? Have you seen anybody who's like not really a coder ship an app yet?

Vidy Thatte: I haven't, but we might do it ourselves.

Nicholas: Yeah. I think, I think it's a great, it's gonna make a great YouTube short. Yeah.

Vidy Thatte: Yeah. Pretty much. I was gonna rebuild collage, uh, and, and stream it or something. Um, using trace. Yeah.

Nicholas: Yeah. That's pretty funny. I, I, I bet also would get clicks is like rebuilding some really super hyped, very highly invested in company in like five minutes or whatever.

Vidy Thatte: Yeah. That would be hilarious.

Nicholas: So what is it really good at doing right now? It's really good at building UIs, basically little, little UIs that are things that have kind of been done before that are in your library.

Vidy Thatte: Yes. Yeah. Yeah. So if you wanna ideally like, you know, something that, uh, something that's not like crazy, like if you want a settings page. Or like onboarding screen, it'll do that really quickly. Uh, and then, and then you can drop it into your project and, and build on top of it.

Nicholas: So you kind of want to almost tailor your prompt if it's a visual prompt, a canvas drawing to kind of what the library of options that it's aware of can do easily. So it's not stretching so much.

Vidy Thatte: I mean, you could try stretching it, but you may not like the results, but yeah, if you want the best results. Yeah. Ideally.

Nicholas: So what makes for a good UI prompt? Yeah.

Vidy Thatte: I think the, you gotta be less like, like less ambiguous. Um, you gotta ha you gotta tell it exactly what you need. Um, you don't have to use any technical language. You can, you can keep it simple. Uh, and so I think, you know, one prompt could be like, make a onboarding, uh, you know, make three pages for an onboarding screen, which has like image text and then description in each page. I think something like that would, would do pretty well. I think. Just be more descriptive. Yeah. More descriptive.

Nicholas: Um, but can you go very, very long? Like if I wrote a whole page of text, would it be able to interpret it?

Vidy Thatte: I don't know about the whole page. I think there's a con token limit. Um, we haven't, we haven't implemented like the word counter thing. We should, uh, character character limit, but, but yeah, there's a, there's a token limit, but, um, but you can go pretty long. I think you're right. Like one or two paragraphs. And, but there's always, you know, the better, yeah, the better approach is to do iterative, do it iteratively where you, you start with a, you know, a, a, a, a, a, a, a, a, a, a, a, a, a little bit with one small thing and you kind of build on top of it.

Nicholas: That's what I was going to ask about. So it is kind of like a conversational UI. as much as it presents us to like, just type it in and do it, but actually you're talking to it and refining what it's given you. Right?

Vidy Thatte: Yeah, exactly. Exactly. And, and we have, you can see all the versions. Um, so every, every time you submit a new prompt, um, it builds on top of the previous version. So you can go back if you don't like something or you can see the progress as well.

Nicholas: Do you have any tips for doing that iterative process?

Vidy Thatte: Yeah. I mean. I think, yeah, it's like start-- when you want to make something, you start with-- you break down whatever you want to build into little chunks. So if I want to make a settings page, start with make a-- honestly, for settings page, you can start with make a settings page. But if I was doing the map thing that I mentioned before, where you can just doodle on a map, the best way to build something like that is start by saying make a map view. And then let me drop-- give me-- I should be able to-- if I press-- add a button, add a floating button to the bottom of the map. If I press on the button, it should drop a pin on the map. So that would be your third prompt. And then fourth prompt would be, if I press on the button, it should open a screen. And then fifth prompt, in the screen, add a canvas view so I can draw. And then final prompt could be-- or the next prompt could be, add a cancel and save button to the canvas. And then finally, it can be something like, when I hit save, it should drop the-- it should screenshot the canvas and drop it as a pin on the map, I think. So that's a rough sequence that I just made up on the fly. But yeah, something like that, where you break down a big chunk of UI into smaller parts, and then you build everything together.

Nicholas: Do you imagine that in the future, you could give that as a single list, and then it would kind of auto-GPT its way there? Or do you think that for the foreseeable future, you should do the whole render cycle compilation?

Vidy Thatte: Yeah, I mean, yeah, 100%. In the future, something like that would be incredible. I think we can actually somewhat pull that off with it. We have a few experiments that we're trying out right now, with kind of doing a chain of thought, breaking a bigger prompt into smaller parts. And then building that incrementally. But yeah, I think that makes a lot of sense.

Nicholas: But having human in the loop at each stage of compilation seems to be like-- with the technology as it stands today, it seems to be pretty-- it doesn't seem like you can get rid of it right away.

Vidy Thatte: I think, yeah, with the current state of LLMs, having human in the loop is the best approach that we've seen.

Nicholas: But GPT-5 maybe changes the game.

Vidy Thatte: Maybe, hopefully. I think that would be really cool. Because I've tried all these agents in AutoGPT and stuff. And yeah, they're not there yet. Not even close.

Nicholas: I mean, has anyone made anything that works with those?

Vidy Thatte: I haven't seen any real-- I just-- yeah, it's just cool demos.

Nicholas: Yeah. They make great videos on Twitter, I guess. Exactly. Is it complicated to render the iOS simulator in the browser? Or is that actually known? I mean, you're using somebody else to do that. But is the whole-- was that difficult to build? It was interesting.

Vidy Thatte: I think it was kind of a hack. The way we do it is we compile the software-- compile the Xcode project, and then send the build file to another service that renders it. So yeah, it wasn't tricky. But I mean, I actually put it together in two days or something. But it was something we didn't know was possible. And then once we got it working, I was like, oh, crap. This is cool. Yeah.

Nicholas: Totally.

Vidy Thatte: I don't think many people have done that specific workflow.

Nicholas: No, I mean, I think the only-- yeah, I don't know. I guess I don't know the iOS scene as well. So I don't know what else is out there. But it's pretty--. I mean, honestly, I think that is my favorite part of it. I don't even need it on my phone just to be able to try it and have immediate feedback right in a browser without having any kind of--. Because the installation process for the test flight thing is very, very complicated at present. Do you think that's going to get easier? Or do you think you're going to mainly focus on making the web experience slicker?

Vidy Thatte: I mean, when I'm making apps, I never use a simulator on the screen. I always run it on the phone all the time.

Nicholas: Not for each iteration.

Vidy Thatte: Actually, for every iteration. Like, when I'm using Xcode, not Trace. And it's pretty fast, because you can just plug your phone into your computer. And the thing about it is Apple has this weird system where your phone has to be registered to your developer account for it to run stuff on the fly. So we had a feature where you could scan a QR code on Trace and then load the app on your phone. But that was just-- it was super hard. Like, the setting up process was really hard. You had to buy a developer account. And most people just ran into issues with it. So we just deleted that feature last week. So what's left?

Nicholas: You can still do something like that? Or there's no pushing it to the phone at all anymore?

Vidy Thatte: Not at the moment, but I want to figure it out. So that you can just-- anything you make on Trace, you should be able to run it on your phone.

Nicholas: Can you run a simulator in a web view of a PWA or something, or an app on the phone that would make it feel like it's the same? I guess not really.

Vidy Thatte: I've thought about that, but it's not the same.

Nicholas: You want that scroll speed, the gesture speed.

Vidy Thatte: Exactly.

Nicholas: No delay, no latency. What are you most excited about doing with this? I mean, you've built so many apps. You have so many ideas and the like to build them yourself. So are you doing anything fun with it, or have you not had time yet?

Vidy Thatte: I'm actually building a calendar app on the site. I also had an idea yesterday. I mean, this isn't an original idea, but for calorie tracking, I should be able to take pictures and have a nice little Streaks view. Yeah, I think I'm going to build all of that in the next few months.

Nicholas: Very cool. That must be pretty exciting.

Vidy Thatte: It's fun. It's cool, yeah.

Nicholas: Amazing. Well, Vidy, this has been a wonderful conversation, and it's great to get to finally talk to you after falling asleep. I've been following you for so many years on Twitter.

Vidy Thatte: Yeah, likewise.

Nicholas: Are you-- I guess, what's next for Tracer? What do we have to look forward to?

Vidy Thatte: I mean, just use it. Try it out. Let us know what you think. We're going to be-- we're going to keep shipping new updates. And yeah, I think that's pretty much it.

Nicholas: That's pretty exciting.

Vidy Thatte: Yeah.

Nicholas: More of the same, but the same is pretty exciting. I'm looking forward to it. Thanks so much for coming on Web3 Galaxy Brain. This is great to talk to you.

Vidy Thatte: Thanks so much for having me. You're welcome.

Nicholas: I'll talk to you soon. Hey, thanks for listening to this episode of Web3 Galaxy Brain. To keep up with everything Web3, follow me on Twitter @Nicholas with four leading ends. You can find links to the topics discussed on today's episode in the show notes. Podcast feed links are available at web3galaxybrain.com. Web3 Galaxy Brain airs live most Friday afternoons at 5:00 PM Eastern time, 22:00 UTC on Twitter spaces. I look forward to seeing you there.