YouTube2Text

أعطني 20 دقيقة.. وسأمنحك ميزة تفوق (غير عادلة) على الجميع بالذكاء الاصطناعي! — Transcript

by Yasser Seleem · 4,985 words · 742 segments · language en · Watch on YouTube

Full transcript

  1. 0:00Most people today think they are using
  2. 0:01artificial intelligence well. When they
  3. 0:03type a question into ChatGPT or Claude.
  4. 0:05And get a nice-looking, organized
  5. 0:07answer. Let me be honest with you, my
  6. 0:08friend. That was barely enough to get
  7. 0:10by six months ago. This approach has
  8. 0:11now become technically primitive. We
  9. 0:13are living in an era where the industry
  10. 0:15changes every three months. Due to the
  11. 0:16extremely rapid evolution taking place.
  12. 0:18And the next shift that is currently
  13. 0:19happening in the market. Is called AI
  14. 0:21Agents, or digital employees. The gap
  15. 0:24being created right now between someone
  16. 0:26who understands what an agent is. And
  17. 0:28knows how to build and manage one. And
  18. 0:29someone who still thinks AI is just a
  19. 0:31chatbot to chat with that helps with a
  20. 0:34few things. This gap will be very
  21. 0:36costly and expensive for your financial
  22. 0:38future. Your job, your business, or the
  23. 0:40side hustle you are trying to build.
  24. 0:42Let’s clarify something important
  25. 0:43before we begin. The figures and
  26. 0:44results mentioned in this video are
  27. 0:45educational examples. Of what can be
  28. 0:47achieved in the real market. And are
  29. 0:48not a promise of profit or a guarantee
  30. 0:50of specific results. What you might
  31. 0:51achieve could be higher or lower than
  32. 0:53the figures mentioned in this video. By
  33. 0:55virtue of my work for years as a CEO.
  34. 0:56And launching more than one company in
  35. 0:58the field of digital marketing.
  36. 0:59Including my digital marketing agency.
  37. 1:01Through which we have managed
  38. 1:02advertising budgets of over $ 10
  39. 1:04million. And AI was a key player in all
  40. 1:06the ad campaigns we managed over the
  41. 1:08last two years. And recently, my launch
  42. 1:11of Soda Marketing AI. Which has become
  43. 1:13one of the best and most important
  44. 1:15tools for creating ad content using AI.
  45. 1:17I’m here to reassure you today and
  46. 1:18tell you that it’s much simpler than
  47. 1:20what you hear outside. This video will
  48. 1:22show you exactly how these systems work
  49. 1:24. And when to implement them. And how
  50. 1:25to build your first digital employee
  51. 1:27from scratch in minutes and get ahead
  52. 1:29of everyone. But before we reach out
  53. 1:30and open the laptop to create your
  54. 1:32first digital employee. There is a
  55. 1:33terrible confusion right now that we
  56. 1:35must break immediately. People are
  57. 1:36mixing up three things that are as far
  58. 1:38apart as possible. Prompts and
  59. 1:39automation. That is, the automation you
  60. 1:42do on platforms like n8n or Make. And
  61. 1:44the real AI Agent. If you don't
  62. 1:46understand the fundamental difference
  63. 1:47between these three. You are wasting
  64. 1:49your time and managing a blind system.
  65. 1:50Focus with me closely and let's crack
  66. 1:52this code together. First: The Prompt (
  67. 1:54the traditional command). This prompt
  68. 1:56is like sitting in a driver's ed car.
  69. 1:58With a student still learning sitting
  70. 1:59next to you. You have to stay awake and
  71. 2:01focused every single second. Telling
  72. 2:03them: turn right, turn left, brake here
  73. 2:05, watch out, or be careful. The chatbot
  74. 2:07executes and then stands waiting for
  75. 2:09your next command to move step by step.
  76. 2:11If you leave it, it will stop in its
  77. 2:12place. It is merely an answer engine.
  78. 2:14Second: Automation or traditional
  79. 2:16automation. Like the automation you can
  80. 2:18create with programs like n8n or Make.
  81. 2:19This automation is like solid steel
  82. 2:21train tracks. The system here is
  83. 2:23designed to execute rigid and blind
  84. 2:25steps. If I get an email, take it, send
  85. 2:27it to AI to summarize, then put it in
  86. 2:29an Excel sheet. You might tell me, "But
  87. 2:31Engineer, we've already put AI in the
  88. 2:32middle and it's doing summaries." I'll
  89. 2:33say, "That's correct, but it's still
  90. 2:35not an agent." Why? Because it's
  91. 2:37obedient and blind automation, obedient
  92. 2:39work. So if a client sends an email in
  93. 2:41a slightly strange format. Or if a
  94. 2:42column gets deleted from the sheet you
  95. 2:44created. That train will go right off
  96. 2:45the tracks. And the whole system will
  97. 2:47shut down and blow up in your face.
  98. 2:48Traditional automation doesn't think.
  99. 2:50It doesn't know how to act if the set
  100. 2:52path has any error or change, even a
  101. 2:54minor one. Third, the real AI Agent (
  102. 2:57the digital employee). This agent isn't
  103. 2:59a student sitting beside you, nor is it
  104. 3:00blind train tracks. This agent is a
  105. 3:02professional, expert driver. You hire
  106. 3:04him, and you sit in the back seat in
  107. 3:05total comfort. You give him only the
  108. 3:07final address, which is the destination
  109. 3:09. And you hand him the car keys. And he
  110. 3:11alone determines the fastest route. He
  111. 3:13is the one who deals with traffic
  112. 3:14congestion. And if he finds a closed
  113. 3:16road or a "no entry" sign. He doesn't
  114. 3:17stop the car and start crying. He opens
  115. 3:19the map, makes an immediate decision,
  116. 3:20and takes an alternative route. He gets
  117. 3:22you to the goal you specified exactly.
  118. 3:24Without even waking you up from your
  119. 3:26sleep. The agent has the same AI brain
  120. 3:28inside. But it's surrounded by four
  121. 3:30digital employees. Working inside his
  122. 3:32mind at the same moment. One is the
  123. 3:34Analyst, the Planner, the Operator, and
  124. 3:37the Auditor. One reads the pattern. And
  125. 3:40the second sets a contingency plan if
  126. 3:41the system breaks. A third executes the
  127. 3:43procedures. And a fourth reviews
  128. 3:45quality and corrects errors. Before the
  129. 3:47final output reaches your desk. The
  130. 3:49question here is, how does this
  131. 3:50professional driver know how to make
  132. 3:52decisions alone? When the traditional
  133. 3:54tracks break. And what is the secret
  134. 3:56system inspired by fighter pilots in
  135. 3:58wars? That makes the agent outperform
  136. 4:00any other automation system. Grab a cup
  137. 4:02of coffee and focus with me on what's
  138. 4:04coming next. Because we are diving into
  139. 4:06the secrets of what happens under the
  140. 4:07hood. Under the bonnet, my friend. And
  141. 4:09no one will tell you what I'm going to
  142. 4:11tell you in this video just like that.
  143. 4:13To understand the genius that makes an
  144. 4:15agent act like a real human employee.
  145. 4:17We must go back in time to 1950, behind
  146. 4:19the scenes of the Korean War. There was
  147. 4:21a colonel in the US Air Force named
  148. 4:23John Boyd. Not Joe Biden, John Boyd.
  149. 4:25Boyd was studying a very strange puzzle
  150. 4:27that baffled all military engineers at
  151. 4:29the time. American pilots in F-86s were
  152. 4:32always winning air battles against
  153. 4:35Soviet MiG jets. "What brought American
  154. 4:37and Soviet planes into the Korean War,
  155. 4:39Engineer?" Bravo, my friend, you're
  156. 4:40paying attention now. Politically and
  157. 4:42officially on paper, the war was
  158. 4:44between North Korea, backed by China
  159. 4:45and the Soviet Union, against South
  160. 4:47Korea, backed by the UN and America. It
  161. 4:49was a proxy war so that a world war
  162. 4:51wouldn't ignite between these two
  163. 4:53superpowers. But behind the scenes of
  164. 4:55the air war, specifically in the MiG
  165. 4:57Alley region, the war was a hot, direct
  166. 4:59, face-to-face confrontation between
  167. 5:01American pilots in American planes and
  168. 5:03Russian pilots in Russian planes. The
  169. 5:05strange part of the puzzle is that the
  170. 5:07MiG jets, from a technical and
  171. 5:08engineering standpoint, were much
  172. 5:09faster. And they could fly at much
  173. 5:11higher altitudes than the American jets
  174. 5:12of that time. So, on paper, the Soviet
  175. 5:14plane should have won and dominated
  176. 5:16every single time. John Boyd kept
  177. 5:18analyzing until he reached the secret
  178. 5:20that changed the understanding of
  179. 5:21strategy worldwide. He discovered that
  180. 5:23American planes had two features beyond
  181. 5:25speed and engine specs. First, their
  182. 5:27cockpit was a rounded bubble glass. It
  183. 5:30allowed the pilot to see everything
  184. 5:31around them at a 360-degree angle,
  185. 5:33unlike the cramped MiG cockpit. And
  186. 5:34second, the stick or lever used to fly
  187. 5:37the plane was hydraulic and very
  188. 5:39responsive. What was the result? The
  189. 5:40American pilot could see faster and
  190. 5:42adapt to maneuvers more quickly. So,
  191. 5:44they would get inside the
  192. 5:45decision-making cycle of the Soviet
  193. 5:47pilot. And beat them before the other
  194. 5:49pilot even understood what was
  195. 5:50happening. John Boyd called this cycle
  196. 5:52the OODA loop. That is: Observe, Orient
  197. 5:57, Decide, and Act. In the world of
  198. 6:00digital employees, this is exactly what
  199. 6:02happens. This is the real test that
  200. 6:04separates an Agent from standard, blind
  201. 6:06automation. When the primary path the
  202. 6:08system is following breaks or gets
  203. 6:09blocked, will the machine stop and fail
  204. 6:11? Or will it enter its own OODA loop
  205. 6:13and choose an alternative path? Let’s
  206. 6:16take a practical example so the picture
  207. 6:17becomes completely clear to you.
  208. 6:19Suppose, for instance, you have a
  209. 6:21social media marketing agency that
  210. 6:22operates fully on AI. Among the systems
  211. 6:25you built for the AI to handle all the
  212. 6:27hard, laborious, and skilled work for
  213. 6:29you is a traditional automation
  214. 6:31workflow. Its function is as follows:
  215. 6:33every Friday, it collects emails and
  216. 6:35data for 50 e-commerce stores in a
  217. 6:36specific country. So you can look at
  218. 6:38them and start determining a way to
  219. 6:40contact them to offer your digital
  220. 6:42services. And because it is a
  221. 6:44traditional automation system, you have
  222. 6:45defined exactly where and how it should
  223. 6:47search. This automation runs like
  224. 6:49clockwork. Until one week, the platform
  225. 6:51you pull data from makes any update, a
  226. 6:53simple update to its code. What will
  227. 6:55happen to the traditional automation
  228. 6:56here? It will stop immediately. The
  229. 6:57iron rails are broken. The system will
  230. 6:59show you a red error sign. It will stop
  231. 7:02all work until you wake up and go in to
  232. 7:03edit the code yourself. And you will
  233. 7:05waste your whole day maintaining and
  234. 7:06fixing the system. But an Agent acts
  235. 7:08with a completely different mindset
  236. 7:10because it goes through the OODA loop.
  237. 7:12It will notice that the platform
  238. 7:13changed the code and the usual path is
  239. 7:15blocked. So it won't stop the work. It
  240. 7:16will start to analyze and orient itself
  241. 7:19(Orientation). It will assess the
  242. 7:20situation in its mind. It will
  243. 7:21understand on its own that a minor
  244. 7:23change occurred in the platform's code.
  245. 7:24It will immediately make a decision,
  246. 7:25which is the Decide phase. It will
  247. 7:27begin to adjust the method it uses to
  248. 7:29retrieve data, based on the changes in
  249. 7:30the platform's code. It will start
  250. 7:32moving along the new path, pulling, and
  251. 7:35executing, which is the Act phase. It
  252. 7:37will close the required folder for you
  253. 7:38and send the report to your desk in the
  254. 7:40morning, with a small note saying the
  255. 7:42first path. Was blocked. I took the
  256. 7:43initiative and solved the problem using
  257. 7:45another path on your behalf, boss.
  258. 7:46Traditional automation follows a script
  259. 7:48. The agent follows a goal. That's why
  260. 7:50when you hear someone in 2026 saying
  261. 7:52they've built an agent, ask them one
  262. 7:54question to expose them. If the first
  263. 7:56path and the first step in the code
  264. 7:58break, will your machine follow the
  265. 8:00blind script and stop? Or will it
  266. 8:02improvise and create a new path on its
  267. 8:03own? But there is a fundamental and
  268. 8:05critical question you must ask yourself
  269. 8:07here. Since the agent possesses the
  270. 8:08freedom to make decisions and improvise
  271. 8:10with such terrifying capability. Why do
  272. 8:11we still see so many agent projects and
  273. 8:13systems failing in reality? And
  274. 8:15literally messing up the work. The
  275. 8:17reason coming up will truly amaze you.
  276. 8:18Because it is not a technology problem.
  277. 8:20It is a human problem first and
  278. 8:21foremost. Focus with me on the next
  279. 8:23part. The real danger in AI agents, my
  280. 8:26friend. Is that they do the wrong
  281. 8:28things faster and with a million times
  282. 8:30more confidence than you could ever do
  283. 8:32yourself. The agent is not magic. The
  284. 8:34agent is a multiplier, or leverage. A
  285. 8:38lever for the power and clarity of your
  286. 8:40own thinking. Let me tell you a secret.
  287. 8:41From my own personal experiences. Most
  288. 8:43AI problems are, at their core,
  289. 8:45recurring human problems. For example,
  290. 8:48when you give the agent vague goals.
  291. 8:49And confused instructions. Without a
  292. 8:51clear context. And expect quality
  293. 8:53results based on that, the system will
  294. 8:54drive the car and hit the first tree it
  295. 8:56sees, but it will do so at rocket speed
  296. 8:58and with absolute confidence. In fact,
  297. 9:00AI is quite funny here in its
  298. 9:01confidence and speed in executing a
  299. 9:03completely wrong decision it is making.
  300. 9:04Artificial intelligence does not fix
  301. 9:06broken thinking. It exposes and
  302. 9:08reinforces it. Even faster; that is why
  303. 9:10before you automate anything or launch
  304. 9:13your digital employee, you must perform
  305. 9:15what we can call the GPS check. The G
  306. 9:17is for Goal. Can you define the agent's
  307. 9:20goal in one clear, concise sentence
  308. 9:22without beating around the bush? The P
  309. 9:24is for Proof of success. How will you
  310. 9:26know that the agent produced a correct
  311. 9:27and excellent result? What is the
  312. 9:29quality standard the machine will use
  313. 9:30to measure the quality of the resulting
  314. 9:32output? So you can add a very important
  315. 9:35stage in any proper agent, which is QC,
  316. 9:37or Quality Control. And the S is for
  317. 9:39Steps. Can you explain the steps to the
  318. 9:42machine in order without jargon or
  319. 9:44vague words? If you cannot do these
  320. 9:46three things clearly, your agent might
  321. 9:47not work 100%correctly. And now let's
  322. 9:49open the laptop together so I can show
  323. 9:51you myself the easiest and fastest way
  324. 9:53to build your first digital employee
  325. 9:54from A to Z. The subject is much easier
  326. 9:56than you think. We are currently going
  327. 9:58to make an employee whose job is to
  328. 9:59check. Emails and reply to them
  329. 10:00automatically in a very professional
  330. 10:02manner. We are going to build a very
  331. 10:04smart agent. Acting as a sniper and
  332. 10:05manager for your emails. It checks,
  333. 10:07analyzes, and prepares responses for
  334. 10:10your side business or the work emails
  335. 10:12you have in your job. And all of this
  336. 10:14will happen while you sleep. The whole
  337. 10:15thing is in three simple steps. Step
  338. 10:17one: Connecting the main artery. All
  339. 10:19you have to do is go to Cloud Desktop.
  340. 10:21And then to the Connectors menu inside
  341. 10:23the system. And click on the Gmail icon
  342. 10:25. And click authorize or grant access.
  343. 10:27With that, you've connected Claude
  344. 10:29Desktop to your email so it can read
  345. 10:31and even reply to your emails. Step two
  346. 10:34: Drafting the assignment for your
  347. 10:35smart employee. Now, go to Claude Work.
  348. 10:38And open a new task. Instead of writing
  349. 10:40a normal prompt, you will write this
  350. 10:41exact task. Watch what I write to it:
  351. 10:43Every day at 9:00 AM, check for new
  352. 10:45emails regarding advertising
  353. 10:47collaborations and sponsorships, like
  354. 10:48sponsorships and YouTube brand
  355. 10:50collaborations. Analyze the email text
  356. 10:52and extract the brand name and their
  357. 10:54website link. Search the website and
  358. 10:55clearly identify the size of the
  359. 10:56company. Then, act as my business
  360. 10:58manager. First, examine all my previous
  361. 11:00replies to similar emails, the deals
  362. 11:02that were closed, and the deals I
  363. 11:03rejected or decided not to answer.
  364. 11:05Learn my pricing and negotiation style
  365. 11:07and improve upon it. Based on all of
  366. 11:08the above, create a professional email
  367. 11:10reply for every offer. Prepare them as
  368. 11:12a draft ready for my review, and
  369. 11:13immediate response from within the
  370. 11:15email. The agent will now ask me: Do
  371. 11:16you want me to start retroactively, or
  372. 11:18start from tomorrow at 9:00 AM? I will
  373. 11:19say. Tell it, for example, to start
  374. 11:21with everything from the past. Then it
  375. 11:22will confirm the 9:00 AM timing with me
  376. 11:24. Look at the brilliance and the speed.
  377. 11:26Without you opening a single email
  378. 11:27yourself. Without wasting your time
  379. 11:28reading them. And without missing an
  380. 11:30email that might be important and a
  381. 11:31good opportunity. The agent will work
  382. 11:33and perform this task to the fullest.
  383. 11:35From this moment on, even while I'm
  384. 11:37sleeping, even when I'm out. This
  385. 11:38digital employee is working in the
  386. 11:39background. It opens the emails,
  387. 11:41analyzes them, and prepares the replies
  388. 11:42. All you have to do is open the email
  389. 11:44from your phone, read the draft, and
  390. 11:45hit send. Meaning, all the heavy work
  391. 11:47is ready in front of you on your device
  392. 11:49, at the speed of light. I want you to
  393. 11:50pause before you just breeze past this
  394. 11:51part. What I just told you is very
  395. 11:53serious. This is capable of multiplying
  396. 11:54your productivity many times over. And
  397. 11:56developing your work level to
  398. 11:57incredible degrees. And setting you
  399. 11:58apart from all your colleagues or
  400. 12:00competitors, whatever your field of
  401. 12:01work. I just used a simple example that
  402. 12:03applies to one actual task from my own
  403. 12:05work tasks. But in reality, what I want
  404. 12:07you to do is look at all the repetitive
  405. 12:09tasks you have in your work and your
  406. 12:11life as well, and create an agent like
  407. 12:12this for them. Let's give an example.
  408. 12:14You can create an agent like the one I
  409. 12:16made, but it looks at emails coming
  410. 12:17from your manager or your manager's
  411. 12:19managers at work, to always prepare
  412. 12:20professional replies for you. Keep in
  413. 12:22mind, we told the agent here to learn
  414. 12:24first and understand how you reply. So
  415. 12:26if you've been in this job, for example
  416. 12:28, for a year or two, and you gave it a
  417. 12:29task like this, in minutes this agent
  418. 12:31will review all the correspondence that
  419. 12:32took place between you and your
  420. 12:34managers over the past two years. It
  421. 12:35will learn you and learn your manager.
  422. 12:37And it will know what works and what
  423. 12:39doesn't. And you will be amazed by the
  424. 12:40result and the replies it creates for
  425. 12:41you. And it’s not just email—we are
  426. 12:43only simplifying things. No, you can
  427. 12:45also do something a bit more advanced.
  428. 12:47If you work in sales, for example, you
  429. 12:48could create an agent that reviews. All
  430. 12:50the communications that took place
  431. 12:51between you and clients over the last
  432. 12:53three months. And it makes you an Excel
  433. 12:55sheet or a Google Sheet. Containing all
  434. 12:57the data for every client. And every
  435. 12:58deal, its status, and its current stage
  436. 13:00. And the next step, and the
  437. 13:01probability of closing or losing the
  438. 13:03deal. It even prioritizes them for you.
  439. 13:05Who you should start with, who you
  440. 13:07should leave for last, and so on, my
  441. 13:08friend. For example, I no longer sit
  442. 13:10and reply to revision requests sent by
  443. 13:12a client for an ad video I delivered.
  444. 13:15No, I do something much smarter. I
  445. 13:17already have an agent that produces
  446. 13:18these ad videos with me from start to
  447. 13:20finish. And when the client replies, I
  448. 13:21don't sit and analyze what they want to
  449. 13:23change or how we'll fix it. I go to the
  450. 13:25agent that was working on the project
  451. 13:26with me from the beginning. And I send
  452. 13:28a message like this: the client replied
  453. 13:29to the delivery message. Check the
  454. 13:31reply and download. All the images and
  455. 13:32videos they sent, and analyze them. And
  456. 13:34let's see how we will implement these
  457. 13:35changes. Focus with me here. To know
  458. 13:38how something simple like this can take
  459. 13:40your work and income to a whole new
  460. 13:42level. If you are an employee and you
  461. 13:43become a pro at using and creating
  462. 13:45these agents, while your colleagues are
  463. 13:47not. Imagine the difference. The
  464. 13:48difference in the speed, quality, and
  465. 13:49professionalism of your work that will
  466. 13:51clearly show to your managers. And how
  467. 13:53that will affect your position and your
  468. 13:54financial income. And if you are a
  469. 13:56business owner and you are a pro at
  470. 13:57using and creating these agents, while
  471. 13:58your competitors are not. Imagine how
  472. 14:00much capacity you'll have to outperform
  473. 14:02and crush them in the market. Or if
  474. 14:03you're an employee with a side hustle,
  475. 14:05like doing freelance work. See also how
  476. 14:07these AI agents won't just make you
  477. 14:09stand out in your job and your side
  478. 14:11hustle. No, but also how they will save
  479. 14:13you time and effort, letting you take
  480. 14:15more rest. Much more, while at the same
  481. 14:16time doing more work and making more
  482. 14:18money than you were before. If I kept
  483. 14:19giving you examples of agents you could
  484. 14:22make to transform your life and work,
  485. 14:24this video would be several hours long.
  486. 14:26But I'll tell you something better to
  487. 14:27do. Go to Claude or ChatGPT and tell it
  488. 14:29: conduct an interview with me right
  489. 14:30now. Your goal is to learn everything
  490. 14:33about me, my life, my work, and my
  491. 14:34goals. Ask me one question at a time so
  492. 14:36I can answer each one individually. And
  493. 14:38after you finish these questions and I
  494. 14:40give you all the answers, I want you to
  495. 14:42give me all the ideas for AI agents I
  496. 14:44can build to improve my work and life,
  497. 14:45increase my income, and increase my
  498. 14:47rest time. That's it. The AI will tell
  499. 14:49you about dozens of AI agents you can
  500. 14:51create that will significantly change
  501. 14:53your life, and it will also be able to.
  502. 14:55Help you with how to implement any one
  503. 14:56of them. But wait. The market in 2026
  504. 14:59won't stop here. We are entering an era
  505. 15:01where the number of AI agents on the
  506. 15:02planet will be greater than the human
  507. 15:04population. And this means that every
  508. 15:06software company that exists today will
  509. 15:07be faced with an agent trying to
  510. 15:09displace and push it out of the market.
  511. 15:11The companies and freelancers who will
  512. 15:13make millions in the coming period are
  513. 15:15not those building general and broad
  514. 15:17systems. They are the ones who will
  515. 15:19focus on a single, specific, and lethal
  516. 15:21niche. And solving a real problem and
  517. 15:23pain for a large customer base. And
  518. 15:25this problem must also be urgent. Let
  519. 15:27me tell you in the next part exactly
  520. 15:29where the money is hidden. And how AI
  521. 15:32agents can change your relationship
  522. 15:34with time and income forever. Let me
  523. 15:38tell you the core truth that represents
  524. 15:40the big shock in our current era.
  525. 15:42Artificial. Intelligence, specifically
  526. 15:44AI agents, is the greatest decoupling
  527. 15:47machine in human history. For the first
  528. 15:49time since we began recording modern
  529. 15:51history, the link between our income
  530. 15:53and the hours we work is being
  531. 15:54completely broken. In the past, even if
  532. 15:57you were at the top of the corporate
  533. 15:58ladder or a business owner, you were
  534. 16:00always trapped in an equation where you
  535. 16:01sell your time, effort, and decisions
  536. 16:03for money. And the day you get tired,
  537. 16:05this machine might stop. Digital
  538. 16:06employees have broken this link
  539. 16:08completely and forever. The machine is
  540. 16:10what will handle the actual execution
  541. 16:11and the back-and-forth work. And your
  542. 16:13role will expand and multiply in only
  543. 16:15one area: your strategic judgment and
  544. 16:17human taste. Understand the economic
  545. 16:19logic for 2026, my friend. You are
  546. 16:22entering an era where output will be
  547. 16:24infinite. Codes will become very cheap.
  548. 16:25Writing content and scripts will be
  549. 16:27free. And financial analyses and
  550. 16:29reports will fill the servers. And when
  551. 16:31intelligence, information, and skills
  552. 16:32become available for free, the only
  553. 16:34thing. That will skyrocket in price and
  554. 16:36become very expensive is human taste
  555. 16:38and the ability to curate. When outputs
  556. 16:40are infinite, taste becomes the rare
  557. 16:43currency that there is none of. The
  558. 16:45richest, most valuable, and
  559. 16:46highest-earning person will not be the
  560. 16:48one who thinks fastest or writes the
  561. 16:50most. It will be the person who knows
  562. 16:52how to define what excellent quality
  563. 16:53work is, and what poor, useless work is
  564. 16:55. And when to trust the digital
  565. 16:57employee and let it drive. And when my
  566. 16:58human hand must intervene to correct
  567. 17:00the path. These are the skills that no
  568. 17:02artificial intelligence will ever be
  569. 17:03able to imitate. Where is this
  570. 17:04immediate money, man? Let me tell you.
  571. 17:05If you want to start a side hustle, a
  572. 17:07simple project alongside your main job,
  573. 17:08that could bring you more money than
  574. 17:10your main job even. And succeed in the
  575. 17:12chaos we are living in, you will go and
  576. 17:14look for one task. A specific thing
  577. 17:15business owners hate doing themselves
  578. 17:17and are forced to repeat every day like
  579. 17:19machines; take this part, and build its
  580. 17:22dedicated agent. And go sell them this
  581. 17:24solution. And I don't mean here that
  582. 17:25you sell. Them the AI agent. No, you
  583. 17:26will sell them this AI agent as a
  584. 17:28service. And this is a new term you
  585. 17:30will hear here and only here. And for
  586. 17:31the first time, AI Agent as a Service.
  587. 17:33What does this mean, man? Come on. Let
  588. 17:35me give you an example. Owners of
  589. 17:36retail shops and online stores, they
  590. 17:38all need to post strong daily content.
  591. 17:41From photos and videos to offers of
  592. 17:43various kinds, across all their social
  593. 17:45media pages. If you go to sell him the
  594. 17:47AI agent, many of them will worry about
  595. 17:49their ability to manage the AI agent.
  596. 17:50But if you go and ask him, who
  597. 17:52currently handles the production and
  598. 17:54publishing of your social media and
  599. 17:55advertising content? And how much does
  600. 17:57it cost you? And how many pieces of
  601. 17:59content do you post daily? He will
  602. 18:00reply and tell you, "It costs me about
  603. 18:02$ 1500 a month.""And a marketing agency
  604. 18:04handles it for us.""They post about 20
  605. 18:06pieces of content monthly." You say to
  606. 18:08him, "Okay.""I will create and post all
  607. 18:10this advertising content.""At a rate of
  608. 18:1230 posts a month, not 20, including at
  609. 18:14least 20 professional videos monthly.""
  610. 18:16And the videos will be very powerful,
  611. 18:18too.""Featuring UGC styles and motion
  612. 18:20design.""And I will do all this, plus
  613. 18:22the posting, for only $ 700 a month.""
  614. 18:24Here is a sample of the work.""The
  615. 18:26entire month's content will be
  616. 18:28delivered within just 72 hours of our
  617. 18:30agreement." You have created an
  618. 18:32irresistible offer here. The constant
  619. 18:34demand in any offer: best price,
  620. 18:35fastest delivery, highest quality. The
  621. 18:37offer that no one refuses. When you
  622. 18:38repeat this process with five clients
  623. 18:39like that. And each one pays you $ 700
  624. 18:41a month. You do the math on how much
  625. 18:43you will be making. And of course, my
  626. 18:44friend, as we clarified at the
  627. 18:45beginning of the video, this is not a
  628. 18:47magic formula for guaranteed profit
  629. 18:48while you sit back and relax. These are
  630. 18:49realistic operational calculations of
  631. 18:51what actually happens in the market.
  632. 18:52When you put in real effort and execute
  633. 18:53correctly, the results you achieve may
  634. 18:55be higher or lower than these figures,
  635. 18:56depending on your skill and commitment.
  636. 18:58By the way, I explain this business
  637. 19:00model in detail in this training
  638. 19:02program, "AI Agency Accelerator: The 48
  639. 19:04-Hour Launch." How to build your social
  640. 19:05media marketing agency, which operates
  641. 19:07almost entirely with artificial
  642. 19:08intelligence. Meaning the AI is what
  643. 19:10hunts for potential clients, prepares
  644. 19:12messages for them, and helps you close
  645. 19:13the sale. And most importantly, it is
  646. 19:15what executes the work. Meaning it is
  647. 19:16what produces and delivers the awesome
  648. 19:18advertising content to clients. So, if
  649. 19:19you want to create your own social
  650. 19:20media marketing agency, check out the
  651. 19:22training program; you'll find the link
  652. 19:23in the video description. But keep
  653. 19:24watching the video with me because what
  654. 19:25comes next is very important. Because,
  655. 19:26my friend, if you leave this video
  656. 19:28thinking that as soon as you build 5 or
  657. 19:3010 agents for your clients or yourself,
  658. 19:32you can let them drive the business car
  659. 19:34100%automatically without supervision,
  660. 19:36let me curb your enthusiasm and tell
  661. 19:38you that you might be committing
  662. 19:39commercial suicide and destroying your
  663. 19:41reputation in the market. "But
  664. 19:43engineering-wise, didn't you just say
  665. 19:44that the agent is a professional driver
  666. 19:46who knows what to do if the road is
  667. 19:47blocked?" That's right, my friend. The
  668. 19:49driver is professional, but even a
  669. 19:50professional can make a mistake. The
  670. 19:51agent is ultimately a probability
  671. 19:53engine, and it might have a moment of
  672. 19:54technical "hallucination," for example,
  673. 19:56writing incorrect information to a
  674. 19:57client. The elite in the world don't
  675. 19:59let the machine deal with live data out
  676. 20:01in the open like that without a safety
  677. 20:04net. Instead, they apply a strict
  678. 20:05protocol called "Human-in-the-Loop,"
  679. 20:07human intervention within the control
  680. 20:09loop. And this protocol simply means:
  681. 20:11AI is the muscles, the execution, and
  682. 20:13the skills, but you are the
  683. 20:14consciousness and the final safety
  684. 20:16valve. When you build the email agent
  685. 20:18we created together, or the video
  686. 20:20creation agent, never give it "
  687. 20:21auto-send" or "auto-publish" permission
  688. 20:23. You build something inside the system
  689. 20:26we call a "Draft Gate," a gateway for
  690. 20:28drafts. Meaning the agent enters to
  691. 20:29read, analyze, write code, design the
  692. 20:31video, and prepare the full
  693. 20:32professional response. But instead of
  694. 20:33launching it publicly, it waits
  695. 20:35politely in the waiting room and sends
  696. 20:37a notification to your phone, saying: "
  697. 20:39Sir, I have prepared the complete work
  698. 20:41for five clients, waiting for your
  699. 20:42review and approval." See the magic
  700. 20:44here; you've transformed from someone
  701. 20:46overworked, spending 10 hours a day
  702. 20:48grinding and producing, into an
  703. 20:50operational maestro. All you do is open
  704. 20:52your dashboard while drinking your
  705. 20:54morning coffee, and use your human
  706. 20:56intelligence and taste to look at work
  707. 20:58that is 99%complete. A quick
  708. 21:00five-minute review, and you press one "
  709. 21:03Approve" button. The system immediately
  710. 21:05flies to execute in seconds. You’ve
  711. 21:06protected your work from any errors,
  712. 21:08while maintaining super-productive
  713. 21:10power that lets you manage 20 clients
  714. 21:12or become the most outstanding in your
  715. 21:14job with incredible quality and without
  716. 21:16any stress. A very important point: if
  717. 21:19you don’t start right now to use and
  718. 21:21harness digital employees (agents) to
  719. 21:24help you in your work and your life,
  720. 21:26your colleagues or competitors will
  721. 21:28surely do so. And the one who will win
  722. 21:30in the coming days is the one who gets
  723. 21:31ahead and does this with mastery and at
  724. 21:33the highest level. It is not too late
  725. 21:35yet, but don't wait until it is. So
  726. 21:37start now, and begin building your own
  727. 21:39army of digital employees. That’s it,
  728. 21:41my friend. Don't forget to check the
  729. 21:42links in the video description. You'll
  730. 21:44find things there that will help you
  731. 21:45greatly on your journey. And don't
  732. 21:46forget to watch the previous videos,
  733. 21:47and the upcoming ones too. If you liked
  734. 21:48the video, like and share it so
  735. 21:49everyone can benefit. And if you
  736. 21:51haven't subscribed to the channel yet,
  737. 21:52subscribe now. Turn on the
  738. 21:52notifications bell to get very valuable
  739. 21:54videos you won't find anywhere else.
  740. 21:55Thank you very much. Yasser Selim was
  741. 21:56with you. I'll see you well in a new
  742. 21:57video. And peace be upon you all.

About this transcript

This page contains the full transcript of أعطني 20 دقيقة.. وسأمنحك ميزة تفوق (غير عادلة) على الجميع بالذكاء الاصطناعي! by Yasser Seleem, generated from the public captions YouTube serves with the video. The transcript has 4,985 words across 742 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.