Transcriptly

Kortex: The $3 Million Dollar Mistake

Dan Koe54:00Added Aug 31, 2026

Kortex, our baby that we've been working on for 2 years, failed. Due to reasons discussed in the podcast, we needed to make a drastic change. So, we decided ...

Watch on YouTube →
Contributed by 刘嘉琪

Transcript

Transcript format

  1. Kortex: The $3M Mistake

  2. 00:00Cortex, the $3 million mistake. One of the things I started thinking about is like what are the problems with
  3. 00:06Cortex? Like we can't make any step forward because [music] we're just like marching
  4. 00:11in place. Then the comments coming in and everything else. It's like what are we doing? Let's set these goals. Let's make them
  5. 00:16really aggressive [music] and see what happens. And we did that. And we saw that the whole car fell
  6. 00:22apart. [laughter] We need desktop. We need light mode. Light mode. Yes. cuz we could like
  7. 00:28that's how that's how slow we were going is like it was difficult to get light mode of all things and so we threw away the one and only
  8. 00:34advantage that we really had as a startup which is moving fast. [music] We haven't tried. Why don't we just try?
  9. 00:40Like surely if we do those things we can rebuild the entirety of Cortex and more
  10. 00:45in [music] much shorter time. Literally 3 3 weeks into being in Japan that hypothesis turned out to be [music] true.
  11. 00:50In terms of like rebranding from Cortex to Eden, what comes [music] to mind for you? It's like a search engine for your
  12. 00:57memories. I just went in and searched pink sweater and it resurfaced every single frame where someone [music] was
  13. 01:03wearing a pink sweater at the exact timestamp. This is insane. I haven't seen this [music] before. The second
  14. 01:10that you think you know what you're doing, you're kind of [ __ ] And it's like, I don't know how to freaking manage a team, dude. I don't know the
  15. 01:16first thing that goes [music] into this. So, like me going into it, as long as I saw some form of progress, I'm like, ah yeah, we're actually doing something.
  16. 01:22success is just like a a series of perhaps well-calulated risks, but like a
  17. 01:27series of big risks. Because as humans, we're very prone to [music] remaking the same
  18. 01:33mistakes over and over again. Well, yeah, we're here to talk about why

  19. How We Met: From Strangers to Co-Founders

  20. 01:38Cortex failed. Yeah. $3 million mistake. That's the title of this video. Very attention grabbing and polarizing, but
  21. 01:45yeah. Yeah. Uh on the topic of cortex being a mistake um I don't think it was actually
  22. 01:51a mistake but I want to hear like your take on that right you I mean you know my philosophy mistakes you have
  23. 01:58to make them in order to actually improve. Yeah. So do you see it as a mistake? I I I see it as a series of necessary
  24. 02:06mistakes for us to get where we need to go. Um, and fundamentally like I don't think the idea of building Cortex, the
  25. 02:13second brain app, was a bad one, but there were a lot of mistakes that we made as like firsttime founders that
  26. 02:20went into Cortex that uh definitely can be categorized as mistakes. Yeah,
  27. 02:26if you follow me, you know I talked for about Cortex non-stop for however many
  28. 02:33months or even years and then just stopped all of a sudden. And then people
  29. 02:38are like, "Oh, what's happening to Cortex? Where's Cortex?" Uh, a lot of
  30. 02:43things happened and we decided to pivot and build a new app, Eden. So, that
  31. 02:49sparks a ton of questions of like, okay, why didn't you stay building Cortex? Why are you doing Eden? Why did you rebrand?
  32. 02:55The first thing there is like where we stopped with Cortex wasn't our vision for it. We kind of
  33. 03:02just stopped there. Yeah. Because we were forced to for the reasons we're going to talk about. Yeah,
  34. 03:07but to frame even further and to say this early on in the video, Cortex users, you probably already know what's
  35. 03:13happening. You've gotten emails about it. If you don't know what's happening, starting Monday, the week right before
  36. 03:19the Monday right before Black Friday, we're going to start rolling out Eden to you. So, expect an email on that. Log
  37. 03:26into the app if you don't get an email cuz you'll get a banner in the app to transfer everything over if you want to.
  38. 03:32But for people who have not used Cortex, have been hearing about Eden, are very excited about it, Black Friday weekend
  39. 03:39from Friday to Monday, we're going to try and open it to the public and see if things are okay.
  40. 03:46Please take note, [laughter] this is early access. This isn't like any kind of public launch. Cortex users, it's not
  41. 03:51going to be up to par like having desktop app and mobile and everything else with Cortex. So, just please keep
  42. 03:57an eye on all the information. But what was Cortex? Let's start at the
  43. 04:02beginning. Like the very beginning. Yeah, the very beginning. How did we meet? For context, for people who have no idea
  44. 04:07who we are, I'm Matt, one of the co-founders of Cortex and now Eden. I'm Ari. Same story.
  45. 04:13Yeah, he's the technical guy. Deep deeply technical guy, our CTO. I'm the product guy. And we are both well for me
  46. 04:21were was but we we're both students at the University of Toronto. Um I I was in computer engineering. He's in computer
  47. 04:27science. And we both kind of have a background of like wanting to do something entrepreneurial or wanting to
  48. 04:33do something that's like not the conventional route. For me, I I went through the first couple of years of uh
  49. 04:39school in university, kind of just like pretty disillusioned with studying and going down the traditional computer
  50. 04:45engineering route, like getting a degree and then just getting a job. Pretty early on in university, I I decided to
  51. 04:50go part-time and then take on a bunch of odd jobs and do doing like random things
  52. 04:57outside of school. Long story short, I ended up hopping through a series of different startups. First, initially
  53. 05:04doing content and then contracting as someone doing demo videos for this other startup. Um, and that was right before I
  54. 05:11had met you. story for how that happened was that when I was at the last startup
  55. 05:18working as a front-end person, I realized that I did not want to do that for the rest of my life. I decided 2
  56. 05:24days before meeting you, I'm going to write an Apple note. And in that in that note, okay, I'll start with
  57. 05:29ghostwriting. I'll start with like finding clients online and then eventually I'll systemize this process into like a software because I had
  58. 05:36learned some software skills and eventually I'll probably partner with Dan Co. And then your friend Dakota
  59. 05:42Robertson happened to be running a ghostwriting cohort. Halfway through that call, you basically said
  60. 05:48off-handedly that you were building a software. And then I was like just in my brain I'm like there's no there's no
  61. 05:54way. There's no way, right? Like and then and then I unmuted myself in front of everybody. There were like 50
  62. 06:00strangers there. I just like decided, you know what? Like I might as well take a shot. Nothing but bad can come out of
  63. 06:07this. And then I just asked like what software are you building? Yeah. And it turns out you you were working on
  64. 06:13an idea that was very very very similar. So then I'm I just like start shamelessly barging you in the Zoom call
  65. 06:19with like oh like uh here here's like everything I've worked on before. I can help you with that. Like like here here
  66. 06:25are like some previous links for things that I've worked on. Well the thing is like sorry there is
  67. 06:30that entire time I was thinking like okay okay like I've heard this before like [laughter]
  68. 06:35I have a lot. Yeah, I'm sure you have. A lot of it doesn't bear any fruit. And so, like me being
  69. 06:41in the spotlight on the call, I was just like, "Okay, yeah, I'll give you my telegram and we'll talk after there."
  70. 06:47Yeah. I remember Dakota, I'm I'm so sorry if you're listening to this, Dakota. I'm so sorry for interrupting
  71. 06:52your thing. Um, and I'm pretty sure he probably got a little annoyed when I was doing that as well, cuz he was like,
  72. 06:58"All right, I'll connect you guys on Telegram later. You can you can like sort it out offline." He was kind of just like, "All right, this guy's like
  73. 07:03derailing the conversation. [laughter] like let's get this back on track. But then me having gotten your telegram, I was
  74. 07:09like, "Oh, [ __ ] Yes, this is an opportunity." So then I remember like right after the call, I just like
  75. 07:15I can literally pull it up right now. Like if we scroll all the way back up, like I just barged you with messages
  76. 07:21and then you were like not interested at first. [laughter] Well, that Yeah. Then I I sent you cuz I
  77. 07:26thought we were doing everything right. Like this was our mine and Joey's first dive into software. Yeah. and we just
  78. 07:34offshed this entire process, this entire podcast
  79. 07:40is going to show like what iteration actually is and it's a lot.
  80. 07:46Yeah. But yeah, you you I sent you that and what we've been doing and then I kind of
  81. 07:52just to summarize what the message back was was like, yeah, you need to burn
  82. 07:57this all down. It sucks. And then I'm like, "Oh, okay. Now I'm listening."
  83. 08:03Over that weekend, you and Joey had decided to fire that guy, bring me on. And then I think
  84. 08:09at that point, maybe you started to take it more seriously, and then you you incorporated the company
  85. 08:14and then the the name Cortex came about and then you injected like the first amount of basically seed funding. You
  86. 08:20brought me onto the call. I saw what the guy had and then TLDDR for the for the technical people that guy had spent like
  87. 08:275 months building basically a CRUD app which Ari can pull off in like two hours and now Ari could do in like 30 seconds.
  88. 08:33Yeah, that's just what I [laughter] did now. So yeah, to to give Ari some limelight here, this is
  89. 08:41the most cracked engineer I know. Well, I don't know many. No, no, no. 100%. I I I think it's like
  90. 08:47I I know some crazy cracked people. Like I know pe I know people working at
  91. 08:52autopilot and Tesla like meetings with Elon Musk and so on. Like I think he is
  92. 09:00at least on the same we need to keep him quiet. he's going to be getting a bunch of emails with [laughter]
  93. 09:05I'd say like technical skills-wise I wouldn't say I'm like super cracked or anything but the thing that I do have
  94. 09:11going for me is I would disagree but is like the thing that really makes a
  95. 09:16difference especially in startups is just persistence like going into it like my first attempt at startups like I had
  96. 09:23zero knowledge about any sort of like web development or front-endf facing things yeah and then you decided to build a
  97. 09:29text editor yes that's an insane first project to take on. Yeah, that's kind of how we met. My conditioning from a very young age
  98. 09:36because of my parents was that, you know, anything's possible. You can do whatever you want. Total opposite what my parents told me.
  99. 09:42[laughter] Props to all your parents. Yeah. Yeah. They they always made it seem like, okay, you always have a fallback
  100. 09:48with us. So, don't worry about being safe. Try experimenting. Try something
  101. 09:53crazy cuz you know, you always have a home with us. So, why not? A that's sweet, right? And my dad always put this
  102. 10:00business aspect into my head where he was like, you know, even if you like sell a toothpick, if you can sell it to
  103. 10:05enough people, you'll make a lot of money. So like [laughter] that's a great analogy. Start small. Start with whatever you'll
  104. 10:11you'll make it work somehow, right? I went into a incubator program and the
  105. 10:17incubator program told us to structure things in a way that I didn't agree with where they wanted us to like think about
  106. 10:23the market think about right like the business proposal and project revenue before you had projections without a product ever
  107. 10:30[laughter] being built which I was I did I do not agree with that philosophy but maybe it works for
  108. 10:36different kinds of startups. We ended up just building a prototype, brought it to them and they just told us, you know
  109. 10:42what, you're too young. You should like come back when you're like about to graduate. And I'm like, and the prototype that you had built was
  110. 10:47embed notes, right? It was a note-taking app. It was a note-taking app. That was my first ever attempt at building something
  111. 10:53outside of like school just in the hopes of turning it into a real product. And when you get a feedback from people who
  112. 10:59are like much senior than you, that you know, you should just come back when you're like older, you're like too
  113. 11:04young, it's it's kind of like it feels bad. It feels really bad. It's like, what do you mean I'm too young? That was
  114. 11:11that it was because you weren't a fourth year student. Yes. I wasn't graduating, right? And to me that just seems like dumb. Like there
  115. 11:17shouldn't be any age uh requirement to like going and doing stuff. So I took that advice and I said, "Okay,
  116. 11:24you know what? You want me to be a fourth year student?" I combined my second and third year education into one.
  117. 11:29Yeah, he he took all eight credits. Eight credits in one year, which is
  118. 11:34ridiculous. Finished as much as I could. I'm like, "Okay, I'm back. I'm a fourth year now." [laughter]
  119. 11:40And that was a few weeks before you met us. Yes. Uh I was doing this while I was
  120. 11:45working at a different company called Flip and I did software engineering for them for a year just to see what like
  121. 11:51industry feels like. And towards the tail end of that like a co-op program, right? Yeah. A co-op program. Towards the tail
  122. 11:57end of that, Ian was already working with you guys. Yeah. So for context, we have a team member called Ian. He um he was one of

  123. The Fatal Team Structure Problem

  124. 12:04the first people that I had like kind of like brought on. Ian was initially responsible for building out the first
  125. 12:10couple of editor features for Cortex. And the problems that he would run into in
  126. 12:16trying to build an editor, which is a lot more complicated than it it sounds, he he would often go to Ari because Ari
  127. 12:23had worked on an editor project like for for his for his first startup. So then after consulting Ari like several times,
  128. 12:30Ian just thought, wait, why don't you just join the company? I guess like this is where we start to get into what were the mistakes that we made with Cortex.
  129. 12:36The very first mistake we made was how we structured our team. I decided like oh like let's let's let's bring on
  130. 12:43this person and then we'll we'll make him front and lead and let's play on. Well, the the thing about like one I
  131. 12:49come from the quote unquote one person business background or like me and my editor after that and it's like I don't know how to
  132. 12:54freaking manage a team, dude. I don't know the first thing that goes into this. So like me going into it, as long as I saw some form of progress, I'm
  133. 13:01like, "Ah, yeah, we're actually doing something." It was compartmentalized based on tech stack. And that was fundamentally the
  134. 13:08wrong approach because years into a year and or or two into doing that, we
  135. 13:14realized that way of doing things at an early stage startup is so slow. like
  136. 13:21you're structuring your team the way big tech companies structure their teams for
  137. 13:27the wrong reasons. It makes much more sense for every single person on the team to be laser focused on product to
  138. 13:34be laser focused on like what is it that we're actually trying to build and can we get product market fit and what that
  139. 13:41translates to is everybody should be full stack. Everybody on the team should be working on features not parts of a
  140. 13:48feature to serve another person on the team. Let me make this like tangible uh to the non tech person. Yeah. Is like
  141. 13:56what this means in a sense is okay as a startup or a small team who
  142. 14:02don't really know what they're doing yet especially like coming from our backgrounds is like iteration is one of
  143. 14:08the most important things. We need to change things. We need to test things. We don't know what's the theme of this
  144. 14:15podcast or you can't improve without mistakes like we had to put something out there
  145. 14:21see how people used it see how we used it and then from there it's like okay this isn't really going to work let's
  146. 14:27rip it out. Yeah. But then the person that built that is like okay we can't keep doing
  147. 14:32this because then it throws everything else off and introduces all these other bugs that is like cross domain rather
  148. 14:38than the thing that the single person is working on and it just slows us down rather than what we're trying to do
  149. 14:44which is speed up to get to a point where we have something that's good. Right? That's the entire thing.
  150. 14:50As a startup the number one advantage that you have is that you're small, you can move fast. Um, and it's like why
  151. 14:56will these big players not catch up to you or like what what what do you have that they don't? It's speed. And
  152. 15:02fundamentally we structured our team and our technology which you can get into in a way that was antithetical to moving
  153. 15:09fast. Um, and so we threw away the one and only advantage that we really had as a
  154. 15:15startup which is moving fast. Yeah. So the AC cortex, it works. There's some bugs. Sure, those bugs get
  155. 15:21fixed, but then like we get the emails of like, hey, you guys said this update
  156. 15:28was coming. Yeah. Where is it? And that's a month later. And we made a big mistake of like
  157. 15:34we we had this like ambitious uh goal that we set cuz we were getting to this
  158. 15:40point where it's like we're not moving fast enough. We need to like push harder with the structure that we had, which
  159. 15:47wasn't going to work. But we're like, "Okay, we need mobile. We need desktop.
  160. 15:52We need desktop. We need light mode." [laughter] Light mode. Yes. Like that's how that's how slow we were
  161. 15:58going is like it was difficult to get light mode out of all things. And there were like a few other things,
  162. 16:05but like the big ones is like mobile, desktop, and mobile's a huge one for a note-taking appuge. That's another thing
  163. 16:11with Cortex is like we had 80,000 users, but we were so siloed that the amount
  164. 16:19the the speed at which we could build things. Our mobile experience, which is like how
  165. 16:2480% of people are going to find you if they're watching YouTube on your phone or they see a tweet of mine or Instagram, they go to the uh Cortex
  166. 16:31website, they sign up and then they're like they log into the Cortex web version and
  167. 16:37they're like, "Okay, this is crap." And it's like sure we have this many
  168. 16:42users but how many did we lose along the way? We started out only free and then
  169. 16:47introduced paid tier and the paid tier like didn't even make sense because we were a note-taking app only. Only did we
  170. 16:53start like actually seeing a good amount of revenue when we introduced the AI feature but the AI
  171. 16:58feature is like the USP of that is oh you can use all models. Yeah. In there you can use any model. So you
  172. 17:05can technically quote unquote replace chat GPT, Claude, Gemini, etc. And you can reference notes and it's like that
  173. 17:11is a big selling point, but it's nothing like fancy. It's nothing uh it's nothing that you don't get in like
  174. 17:17say notion or even in claude built in now. Yeah. Yeah. Exactly. And so
  175. 17:22we're at this point where we're like we aren't a compelling app. Like it's good. It works. I do all my writing in it. I
  176. 17:29like it. I couldn't see myself living without it. But it's like there's nothing spe this doesn't feel like a
  177. 17:36startup that we're building. It feels like we're just rebuilding Obsidian with an AI chat latched on. And I didn't want
  178. 17:43to face that for the longest time. People would comment that. Yeah. And I just be like, "No, that's not what it is." Because I have this different
  179. 17:49image of it in my head and I was I spent too much time justifying that. That was my mistake is we should have
  180. 17:55called it quits earlier. Yeah. People see Cortex, but then they're like, "Okay, what about mobile?

  181. Why We Couldn't Ship Mobile or Desktop

  182. 18:01What about desktop? Hey, I haven't heard from you in a month. Where's the update?" Yeah. And it's like we're just stuck like on
  183. 18:08maintenance and bug fixes. Like we can't make any step forward because we're just like
  184. 18:13marching in place. There were also many many deeper underlying like technical decision problems.
  185. 18:18Yes. for example, literally the choices that we made for authentication for for
  186. 18:23for how we scaled hosting etc. I think you should like probably get into it better than I can.
  187. 18:29I came into Cortex a couple months after uh it was already set up
  188. 18:34and I was just brought on for initially engineering like just part-time engineering work.
  189. 18:40Yeah. So I come in and the the way the stack is set up looked very similar to what I've seen at other big tech
  190. 18:46companies where we had like our own open- source um authentication system
  191. 18:51hosted and our own like EC2 instances and we have guys like running around
  192. 18:57with Terraform scripts and a lot of other stuff like that where a lot of things that we decided to build
  193. 19:02ourselves that was the fundamental problem and I've seen like even midsize companies you have like groups of five
  194. 19:08to six people whose full-time job is handling all of that stuff and we're a team of five to six people doing that
  195. 19:15plus building the product. So to me looking at that I just went okay this is
  196. 19:20going to be a bit of a challenge. Yeah it's tough and you can't back out at that point. Yeah, you can't back out because people
  197. 19:25start using it and it's like it's like um yeah, just just to like nail into like one
  198. 19:31example, right? Um and for for the nontechnical people like authentication was one of these problems. For the nontechnical people, authentication is
  199. 19:37just logging in signing up very very basic stuff like you expect any app to just have you just log in and sign up.
  200. 19:43Like what could what could be so hard with that? Sign up with Google. What could be so hard about that, right? And like the [snorts] answer is it shouldn't
  201. 19:49be that hard for simple. It should be really really simple unless you try to build it yourself. [laughter]
  202. 19:56And I think it's a mistake that a lot of first-time founders make. Um where you
  203. 20:01think you think to yourself it would be very valuable if I were if we as a
  204. 20:07startup were to own all of this IP if we were to control and own everything. But
  205. 20:12all of that and uh and trying to scale prematurely or trying to trying to worry
  206. 20:18about like, oh, I want to own the IP so I'm not relying on third party vendors way too early on translates to syncing
  207. 20:26literally thousands of dev hours on building those things yourselves and not spending that time on working on the
  208. 20:33actual product. Yeah, that's the that's the thing is like when I hear you guys say this, it's still in
  209. 20:39my head as someone who's never seen the code or done anything with the code is like, okay, well, how hard can it be to
  210. 20:44set up? I think that puts it into perspective is people hear this and it was like, oh, well, that's a like a stupid mistake. Like it it
  211. 20:51shouldn't have been that much of a problem. That's the thing is like entire companies are built solving
  212. 20:58authentication. Yeah, that shows how much of a problem it can be if you do it yourself is you're
  213. 21:03in essence like you become an authentication company somewhat at that point, but now you're split between what
  214. 21:08you're trying to build and building authentication. And that's just one of the pieces. Yeah. Well, it's like if the product is
  215. 21:14a second brain or what we envisioned it to be beyond that and now we're doing something that another company does,
  216. 21:20especially as a small team. Yeah. Then it's like, what kind of company are we? We say
  217. 21:26we're a second brain or whatever like does our time reflect that?
  218. 21:33No. So that's like the problem of it. Yeah. And and and like authentication is like one of many many examples. Another
  219. 21:39one is like trying to build a sync engine ourselves. Um yeah, and and like
  220. 21:44off offline mode was the other thing we requested because uh since we envisioned it as so much
  221. 21:50more than just a second brain by the time we had the second brain portion of it done, people are like oh well
  222. 21:57obsidian has offline. Like well thanks for just like ruining our [laughter] entire day. Like
  223. 22:02like we know like come on like that's not the only selling point of this but
  224. 22:07there's a point there and it's like it shouldn't have been should it would have
  225. 22:13been difficult but it shouldn't have been that difficult shouldn't have taken like notion as an example everyone has
  226. 22:19been like crying to them about offline mode and what four years later now they finally have it and it's like the way
  227. 22:26that we were building it we were heading down that direction as well. Yeah. And the aggressive deadlines thing that that was my bad.
  228. 22:32Well, I was all for it is the thing is like, oh, okay, like we're going to make some progress here. Yeah. The way I saw things moving was
  229. 22:38like, okay, maybe we just don't have enough pressure on us. Maybe we're just like too comfortable because one of the
  230. 22:44things like as a result of us structuring things in a more big tech way was the mindsets of our team became
  231. 22:51more big tech where it's like, okay, it's a 9 to5. Yeah. What's wrong is at a big tech company, the product market fit is
  232. 22:58guaranteed. It's the guaranteed profit's guarant.
  233. 23:08Your number one mission as a startup is to find at least in the early stage is to find product market fit so that people so
  234. 23:15that you actually make something people want. And that's the fundamental thing that like we did not spend our time on
  235. 23:22like focusing on the fact that we haven't yet built something that people actually want. And so I thought, okay,
  236. 23:28we need to do something different, right? Let's set these goals. Let's make them really aggressive and see what
  237. 23:34happens. Because then once you once you like push it to the speed limit, you start to see what things are breaking,
  238. 23:39what can break. Yeah. Yeah. And we did that. And we saw that the whole car fell apart. [laughter]
  239. 23:46Like what's the worst thing imaginable for a note-taking app? You write stuff and it disappears. Yeah. And we caused a lot of that by trying to
  240. 23:53build our own sync engine. Yeah. And it the the other thing is like we were confident in that not happening
  241. 23:59and then it happened and we're like oh my god like we don't like this is like the cardinal mistake that we shouldn't

  242. More Of What Made Us Fail

  243. 24:06have made. And so those come in and we split attention again and it's like as we continue to change things and we're
  244. 24:13digging ourselves deeper into this tech hole that is already causing problems. It's like they just keep coming up. And
  245. 24:21going back to like the very beginning, it's like that wasn't our vision for the app. So, we're further solidifying
  246. 24:27ourselves as this thing that we didn't mean to be. I guess all of this leads up to like um me and Ari getting really
  247. 24:34frustrated over time with like just not being able to make actual progress on the product itself and we decided to
  248. 24:41take a trip to to to Japan. Leading up to the trip, we've had several conversations. We had many many
  249. 24:48conversations and between ourselves and with other people on the team like hey what do you think about rebuilding?
  250. 24:55What do you think about like just like throwing everything not throwing everything away but like just like
  251. 25:00putting everything aside for a bit and try rebuilding in a drastically different way that is much leaner much
  252. 25:05quicker to to to to iterate on and not like entrenched in all these open-source
  253. 25:12or like build-yourself technologies that take insanely long to iterate on. But then Ari and I were both on that trip
  254. 25:18and then we decided you know what be because it's like kind of like a vacation to everybody on the team. It
  255. 25:25just looks like oh we're just flying out to like have fun for like a month or whatever. But then to us it was like you know what because we're we have this
  256. 25:31opportunity anyways while we're overseas let's just in secret from the rest of the team try to rebuild it like we
  257. 25:39haven't tried. Why don't we just try? Yeah. Surely it can't be that hard if we
  258. 25:45were to choose technologies that we're more familiar with. Outsourcing a lot of the problems and abstracting away a lot
  259. 25:51of the problems to like third party providers. Like surely if we do those things we can rebuild the entirety of
  260. 25:56Cortex and more in much shorter time. Like we just started out with that hypothesis and literally three three
  261. 26:03weeks into being in Japan that hypothesis turned out to be true. So it went to show that like we should have
  262. 26:09scrapped things way way way way earlier because if we can do it in three weeks at that point, what's stopping someone
  263. 26:15else from doing it in a month as well? Yeah. Yeah. Well, that's the other thing is we would see all of these like great apps
  264. 26:20coming out and we'd send it in the group chat and be like oh this is cool. Yeah. And like we'd get inspiration, like a
  265. 26:26short hit of inspiration, but then it's like uh we can't we can't do it really compete. Like we just knew every
  266. 26:32time we saw an app come out and do really well that it's like oh we were planning on doing that but like we we're
  267. 26:39stuck in the mud like it's going to be 6 months before we can get on that. It wasn't originally planned to be Eden. We
  268. 26:46were just planning on I guess replacing Cortex. Exactly. Yeah. So aside from the
  269. 26:54technical issues with Cortex, we were kind of in that
  270. 27:00a lot there were there are a lot of tech companies that have the blurp branding
  271. 27:06[laughter] like blue/p purple. Yeah. Glowy gradients. Yes. Uh dark theme. I wanted that because my
  272. 27:14branding is black and white and so I'm like we're going to be launching it to our audience and all of this other stuff
  273. 27:21and like we once we decided, oh, we're rebuilding it. We actually got to take a
  274. 27:27good look at what we had built and be like, okay, Cortex doesn't look that
  275. 27:32good. People like it because it resonates with them and they're more of like that dark techy vibe, but we're not
  276. 27:39reaching the audience for the vision that we were going to achieve. So that led to like, okay, maybe this isn't just
  277. 27:46a replacing Cortex. Maybe this is a rebrand as a whole, but it's an entirely
  278. 27:51different tech stack. So why don't we actually start over? Like this is us starting a new company,
  279. 27:57clean slate. But this also imposed some limitations is like

  280. The Japan Trip That Changed Everything

  281. 28:03the Cortex users trusted us so much. How do we, I guess, make it up to them? How do we
  282. 28:09get everything that they've been looking forward to for so long into this like incredible experience? And we may be
  283. 28:15hyping it up now as a reminder when we roll out to Cortex users and open for
  284. 28:21Black Friday weekend, just know that that's like early access. Treat it as beta. It's only going to be web.
  285. 28:27Mobile will come out when it comes out. We'll say [snorts] that rather than giving a date, which is
  286. 28:33something that we've learned is if you have a deadline or if you have a public date you want to give out, push it back
  287. 28:39by like two weeks. Just give yourself some breathing room so you're not Yeah, Eden we think is really cool, really
  288. 28:46good. Like we'll go over features, but with that there's obviously going to be things that can be improved. So, if
  289. 28:52you're joining during early access, give us feedback and know that you're coming
  290. 28:58in on something that isn't its full potential, but is much further along the
  291. 29:05path than Cortex ever was. So, do you in terms of like rebranding
  292. 29:10from Cortex to Eden? What comes to mind for you? What were the decisions behind that? One of the things that comes to mind is
  293. 29:16it's really hard to get across what a second brain is to anybody who's not in like the productivity niche. It makes
  294. 29:21sense as a concept, but as like a process, it doesn't. Yeah. Right. Well, one
  295. 29:26I guess good bit of context here is that while we were building Cortex, like I
  296. 29:32was trying to do what I do best in building community. Yeah. Uh doing writing cohorts to educate
  297. 29:39people to use Cortex. uh creating tutorials on the Cortex YouTube and a lot of them were fairly
  298. 29:46popular but when people would watch them they we'd get questions and support a lot of like okay well like I have to use
  299. 29:54tags I have to use backlinks I have to connect things like I'm not really
  300. 29:59getting what I have to do here part of the original vision for me was also just like what is a second brain if you
  301. 30:06actually boil it down to like what what function it serves well we think about like well you can think about what a
  302. 30:12brain your first brain does. It it stores and retrieves information right that that's ultimately what a brain is
  303. 30:18good at doing like a storing and retrieving information really really quickly. You store information so that you can quickly retrieve it later so
  304. 30:24that you can use it in your projects so that you can use it to create content or so that you can use it to write books so
  305. 30:29that you can write you can you can write research papers write video scripts um use it for any kind of creative or
  306. 30:37information related endeavor. Another thing with that is like it's a curated space. Yeah. It's not like yeah you can
  307. 30:43Google search, you can ask AI, you can tap into the entire like water hose that
  308. 30:48is the internet of information, but this is the place of like your favorite ideas, the ones that mean something to
  309. 30:54you, the one that you saw and you're like, I need to save this for a project I use later. Yeah. But if that's not findable or
  310. 31:01usable, then you're back at ground zero and you either need to come across that idea again or you need to somehow type
  311. 31:07in the right keyword into Google or AI in order to get that idea back to you. And another way to put it would in my
  312. 31:12mind was just like it's like a search engine for your memories. It's like you you can just really really easily
  313. 31:19whenever you come across a video that you like, someone says something that you like or a podcast that you like, a blog post that you like, just hit a
  314. 31:26single shortcut, not have to think about anything and it's just saved somewhere. And then when you want to resurface the
  315. 31:31ideas, you can just save, you can just search by exactly what you remember. So for example, oh I remember like Dan Ho
  316. 31:38said that one thing about one person business and I just hit one keyword shortcut search up like Danco one person
  317. 31:44business or whatever and it's the exact frame or the exact moment from the video
  318. 31:49that I remember not just every moment from the internet where Danco talks about this. It's the one where I saved
  319. 31:56because the way that he worded it there is the way that like really made it click for me and the way that I want to

  320. Eden's Search Feature

  321. 32:02reference when I'm writing. So it's like almost like the search engine for your memories. And I think about that and I
  322. 32:07compared to what Cortex was and I was like it's not there. It's so different. It's manual process. It's tags and back.
  323. 32:14And and it just it just felt like if this was the original vision then why
  324. 32:19are we still stuck on this? So Eden for me is like okay well what do creative what sparks the word creative
  325. 32:26or artistic or some other things of that nature? And we started brainstorming
  326. 32:32names. But the branding we knew had to be brighter. It had to like fall into that space and it had to be we had to be
  327. 32:38able to see it getting into millions of people's hands. Like Cortex for us, even though Obsidian
  328. 32:44did great, it's like we don't really see that like Cortex. Like do you hear that as like a household name? Oh, what do
  329. 32:51you take notes in? Cortex. It's like I don't really want to use that. It's it's also it's also just like every single
  330. 32:57time I put it in front of someone like put the actual app in front of someone and just anecdotally especially when I
  331. 33:04did this with like some of my my uh friends who are girls like if I would put it in front of them immediately their reaction is like oh this doesn't
  332. 33:11look like something I would use. Well we're we were trying to build something special as well like if we would have had the mobile app and would
  333. 33:17have been called it like Cortex notes. It's a note-taking app. Go in take notes. Oh cool AI chat tacked on. But
  334. 33:23since we were like in this state of dissonance of like we're not a note-taking app, but our product reflects that, our messaging around it
  335. 33:30is like, oh, it's this thing, really cool system. Yeah. You store all your knowledge. We
  336. 33:36over complicated it to make it more novel than it was because we knew we needed to build something more novel and
  337. 33:43we were marketing the product as more novel when it wasn't. So that was like a point of confusion.
  338. 33:48Yeah. Which leads to what's actually novel about Eden. Yeah. I think. Yeah. So, uh, one other thing is you mentioned
  339. 33:55in the second brain process, uh, like being able to capture or store information and retrieve information.
  340. 34:01Like you're big on the search feature. I'm big on the canvas feature because I'm always viewing things from
  341. 34:06the lens of like how am I going to use this in my workflow? And if my like if I'm a
  342. 34:11if I'm not able to write in it every morning Yeah. and like navigate that seamlessly with search
  343. 34:17involved. Yeah. It's like, okay, well, would I use this? Yeah. because I don't use notes to write in because it doesn't serve that
  344. 34:23function. Uh so yeah, we didn't really explain explain features. You said
  345. 34:30capture anything with keyboard shortcut, retrieve anything with Yeah. Yeah. So I think like fundamentally for me what it boiled down
  346. 34:36to like the vision I originally had for Cortex and now that we are actually carrying forward to Eden was like just a
  347. 34:44really seamless way to capture anything at a moment's notice and then recall exactly what you want later. Thinking
  348. 34:51about that problem I boiled it down to well why don't we also just build like aside from like a really good capture
  349. 34:57system a really really good search system. Um, so and and so that's one of the uh really cool new things that we
  350. 35:03we've built with Eden, which is a really really really powerful search engine. So
  351. 35:08to give more context, what the search allows you to do is it allows you to search across not just like text, which
  352. 35:14is what most search engines do. It allows you to search across literally specific visual frames in any footage
  353. 35:20that you've ever captured. So, for example, um the the the the prototype
  354. 35:25that Ari showed me, he uploaded a bunch of episodes of random anime, and I just went in and searched pink sweater, and
  355. 35:33it resurfaced every single frame where someone was wearing a pink sweater at
  356. 35:38the exact timestamp. And I was just like, "Holy [ __ ] this is insane. I haven't seen this before." Like, YouTube
  357. 35:46doesn't even do that. Google search doesn't do that. like no no no finder
  358. 35:51app or or or search app or note-taking app for that matter like has a search
  359. 35:56this granular that takes me to the exact point that I want. There are some that we discovered after
  360. 36:02we built it, but we're not going to name them. [laughter] Yeah. But and and they're $600 a month.
  361. 36:07Oh yeah, that's ridiculous. Terabyte per terabyte. Yeah, it's ridiculous pricing. But but even even if
  362. 36:14they do exist like usually the companies like usually each
  363. 36:20company solves like one fragmented part of the entire creative workflow, right? Like search is only one of the parts.
  364. 36:26There's also capturing anything really easily which is like uh I just want like a single keyboard shortcut where I can
  365. 36:31like I'm looking at a YouTube video, boom, press a keyboard shortcut, it's done. Or I just like if if I'm not
  366. 36:37someone who likes using keyboard shortcuts, I can just copy the link, go into the app, paste it in, boom, it's done. the the phrasing here that I think
  367. 36:44a lot of people will resonate with, but we may not say very much is like just that it's a drive for
  368. 36:51anything. It's a drive for all media that's transcribed. Like as a creator, that is such a magical thing
  369. 36:57is like any YouTube video I paste is not only downloaded, which you have to go and get a sketchy app to do that right
  370. 37:04now, [laughter] but it's also transcribed and you can read it right there. Yeah. And you can search
  371. 37:09the frames of it and the transcript of it. Now, that's just YouTube. Like, think 3-hour long podcast. Incredible
  372. 37:15tool. Like, that's all that could be an app in and of itself. Yeah. The second is like Instagram reels.
  373. 37:21Where can you actually download those and watch the video outside of Instagram and have them transcribed?
  374. 37:27Yeah. Tweets, Substack articles. For for context there, for like our mobile app, one thing that you can do is
  375. 37:32like you're watching an Instagram reel. You just click share and you send it to the Eden app. Yeah. and then it'll save
  376. 37:37the video and download it into your workspace in whatever folder you choose. And then everything in that video will
  377. 37:43be transcribed and searchable. So if even if it's like a specific style like a specific visual style like you search
  378. 37:51like cinematic moody, it'll take you to the specific styles that you're looking for later when you're when you're like
  379. 37:57actually doing work in the workspace. Yeah. But now think of it like this is like if you use AI, you know that one of

  380. Canvas: Spatial AI for Creatives

  381. 38:03the big problems is context. And when you're trying to help it, help you like
  382. 38:09brainstorm stuff like how a landing page is structured or teach you like you uh save a reel because you're like, "Ah,
  383. 38:15this is great. I want to emulate this structure and pull some things from it." Now you have like I can create a folder
  384. 38:22of best reels and I can literally at@mention that folder in a chat. Yeah. And be like break down what works about
  385. 38:27all these reels. Yeah. You can't chat with a reel anywhere else or a YouTube video. And it's like that
  386. 38:33is like a huge missing piece with AI. And when you can, usually you're only chatting with the transcript. You're not
  387. 38:38chatting with the visual information. Whereas the AI chat, and this is something that Ari worked on recently as
  388. 38:44well, like what it allows you to do is it not just you're not just in our chat, you're not
  389. 38:50just chatting with the the transcript of the videos, you're chatting with the visual information. So you can do things
  390. 38:56like, oh, break down the shots in this YouTube video and explain like why it
  391. 39:04worked or or why it's visually appealing. Why is the hook in this video or in this reel visually appealing?
  392. 39:11That's where canvas comes into play is like so canvas you think miro, you think mind mapping, you think collage, you
  393. 39:18think other things, but um the chat interface is like onedimensional. like
  394. 39:24you're having one chat, you have to copy paste things in and out of it. Uh if you want to clear the context and have
  395. 39:30something else, then you create another chat. Um so a canvas feature where you can drag PDFs, YouTube videos, reels,
  396. 39:38anything in your own MP4 files. Yeah. If you're a YouTube like it's a Dropbox replacement, so you store all of
  397. 39:44your video files in there. If you need to find B-roll, you search for a specific shot, boom, it pops up. You clip that out, send it to your editor,
  398. 39:50good to go. But you upload those as well. So, imagine that I have like my
  399. 39:56raw filmed YouTube video. I put that on the canvas. I put uh one of the better
  400. 40:02YouTube videos of mine on the canvas. I put a few reels on there. Uh I put my brand guidelines, other things. I can
  401. 40:08connect all of those to an AI chat so they're pulled in as context, ask it whatever I want, like break down this
  402. 40:13video and replace it with something from this video, so on and so forth. And then I can connect any AI chat to any
  403. 40:20individual item, right? It's kind of free form visual spatial AI which for
  404. 40:26creatives I think that's a big thing. I've been using it every week for my newsletters and it's like kind of
  405. 40:32liberating to not have to open a pane now because like if I just want to take a little idea boom create a note in
  406. 40:38there jot it down there. Yeah. Drag that around to where I'm writing the newsletter in the canvas.
  407. 40:44If I have to research something with AI add an AI node ask it. Um,
  408. 40:50the other thing that it's helpful with is I don't have to take much manual notes anymore because a lot of my ideas
  409. 40:56come from YouTube videos or audiobooks or other things of that nature. But if it comes from a YouTube video, then I
  410. 41:02just add that YouTube video to the canvas and if there's a specific idea that I wanted, I connect that to an AI chat and
  411. 41:07I'm like, what was the idea about this? Or explain this concept to me in the way the video does. Right. Boom. That's my research for the
  412. 41:14newsletter. Right. Right. I'm not asking the AI to write it all for me. I'm just like, I have this rough idea of what I want. I have these
  413. 41:20sources all in my workspace. How do I get the information out of those sources
  414. 41:27uh in like a synthesized way because that's a different use case from like finding the exact part of it? Yeah.
  415. 41:32And then drag that over right next to where I'm writing the newsletter and it's like I have my outline, I have my
  416. 41:38research, I have an idea I threw over here, I have an image. Yeah. Yeah. I think like a drive for media,
  417. 41:44all sorts of media. Yeah. All transcribe is a really good way of putting it. The the change from Cortex to Eden, it's
  418. 41:50like the scope of what we're building is like 10x bigger than what we've done before. It's like we went from a note-taking app to now building Google
  419. 41:58Drive, Miro, Poppy, um like Frame.io, all downloaders from like different

  420. Why We Chose Eden Over Kortex

  421. 42:05websites that like sketchy websites that let you download Instagram reels or YouTube videos. All of that is now Eden.
  422. 42:12And and we're able to do that because of the because of all the mistakes that we made
  423. 42:17and the what we've learned from them. Like the way that we structure our team now is that every single person has to
  424. 42:23be laser focused on a project ra and and has to be full stack rather than like
  425. 42:28focusing on one compartmentalized part of of the app. And the the way that the the tech is structured to begin with
  426. 42:34allows for us to actually build these things really really really quickly and move really quickly. And so I think like
  427. 42:41like definitely yes, Cortex was a mistake, but it was it was a necessarily
  428. 42:46learning process for for us to be able to like get here. Yeah. All right, let's wrap up with how
  429. 42:53we came about Eden as a name and like the branding. Yes.
  430. 42:59Branding behind it, but like the brand messaging. Yeah. So like what we first
  431. 43:05did when we decided, okay, we're actually doing Eden. we're doing something new is like we had to rethink
  432. 43:11and clarify exactly who we're for. Um what what how are we going to like what
  433. 43:18do we want to be known as? Yeah. Is the main thing is like you think of the company 10 years from now who do you
  434. 43:24want to serve? Who's attracted to it? So on and so forth. And we were initially we started with
  435. 43:29like who are we for? Initially with Cortex, we're for creators, which is still kind of who we're for, but what
  436. 43:36kind of creators? And then we think deeper and it's like people who still deeply care about their craft. Yes.
  437. 43:42Two is like, okay, well, what's the opposite of AI slop? And it's craft or
  438. 43:48like artisan work. And then we started to draw that distinction of like okay we're more for artisans or people who
  439. 43:54want to be artisans because in the future when AI slot become so noisy like
  440. 44:00the artisans are going to be the one that stands out. So we're taking a bet on the the future of the artisan so to say.
  441. 44:06All of that started to come together and we started to get this really cohesive brand vision where it's like everything we put into this app, every minor detail
  442. 44:13we add has to be this like focused around craft and artisan level work because the
  443. 44:20people that we become and build for have to be attracted to this thing, right? They
  444. 44:25have a very high expectation of quality and polish. We went through a bunch. We're like, "Okay, we're for craftsmanship. What about like Da Vinci?
  445. 44:31There's Da Vinci Resolve. [laughter] What about this or this or this?" And we just went through Greek names,
  446. 44:38Renaissance names, artistic names, etc., etc. And I think I threw out just like
  447. 44:44Eden. Yeah. You randomly said Eden. Yeah. And like thought about it. Yeah. We were kind of we were at we
  448. 44:51always hit this point. There's a cycle in the ideas that we have when we're trying to refine a feature or we're
  449. 44:56trying to refine the name or whatever part. And it's like we run through so many ideas to the point of nearly giving
  450. 45:02up and just like forgetting about it and then we throw one out and it's like, "Oh, okay. Yeah, that one's good."
  451. 45:08Yeah. Yeah. We thought about Eden for a while. There were there was a number of reasons why it's good. Oh my god. The logo
  452. 45:14was Oh, wow. a pain to get to. So, so for those of you who follow me but don't know that
  453. 45:20Matt and Ari have their own channel, go watch it because it's a very different
  454. 45:25vibe from mine. It's like I watch it. It's fun. [laughter] But uh thank you.
  455. 45:30Yeah, the thing there with the logo. Props to Jason. Oh yeah, Jason videographer. He's not in
  456. 45:35frame, but maybe he'll face Has he face revealed yet? He has. He has. Okay,
  457. 45:41[laughter] we got Jason. Uh the thing with the logo
  458. 45:46is like we had this really good logo that we really liked and wanted to keep and it looked good in the mo like the
  459. 45:53app icon and every it was just perfect and we posted it or they included it in
  460. 46:00one of their first video on their YouTube channel at the end and then someone commented like
  461. 46:05isn't this the same as like Flora? Yeah, isn't this the same logo as Flora which is another canvas based or
  462. 46:10nodebased app? And then we looked at it exact same and it's like how did we even come to that logo and it's kind of
  463. 46:17simple. It's like four pedals. Yeah. And it fit our name perfectly. So we're
  464. 46:22like do we have to change it? So if anyone from floor is watching this, we
  465. 46:27changed it and we did change it. Now we have something that I I think it it's grown on me a lot. And on the note of like the name Eden,
  466. 46:34the reason why we thought that the name was good is because Eden was the garden of knowledge, right? It's like
  467. 46:40Yeah. It's it's the place where knowledge began for humanity. Um, and we
  468. 46:46want our app to be the place for that is like your home as a creative. And the old logo, it was duplicated
  469. 46:52across so many startups. [laughter] Yeah. Every single day there was another another company that had the same logo.
  470. 46:58Yeah. I think we have something more unique now. Yeah. We have something extremely unique now. Yeah. All right. Let's end strong. Biggest

  471. The Logo Crisis and Rebrand

  472. 47:05either lesson or failure. I think the biggest thing is a lot of times people
  473. 47:10will like at least for me like a lot of times people will say oh this is too hard you shouldn't do it because it's
  474. 47:16too hard and until you actually try giving up at that point is like it it's
  475. 47:24meaningless like we wouldn't be here if we just listen to people saying oh it's too hard you should just not do it like
  476. 47:31even Eden wouldn't exist if we if we kept to that advice yeah there there there
  477. 47:37It's actually a common reaction that we get like when we when we meet like VCs or random founders or investors and we
  478. 47:44explain what we're working on and the fact that we're trying to build something that's better than Google Drive.
  479. 47:50[snorts] The initial reaction that a lot of people give us is like how are you going to do
  480. 47:56that? Like yeah, it's too hard. Why would you do that? Why would you do that? Yeah. And several months later, we're at a
  481. 48:01point now where I actually think we have a better product. Yeah. Yeah. So, it's possible. Just have to try.
  482. 48:07Yeah. You just have to try. If I were to like zoom out and and and see what I've learned big picture, it would just be
  483. 48:14that like h you
  484. 48:19there's no point in not taking risks. That's a poor way to frame it. But
  485. 48:26everything that I I've done is uh and everything successful that that came in
  486. 48:31my life came as a result of taking like a big risk. For example, going part-time in university, contrary to what my
  487. 48:38parents suggested, in order to pursue shitty random jobs outside of school
  488. 48:43that were like contentbased and like paid like pennies. But if I didn't do all that, it wouldn't have led me to
  489. 48:49meeting you. And being on that call, taking the risk of just like public humiliation by by by trying to pitch you
  490. 48:56in front of 50 strangers. If I hadn't done that, we wouldn't be here. And then taking the risk of flying to Japan and

  491. Biggest Lessons From Our Failures

  492. 49:03rebuilding the thing when we had the existing app to worry about and when we when everybody in the team told us, you
  493. 49:09know, rebuilding is probably not a very good idea right now. If we didn't take that risk, we also wouldn't be here. So,
  494. 49:14I think like success is just like a a series of perhaps well-calulated risks,
  495. 49:20but like a series of big risks. And I think anybody who makes it big does
  496. 49:25things that seem conventionally like a bad idea. Yeah. But you know that it's a good idea and
  497. 49:31can lead to potentially something great. Yeah. Oh, mine is the second that you think you know what
  498. 49:37you're doing, you're kind [laughter] of [ __ ] where like I thought that a lot
  499. 49:45of the skills and things that I had learned from my whole content creator journey and freelancing and all these
  500. 49:52different types of products would transfer over like onetoone to a startup. It's like I'm having to relearn, not
  501. 49:58relearn but refine a lot of those things cuz I I have to refine them to the point of like revealing what the principle
  502. 50:05actually was. I thought I knew what the principle was, but now it's like things are thrown on top of their head because
  503. 50:11it's like, okay, set a deadline pressure that kicks people into action. And when I'm applying it to myself, it's like I
  504. 50:17can I know what I'm trying to do. So, I have all of the the different variables
  505. 50:23accounted for in that situation. But when trying to apply that to a team, it's like you don't really know how
  506. 50:28that's going to work. And when it doesn't work, you're like is like this fundamental
  507. 50:33framework principle. Yeah. is this framework of how I work wrong as a whole and then
  508. 50:40it's like okay well how then what's the actual deeper underlying truth there and and that's the thing is like
  509. 50:46how do I apply my worldview to a team like not that I'm trying to
  510. 50:51impose my worldview but how do I make sense of this team in a way that I can best like guide or lead you know what I
  511. 50:57mean because I can lead myself just fine I think sure I'm an idiot sometimes but like like [laughter] for a team it Like
  512. 51:05I think about that a lot is and it kind of frustrates me is like I have like
  513. 51:11these people that I feel like I need to guide but it's also a value of mine to let people do their own thing.
  514. 51:17Yeah. And so then it's like where do I draw the boundary between that right? It's like I want everyone on our team whether
  515. 51:25they want it or not is another problem to be able to take feel as if they are
  516. 51:32like partly owning this thing. They have the the ownership mindset to the point where they can make decisions solve
  517. 51:39problems on their own. So it's just a ch it's a very like interesting and rewarding challenge that
  518. 51:46I am trying to new sets of problems. Yeah. Yeah. better sets of problems.
  519. 51:51Yeah. And I think one thing that we need to look out for in the future is we we've
  520. 51:57made mistakes in the past. That's for sure. That's this video. A big $3 million of mistakes. Yeah.
  521. 52:02Yeah. We will continue to we will continue to. But I think what sets people apart is whether you can
  522. 52:08take your mistake and not repeat it. Because as humans, we're very prone to remaking the same mistakes over
  523. 52:16and over again. And keep in mind that we will continue to make new mistakes and we we'll have to watch out for them and
  524. 52:21learn from them the same way that we made all the mistakes that led to Cortex. Yeah. And you guys will see it in real
  525. 52:28time [laughter] and you will be the ones to inform us of these mistakes. So with that said,
  526. 52:33again, rolling out to Cortex users from highest payment tier just because
  527. 52:39there's less people to the free tier. That's going to happen the Monday before
  528. 52:44Black Friday through Thursday. This video will be out on Thursday. Friday through Monday is the only time the
  529. 52:51early access will be open to the public. So, if you're watching this within 4 days of it going out, go sign up and try
  530. 52:58it. And uh after that, we
  531. 53:04will do a full public launch at some point in 2026 that will be with desktop,
  532. 53:10mobile, etc. Yeah. And then um in between that time, we may try something. So if you're
  533. 53:16watching this in between that like dead zone, uh you may be able to reach out to
  534. 53:21an Eden user and get them to invite you to the platform if that's a thing. Leave
  535. 53:27a comment and maybe someone can help you. Comment on someone who actually got in. Go on Twitter, go on Instagram, find
  536. 53:33people that have gotten into Eden and they should be able to send you an invite link and they'll only have a limited amount, like three to five.
  537. 53:40Yeah. So, it'll be a fun way to see how quickly eating grows and how much people actually want to use it. So, with that,
  538. 53:47thanks for coming on. I guess we're a safe channel, but uh yeah, this is the Eden
  539. 53:55co-founders. Enjoy Eden. Thank you for your time.
Kortex: The $3 Million Dollar Mistake — Transcriptly