The Dirty AI lie : How the GREATEST bet in human history started to crack in June 2026? — Transcript

Full transcript

1. 0:02How do you think about this bubble talk
2. 0:03that has been going on for the last few
3. 0:05months especially?
4. 0:09>> I mean I I think it's quite possible.
5. 0:12>> Ladies and gentlemen, on 25th of June
6. 0:142026, Apple did something that it has
7. 0:16never done in its history. In the middle
8. 0:19of the year for no new product, it just
9. 0:22raised prices. MacBook Air is up by 18%,
10. 0:25iPad Pro is up by 20% and Apple TV is up
11. 0:29by 54%.
12. 0:33Apple said yesterday it is immediately
13. 0:35raising prices on the products.
14. 0:36>> The company says soaring memory chip
15. 0:38prices are driving up costs and the
16. 0:40[music] AI boom is a major factor behind
17. 0:42the surge.
18. 0:42>> The AI trade is leading to real
19. 0:45near-term inflation.
20. 0:48>> And when asked why, Apple said something
21. 0:50remarkable. They said, "We have never
22. 0:53seen a competent price increase this
23. 0:55much this quickly." And the reason your
24. 0:57laptop got more expensive is because of
25. 1:00a war being fought over tiny memory
26. 1:02chips thousands of kilometers away. And
27. 1:04look at this graph. In 2020, before Chad
28. 1:07GBD existed, the four biggest US tech
29. 1:10companies spent combined $90 billion on
30. 1:13capeex. In 2023, they spent $147
31. 1:16billion. In 2025 that number went up to
32. 1:20$410 billion and then in 2026 it is up
33. 1:25to $725 billion. So in 6 years the capex
34. 1:30has grown 8x and all of this is coming
35. 1:33just from Amazon, Meta, Google and
36. 1:36Microsoft. At the same time, the stock
37. 1:38market was going so crazy over the AI
38. 1:40wave that on 2nd of June 2026, Nvidia
39. 1:43was worth $5 trillion and analysts were
40. 1:46screaming to buy AI stocks.
41. 1:48>> The AI will be a pretty [music] good
42. 1:50thing to invest in.
43. 1:51>> It is going to be, I think, just a
44. 1:53booming year for AI and tech stocks,
45. 1:55[music] especially in the first half of
46. 1:57the year.
47. 1:57>> But just 3 days later, something started
48. 2:00cracking. Nvidia lost $320 billion in
49. 2:04market cap. By 24th June, Micron was
50. 2:06down by 13%, SanDisk was down by 10.59%,
51. 2:11Apple fell by 6.1% and Soft Bank tanked
52. 2:1412%. On top of that, OpenAI delayed its
53. 2:17IPO and slowly warnings are coming from
54. 2:20the smartest people on earth are
55. 2:22>> right now rising close to the same level
56. 2:25in 2010.
57. 2:26>> So what's the end? Is it a bubble that
58. 2:28bursts eventually?
59. 2:29>> I think it is. Yes. The problem with the
60. 2:31AI capex boom is not only is it immense
61. 2:34but a big chunk of it is funded with
62. 2:36debt and that pain doesn't stay
63. 2:38restricted. It spills over into the rest
64. 2:42of society. The second thing that
65. 2:43happens when people get very excited as
66. 2:46they are today about artificial
67. 2:47intelligence for example is every
68. 2:50experiment gets funded. This is a kind
69. 2:53of industrial bubble as opposed to
70. 2:55financial bubbles. Now looking at this
71. 2:57madness, I went back to understand all
72. 2:59the bubbles in history. I read the Wall
73. 3:01Street Journal, the Financial Times, the
74. 3:03CNBC transcripts and even the JP Morgan
75. 3:05gap analysis to understand why does Ray
76. 3:08Dalio call this a textbook example of a
77. 3:10bubble. And by the end of this video,
78. 3:11you will understand better than 99% of
79. 3:13Indian investors, whether this is the
80. 3:15greatest business bet in human history
81. 3:16or the greatest bubble ever inflated.
82. 3:18Why are Michael Bur, Ray Dalio, and Jeff
83. 3:20Bezos implying that this is a bubble?
84. 3:23And what happens when this bubble
85. 3:24bursts? Before we move on, if you're a
86. 3:27manufacturer or if you run an e-commerce
87. 3:29store, ODO has something very special
88. 3:31for you. ODO understands that as a
89. 3:32business owner, you are already juggling
90. 3:34between three different tools. One for
91. 3:35inventory, one for accounting, and one
92. 3:37for sales. And every single day,
93. 3:39something slips through the cracks even
94. 3:40with your best efforts. Some order goes
95. 3:42missing, the stock count doesn't match,
96. 3:44the customer calls get lost between two
97. 3:46teams, and hours are wasted just keeping
98. 3:48these disconnected systems alive. Now,
99. 3:50imagine if all of this ran from one
100. 3:52place. This is exactly where ODO comes
101. 3:54in. Odo is an all-in-one business
102. 3:56management software that brings together
103. 3:5845 easy to use applications under one
104. 4:01roof. So whether you are managing sales,
105. 4:02invoicing, inventory, projects, or even
106. 4:04building your own website, UDO runs
107. 4:06everything from a single platform. So no
108. 4:09more juggling and no more messy
109. 4:11integrations. And the real magic is that
110. 4:13every app talks to each other. So once
111. 4:15you make a sale on your e-commerce
112. 4:17store, the invoice is automatically
113. 4:18created and your stock is updated in
114. 4:20your inventory. So you always get a
115. 4:23real-time view of your business in one
116. 4:25place. And as your business grows, ODO
117. 4:27grows with you. And the best part is you
118. 4:29can get started with ODO for free today.
119. 4:31And as your business scales, you can
120. 4:33access the full suite for just 580
121. 4:35rupees per user per month. So if you
122. 4:37want to grow your business in a flash,
123. 4:39click the link in the description and
124. 4:40start your journey with ODO today.
125. 4:44>> [music]
126. 4:46>> This is the story of one of the greatest
127. 4:48bets in human history. Before we go
128. 4:50anywhere, let me install a mental model
129. 4:52in your head. Because if you don't
130. 4:53understand what a data center actually
131. 4:54is, none of the numbers will make sense.
132. 4:56Imagine your phone. When you type a
133. 4:58question into Chad Gupty, your phone
134. 5:00doesn't just answer by [music] itself.
135. 5:02Your phone is just a screen with Wi-Fi.
136. 5:04The actual thinking happens somewhere
137. 5:06else in a warehouse. A giant windowless
138. 5:09industrial warehouse which is filled
139. 5:11with metal racks. And each rack is
140. 5:13installed with thousands of these
141. 5:15[music] chips. That warehouse is called
142. 5:17a data center. Each one of these
143. 5:19buildings can hold 100,000 Nvidia GPUs.
144. 5:22Each Nvidia GPU cost 30 to $40,000.
145. 5:26So one building holds 3 to4 billion
146. 5:29worth of chips. One building just holds
147. 5:32[music]
148. 5:333 to4 billion worth of chips. This 3 to4
149. 5:37billion is [music] just for chips. On
150. 5:40top of that, you have to power them,
151. 5:42cool them, connect them with [music]
152. 5:43high-speed cables and build them with
153. 5:46concrete, security, and fire
154. 5:47suppression. All in all, a single large
155. 5:50AI data center cost 10 to 25 billion to
156. 5:53build. [music] And this is where the
157. 5:55race is happening. Like I told you in
158. 5:57the data center case study, if you look
159. 5:58at this graph, in 2010, the world
160. 6:01created or replicated two zetabytes of
161. 6:03data. That's roughly 2 trillion GB. But
162. 6:06fast forward to today, something
163. 6:08terrifying is happening. By 2026, the
164. 6:11[music] world is projected to generate
165. 6:13221 zetabytes of data. That's over 100x
166. 6:16more than in 2010. So, the world is
167. 6:20producing more data in a month than it
168. 6:22did in all of history until 2010. This
169. 6:25is the reason why the investment in data
170. 6:27centers has shot up from $90 billion to
171. 6:29$725 billion in just the last 6 years.
172. 6:33Now, here's a number that made me
173. 6:35question everything that is happening.
174. 6:36The PIMCO report says that big tech
175. 6:39capeex will consume 94% of operating
176. 6:42cash flows. I repeat 94% of operating
177. 6:46cash flows over the next 2 years. You
178. 6:49know what that means? If big tech earns
179. 6:52$100, they will spend $94 back into
180. 6:55building AI infrastructure. Only $6 will
181. 6:58be left for dividends, buybacks, salary
182. 7:01hikes, innovation, and everything else.
183. 7:04In 2023, that same ratio was just 40%.
184. 7:07And now it stands at 94%. So do you
185. 7:10realize big tech is betting 94% of all
186. 7:14its money into just one assumption. And
187. 7:16the assumption says that in just 5
188. 7:18years, the world will need so much AI
189. 7:21compute that every dollar being spent
190. 7:23right now will practically look like a
191. 7:25bargain. Sounds unstoppable, right?
192. 7:27After all, we are producing so much
193. 7:29data. Well, here's where it gets
194. 7:31dangerous. Now, let's forget economics
195. 7:33for a second and just imagine that you
196. 7:34are a business owner. Let's say you
197. 7:36spend $10 million building a coffee
198. 7:38machine factory. Now, imagine that after
199. 7:40all that, your factory only sells
200. 7:42$400,000 worth of coffee machines in a
201. 7:45year. So, you just make $400,000 from a
202. 7:48factory that cost you $10 million. Is
203. 7:51that good, bad, or terrible? You tell
204. 7:53me. It's terrible, right? Why would you
205. 7:56build another factory if your current
206. 7:57factory doesn't make any money? Now take
207. 8:00that exact same example and apply it to
208. 8:02AI. Now let me show you the math. JP
209. 8:05Morgan sat down and did this calculation
210. 8:07and the logic is pretty simple. If
211. 8:09you're an investor and you put money
212. 8:10into something, you would at least
213. 8:12expect a 10% return. That's bare minimum
214. 8:14any serious investor demands on a risky
215. 8:16bet like this. So JP Morgan said for AI
216. 8:18giants to justify all the money that
217. 8:20they're spending, how much money does AI
218. 8:22actually need to bring in every year?
219. 8:24The answer was $650 billion every single
220. 8:28year. Okay, now remember this figure,
221. 8:31$650 billion. Now, do you know how much
222. 8:34AI is actually earning right now? Let's
223. 8:36add it up. OpenAI, the makers of Chad
224. 8:38GBT, make $25 billion a year, and
225. 8:40they're losing $14 billion a year.
226. 8:43Anthropic is set to make $47 billion at
227. 8:45best if their current run rate goes on
228. 8:47for one year. As of now, the target for
229. 8:50Anthropic is about $26 billion by the
230. 8:52end of this year. And let's say Gemini
231. 8:54also makes $25 billion. So every major
232. 8:57AI model company combined make around
233. 9:00$75 billion with OpenAI losing 14
234. 9:03billion and Anthropic losing 3 billion
235. 9:06in 2025 alone. Now put these three
236. 9:08numbers side by side. Money that AI
237. 9:10needs to earn to make sense $650
238. 9:12billion. Money AI is actually earning
239. 9:16$75 billion. Money that AI is losing is
240. 9:19minimum $17 billion. But the money that
241. 9:21the giants are spending on top of all of
242. 9:23this is $725 billion. That difference
243. 9:28between what they earn and what they
244. 9:30need to earn is about 9 to 10 times. Now
245. 9:34read that one more time slowly. For
246. 9:36every single dollar that the AI industry
247. 9:38is bringing in, the tech giants are
248. 9:41spending 9 to 10 times more than they
249. 9:44earn. This is why SEOA's David Khan
250. 9:46calls this the $600 billion question. a
251. 9:49$600 billion annual revenue deficit that
252. 9:52nobody knows who will fill. Now the
253. 9:53single biggest argument against this
254. 9:55crazy number is Ganesh enterprises will
255. 9:57pay money. Every single one of these
256. 9:59companies will become profitable and
257. 10:01investors will make money when the
258. 10:03enterprises will pay money because AI is
259. 10:06making all enterprises very very
260. 10:07efficient at dirt cheap cost. Okay.
261. 10:11Well, that is not the right argument
262. 10:14because even I thought the same and then
263. 10:15I found the service. McKenzie says 73%
264. 10:18of enterprise AI deployments are failing
265. 10:20to achieve projected return on
266. 10:22investment. BCG says only 5% of
267. 10:25companies are seeing substantial ROI
268. 10:26from AI. MIT says there is a 95% failure
269. 10:29rate in achieving measurable financial
270. 10:31returns. Only 29% of the executives can
271. 10:33even measure their AI return on
272. 10:35investment. And this is where the story
273. 10:38gets its first phase. Meet Flo. He runs
274. 10:40an AI startup in San Francisco called
275. 10:42Lindy. They have about 25 employees. In
276. 10:45June 2026, he did an interview with CNBC
277. 10:47that shook the AI industry. His team was
278. 10:50spending more on Anthropic Cloud API
279. 10:52than on their entire payroll. So, you
280. 10:55know what Flo did? Flo switched 100% of
281. 10:57his traffic to Deep Seek and his cost
282. 10:59dropped by 90%. And then Uber CTO
283. 11:01admitted publicly that Uber had blown
284. 11:03its entire annual AI budget in just 4
285. 11:06months. And that ladies and gentlemen is
286. 11:09the twist because everyone assumed that
287. 11:11enterprises would keep paying more and
288. 11:13more for AI tokens forever. That was the
289. 11:16whole model. That is why OpenAI is worth
290. 11:18$850 billion. That is why Anthropic is
291. 11:21worth $965 billion. But in June 2026,
292. 11:24Enterprise started doing something that
293. 11:25the market did not expect. They started
294. 11:28looking for cheaper alternatives. Which
295. 11:30is why Alex Karp, the CEO of Palanteer
296. 11:33went on CNBC and said this on 1st of
297. 11:35July. Every single enterprise I deal
298. 11:38with, they're like, I am paying for
299. 11:39tokens that create no value. These
300. 11:42people are stealing the weights and
301. 11:43alpha of my business and they're
302. 11:45creating a wealth tax. And the reason
303. 11:46for it is because [music] these models
304. 11:49have been completely over irresponsibly
305. 11:51oversold. And Palanteer, if you saw our
306. 11:53previous case study, is one of the
307. 11:55biggest software enterprise companies on
308. 11:57earth. They sell to the CIA, the US
309. 11:59government, Airbnb, JP Morgan, and god
310. 12:02knows how many large companies. and the
311. 12:05CEO of that company is telling you that
312. 12:07something has completely gone wrong.
313. 12:09Now, at this point, I know exactly what
314. 12:11you're thinking. You must be thinking,
315. 12:12"Yeah, Ganesh, this is a rich man's
316. 12:14problem. Nvidia losing 500 billion, Sam,
317. 12:16Dario, Sundar, they're all billionaires.
318. 12:18How am I getting affected by all of
319. 12:20this?" Well, let me take you to South
320. 12:22Korea and show you how.
321. 12:26This is a factory in South Korea that is
322. 12:27owned by Samsung. This factory makes a
323. 12:30very specific kind of memory chip called
324. 12:32DM. the same DAM that goes into your
325. 12:34laptop, your smartphone, your Xbox, and
326. 12:37even your washing machine. In 2024,
327. 12:39Samsung had a choice. It could sell its
328. 12:41DAM to consumer companies like Apple,
329. 12:43HP, or Dell. Or it could sell a special
330. 12:47extremely expensive version called high
331. 12:50bandwidth memory to AI data centers. And
332. 12:53guess which one pays more? The AI data
333. 12:56centers paid 10x more per module. So
334. 13:00Samsung, SKH Highix and Micron, the
335. 13:02three companies that control 90% of the
336. 13:04world's memory chip supply, did what
337. 13:06[music] any factory would do. They
338. 13:08shifted 93% of their production towards
339. 13:11AI memory because that is a rule of
340. 13:13capitalism, right? Capital always flows
341. 13:15to the highest bidder. Now watch what
342. 13:17happens at bigger scales. DM prices are
343. 13:20up by 171% year-over-year as of March
344. 13:232026. DDR5 memory are up 4x since
345. 13:27September 2025. A contract price for PC
346. 13:29memory is up by 105 to 110% in one
347. 13:32quarter. In fact, Dell CEO said that the
348. 13:34price of 1 GB of DAM went from 0.43 to
349. 13:37$2.39 in just 6 months. That is a 5 and
350. 13:41a half times increase in price. In fact,
351. 13:43that is why on 25th of June 2026, Apple
352. 13:46did something that it had never done
353. 13:47before. In the middle of a product year,
354. 13:50Apple simply raised their prices. This
355. 13:52is the reason why they said, "We have
356. 13:54never seen a competent price increase
357. 13:56this much this quickly. We've shielded
358. 13:58our customers from these increases so
359. 14:00far. But now we've reached a point where
360. 14:02we need to begin raising prices. Now
361. 14:04that is Apple telling you this guys. The
362. 14:06richest most vertically integrated tech
363. 14:08company in the world is telling you that
364. 14:10they cannot absorb this cost. That is
365. 14:13how you are paying the AI tax. But this
366. 14:16is where a scary question arises. If
367. 14:19enterprises are moving off claw to save
368. 14:2190%. If Apple cannot absorb cost
369. 14:24anymore, if the ROI is broken, then why
370. 14:27are Amazon, Microsoft, Google, and Meta
371. 14:29still spending more? Well, the answer is
372. 14:32one of the most fascinating concepts in
373. 14:34economics, and it explains every single
374. 14:36bubble in human history. It's called the
375. 14:38capital cycle. [music] In this cycle,
376. 14:40there are four steps. Step number one,
377. 14:42high returns attract capital. Step
378. 14:44number two, capital keeps flowing until
379. 14:47over capacity is built. Step three,
380. 14:49return over capacity eventually results
381. 14:51into collapse. And step four, everyone
382. 14:54dies except a few survivors who
383. 14:56eventually make a fortune when demand
384. 14:58catches up. And every bubble in modern
385. 15:00history has followed this exact same
386. 15:02pattern. Let me show you how. In 1996,
387. 15:05the US passed the Telecommunications Act
388. 15:07because just like AI, the internet back
389. 15:09then was a life-changing technology
390. 15:11which was exploding in demand. The story
391. 15:13was so intoxicating because it was clear
392. 15:15to the world that internet was the
393. 15:17future. Data traffic was exploding and
394. 15:19everybody just knew that bandwidth
395. 15:21demand would grow forever. Some founders
396. 15:24even believed that internet traffic
397. 15:26would double every 3 months. So money
398. 15:28came pouring in to build the fiber optic
399. 15:31cables. And then came the flood. Several
400. 15:33companies raised to lay fiber optic
401. 15:34cables across the country. And in just 5
402. 15:37years after that act, telecom companies
403. 15:39poured more than $500 billion into
404. 15:41cables, switches and networks. And if
405. 15:43you look at the financials of these
406. 15:44companies, you will see why the AI
407. 15:46bubble is very similar. A company called
408. 15:48Global Crossing went from a small equity
409. 15:50check to a $47 billion valuation without
410. 15:54ever making a single year of profit.
411. 15:56Corvis, a fiber equipment startup,
412. 15:57pulled off a $1.1 billion IPO with
413. 16:00literally 0 in revenue and carried a $32
414. 16:03billion market cap. And just when
415. 16:05everyone thought they'll become
416. 16:06millionaires and billionaires, the
417. 16:08collapse happened. You know what
418. 16:10happened? Everyone thought that the
419. 16:12internet will explode by 1,000% year on
420. 16:14year, but the internet traffic only
421. 16:16exploded by 100% year on year, which was
422. 16:18great, but not great enough to justify
423. 16:21the cost of investment. You know how
424. 16:23much of this installed fiber was
425. 16:24actually utilized? Take a guess. 50%,
426. 16:2820%, [music]
427. 16:3010%, at least 5% must have been
428. 16:32utilized, right? Well, guess what? By
429. 16:34early 2000s, as little as just 2.7% of
430. 16:38the installed fiber was actually
431. 16:40carrying data. Over 95% sat unused
432. 16:43underground. That is how trillions of
433. 16:45dollars of cable got buried without
434. 16:46earning anything. So when there was no
435. 16:48revenue, bandwidth prices collapsed by
436. 16:50up to 90% and the giant started failing.
437. 16:53WorldCom, after hiding $3.8 billion of
438. 16:55expenses to fake profits, filed the
439. 16:57biggest bankruptcy in US history. Global
440. 16:59Crossing, that $47 billion darling went
441. 17:01bankrupt. And in total, the telecom
442. 17:03crash wiped out $2 trillion of market
443. 17:05value with stocks going down by 95%. And
444. 17:08then [music] came step four, the
445. 17:10survivors. Now, here's where the twist
446. 17:13comes in which makes it the perfect
447. 17:14mirror for AI. Those fiber optic cables
448. 17:17did not vanish. They stayed in the
449. 17:19ground and within a few years, demand
450. 17:21finally arrived. YouTube happened,
451. 17:23streaming started, cloud storage became
452. 17:25a real thing, and [music] smartphone
453. 17:27became popular. And suddenly the world
454. 17:29needed exactly what had been overbuilt.
455. 17:32So the survivors bought the wreckage for
456. 17:34dirt cheap prices and that wasted cable
457. 17:36became the physical backbone of the
458. 17:38modern internet. The same infrastructure
459. 17:40that made Google, Netflix [music] and
460. 17:42AWS possible. So do you realize that
461. 17:45technology was real? The internet did
462. 17:47change everything but the bubble still
463. 17:49burst. Why? Because the demand was
464. 17:51exploding but not so much to justify
465. 17:53over capacity. So the technology
466. 17:55survived but the companies that built
467. 17:57did not. Now, here's what the pattern
468. 17:58looks like. Britain in 1846 authorized
469. 18:019,500 miles of track and one/ird of it
470. 18:03never got built. And then the bubble
471. 18:05burst. America in 2000 laid millions of
472. 18:07miles of fiber and 97% of it was unused
473. 18:10and eventually the bubble burst. In
474. 18:122026, America alone is building 725
475. 18:16billion of data centers per year and we
476. 18:19don't know how much of it will actually
477. 18:21be used. So the question is, will it all
478. 18:22be worth it and become the greatest tech
479. 18:24story ever told? Or will it go down as
480. 18:27the greatest bubble in world history?
481. 18:29Only time can give us the answer. So is
482. 18:32this definitely a bubble? Well, we don't
483. 18:34know that yet. Why? Because the
484. 18:35companies in the telecom bubble were
485. 18:37funded by debt and they were losing
486. 18:38money. But Nvidia earned $120 billion in
487. 18:41net income last year. And Microsoft,
488. 18:43Google, and Amazon are literally the
489. 18:44most profitable enterprises in human
490. 18:46history. So they won't collapse like
491. 18:47other weak companies. Similarly, at the
492. 18:492000.com peak, the NASDAQ 100 forward PE
493. 18:53was about 60x. Today, it's around 26x.
494. 18:56It's higher than normal, but nowhere
495. 18:58near the insanity of 1999. So, if anyone
496. 19:01tells you for certain that this is a
497. 19:02bubble, they're lying to you. And anyone
498. 19:04tells you that it is definitely not a
499. 19:06bubble is also lying to you because the
500. 19:08truth is uncomfortable and it's
501. 19:09somewhere in between. There is a very
502. 19:11high possibility of a bubble, but not a
503. 19:13certainty. The technology is real, the
504. 19:15revenue is real, and we're not betting
505. 19:17on whether AI changes the world or not.
506. 19:18We are betting on whether the price for
507. 19:20it actually makes sense or not. So now
508. 19:23the question is what exactly is going to
509. 19:24happen if the bubble burst? And what if
510. 19:26it doesn't? Well, there are two
511. 19:29possibilities. Path one, the bubble
512. 19:31pops, jobs are lost, the NASDAQ crashes,
513. 19:34and every big tech company slams the
514. 19:36brakes on spending, and that spending is
515. 19:38what feeds our Indian IT and service
516. 19:40sector. So your cousin's first coding
517. 19:42job disappears before the boom can catch
518. 19:44him. Path two is that the bubble doesn't
519. 19:46pop. Instead, to justify those trillion
520. 19:48dollar valuations, the company will try
521. 19:49to race towards profit. So the price of
522. 19:51AI, as in the token cost will shoot up
523. 19:53and suddenly only the giants will be
524. 19:55able to afford AI. So the cheap AI tools
525. 19:57that you use today will eventually
526. 19:59become a luxury. So a lot of AI products
527. 20:01might die not because the tech failed,
528. 20:03but because it just got too expensive to
529. 20:04run. Or lastly, we could expect a
530. 20:07miracle that will drop down the token
531. 20:09cost, will make enterprises pay, and
532. 20:11everyone will make money. But that, my
533. 20:14dear friends, is a teeny tiny
534. 20:16possibility. This, my dear friends, is
535. 20:18the story of the AI bubble. Now, you
536. 20:19tell me in the comments what do you
537. 20:20think about the situation with the
538. 20:22trillion dollar valuation that we
539. 20:23seeing. Is this really a bubble or is
540. 20:25this the greatest bet humanity has ever
541. 20:27taken? That's all from my side for
542. 20:29today, guys. If you learned something
543. 20:30valuable from this case study, please
544. 20:32hit the like button to support our work.
545. 20:34And for more such business and political
546. 20:35case studies, please subscribe to our
547. 20:37channel. Thank you so much for watching.
548. 20:38I will see you in the next one. Bye-bye.
549. 20:50>> [music]