1
00:00:00,115 --> 00:00:03,577
There's no easy way to say this, but we might be witnessing the collapse of

2
00:00:03,577 --> 00:00:08,017
the AI bubble in real time. Over the last two weeks, we've seen over $2 trillion

3
00:00:08,437 --> 00:00:09,740
wiped out in the stock market.

4
00:00:10,053 --> 00:00:13,977
Amazon's down 12%, wiping off half a trillion dollars of its market cap.

5
00:00:14,227 --> 00:00:17,060
Google is the same, wiping off another half a trillion dollars.

6
00:00:17,304 --> 00:00:23,117
And Intel has seen a drawdown of 30% last Friday in a single day in the opening

7
00:00:23,117 --> 00:00:25,697
hours, its biggest drawdown in history.

8
00:00:25,697 --> 00:00:29,297
And it's all because people are losing faith with the hundreds of billions of

9
00:00:29,297 --> 00:00:34,657
dollars that are being invested in AI CapEx to buy GPUs and to build these new models.

10
00:00:34,657 --> 00:00:38,747
It's getting very costly and people don't see the return on that investment.

11
00:00:38,747 --> 00:00:40,953
They don't see how AI is going to make money.

12
00:00:41,214 --> 00:00:44,157
And this isn't just something happening in the West. China's stock market is

13
00:00:44,157 --> 00:00:48,427
down 40% over the last couple of months, the biggest crash they've ever seen.

14
00:00:48,427 --> 00:00:52,367
And the big question on my mind is, are we in another dot com era bubble that

15
00:00:52,367 --> 00:00:55,557
is just a hundred times bigger and is going to result in a much

16
00:00:55,897 --> 00:01:00,747
worse crash well the thing is someone's already answered that question for us

17
00:01:00,747 --> 00:01:06,067
the hyperscalers amazon microsoft meta google and all the ai labs have made

18
00:01:06,067 --> 00:01:07,457
their decision and it's clear

19
00:01:07,882 --> 00:01:12,037
they're doubling down. Next year, we're planning to spend over $1 trillion on

20
00:01:12,037 --> 00:01:15,607
compute, buying GPUs, building data centers, investing in power, etc.

21
00:01:16,047 --> 00:01:21,727
To stand up these next gen AI models that we still haven't quite seen the ROI from.

22
00:01:21,727 --> 00:01:28,477
The craziest part is all of this is being basically paid for by two companies, Anthropic and OpenAI.

23
00:01:28,477 --> 00:01:31,927
So the big question mark is, are these companies, one, going to be able to make

24
00:01:31,927 --> 00:01:35,487
enough money and two, continue growing at the rate that they are?

25
00:01:35,487 --> 00:01:39,387
Well, the answer is there's a lot of nuance. It's partly like the dot-com era

26
00:01:39,387 --> 00:01:42,437
bubble, and there's certainly things that we need to watch out for and manage

27
00:01:42,437 --> 00:01:44,157
risk, but it's not entirely true.

28
00:01:44,157 --> 00:01:47,447
If you look at any of the earnings reports from any of these companies that

29
00:01:47,447 --> 00:01:51,867
are tanking over the last week, you'll see that they're making money and they're

30
00:01:51,867 --> 00:01:56,735
having record quarters making more money than they have in the prior 10 years.

31
00:01:56,979 --> 00:02:00,077
So the question we have to ask is, why is the market worried?

32
00:02:00,077 --> 00:02:02,320
What do they see that companies don't?

33
00:02:02,506 --> 00:02:05,837
And why they're still questioning that the money that's being made isn't really

34
00:02:05,837 --> 00:02:08,927
real. So we'll get into all of this and I'm going to break it down for you on

35
00:02:08,927 --> 00:02:10,758
today's episode brought to you by Qualcomm.

36
00:02:14,513 --> 00:02:18,762
So first, some quick context. This drawdown really started around two weeks ago.

37
00:02:18,931 --> 00:02:22,431
And it was from one single tweet from Elon Musk, where he responded,

38
00:02:22,681 --> 00:02:27,821
yikes, to the over leveraging of debt seen in neoclouds, neoclouds being the

39
00:02:27,821 --> 00:02:32,097
ones that buy up this compute GPUs and serve it to some of the AI labs.

40
00:02:32,288 --> 00:02:36,661
And the company in focus at that time was CoreWeave. CoreWeave subsequently

41
00:02:36,661 --> 00:02:39,916
then dropped down around 30 to 35% in its stock price,

42
00:02:40,142 --> 00:02:44,161
as well as other competitors that were close to it, like nebius and iron and

43
00:02:44,161 --> 00:02:47,321
so people really started to double down and ask the question are we spending

44
00:02:47,641 --> 00:02:52,771
way too much money on gpus and all of these gpus actually being used do we have

45
00:02:52,771 --> 00:02:57,201
the power and the electricity to power them now that then tied into last week's.

46
00:02:57,541 --> 00:03:02,061
Record earnings week where basically the majority of the companies that being

47
00:03:02,061 --> 00:03:07,321
hyperscalers like microsoft amazon came out with their quarterly earnings and

48
00:03:07,321 --> 00:03:11,321
it gives you an insight into the numbers as to what's actually going on behind the scenes.

49
00:03:11,588 --> 00:03:16,348
Three things were very clear. One, they're spending more money than they can afford.

50
00:03:16,603 --> 00:03:20,781
Their free cash flow has officially gone negative for the first time for all

51
00:03:20,781 --> 00:03:25,071
three of these companies, Google, Amazon, and Microsoft, for the first time in a decade.

52
00:03:25,071 --> 00:03:28,921
So they're technically in the red, and they're borrowing a lot of money and

53
00:03:28,921 --> 00:03:32,871
using profits from other business lines of their business to double down on

54
00:03:32,871 --> 00:03:35,492
this bet that we still haven't seen play out.

55
00:03:35,776 --> 00:03:43,851
Then number two, the ROI on that investment is starting to be seen from them selling compute.

56
00:03:44,124 --> 00:03:49,051
So they would sell compute to say like Anthropic or OpenAI, and they would make a profit on that.

57
00:03:49,051 --> 00:03:53,571
And all of their cloud service providers, so AWS, Google Cloud,

58
00:03:53,571 --> 00:03:56,751
Microsoft Azure, posted record earnings.

59
00:03:56,751 --> 00:04:00,383
Then that leads me to point three, which is, I guess, why the market it is tanking,

60
00:04:00,755 --> 00:04:06,914
the fundamental business model of the internet, which is basically search and ads.

61
00:04:07,403 --> 00:04:10,607
Is crumbling and Google is kind of like the leader here.

62
00:04:10,874 --> 00:04:15,340
The point is that Google has driven ads and sales and revenue for the entire

63
00:04:15,340 --> 00:04:17,614
Internet economy since the start of it.

64
00:04:17,881 --> 00:04:22,192
So the idea here is any company that sits on top of Google Search,

65
00:04:22,489 --> 00:04:26,390
basically relies on Google Search and search is starting to break.

66
00:04:26,390 --> 00:04:30,430
The main reason why it's starting to break is people aren't going to google.com

67
00:04:30,430 --> 00:04:34,410
anymore. They're not searching things up. They're not scrolling down a list

68
00:04:34,410 --> 00:04:36,241
of links and trying to select which one.

69
00:04:36,617 --> 00:04:39,532
They're just using AI to do all of that for them.

70
00:04:39,532 --> 00:04:44,342
They're outsourcing it to AI. And AI doesn't have a monetizable search revenue.

71
00:04:44,342 --> 00:04:49,012
And most importantly, it doesn't belong to that system that Google set up.

72
00:04:49,012 --> 00:04:54,962
It belongs to Anthropic. It belongs to OpenAI and any other competitor that has an AI chatbot.

73
00:04:54,962 --> 00:04:59,022
And so that's why you're starting to see this affect companies that aren't necessarily

74
00:04:59,022 --> 00:05:02,042
affiliated with the AI bubble. Let's take Reddit, for example.

75
00:05:02,042 --> 00:05:05,432
They announced their earnings. Again, you guessed it, record earnings,

76
00:05:05,432 --> 00:05:07,691
all positive, no reason why the stock would crash.

77
00:05:07,981 --> 00:05:13,479
And yet it crashed around 26% in the opening hours of its trading. Why?

78
00:05:13,804 --> 00:05:15,562
Because they noted two things.

79
00:05:15,902 --> 00:05:22,011
One, search ad revenue wasn't driving enough traffic to their site as usual.

80
00:05:22,191 --> 00:05:25,312
And that's because, as Google mentioned a day earlier in their earnings,

81
00:05:25,312 --> 00:05:29,992
we're just losing kind of our grip on this because AI search doesn't monetize

82
00:05:29,992 --> 00:05:33,532
in the same way. But there's also a second thing happening, which is arguably

83
00:05:33,532 --> 00:05:36,506
more concerning. And I'll stay on Reddit as an example.

84
00:05:36,628 --> 00:05:42,802
Reddit is starting to build up walls around its service to stop people accessing

85
00:05:42,802 --> 00:05:47,252
it for free, which is crazy because the internet economy, the way that it was

86
00:05:47,252 --> 00:05:51,302
able to prosper in the first place, was because we ended up becoming more open.

87
00:05:51,302 --> 00:05:54,842
It started off very closed with the intranet, and then we moved over for it

88
00:05:54,842 --> 00:05:58,272
becoming much more open. And society and the world benefited a lot more.

89
00:05:58,272 --> 00:06:01,194
We had all free access to information.

90
00:06:01,572 --> 00:06:05,842
And Reddit has been kind of like the bedrock and core of the cultural pulse

91
00:06:05,842 --> 00:06:09,606
on society, if that makes sense. And now if you want to access Reddit information,

92
00:06:09,896 --> 00:06:12,276
you need to create an account and log on to it.

93
00:06:12,479 --> 00:06:16,602
And the main reason why they're doing this is they're concerned that their entire

94
00:06:16,602 --> 00:06:20,282
company in Moat will get replaced by AI.

95
00:06:20,582 --> 00:06:25,402
Reddit basically has an enormous trove of data from people's comments,

96
00:06:25,402 --> 00:06:29,612
subreddits, threads, et cetera. And so this information is very valuable if

97
00:06:29,612 --> 00:06:31,524
it's working to inform you to.

98
00:06:31,919 --> 00:06:35,023
Make a purchasing decision or research something or whatever that might be.

99
00:06:35,023 --> 00:06:37,277
News often breaks on Reddit.

100
00:06:37,648 --> 00:06:42,676
And to date, they've been given that away for free. But what AI is doing is

101
00:06:42,948 --> 00:06:47,883
they're taking this data, they're training models to then become smart on this

102
00:06:47,883 --> 00:06:51,551
data, to effectively become your own personal Reddit.

103
00:06:51,865 --> 00:06:54,936
And then Reddit's not getting paid for any of this. Now, traditionally,

104
00:06:55,214 --> 00:06:58,941
Google would drive all of the traffic to Reddit.

105
00:06:59,202 --> 00:07:01,043
Reddit can display ads, et cetera, et cetera.

106
00:07:01,368 --> 00:07:05,643
But if you're just communicating to the internet via your AI agent or your AI

107
00:07:05,643 --> 00:07:10,882
chatbot, why would you go to any other website bar just your chatbot LLM interface,

108
00:07:11,085 --> 00:07:14,213
such as Claude ChatGPT or your personal agent interface?

109
00:07:14,213 --> 00:07:18,453
And so that's one of the main reasons why I think we're seeing a negative correction

110
00:07:18,453 --> 00:07:19,930
towards the entire AI bubble.

111
00:07:20,243 --> 00:07:25,673
But the second thing is equally as important, and that is a potential overbuild

112
00:07:25,673 --> 00:07:30,785
in AI capex. So right now, we have invested on order of, I believe...

113
00:07:31,220 --> 00:07:35,940
Just over a trillion dollars buying GPUs, standing up data centers,

114
00:07:36,155 --> 00:07:41,344
connecting it with power lines, all to train AI models that we don't quite know

115
00:07:41,542 --> 00:07:45,849
how it's going to end up benefiting society in the long term, right?

116
00:07:46,104 --> 00:07:49,720
There's a lot of question marks around, okay, what's the necessary ROI from this?

117
00:07:49,720 --> 00:07:52,937
In fact, we had a report from the chief economist of Apollo,

118
00:07:53,337 --> 00:07:57,000
I think a few weeks ago, basically saying, we haven't seen where the money's

119
00:07:57,000 --> 00:08:01,581
being made outside of the seven companies, the seven top companies in the world.

120
00:08:02,016 --> 00:08:05,620
Seven top companies in the world are these hyperscalers who are buying these

121
00:08:05,620 --> 00:08:08,912
GPUs to then sell to two companies, Anthropic and OpenAI.

122
00:08:09,191 --> 00:08:12,690
So you can see how the circular economy is being criticized right now,

123
00:08:12,690 --> 00:08:15,180
basically saying, I don't know where the money is coming from.

124
00:08:15,180 --> 00:08:16,951
But it's not just where the money is coming from.

125
00:08:17,177 --> 00:08:21,014
It's also the ability to deliver on this. Two points here.

126
00:08:21,299 --> 00:08:25,910
Microsoft's intended spend for next year is around $170 billion.

127
00:08:26,523 --> 00:08:32,670
That's huge. but a chunk of that comes from this year where neoclouds or third

128
00:08:32,670 --> 00:08:36,900
parties which they engaged with to deliver and stand up these data centers weren't

129
00:08:36,900 --> 00:08:39,097
able to deliver on it and the main reason is,

130
00:08:39,567 --> 00:08:42,630
they couldn't construct it fast enough there was too much regulatory red tape

131
00:08:42,970 --> 00:08:47,503
and it was just hard to get their hands on certain credit financing and gpus

132
00:08:47,787 --> 00:08:51,450
and the second point is they're realizing that it's so much harder

133
00:08:51,770 --> 00:08:55,840
to build these data centers than they initially projected and it's not a matter

134
00:08:55,840 --> 00:09:00,123
of, oh, it'll get delayed by a few weeks or maybe a couple months in the worst case scenario.

135
00:09:00,459 --> 00:09:04,580
We're looking at potentially years. So there's two forces at play here.

136
00:09:04,580 --> 00:09:08,690
One, who's paying for all of this? Right now, it's two companies,

137
00:09:08,690 --> 00:09:13,590
Anthropic and OpenAI, and a heck ton of financing and loans.

138
00:09:13,590 --> 00:09:17,000
Just to put this into perspective, I think I saw a JP Morgan report,

139
00:09:17,000 --> 00:09:22,160
which basically said that around $2.9 trillion is planned on being spent by

140
00:09:22,160 --> 00:09:28,363
some of the top, I think, 10 companies in the world over the next couple of years, up until 2028.

141
00:09:28,898 --> 00:09:33,529
Half of that is being financed by the hyperscalers and these companies themselves.

142
00:09:33,529 --> 00:09:38,049
So they will pay for it off their own bank balance sheet as well as any money

143
00:09:38,049 --> 00:09:39,190
that they make in the future.

144
00:09:39,660 --> 00:09:44,789
But the other half is being financed. And that financing is getting quite expensive right now.

145
00:09:44,789 --> 00:09:48,399
We saw last week where CoreWeave's borrowing debt basically rose to around,

146
00:09:48,399 --> 00:09:51,649
I think, 16% when it was 10% literally two months ago.

147
00:09:51,649 --> 00:09:52,808
And we're starting to see companies,

148
00:09:52,977 --> 00:09:55,949
even the top labs like Anthropic are struggling to get financing.

149
00:09:55,949 --> 00:10:01,499
They recently tried to close a, I think, $63 billion deal, which Hocktan from

150
00:10:01,499 --> 00:10:07,229
Broadcom had to meet with the bankers and really push them to give them better rates.

151
00:10:07,229 --> 00:10:10,309
Otherwise, the deal wasn't going to go ahead. So there's a lot of pressure in

152
00:10:10,309 --> 00:10:13,247
this bubble when it comes to money. The money is effectively thinning.

153
00:10:13,525 --> 00:10:19,969
And the second force is the business model for AI relies on these models being

154
00:10:19,969 --> 00:10:24,328
able to train on the open data of the web.

155
00:10:24,566 --> 00:10:28,589
So we're talking about all the data in the web that can allow it to get smarter

156
00:10:28,589 --> 00:10:31,857
and smarter with every model generation. But if we have companies like Reddit,

157
00:10:32,333 --> 00:10:35,195
which owns a large portion of the data on the web.

158
00:10:35,700 --> 00:10:41,227
Putting up walls and banning access for AI labs or models to kind of get smarter,

159
00:10:41,662 --> 00:10:46,115
you can see how all of this money being invested in CapEx and GPUs might actually

160
00:10:46,115 --> 00:10:51,025
not be put to good use because we're going to see a slowdown in models getting smarter.

161
00:10:51,025 --> 00:10:54,895
And on top of that, it also makes it harder to build a really cool product.

162
00:10:54,895 --> 00:10:58,955
I use the example of Metamuse, which is like one of my favorite personal agents

163
00:10:58,955 --> 00:11:03,055
and honestly, favorite AI products of the year. And I have no vested interest

164
00:11:03,055 --> 00:11:06,425
in meta at all. I just enjoy the product. It's so good because I could just

165
00:11:06,425 --> 00:11:09,506
talk to it and it goes off and does things for me, right?

166
00:11:09,738 --> 00:11:14,515
But one of the main enablers allowing it to go do things for me is having access

167
00:11:14,515 --> 00:11:17,115
to these different types of information. And one of the things that it does

168
00:11:17,115 --> 00:11:21,831
for me is it can do some of my Amazon shopping or Instacart shopping or DoorDash ordering, right?

169
00:11:22,290 --> 00:11:27,645
But Amazon recently shut off access to Metamuse for the exact same reason that

170
00:11:27,645 --> 00:11:30,045
Reddit has, which is I don't want you.

171
00:11:30,605 --> 00:11:35,875
Meta, being the interface that kind of drives traffic towards our website that

172
00:11:35,875 --> 00:11:37,730
drives more purchasing utility.

173
00:11:37,975 --> 00:11:42,972
I want it to be the old model where I'm in control and I'm able to figure out

174
00:11:43,251 --> 00:11:45,615
what customers want and what they don't want.

175
00:11:45,615 --> 00:11:50,055
In the traditional mode, you and I would go to the Amazon website and we would

176
00:11:50,055 --> 00:11:53,035
scroll and Amazon would see that. They would see what we click on.

177
00:11:53,035 --> 00:11:54,765
They would see what we review, etc, etc.

178
00:11:55,285 --> 00:11:59,185
If it's instead an AI agent doing that, if it's instead an interface that I

179
00:11:59,185 --> 00:12:00,805
interact with that has nothing to do

180
00:12:01,205 --> 00:12:05,835
with Amazon's website, then they lose a lot of important customer behavior and

181
00:12:05,835 --> 00:12:09,965
analytics that will inform how they are to build billions and millions of dollars

182
00:12:09,965 --> 00:12:12,002
more value for their company.

183
00:12:12,269 --> 00:12:16,605
So the conundrum here is if they build up walls, if a company as big as Amazon

184
00:12:16,605 --> 00:12:19,380
builds up walls, then every other company is going to do the same thing.

185
00:12:19,542 --> 00:12:23,415
And what that world looks like, AI agents interacting with AI agents is still

186
00:12:23,415 --> 00:12:28,445
unclear. Okay, so are we in a dot-com era bubble? Let's answer that question

187
00:12:28,445 --> 00:12:30,170
on today's episode brought to you by Qualcomm.

188
00:12:34,180 --> 00:12:38,833
Firstly, the bear case is, let's see, there's a concentration risk.

189
00:12:38,833 --> 00:12:43,003
So seven of the top companies in the world are the ones that are spending all

190
00:12:43,003 --> 00:12:47,093
of this money to fund this CapEx build out. And it's important to note that

191
00:12:47,093 --> 00:12:53,058
if it wasn't for these seven companies, the entire Fortune 500 index would be down this year.

192
00:12:53,250 --> 00:12:56,913
A lot of companies on average performed badly. It was only these seven companies

193
00:12:56,913 --> 00:13:02,596
that were keeping the index profits and chart go up mentality alive in the first place.

194
00:13:02,822 --> 00:13:07,663
A lot of the money is being borrowed, around 50%. Again, of $3 trillion that

195
00:13:07,663 --> 00:13:11,293
are being committed over the next couple of years, half of that is being borrowed

196
00:13:11,293 --> 00:13:12,720
at pretty high interest rates.

197
00:13:13,121 --> 00:13:16,153
The question is, can they be able to pay all of that back? Well,

198
00:13:16,153 --> 00:13:18,850
that depends on AI making money in the first place.

199
00:13:19,251 --> 00:13:22,583
And that leads me to number three, which is, if you believe what Chamath is

200
00:13:22,583 --> 00:13:28,833
saying, around 75% of every dollar spent in AI is through VC subsidized dollars.

201
00:13:28,833 --> 00:13:33,233
So it's this whole circular economic loop. And the fourth final one is companies

202
00:13:33,233 --> 00:13:34,963
are now starting to build up walls.

203
00:13:34,963 --> 00:13:39,083
So we're doing the opposite of what we hoped or intended, which is AI just goes

204
00:13:39,083 --> 00:13:42,763
out and does everything for you to AI can go out, but then they're just going

205
00:13:42,763 --> 00:13:46,579
to get blocked by a bunch of walls because these companies don't want to give access to them.

206
00:13:46,875 --> 00:13:50,623
So that's the bear case. But I want to focus on the bull case because this is

207
00:13:50,623 --> 00:13:54,343
ultimately how I view it and how I feel things are going to play out over the

208
00:13:54,343 --> 00:13:56,643
next six months to a couple of years.

209
00:13:56,643 --> 00:14:00,013
Firstly, if you want to compare this to the dotcom era bubble,

210
00:14:00,013 --> 00:14:01,945
there's one very obvious distinction.

211
00:14:02,229 --> 00:14:05,486
In the dotcom era bubble there was a lot of supply being built up.

212
00:14:06,207 --> 00:14:12,124
Zero demand, like literally not enough demand to satisfy even 20% of what was

213
00:14:12,124 --> 00:14:14,659
being built. It was a massive overbuild.

214
00:14:15,042 --> 00:14:19,284
We are seeing the opposite here. Back then, it was promises.

215
00:14:19,284 --> 00:14:21,480
It was like, yeah, we're going to pay you this amount of money.

216
00:14:21,817 --> 00:14:26,624
Over here, we're seeing quarterly earnings result of money, cash in hand,

217
00:14:26,624 --> 00:14:30,804
being paid to these companies that are investing a lot of money into this capex.

218
00:14:30,804 --> 00:14:35,344
So they have demand of people actually buying these products using the models.

219
00:14:35,344 --> 00:14:39,243
If you look at Anthropik's recent S1, you'll see that they are making not only

220
00:14:39,597 --> 00:14:43,024
a lot of revenue, which is the fastest growing revenue of any company that we've

221
00:14:43,024 --> 00:14:47,364
ever seen in the past, but the revenue is also lean, meaning the amount that

222
00:14:47,364 --> 00:14:52,084
they invest to get that revenue is now overwhelmingly positive.

223
00:14:52,084 --> 00:14:56,724
Number two, the people or companies that are spending in this case versus the

224
00:14:56,724 --> 00:15:00,544
dot-com era bubble are companies that can actually afford it.

225
00:15:00,544 --> 00:15:03,680
I'm talking about Amazon, Microsoft, Meta, Google, and the likes,

226
00:15:04,005 --> 00:15:07,684
which have such a huge amount of money on their cash balance sheet.

227
00:15:07,684 --> 00:15:12,244
And yes, they may now be going into cash flow negative. But the idea is they

228
00:15:12,244 --> 00:15:16,509
have other business lines that are generating billions and billions of dollars

229
00:15:16,753 --> 00:15:19,870
every single month, every single quarter, every single year.

230
00:15:20,085 --> 00:15:23,694
So they feasibly have a way to be able to fund this. If you compare that to

231
00:15:23,694 --> 00:15:26,888
the dotcom era bubble, that wasn't the case. Okay, but then you might add,

232
00:15:27,057 --> 00:15:32,786
how is Anthropic and OpenAI, their startups, that's where a lot of this purchase demand comes from.

233
00:15:33,105 --> 00:15:36,994
That is right. And Anthropic has committed to around half a trillion to $750

234
00:15:36,994 --> 00:15:41,016
billion worth of spending over the next five years. A heck ton of money.

235
00:15:41,487 --> 00:15:45,964
I looked at the S1 earlier this week and I saw, well, the main thing being that

236
00:15:45,964 --> 00:15:50,130
they have around $132 billion on their cash balance sheet. But most importantly,

237
00:15:50,362 --> 00:15:53,578
they only spent $7 to $9 billion on compute.

238
00:15:53,747 --> 00:15:56,674
The rumor had it was that they were overspending, that they would be in the

239
00:15:56,674 --> 00:16:00,284
loss. But actually, if you remove that and some of the share equity positions,

240
00:16:00,284 --> 00:16:03,114
which was recorded as a loss on their balance sheet, they would actually be

241
00:16:03,114 --> 00:16:05,705
turning a profit. Now, it doesn't matter if you agree or disagree with that.

242
00:16:05,827 --> 00:16:08,474
That's just the direction that they're heading. They're eventually going to

243
00:16:08,474 --> 00:16:10,214
be turning quite a big profit.

244
00:16:10,214 --> 00:16:14,024
So the idea here is if OpenAI is also working at a similar rate,

245
00:16:14,024 --> 00:16:17,484
which my inclination leads me to believe, then we end up with these companies

246
00:16:17,484 --> 00:16:20,924
being able to pay for a lot of the GPU compute that they're paying the likes

247
00:16:20,924 --> 00:16:22,154
of Amazon, Microsoft, etc.

248
00:16:22,514 --> 00:16:25,804
To pay for. And remember, Amazon, Google, and Microsoft aren't just selling

249
00:16:25,804 --> 00:16:30,324
to these two companies. They have a heck ton of enterprise contracts of which

250
00:16:30,527 --> 00:16:34,004
around 50% of Anthropics revenue actually feeds through.

251
00:16:34,184 --> 00:16:39,074
So Google Cloud, Microsoft Azure, AWS all have enterprise customers that are

252
00:16:39,074 --> 00:16:43,074
paying those companies in order to get access to Claude and ChatGPT.

253
00:16:43,434 --> 00:16:47,024
That's not direct to open air. So the idea is these companies are generating

254
00:16:47,169 --> 00:16:50,034
quite a heck ton of money. And so when I look at these enterprise companies,

255
00:16:50,034 --> 00:16:54,199
I'm like, okay, well, a lot of these enterprise companies are in the Fortune 100.

256
00:16:54,413 --> 00:16:57,864
They're not VC backed. So this is coming from money that they're making,

257
00:16:57,864 --> 00:17:03,562
and they're not going to be spending this money if it doesn't end up with them making more money.

258
00:17:03,812 --> 00:17:08,792
And when I look at the net revenue retention, which is a NRR stat,

259
00:17:09,019 --> 00:17:10,934
basically tells you how much money.

260
00:17:11,543 --> 00:17:14,932
Someone spends at the start of the year using a product versus how much they're

261
00:17:15,292 --> 00:17:16,895
estimated to spend at the end of the year.

262
00:17:17,110 --> 00:17:20,187
For Anthropic Claude, at least, it's 900%.

263
00:17:20,697 --> 00:17:23,617
So you end up spending nine times more because the product is that good.

264
00:17:23,728 --> 00:17:26,332
I don't think that's made up. And the final point I'll make,

265
00:17:26,332 --> 00:17:30,332
which I think is arguably the most important and biggest point to make,

266
00:17:30,332 --> 00:17:35,418
the internet error still ended up panning out pretty damn well.

267
00:17:35,609 --> 00:17:39,692
Yes, we saw a massive crash in correction, but that also led to the birth of

268
00:17:39,692 --> 00:17:43,562
some of the biggest, most important, relevant AI companies that back the entire

269
00:17:43,562 --> 00:17:47,991
economy. Forget about the internet economy. The internet economy is the economy today, right?

270
00:17:48,258 --> 00:17:53,082
And so whilst we may see, yes, a huge inflation in this AI bubble,

271
00:17:53,082 --> 00:17:56,205
and I definitely think we need to have cause for concern, call out the risks,

272
00:17:56,420 --> 00:17:58,800
monitor the risk, and make sure that we don't overstep,

273
00:17:59,282 --> 00:18:04,036
we should also expect some kind of deflation or correction of this entire bubble

274
00:18:04,431 --> 00:18:07,728
and i would actually argue that that is what we're seeing right now

275
00:18:08,042 --> 00:18:13,432
and a main reason why we probably won't result in an overbuild or dot-com error

276
00:18:13,432 --> 00:18:17,184
type bubble is because it's really hard to overbuild right now

277
00:18:17,446 --> 00:18:22,152
in the dot-com error bubble you could build out the telecom lines and like over

278
00:18:22,152 --> 00:18:25,091
invest in that and that's what ultimately led to the collapse here.

279
00:18:25,631 --> 00:18:29,871
You're restricted by a lot of electricity production and the ability to stand

280
00:18:29,871 --> 00:18:32,441
up power to fund these GPUs and data centers.

281
00:18:32,441 --> 00:18:35,901
So I think inherently, a lot of these big hyperscalers and AI companies are

282
00:18:35,901 --> 00:18:38,661
seeing this and thinking, yeah, as much as I want to double down,

283
00:18:38,661 --> 00:18:44,019
we aren't going to be able to spend money as fast as we would like because we just can't.

284
00:18:44,315 --> 00:18:49,481
And that works as a indirect risk manager, I guess, in a way,

285
00:18:49,667 --> 00:18:51,821
where it prevents us from inflating the bubble too much.

286
00:18:51,821 --> 00:18:55,091
But I do believe that the market is in two buckets right now.

287
00:18:55,091 --> 00:18:58,061
Some people that are thinking long-term about this and are betting that this

288
00:18:58,061 --> 00:19:01,961
AI thing is gonna pan out pretty impressively over time. It's gonna consume

289
00:19:01,961 --> 00:19:05,393
anything and everything and generate a huge amount of wealth for the world.

290
00:19:05,561 --> 00:19:09,681
I sit in that bucket. And if you're a long-term investor, I'm just buying some

291
00:19:09,681 --> 00:19:12,901
of these dips and just waiting to see how the market kind of pans out and making

292
00:19:12,901 --> 00:19:15,348
sure that I don't recklessly throw money into this market.

293
00:19:15,708 --> 00:19:18,564
But if you're a short-term investor and you're kind of like looking at these.

294
00:19:18,982 --> 00:19:22,241
Inflated valuations, you know, Anthropic potentially coming out at a $2 trillion

295
00:19:22,241 --> 00:19:26,070
valuation, SpaceX had a $3 trillion valuation or whatever that might be,

296
00:19:26,471 --> 00:19:29,161
you're kind of looking at this and thinking, well, damn, like this might be

297
00:19:29,161 --> 00:19:31,131
overvalued and you might be right over the next six months.

298
00:19:31,131 --> 00:19:35,531
So take both of these things into account and hopefully this was a good enough

299
00:19:35,531 --> 00:19:40,808
deduction, at least on my internal monologue as to how things are actually panning out.

300
00:19:40,982 --> 00:19:44,701
Like it's easy to look at the markets and panic, but it's also important to

301
00:19:44,701 --> 00:19:48,621
understand the nuance between historical context and what's actually going on today.

302
00:19:49,033 --> 00:19:52,271
So yeah, I guess, you know, let me know your thoughts. I'm sitting here on my

303
00:19:52,271 --> 00:19:55,959
own kind of like working through these things trying to figure out what makes sense and what doesn't

304
00:19:56,226 --> 00:19:58,731
and maybe there are folks out there with different perspectives that might make

305
00:19:58,731 --> 00:20:02,001
sense and change my own perspective so if you're on Spotify or Apple give us

306
00:20:02,001 --> 00:20:06,761
a rating leave us a comment helps us out massively and also DM or comment.

307
00:20:07,621 --> 00:20:10,621
Any kinds of thoughts or feedback that you have any kinds of topics that you

308
00:20:10,621 --> 00:20:14,471
want me to cover I'm all ears and I see all of these things and I really enjoy

309
00:20:14,471 --> 00:20:17,751
making these episodes so yeah I guess that's it thank you so much for listening

310
00:20:17,751 --> 00:20:19,435
to today's episode brought to you by Qualcomm.