YouTube2Text

72 часа зарабатываю на вайбкодинге — получится ли? — Transcript

by SGX · 2,739 words · 395 segments · language en · Watch on YouTube

Full transcript

  1. 0:01I gave myself 72 hours and 20 bucks to
  2. 0:04build an AI app using v0-coding. I
  3. 0:07didn't write any of the code by hand.
  4. 0:09An AI model will write all the code.
  5. 0:11Then I'll have another week to try and
  6. 0:14sell it to someone. At the end of the
  7. 0:16video, I'll show how much I earned and
  8. 0:18what the expenses were. I've already
  9. 0:20been through it all, so I'm telling the
  10. 0:22story in reverse. By the way, making
  11. 0:24the product turned out to be the
  12. 0:26easiest part. Well then, let's go. Here
  13. 0:29is my idea. I used to build bots for
  14. 0:32clients. The client says: "I built it,
  15. 0:35got paid a buck, and moved on." What
  16. 0:37can I say? It was just regular
  17. 0:39freelance work, but with v0-coding. I'm
  18. 0:41talking about these two videos; if you
  19. 0:43haven't seen them, you can check them
  20. 0:45out. But this time, I wanted to create
  21. 0:47my own product. I build it once, and it
  22. 0:50works for me even while I sleep. This
  23. 0:52thing lives right inside a Telegram bot
  24. 0:54. It’s a mini app. No App Store, no
  25. 0:56need to install anything. You just open
  26. 0:59the link and start using it. Let me
  27. 1:01demonstrate my idea. I stopped
  28. 1:04wondering where to write down my
  29. 1:06thoughts and deleted all my task
  30. 1:07managers. That's why I v0-coded my own
  31. 1:10app, GX Planner. Now my entire life is
  32. 1:12in one Telegram bot. You don't even
  33. 1:14have to open Telegram to use it. For
  34. 1:15example, I can hold down the button and
  35. 1:17say: "Tomorrow at 3:00 PM I have a
  36. 1:18meeting with Andrey. For lunch, I had
  37. 1:20borscht. I need to remember to pay for
  38. 1:22the internet. And I also spent 500
  39. 1:24rubles on a taxi." GX Planner
  40. 1:26automatically figures out what's a
  41. 1:28meeting, a meal, a task, or an expense,
  42. 1:30and sorts everything into the right
  43. 1:32sections.
  44. 1:33The system organizes everything itself.
  45. 1:36One message instead of ten apps.
  46. 1:44That’s the product I built, and I
  47. 1:46even sold a bunch of subscriptions for
  48. 1:48a very nice sum, but more on that later
  49. 1:50. So, what do you think? Pretty awesome
  50. 1:53, right? Or is it complete garbage?
  51. 1:56Whatever. Moving on. The rules are
  52. 1:58simple: 3 days to build, 20 bucks for
  53. 2:01Cursor Pro, and 5 bucks for the OpenAI
  54. 2:04API. There are no other investments
  55. 2:06here. Then 7 days of promotion. I'm not
  56. 2:09buying any ads. My TikTok is starting
  57. 2:11from zero. I have my own Telegram
  58. 2:13channel with a small audience. It
  59. 2:14helped, but not by much. I’ll
  60. 2:16definitely show you the stats for all
  61. 2:18social media. I think the idea is clear
  62. 2:20to you now. Now let's rewind to the
  63. 2:22very beginning. Just as the idea was
  64. 2:25starting to form, 10 days ago: 3 days
  65. 2:27to build, a week to sell, a blank
  66. 2:29canvas, and zero code. Hello everyone,
  67. 2:34my favorite v0-coders. Let’s start
  68. 2:36our experiment. This is day one. I’m
  69. 2:39not touching the code at all. To start,
  70. 2:41I’m putting the idea together in
  71. 2:42Figma. I already have an idea, and
  72. 2:44it’s not just some made-up notion,
  73. 2:47but a pain I experience myself every
  74. 2:49single day. Look, it’s simple and the
  75. 2:52same for everyone. I often have great
  76. 2:55thoughts, but nowhere to write them
  77. 2:57down, so they just get lost. Or another
  78. 3:00example: I spend money and forget how
  79. 3:02much. I don’t keep track of my
  80. 3:04expenses at all. I also don’t keep an
  81. 3:06eye on meetings, and sometimes I might
  82. 3:07miss something. That’s why this
  83. 3:09feature needs to be implemented. And
  84. 3:12since I watch my diet and work out—
  85. 3:14well, for athletes, we need a fitness
  86. 3:16tracker. If you try to do this manually
  87. 3:20, you end up with about seven apps, and
  88. 3:22you have to choose where to put each
  89. 3:24entry. But that’s just absurd.
  90. 3:27That’s the choice I’m getting rid
  91. 3:29of. This is how I came up with the SGX
  92. 3:32Planner. One place where you dump
  93. 3:34everything by voice, without even
  94. 3:36opening the app. And the built-in AI
  95. 3:39will automatically figure out what it
  96. 3:41is and where to put it. A couple of
  97. 3:43ideas went straight to the trash, like
  98. 3:45a feed inside the app and various
  99. 3:47streaks. Well, that’s just overload
  100. 3:50and has nothing to do with the problem.
  101. 3:52Then I thought: "How can I make this as
  102. 3:54lazy as possible?" Typing really takes
  103. 3:56too long, so voice it is. And it would
  104. 3:59be even better not to take the phone
  105. 4:00out at all. That’s where two features
  106. 4:02were born. I’m marking them with an
  107. 4:04asterisk. I’ll come back to them a
  108. 4:05bit later. It’s going to be really
  109. 4:07intense. I’ve also already defined
  110. 4:09the pricing plans, how we’ll monetize
  111. 4:11, and in what way. Essentially, my full
  112. 4:14MVP is ready, which is pretty cool. And
  113. 4:18once the picture came together, I
  114. 4:20started turning it all into a prompt.
  115. 4:22To build the prompt, I immediately
  116. 4:24create a mind map with all the app's
  117. 4:26features and work out the architecture
  118. 4:29so the AI understands the structure
  119. 4:31instead of it being a mess. I
  120. 4:33immediately decide where the app will
  121. 4:35live. I’m making it right in Telegram
  122. 4:36. I get the mini-app bot token from
  123. 4:39BotFather, give it to the AI, and then
  124. 4:41it connects everything itself. No App
  125. 4:44Store, no moderation, no waiting weeks
  126. 4:46for a three-day experiment. My tool is
  127. 4:49just the standard $ 20 Cloud Code.
  128. 4:52Although I’ve long been on a more
  129. 4:54expensive tier from Codeium. I got the
  130. 4:56cheaper one specifically for this video
  131. 4:57, the same one most people have. And
  132. 4:59yes, another very important point,
  133. 5:01without which I won't make it. I'm
  134. 5:03talking about skills. These are
  135. 5:04pre-made instructions. I have five of
  136. 5:06them for the front-end so I don’t
  137. 5:07have to explain every time that I need
  138. 5:09a decent design. On its own, AI does it
  139. 5:12just terribly, in the most generic way.
  140. 5:15And I will also use MCP—not exactly
  141. 5:18skills, but direct model access to
  142. 5:20services. It goes to the database
  143. 5:23itself and pulls UI components without
  144. 5:25me. I'll drop the name and links in
  145. 5:27Telegram. Next, I feed the prompt to
  146. 5:30Claude, and it gets to work. And while
  147. 5:32our miracle is being written, I realize
  148. 5:34what I’m actually doing. It’s an AI
  149. 5:37B2C Six7 SAS. Yeah, a lot of jargon
  150. 5:40that just makes you want to barf. If
  151. 5:42you translate it into normal terms,
  152. 5:45it's a project made for everyday users
  153. 5:48with a subscription, all powered by AI.
  154. 5:50But what’s up with the "Six7" part?
  155. 5:52Honestly, I have no idea, I'm just an
  156. 5:55idiot. Everyone and their dog is
  157. 5:56churning out apps like these. And
  158. 5:58almost all of them are losing money. Oh
  159. 6:00well, I’ll make my own money-losing
  160. 6:01project just for kicks. Why not, right?
  161. 6:04At first, I’ll probably be right
  162. 6:05about losing money. But there’s a
  163. 6:07reason people make fun of these things.
  164. 6:09They never validate pain points or
  165. 6:11demand, so the projects don't last a
  166. 6:13week. My approach is different. I’m
  167. 6:15building it based on my own specific
  168. 6:17struggle. But will people pay for it?
  169. 6:19Well, we’ll test it on real people
  170. 6:21via social media. At the same time,
  171. 6:23I’m mapping out on this canvas how it
  172. 6:25should make money. The video leads to a
  173. 6:27bot. Everything in the bot is free, but
  174. 6:30when a user hits the 2-action AI limit,
  175. 6:33they want more and buy the Pro sub for
  176. 6:36290 rubles. Per month. Looks great on
  177. 6:39paper, right? But will even one
  178. 6:41stranger actually buy it? We’ll find
  179. 6:43out in a week. Well, end of day one.
  180. 6:46Claude built me an app, the framework
  181. 6:49is set, features are polished, and a
  182. 6:51marketing plan is ready. Everything
  183. 6:53looks better in the morning. See you on
  184. 6:55day two. Day two. Today’s plan is to
  185. 6:58fix the mess the neural network made
  186. 7:00yesterday. I woke up, ran the app
  187. 7:03quickly, and saw enough errors that
  188. 7:06need fixing. I also need to add the
  189. 7:08design and implement those two "star"
  190. 7:11features. Let’s get started. I’m
  191. 7:13describing the features in two prompts
  192. 7:15to handle it all step-by-step. And I
  193. 7:17sent it all off to Claude. As soon as I
  194. 7:19have some results, I’ll be sure to
  195. 7:21show you everything. Feature one. New
  196. 7:23iPhones have a side Action Button. Most
  197. 7:26people don't even know it exists. I’m
  198. 7:28mapping it to thought capture. I don't
  199. 7:30even need to unlock my phone. A couple
  200. 7:33of seconds after I record something,
  201. 7:35it’s already sorted into sections,
  202. 7:38and the AI categorizes everything
  203. 7:40itself. My philosophy for this app is
  204. 7:42that the most convenient tracker is the
  205. 7:44one you never even have to open. And
  206. 7:47for those who think it’s just for
  207. 7:49iPhones, no—you can record the same
  208. 7:51thing with a regular voice message, you
  209. 7:53just have to open Telegram. The second
  210. 7:56feature is a bit more hardcore. iPhones
  211. 7:58have a "Back Tap" feature. Almost
  212. 8:00nobody knows about it. I’m setting it
  213. 8:02up for quick expense entry. Like, if I
  214. 8:04buy something, I tap the back twice,
  215. 8:06and the expense is already recorded.
  216. 8:09You don't even have to unlock the phone
  217. 8:11. These aren't just some extra buttons.
  218. 8:14Every app competes to make the
  219. 8:17prettiest screen. But I want screens
  220. 8:20you never even have to open. Honestly,
  221. 8:22the best interface is no interface.
  222. 8:25What else is inside? Let me show you
  223. 8:27quickly. We can count calories just by
  224. 8:29snapping a photo of our meal. It’s
  225. 8:31elementary. Or we can track money in
  226. 8:34any currency. It factors in monthly
  227. 8:37limits, plus tips from an analyst. You
  228. 8:41can also track your habits and check
  229. 8:43trends. We have eight sections in total
  230. 8:45here. You also get morning
  231. 8:48notifications with a daily overview,
  232. 8:50plus a bunch of useful dashboards.
  233. 8:53Today was an awesome day, I got a lot
  234. 8:57done, so I can safely go rest and
  235. 9:00finish our app tomorrow. The second day
  236. 9:04really turned out to be the best.
  237. 9:05Everything was going great. I was in
  238. 9:07the zone, but on the third day,
  239. 9:09everything almost fell apart. Okay guys
  240. 9:16, day three, and two red flags at once.
  241. 9:19First: my twenty-dollar subscription
  242. 9:21limits are almost gone. I'm working
  243. 9:23right down to the wire. I really
  244. 9:25thought I wouldn't finish in 72 hours.
  245. 9:27My hands are itching to dive into the
  246. 9:29code, but we’re holding steady,
  247. 9:31brothers. This is a real experiment.
  248. 9:33Second red flag. I gave my vibe-coder
  249. 9:35brother a feel of the—no, not what
  250. 9:38you’re thinking—the app. And I got
  251. 9:41a report almost immediately. And it was
  252. 9:43for the exact feature I was bragging
  253. 9:44about yesterday. He asks the bot, "How
  254. 9:47much did I spend at cafes this month?"
  255. 9:49Yes, we have a built-in neural network.
  256. 9:52And it writes back, "I couldn't find
  257. 9:54anything in your records." Even though
  258. 9:56he has a ton of entries, it’s a
  259. 9:58pretty interesting bug. I need to sit
  260. 10:00down, look into it, and fix it now.
  261. 10:02That’s vibe-coding for you. Just as
  262. 10:04it is. I’m not cutting this out,
  263. 10:06I’m sitting down to fix it. Phew. Man
  264. 10:15, I struggled with that bug. I
  265. 10:17basically wasted the whole day on it
  266. 10:19because I was fixing the wrong sections
  267. 10:20. The neural network couldn't
  268. 10:22understand the problem for a long time.
  269. 10:24I barely managed to fix it. It turned
  270. 10:26out I hadn't properly defined the
  271. 10:28finance section. So it was just
  272. 10:30answering blindly. Anyway, I fixed it.
  273. 10:32Now it calculates from the records
  274. 10:35first, then replies. Same question. And
  275. 10:38we finally have a proper response with
  276. 10:40a category breakdown, so it works
  277. 10:43perfectly. A word on security. In short
  278. 10:45, the keys are on a server with UPass.
  279. 10:47They never touch the app. I didn't want
  280. 10:49to build payment processing myself. I
  281. 10:50connected a payment system through a
  282. 10:53manager, and that's how I’ll handle
  283. 10:55all payments. You don't even need any
  284. 10:57complicated layouts there. Everything
  285. 10:59works out just fine. Well, the 72 hours
  286. 11:02are coming to an end. Our timer is
  287. 11:04already at zero. I made it just in time
  288. 11:06, but the product is alive, that's the
  289. 11:09main thing. It's far from perfect, but
  290. 11:11it's usable. I think I'll display it
  291. 11:13for you in the background. The first
  292. 11:15task is complete. Building it in 3 days
  293. 11:17is realistic. And that's where
  294. 11:19something I was completely unprepared
  295. 11:21for begins. The product exists, but
  296. 11:24there's no one to use it. Absolutely no
  297. 11:27one knows about it. My idea on how to
  298. 11:31sell this miracle. I'm cutting the
  299. 11:33product into short clips, one per
  300. 11:35feature, and posting them everywhere at
  301. 11:37once. And let's see what comes of it.
  302. 11:40If you've enjoyed watching this so far,
  303. 11:42by the way, smash that like button.
  304. 11:44Let's keep going. Our favorite money is
  305. 11:46waiting for us there. Fast-forwarding a
  306. 11:49week ahead. I want to say that the
  307. 11:51clips are already posted. And right
  308. 11:53away, before they start writing in the
  309. 11:54comments, I'll say that I posted from
  310. 11:56several accounts on each social network
  311. 11:58. I tested one clip in different ways,
  312. 12:00changing the post time, description,
  313. 12:02tags, and seeing where the hook would
  314. 12:04hit. So all the reach further is the
  315. 12:06sum across these accounts. The first 2
  316. 12:08days there was silence. I filmed how
  317. 12:10nicely the voice input works. And, well
  318. 12:13, very few people watched it. I already
  319. 12:16thought I’d failed the challenge, but
  320. 12:18then it hit me what the problem was. I
  321. 12:20was showing features. Viewers don't
  322. 12:22wake up in the morning thinking they
  323. 12:24need speech recognition in a tracker. I
  324. 12:26rewrote it specifically for that pain
  325. 12:28point, for example, that I tap my phone
  326. 12:31twice and the expense is recorded
  327. 12:32immediately. Showing it in action. The
  328. 12:35first video was about the function, the
  329. 12:37second about me. And that one took off.
  330. 12:39It got 15,000 views in a day, which is
  331. 12:42just wow. I really felt that my product
  332. 12:45wasn't made in vain. And then
  333. 12:47everything started growing. The results
  334. 12:50of the week turned out to be
  335. 12:52magnificent. I was simply in shock. It
  336. 12:54ended up with 88,000 views. 1,000
  337. 12:57people visited the bot itself. 11
  338. 13:00payments were made from Telegram. That
  339. 13:03is my actual audience. The rest are new
  340. 13:06accounts from social networks. The
  341. 13:08first thing I saw were messages like
  342. 13:10these. It felt very good. And then came
  343. 13:13what it was all started for: after the
  344. 13:16first two days, the first paying users.
  345. 13:18And finally, we've reached the money.
  346. 13:20Remember the blurred payment screen at
  347. 13:22the beginning? Well, here it is. 623
  348. 13:24dollars for the week, or 47,000 rubles.
  349. 13:29A fifth of that amount came from two
  350. 13:30people. They bought lifetime access.
  351. 13:34Our expenses were 20 for the
  352. 13:36subscription, 5 for requests, and about
  353. 13:39598 dollars remained. This is an
  354. 13:42incredible sum that I didn't even count
  355. 13:45on. But 623 dollars isn't even the most
  356. 13:48important figure here. There's a second
  357. 13:50one, smaller, and it will drip into my
  358. 13:52account every month on its own. About
  359. 13:5518,850 rubles. per month. 65
  360. 13:58subscriptions at 290 each. But some
  361. 14:02will say, "People won't pay you for a
  362. 14:04subscription every month." But wait,
  363. 14:06290 rubles is less than a cup of coffee
  364. 14:09. I also factored in churn, because I
  365. 14:12won't be actively uploading videos and
  366. 14:14pushing this theme. By my estimates, if
  367. 14:17there are no new sales at all, it comes
  368. 14:19out to about 135,000 a year. This is
  369. 14:22just a forecast for now. The product is
  370. 14:24only a week old. We’ll see how it
  371. 14:26goes. And I won't be rambling on about
  372. 14:29how anyone can do this. Deconstructing
  373. 14:31a product and managing an agent is a
  374. 14:33skill. I had it before I even started.
  375. 14:35But previously, such an attempt would
  376. 14:37have taken a whole team or six months
  377. 14:39of work, while for me, it took 3 days
  378. 14:41and 20 bucks. Code is no longer a
  379. 14:43barrier. The barrier now is coming up
  380. 14:45with what to build and explaining why.
  381. 14:47Thanks to wipe-coding, the cost of an
  382. 14:49attempt has plummeted. This is enough
  383. 14:51to stop procrastinating and build your
  384. 14:54own project, or go build them for
  385. 14:56clients. I’ll post all the prompts,
  386. 14:58skills, roadmap, and promotion
  387. 15:00breakdown in Telegram. And don't forget
  388. 15:03to smash that like button on this video
  389. 15:05if you enjoyed it, and subscribe to the
  390. 15:06channel. Also, write in the comments
  391. 15:09what product you would test in 3 days
  392. 15:11and why people should even pay for it.
  393. 15:14Truly, it will be interesting to read.
  394. 15:16That's all from me. Go wipe-code
  395. 15:18something already. Hugs to everyone.

About this transcript

This page contains the full transcript of 72 часа зарабатываю на вайбкодинге — получится ли? by SGX, generated from the public captions YouTube serves with the video. The transcript has 2,739 words across 395 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.