Como Criar Um Negócio de Uma Pessoa (Em 2026 ainda) — Transcript

Full transcript

1. 0:01If your goal, or even your dream, is to
2. 0:03have an extremely scalable business
3. 0:05that will give you great freedom to
4. 0:08generate high profits and put money in
5. 0:10your pocket, without the headache of
6. 0:12being enslaved by that business or
7. 0:14worrying about a team of 10, 15, 20, 50
8. 0:17, or even 100 people to maintain those
9. 0:19results or keep them growing, then
10. 0:21you’ve come to the right video. Not
11. 0:25only the CEO of OpenAI, but also the
12. 0:28CEO of Anthropic, the company behind
13. 0:30Claude, bet that we would see a
14. 0:33billion-dollar company operated by a
15. 0:35single person. And that idea brings me
16. 0:38to one single question: how much could
17. 0:40you build if a much larger part of your
18. 0:42business execution were handled by
19. 0:43agents? And today, I’m going to map
20. 0:47out for you a company with three
21. 0:48functions operated entirely by agents.
22. 0:51And of course, I’ve been working for
23. 0:52a while to implement this within my own
24. 0:54businesses, and I’ll tell you where
25. 0:56I’m focusing most of my energy as an
26. 0:57owner. So, if besides generating great
27. 0:59results and actually profiting a lot
28. 1:01from your business, your goal is to not
29. 1:03rely on something extremely complex to
30. 1:05keep it growing, then stay until the
31. 1:07end of this video, because you might be
32. 1:09just a few steps away from achieving
33. 1:11that goal. And today I will use an
34. 1:13example from one of the markets I
35. 1:14operate in: the SaaS and apps market.
36. 1:17But remember that what I’m going to
37. 1:19share here can also be adapted to any
38. 1:21other business, even those in the
39. 1:23physical world. Maybe it’s not
40. 1:25possible to be completely independent
41. 1:27of human labor in that business, but
42. 1:29you might want to delegate about 80 to
43. 1:3190%of the functions you currently
44. 1:32handle yourself. But first, we need to
45. 1:36understand what this prediction means
46. 1:38—the prediction that soon we will see
47. 1:40billion-dollar businesses orchestrated
48. 1:42by a single person. So, that statement
49. 1:46by Dario, the CEO of Anthropic, was in
50. 1:48May 2025, when they were asked when the
51. 1:50first billion-dollar company with a
52. 1:52single employee might emerge. And at
53. 1:55that moment, his answer was somewhere
54. 1:57between 2026 and 2027. Of course,
55. 1:59it’s a prediction, but there is also
56. 2:02a huge difference between a company
57. 2:04being valued at 1 billion dollars and a
58. 2:06company actually bringing in 1 billion
59. 2:08in revenue or putting that kind of
60. 2:10profit into the owner's pocket. So, I
61. 2:13look at this discussion from a very
62. 2:15different angle, one much closer to our
63. 2:18reality. And the question remains: can
64. 2:22a person identify a problem, build a
65. 2:23solution, acquire clients, and maintain
66. 2:26an operation with a much smaller
67. 2:27structure than what was previously
68. 2:29required for all of that to work? And
69. 2:32it's this possibility that I've always
70. 2:34been interested in exploring. And not
71. 2:36just now that we're in the AI boom, but
72. 2:38for those who have known me longer,
73. 2:39I've mentioned many times here on the
74. 2:41channel that my goal with digital and
75. 2:43internet entrepreneurship has always
76. 2:44been to carry a series of
77. 2:45non-negotiables. And among these
78. 2:48non-negotiables, the most important is
79. 2:50efficiency and effort disproportionate
80. 2:52to the result. In other words, I have
81. 2:55always sought to create lean businesses
82. 2:56, businesses that don't depend on great
83. 2:58complexity to generate millions in
84. 2:59revenue and profit. And of course, this
85. 3:02entire goal has become increasingly
86. 3:04refined, thanks to artificial
87. 3:06intelligence and the latest major wave
88. 3:09we are witnessing, which is the wave of
89. 3:11autonomous AI agents. But in light of
90. 3:14all this, I've always noticed a very
91. 3:16big mistake in people who also want to
92. 3:18build scalable businesses—businesses
93. 3:20that generate millions in profit, but
94. 3:22with little complexity, little headache
95. 3:24, and without depending on dozens or
96. 3:26even hundreds of people on the team—
97. 3:28which is what? People want to design
98. 3:31the entire structure before having
99. 3:33something validated. So, before
100. 3:35thinking about how you can use AI
101. 3:38agents, how you can create several
102. 3:40departments orchestrated by agents, how
103. 3:44you can better utilize Cloud Code or
104. 3:46Codex, or how you can use any other AI
105. 3:49tool...you have to understand your "why
106. 3:53" behind all of this very well. The
107. 3:55more validated what you already have in
108. 3:57hand is, the easier this "why" comes,
109. 4:00the easier this other series of answers
110. 4:02and the design you're going to build
111. 4:04becomes. So, there's no point in
112. 4:06wanting to create a one-person business
113. 4:08or wanting to significantly reduce
114. 4:10complexity or increase the efficiency
115. 4:11of my business if I don't have a
116. 4:13business yet or don't have the idea for
117. 4:15this business and don't know if this
118. 4:17idea is validated. So, you don't
119. 4:20necessarily need to have something
120. 4:22already making millions, something with
121. 4:24a past history, but you at least need
122. 4:26to be sure that your idea is validated.
123. 4:29And for the context that I'm going to
124. 4:31bring here for my example, I started by
125. 4:32dividing this example, which is one of
126. 4:34my businesses that I'm going to bring
127. 4:35here for you, into three functions. The
128. 4:37first is delivery, which is the work
129. 4:39for which the client pays. The second
130. 4:42is sales and support, meaning
131. 4:44understanding who arrived to explain
132. 4:46the solution and help them use it. And
133. 4:49the third function is acquisition, that
134. 4:51is, getting new customers to know that
135. 4:53business, that solution. So you notice
136. 4:56that it's a reverse engineering process
137. 4:58. Most people will start by focusing on
138. 5:00that first function, on how to use
139. 5:02agents or how to use AI to sell more.
140. 5:05But that is not the most coherent path
141. 5:07to follow, especially since acquisition
142. 5:09isn't the easiest starting point. And
143. 5:12that is always what I’ve sought to do
144. 5:13as well: think big, but start small. So
145. 5:16, before you think about creating a
146. 5:18business that will be operated and
147. 5:19orchestrated by a single person, in
148. 5:21this case, you, you have to know how to
149. 5:23choose an offer and a delivery model
150. 5:25that fits within this structure or that
151. 5:27you can at least visualize, have an
152. 5:29idea that it will fit. So, in my case,
153. 5:32I already had a ready-made application,
154. 5:34an app with thousands of users, already
155. 5:36being sold and renewed for over 3 years
156. 5:38, with very well-defined sectors within
157. 5:41that delivery model. So, within that
158. 5:44first function I mentioned, which was
159. 5:46the structural part, if you don't yet
160. 5:48have, for example, a deliverable—and
161. 5:51let's use this example: an info-product
162. 5:53or a SaaS app—the first step is to
163. 5:55understand how you can use the AI tools
164. 5:57at your reach to research the main pain
165. 6:00points in a specific niche or market,
166. 6:02to research the main solutions that
167. 6:04exist in that niche or market, and how
168. 6:06you can create a similar solution. Many
169. 6:11people think it's bad to enter a market
170. 6:13where, for example, there are
171. 6:14technically a lot of people already
172. 6:16selling, that already has a product, a
173. 6:18highly scaled solution in that market.
174. 6:21And that is a huge mistake. If there is
175. 6:24a product, if it's scaled, it's because
176. 6:26there is demand. And if there is demand
177. 6:27, it's because people are paying to
178. 6:30solve that problem. So, what stops you
179. 6:33from understanding how that application
180. 6:34works, how that solution is delivered,
181. 6:36improving that solution, for example,
182. 6:37creating a series of new angles for
183. 6:39potential offers for the same solution,
184. 6:41and attacking it that way? And in my
185. 6:43case, the application was already ready
186. 6:46since 2023, so I didn't have to redo
187. 6:48anything there with artificial
188. 6:49intelligence. But there was a very
189. 6:51important point within this delivery,
190. 6:53this first function, that helped me a
191. 6:55lot here with renewals, which was in
192. 6:57the sense of getting people to keep
193. 6:58using the application. So, many people
194. 7:01still don't see the power of LTV, how
195. 7:03much a single customer can generate for
196. 7:05you over time. Today, in all the
197. 7:07businesses I have on the internet, in
198. 7:09none of them do I look at a client and
199. 7:11think about selling to them just once.
200. 7:14Everything revolves around
201. 7:15subscriptions or products for constant,
202. 7:17recurring use. And to succeed with
203. 7:20these business models, the product
204. 7:22needs to be good, the delivery needs to
205. 7:24be good, and most importantly, the
206. 7:26customer needs to use the product. So,
207. 7:28in the app scenario, imagine 1,000
208. 7:30women download the app. Man, of those
209. 7:321,000 women who subscribed and
210. 7:34downloaded the app, you can be sure
211. 7:37that about 800 will actually complete
212. 7:39the sign-up and watch the first workout
213. 7:42program. But of those 800 who watched
214. 7:46the first workout program, 7 days later
215. 7:49only about 200 will open the app again
216. 7:51to continue using the workouts, using
217. 7:54the community, and even buying or
218. 7:56consuming other physical products made
219. 7:59available there. So, with all these
220. 8:02events, like, at what point is the
221. 8:04person getting stuck, why didn't they
222. 8:07complete their sign-up and watch that
223. 8:09first sequence of workouts? How long
224. 8:11does it take them to reach the main
225. 8:13experience? Does it depend on the
226. 8:16specific ad? So, women coming from ad X
227. 8:20tend to consume the app more than those
228. 8:23coming from ad Y, since the angle and
229. 8:26expectations generated by ad Y differ
230. 8:29from ad X. When you study these points,
231. 8:32sometimes you see that ad Y has a
232. 8:35cheaper cost, but it isn't necessarily
233. 8:40bringing in the most qualified customer
234. 8:43—the one who will pay for a more
235. 8:46expensive subscription or renew their
236. 8:48plans for a longer time. With all these
237. 8:53events organized, I was able to ask an
238. 8:55autonomous AI agent, created in
239. 8:57OpenClaw, to compare the groups, show
240. 9:00me who came from each campaign and
241. 9:02funnel, and then show me, in an
242. 9:04extremely segmented way, the possible
243. 9:06reasons why these people were or
244. 9:08weren't moving forward. So, once again,
245. 9:12maybe one campaign brings in many cheap
246. 9:14acquisitions, but from people who
247. 9:15aren't as qualified. Another one might
248. 9:18cost a bit more, but it will bring in a
249. 9:19more qualified user, reaching people
250. 9:21who will actually use that thing for a
251. 9:23longer time. And this is a difference
252. 9:25that also changes the investment
253. 9:27decision. So this is the type of work
254. 9:29that I will initially seek to
255. 9:31systematize to understand what kind of
256. 9:33experience is making the person come
257. 9:35back before bringing more people to
258. 9:37this same experience. Now, imagine how
259. 9:41a person, a human being, or even I
260. 9:43could sit there all day monitoring this
261. 9:45amount of data, monitoring this amount
262. 9:48of information; it would be humanly
263. 9:50impossible. Moving on to the second
264. 9:53function, which is the commercial and
265. 9:55support side, this was perhaps the
266. 9:57easiest to do and maybe also a point
267. 9:59where you can reverse the order, right,
268. 10:01start here. Why? If you already have a
269. 10:04product, man, if you already have all
270. 10:06the deliverables for this product, I
271. 10:08don't see why you should have humanized
272. 10:10support for this product from the first
273. 10:12contact with the customer. So, you can
274. 10:14take all this material that will be
275. 10:16delivered, you can put it, for example,
276. 10:18into an artificial intelligence agent
277. 10:20and you can create a bot with this
278. 10:22agent. This bot can answer customer
279. 10:25questions via email, WhatsApp, or even
280. 10:27Telegram. And then you can set a series
281. 10:31of rules within it that will make it
282. 10:33necessary for a human to be called in
283. 10:35for that support case. So, besides the
284. 10:38commercial part of sales recovery, this
285. 10:41was another second very important point
286. 10:43that I was also able to systematize.
287. 10:45Instead of needing three or four people
288. 10:47in support for an extremely scaled
289. 10:49business, and even people who needed to
290. 10:51speak more than one language because
291. 10:53the app is scaled globally, I managed
292. 10:56to do this with just one artificial
293. 10:57intelligence agent and reduce this
294. 10:59humanized support to a single person.
295. 11:01Third function, the acquisition part.
296. 11:04This part involves a few steps. So, in
297. 11:07a very simplistic way, if we consider a
298. 11:10classic sales funnel for direct traffic
299. 11:12, where we have an ad, VSL, and
300. 11:13checkout, or where we have an ad, quiz,
301. 11:16VSL, and checkout, my first point of
302. 11:18thought was to seek to understand how I
303. 11:20could replace tasks that depended on
304. 11:22human work with AI agents. Once again,
305. 11:25then, starting with metrics analysis,
306. 11:27man, that was the easiest part that was
307. 11:29also already linked to function number
308. 11:31one. Imagine connecting your sales
309. 11:35platform, Meta Ads, and any tracking
310. 11:37tool you use to your agent, asking it
311. 11:39to generate daily performance reports
312. 11:41with highly accurate insights on what
313. 11:44you need to do to scale and get better
314. 11:46results. Imagine how you could break
315. 11:50down an existing ad creation process
316. 11:53and now integrate artificial
317. 11:55intelligence tools and agents to
318. 11:57operate them, creating ads for you at a
319. 12:00much higher speed. I recently showed in
320. 12:04a video here on the channel how the
321. 12:06audiovisual side is becoming very
322. 12:07refined for Codex and also for Claude
323. 12:09Code. So, that is the perspective from
324. 12:12which you need to start seeing how to
325. 12:14divide your entire business into stages
326. 12:17, areas, and functions currently
327. 12:18performed by you or someone else. And
328. 12:21how can you list the tasks within these
329. 12:24functions? And now, how can all these
330. 12:27tasks be performed by agents? Take
331. 12:29benchmarking, for example—knowing
332. 12:31what is working in the market. Man, I
333. 12:34have an agent today that analyzes the
334. 12:36market in this niche every day and
335. 12:38shows me the funnels and the players
336. 12:40that are scaling the most. But not only
337. 12:44that, I have another skill in there
338. 12:46with a copy agent that can create an
339. 12:48entire new funnel, from ad ideas to all
340. 12:51the quizzes used and sales pages. And I
341. 12:54just need to approve it and set a test
342. 12:56to run. So that is the workflow you
343. 12:58should follow and think about to build
344. 13:00a one-person business, the famous "
345. 13:02one-person business." Of course, in
346. 13:04theory, that name sounds great. But in
347. 13:06practice, you see that you have to
348. 13:08apply this knowledge and concept in
349. 13:11stages. It’s not impossible to do,
350. 13:14but I am certain, at least according to
351. 13:16my forecast, that in the coming months,
352. 13:18I will have a business capable of
353. 13:20hitting at least 1 million reais per
354. 13:22month, orchestrated solely by me and a
355. 13:24series of AI agents. So this video was
356. 13:27a sort of roadmap for you to understand
357. 13:29what the CEOs of OpenAI and Anthropic
358. 13:31meant when they talked about
359. 13:32billion-dollar companies run by a
360. 13:34single person. And how you should view
361. 13:37this path from here on out. Not as
362. 13:40something miraculous where you press a
363. 13:42button and suddenly a
364. 13:43multimillion-dollar business appears in
365. 13:46front of you. No, but starting from
366. 13:48validated things you already have,
367. 13:50seeking to replace all human labor with
368. 13:52agentic labor, as well as the
369. 13:53optimization and delivery, of course,
370. 13:55beyond just the work. And for those
371. 13:58interested in learning more about
372. 14:00autonomous AI agents, I have a
373. 14:01mini-course where I teach you how to
374. 14:03create your first autonomous AI agent,
375. 14:05which you can use in your personal life
376. 14:07for more productivity or apply directly
377. 14:09to your business today. I will leave
378. 14:12the link to this training in the video
379. 14:13description. Remember that we are still
380. 14:16in the first batch at the time you're
381. 14:17watching this video, and we might
382. 14:18increase the price tomorrow with all
383. 14:20the implementations being added to this
384. 14:21training. I'll also leave the link to
385. 14:25the AI Express mentorship below, where,
386. 14:27in addition to gaining all this
387. 14:29knowledge related to AI agents and
388. 14:31advanced AI, you also get several other
389. 14:33training tracks and 12 months of
390. 14:35guidance from me and my team, showing
391. 14:37you the behind-the-scenes of all my
392. 14:40businesses and everything I'm doing to
393. 14:42scale them massively online, extremely
394. 14:44effectively, up to nine figures and
395. 14:46beyond. Other than that, subscribe to
396. 14:49the channel, leave a comment, let me
397. 14:51know what you thought of this video,
398. 14:52and leave suggestions for future videos
399. 14:53as well. So that’s it, we’ll wrap
400. 14:56up this video here.