1
00:00:00,080 --> 00:00:03,840
It seems as though every day we 
wake up to the market going down

2
00:00:03,840 --> 00:00:07,560
and a new viral article that's 
being passed around speaking 

3
00:00:07,560 --> 00:00:11,120
doomsday of SAS companies and 
the world as we know it. 

4
00:00:11,240 --> 00:00:14,120
Now, we've seen this happen 
before with articles like Big 

5
00:00:14,120 --> 00:00:17,240
things are happening, right? 
There's ones about big changes 

6
00:00:17,240 --> 00:00:19,120
and disruptions in technology 
companies. 

7
00:00:19,400 --> 00:00:22,480
We've also seen ones that are 
essentially the doom of SAS 

8
00:00:22,480 --> 00:00:24,720
companies. 
That's caused a lot of companies

9
00:00:24,720 --> 00:00:27,440
that we know and are invested in
to go down dramatically every 

10
00:00:27,440 --> 00:00:29,720
single day. 
And there's more and more of 

11
00:00:29,720 --> 00:00:31,400
these articles again, every 
single day. 

12
00:00:31,400 --> 00:00:33,720
It seems like there's just two 
or three more of them that are 

13
00:00:33,720 --> 00:00:38,040
going viral around X, around 
social circles, around 

14
00:00:38,040 --> 00:00:40,160
investment circles. 
But the one that was shared over

15
00:00:40,160 --> 00:00:42,440
the weekend is perhaps the final
boss. 

16
00:00:42,800 --> 00:00:44,760
It's like a video game. 
You're getting to the end, 

17
00:00:44,920 --> 00:00:47,960
You've gotten through all the 
smaller bosses, all the bigger 

18
00:00:47,960 --> 00:00:51,680
challenges and hurdles along the
way, and now we're facing the 

19
00:00:51,680 --> 00:00:54,840
final boss. 
This is Citrone Research. 

20
00:00:54,880 --> 00:00:57,000
Why is Citrone Research the 
final boss? 

21
00:00:57,000 --> 00:00:59,920
Well, because of out of all the 
hundreds and thousands of 

22
00:00:59,920 --> 00:01:02,440
different sub stacks out there, 
out of all the ones that you can

23
00:01:02,440 --> 00:01:05,720
read, all the different 
financial authors giving takes 

24
00:01:05,720 --> 00:01:08,920
and thesises, there's one that's
listened to and paid attention 

25
00:01:08,920 --> 00:01:14,280
to more than any others. 
That is Citrone, the top selling

26
00:01:14,280 --> 00:01:15,880
financial sub stack in the 
world. 

27
00:01:16,040 --> 00:01:19,080
It's based on macro. 
It's based on large events, big 

28
00:01:19,080 --> 00:01:21,160
takes. 
Investors listen to it. 

29
00:01:21,440 --> 00:01:22,840
Hedge fund managers listen to 
it. 

30
00:01:23,200 --> 00:01:26,160
Thousands of investors subscribe
to this and pay good money to 

31
00:01:26,160 --> 00:01:28,240
hear their takes. 
It's so influential that their 

32
00:01:28,240 --> 00:01:30,120
takes routinely land on 
Bloomberg. 

33
00:01:30,120 --> 00:01:32,760
They land on the Wall Street 
Journal, And what they wrote 

34
00:01:32,760 --> 00:01:34,480
over the weekend is no 
exception. 

35
00:01:34,680 --> 00:01:37,120
It is now on the front page of 
the Wall Street Journal. 

36
00:01:37,200 --> 00:01:41,080
This viral doomsday report lays 
bare Wall Street's deep anxiety 

37
00:01:41,080 --> 00:01:44,760
about the AI future. 
Citronee Research thought 

38
00:01:44,760 --> 00:01:47,960
experiment rattles investors 
already wary of tech 

39
00:01:47,960 --> 00:01:49,880
disruptions. 
That's right. 

40
00:01:50,560 --> 00:01:54,480
No, this is not a article that's
on some obscure forms on the 

41
00:01:54,480 --> 00:01:56,320
Internet. 
This is mainstream. 

42
00:01:56,400 --> 00:01:59,080
This is all over CNBC, 
Bloomberg, the Wall Street 

43
00:01:59,080 --> 00:02:02,440
Journal, the Financial Times. 
Citrone's article went hyper 

44
00:02:02,440 --> 00:02:05,400
viral and it caused many of the 
most powerful biggest companies 

45
00:02:05,400 --> 00:02:07,560
in the world their stocks to get
crushed. 

46
00:02:08,000 --> 00:02:10,160
DoorDash is down like 7% on the 
day. 

47
00:02:10,160 --> 00:02:15,080
American Express is down big 3%,
plus we have Blackstone going 

48
00:02:15,080 --> 00:02:17,600
down. 
We have Visa and MasterCard, two

49
00:02:17,600 --> 00:02:19,480
companies that you think are a 
little bit more resilient. 

50
00:02:19,640 --> 00:02:21,040
Yeah, those ones went down as 
well. 

51
00:02:21,600 --> 00:02:24,240
This article is impacting 
investors thoughts. 

52
00:02:24,360 --> 00:02:27,600
It's impacting their decisions. 
Investors are human and they're 

53
00:02:27,600 --> 00:02:30,680
selling as a result of this. 
Now, I didn't think I was going 

54
00:02:30,680 --> 00:02:35,440
to spend all of 2026 reacting to
viral articles online, but here 

55
00:02:35,440 --> 00:02:38,640
we are. 
This is what we're doing, and 

56
00:02:38,960 --> 00:02:41,040
this is what I'm going to be 
doing in this episode. 

57
00:02:41,640 --> 00:02:45,640
We'll be going over the thesis 
of Citronee, why it is so 

58
00:02:45,840 --> 00:02:49,080
engrossing, why investors have 
so much anxiety about what 

59
00:02:49,080 --> 00:02:51,120
they're talking about. 
We'll be reviewing their 

60
00:02:51,120 --> 00:02:54,880
arguments, and importantly, I'll
be tearing some of them apart 

61
00:02:54,880 --> 00:02:58,040
because frankly, some of them 
are ridiculous. 

62
00:02:58,320 --> 00:03:01,440
They deserve to be torn apart. 
Now, it's no blame of Citronee. 

63
00:03:01,440 --> 00:03:04,040
They say that this is all 
hypothetical, but I think even a

64
00:03:04,040 --> 00:03:07,960
hypothetical argument can have 
severe flaws, and I'll be going 

65
00:03:07,960 --> 00:03:09,760
over many of them in this 
article. 

66
00:03:09,760 --> 00:03:13,080
Today I'll be going through in 
detail some of the core thesises

67
00:03:13,080 --> 00:03:16,000
that are causing these stocks to
sell off and addressing what I 

68
00:03:16,000 --> 00:03:19,200
believe are some of the biggest 
logical leaps and mistakes in 

69
00:03:19,200 --> 00:03:20,920
this article. 
This will be a fun episode. 

70
00:03:21,080 --> 00:03:22,840
We do have some other news we'll
be touching on today. 

71
00:03:22,840 --> 00:03:26,560
For example, Warner Brothers 
Discovery received a revised 

72
00:03:26,560 --> 00:03:29,080
offer from Paramount. 
The higher bid, we don't know 

73
00:03:29,080 --> 00:03:31,960
what it is, but we're now in 
this limbo where we're waiting 

74
00:03:31,960 --> 00:03:34,760
for Netflix's next move. 
I'll be giving some thoughts on 

75
00:03:34,760 --> 00:03:37,120
that. 
We have Meta agreeing to buy 

76
00:03:37,120 --> 00:03:41,600
$100 billion more worth of chips
from AMD, and we have the fail 

77
00:03:41,600 --> 00:03:45,160
of the week, which in this case 
is a DJI robot vacuum that had 

78
00:03:45,160 --> 00:03:48,960
security so poor that this man 
remotely accessed thousands of 

79
00:03:48,960 --> 00:03:51,080
them. 
This guy accidentally hacked 

80
00:03:51,120 --> 00:03:53,720
like 7000 robots. 
We'll be going over that as 

81
00:03:53,720 --> 00:03:54,640
well. 
Now, if you're not paying 

82
00:03:54,640 --> 00:03:57,080
attention to every detail, 
you're not listening to every 

83
00:03:57,080 --> 00:03:59,680
little thing going on. 
You may have missed what caused 

84
00:03:59,680 --> 00:04:02,680
this most recent panic, what 
caused many of the biggest 

85
00:04:02,680 --> 00:04:07,080
companies in the world to sell 
off 3 to 8% in the past day, and

86
00:04:07,080 --> 00:04:09,880
that was this viral article. 
This was the cause. 

87
00:04:10,120 --> 00:04:12,680
Many people believe that 
Substack and these articles 

88
00:04:12,680 --> 00:04:14,240
can't 'cause this type of 
effect. 

89
00:04:14,560 --> 00:04:17,519
They can, especially when they 
have the distribution of a 

90
00:04:17,519 --> 00:04:20,640
Citronee Research article. 
They are one of the biggest, if 

91
00:04:20,640 --> 00:04:23,920
not the biggest publisher in the
financial world on an 

92
00:04:23,920 --> 00:04:26,360
independent basis. 
The article that they published 

93
00:04:26,360 --> 00:04:32,800
is called The 2028, 2028 Global 
Intelligence Crisis, A Thought 

94
00:04:32,800 --> 00:04:35,920
Exercise in Financial History 
from the future. 

95
00:04:36,480 --> 00:04:40,880
Now they preface this and they 
say what follows is a scenario. 

96
00:04:41,040 --> 00:04:43,160
It's not a prediction. 
So they give a bit of a 

97
00:04:43,160 --> 00:04:46,040
disclaimer and say they're just 
modeling out a scenario. 

98
00:04:46,640 --> 00:04:49,760
Now, like most disclaimers, this
has been ignored because many 

99
00:04:49,760 --> 00:04:52,080
people were fearful as a result 
of this. 

100
00:04:52,560 --> 00:04:55,760
And I'll note that this scenario
they're modelling out also does 

101
00:04:55,760 --> 00:04:59,360
benefit them personally. 
See, Citrone is not just someone

102
00:04:59,360 --> 00:05:01,720
that publishes takes and 
thoughts on the market, but 

103
00:05:01,720 --> 00:05:04,560
they're also investors and they 
are bearish on many of the 

104
00:05:04,560 --> 00:05:06,520
companies that they're sharing 
these scenarios on. 

105
00:05:06,600 --> 00:05:08,360
They have short positions on 
them. 

106
00:05:08,640 --> 00:05:11,320
We are constantly sort of, you 
know, turning our book. 

107
00:05:11,320 --> 00:05:13,320
And then we certainly had shorts
in some of these businesses. 

108
00:05:13,320 --> 00:05:16,200
We generally have a set of 
shorts out against businesses 

109
00:05:16,200 --> 00:05:17,800
that we think are going to be 
disrupted by AI. 

110
00:05:18,080 --> 00:05:21,560
He mentions right away that we 
are short many of the businesses

111
00:05:21,560 --> 00:05:24,560
that we mention in this article.
So in a way, this is an article 

112
00:05:24,560 --> 00:05:27,240
representing their view, which 
there's nothing wrong with that.

113
00:05:27,240 --> 00:05:29,640
I have bullish takes on 
companies that I'm invested in 

114
00:05:29,640 --> 00:05:31,920
as well. 
The difference is I'm not able 

115
00:05:31,920 --> 00:05:35,200
to move the market in the same 
way that Citrone is, so just 

116
00:05:35,200 --> 00:05:37,560
keep that in mind as you're 
reading this that they are short

117
00:05:37,560 --> 00:05:39,040
sellers on many of these 
companies. 

118
00:05:39,040 --> 00:05:41,040
We won't go over the whole 
thing, but I want to summarize 

119
00:05:41,040 --> 00:05:43,240
what I believe are some of the 
biggest points of this article. 

120
00:05:43,520 --> 00:05:47,280
In a couple of years we'll move 
past vibe coding to agentic 

121
00:05:47,280 --> 00:05:50,520
technology, and agentic 
technology will cause what they 

122
00:05:50,720 --> 00:05:53,240
they say is when friction goes 
to zero. 

123
00:05:53,240 --> 00:05:56,040
And again, this is a future 
hypothetical. 

124
00:05:56,360 --> 00:06:00,120
They say by early 2027, LLM 
usage had become default. 

125
00:06:00,280 --> 00:06:03,200
People are using AI agents who 
didn't even know what an AI 

126
00:06:03,200 --> 00:06:06,680
agent was in the same way people
who had never learned of cloud 

127
00:06:06,680 --> 00:06:08,960
computing was using streaming 
services. 

128
00:06:09,200 --> 00:06:11,640
They thought of it in the same 
way that you think of auto 

129
00:06:11,640 --> 00:06:14,880
complete or spell check. 
Just a thing that they just did 

130
00:06:14,880 --> 00:06:16,720
now. 
The next chain was already 

131
00:06:16,720 --> 00:06:19,880
breaking intermediation. 
Over the past 50 years, the US 

132
00:06:19,880 --> 00:06:23,280
economy built a giant rent 
extraction layer on top of human

133
00:06:23,400 --> 00:06:26,560
limitations. 
Things take time, patience runs 

134
00:06:26,560 --> 00:06:30,400
out, brand familiarity 
substitutes for diligence, and 

135
00:06:30,400 --> 00:06:33,720
most people are willing to 
accept a bad price to avoid more

136
00:06:33,720 --> 00:06:35,520
clicks. 
Trillions of dollars of 

137
00:06:35,520 --> 00:06:38,400
enterprise value depends on 
those constraints persisting. 

138
00:06:38,480 --> 00:06:41,920
It started out simple enough. 
Agents removed friction. 

139
00:06:42,280 --> 00:06:46,240
So they explore the consequences
of agents removing friction. 

140
00:06:46,920 --> 00:06:49,680
Subscriptions and memberships 
that passively renew despite 

141
00:06:49,680 --> 00:06:52,600
months of disuse. 
Introductory pricing that 

142
00:06:52,600 --> 00:06:54,720
sneakily doubled after the trial
period. 

143
00:06:55,080 --> 00:06:58,200
Each one was rebranded as a 
hostage situation that agents 

144
00:06:58,200 --> 00:07:00,640
could negotiate. 
The average customer lifetime 

145
00:07:00,640 --> 00:07:03,360
value, the metric that 
enterprise subscription economy 

146
00:07:03,360 --> 00:07:05,120
was built on, distinctly 
declined. 

147
00:07:05,440 --> 00:07:08,640
Consumer agents begin to change 
how nearly all consumer 

148
00:07:08,640 --> 00:07:11,200
transactions work. 
Humans don't really have the 

149
00:07:11,200 --> 00:07:14,680
time to price match across 5 
competing platforms before 

150
00:07:14,680 --> 00:07:17,320
buying a box of protein bars, 
but machines do. 

151
00:07:17,760 --> 00:07:20,320
So part of their thesis is when 
you're buying something online, 

152
00:07:20,720 --> 00:07:22,680
you're just going to buy like 
the first thing that you see 

153
00:07:22,680 --> 00:07:24,560
because you don't have time to 
go look everywhere to save a 

154
00:07:24,560 --> 00:07:26,320
dollar. 
But if you can do that on an 

155
00:07:26,320 --> 00:07:29,040
agent, if you can say, hey, I 
want to buy this and it can go 

156
00:07:29,040 --> 00:07:31,680
and price match 50 different 
websites, you're going to let 

157
00:07:31,680 --> 00:07:35,080
the agent do that. 
The agent does have the time to 

158
00:07:35,080 --> 00:07:37,160
make those, to make those 
comparisons. 

159
00:07:37,680 --> 00:07:41,440
Travel booking platforms were an
early casualty because they were

160
00:07:41,440 --> 00:07:44,520
the simplest. 
By Q 4/20/26, our agents could 

161
00:07:44,520 --> 00:07:48,040
assemble a complete itinerary, 
flights, hotels, transport, 

162
00:07:48,240 --> 00:07:51,280
loyal loyalty optimization, 
budget constraints, refunds 

163
00:07:51,600 --> 00:07:53,400
faster and cheaper than any 
platform. 

164
00:07:53,880 --> 00:07:57,600
Insurance renewals where the 
entire renewal model depended on

165
00:07:57,760 --> 00:08:00,120
policyholder inertia were 
reformed. 

166
00:08:00,560 --> 00:08:04,360
Agents that reshop coverage 
annually dismantled the 15 to 

167
00:08:04,360 --> 00:08:07,480
20% premiums that insurers 
earned from passive renewals. 

168
00:08:07,960 --> 00:08:11,520
Financial advice, tax prep, 
routine legal work, any category

169
00:08:11,520 --> 00:08:14,720
where the service providers 
value proposition was ultimately

170
00:08:14,720 --> 00:08:17,960
quote I will navigate. 
Complexity that you find tedious

171
00:08:17,960 --> 00:08:21,720
was disrupted as agents found 
nothing tedious. 

172
00:08:22,440 --> 00:08:25,880
Even places we thought insulated
by the human value of 

173
00:08:25,880 --> 00:08:29,520
relationships proved fragile. 
Real Estate, where buyers had 

174
00:08:29,520 --> 00:08:33,159
tolerated 5 to 6% commissions 
for decades because of 

175
00:08:33,159 --> 00:08:36,799
information asymmetry between 
agents and consumers, crumbled 

176
00:08:36,799 --> 00:08:40,280
once the AI agents equipped with
MLS access and decades of 

177
00:08:40,280 --> 00:08:43,720
transaction data could replicate
the knowledge base instantly. 

178
00:08:44,120 --> 00:08:47,160
So now they're saying that the 
real estate market is going to 

179
00:08:47,160 --> 00:08:50,720
be disrupted because the real 
estates, that intermediary layer

180
00:08:51,000 --> 00:08:54,160
will be taken out by agents. 
This is all just fiction. 

181
00:08:54,240 --> 00:08:56,240
They're, they're making this up,
they're exploring this 

182
00:08:56,240 --> 00:08:58,640
hypothetical. 
So this hasn't happened. 

183
00:08:59,120 --> 00:09:02,760
They're saying that real estate 
agents are, are getting 

184
00:09:03,160 --> 00:09:07,720
disintermediated by agents and 
that's just all in the future. 

185
00:09:07,760 --> 00:09:09,880
That's going to happen in the 
future is what they're 

186
00:09:09,880 --> 00:09:12,400
predicting here. 
We had overestimated the value 

187
00:09:12,400 --> 00:09:15,800
of, quote, human relationships. 
Turns out that a lot of what 

188
00:09:15,800 --> 00:09:19,360
people call relationships was 
simply friction with a friendly 

189
00:09:19,360 --> 00:09:21,520
face. 
That was just the start of the 

190
00:09:21,520 --> 00:09:24,000
disruption for the 
intermediation layer. 

191
00:09:24,200 --> 00:09:28,080
Successful companies had spent 
billions to effectively exploit 

192
00:09:28,240 --> 00:09:32,040
quirks of consumer behavior and 
human psychology that didn't 

193
00:09:32,040 --> 00:09:35,520
matter anymore. 
Machines optimized for price and

194
00:09:35,520 --> 00:09:38,720
fit that do not care about your 
favorite app or your website 

195
00:09:39,080 --> 00:09:42,000
that you've habitually been 
opening for the last four years 

196
00:09:42,360 --> 00:09:45,000
nor fill the pole of a well 
designed checkpoint out 

197
00:09:45,000 --> 00:09:47,080
experience. 
They don't get tired or accept 

198
00:09:47,080 --> 00:09:49,800
the easiest option by default to
quote. 

199
00:09:50,080 --> 00:09:53,720
I always just order from here 
that destroyed a particular 

200
00:09:53,720 --> 00:09:57,120
kinds of motes habitual 
intermediation. 

201
00:09:57,840 --> 00:10:01,920
OK, so that sets the premise of 
their main core thesis here. 

202
00:10:02,080 --> 00:10:04,800
Their core thesis is that agents
remove friction. 

203
00:10:04,880 --> 00:10:08,400
They bring friction down to 0. 
So any type of company or 

204
00:10:08,400 --> 00:10:11,720
product that specializes in 
removing friction at a price, 

205
00:10:12,240 --> 00:10:15,240
that company will be essentially
useless. 

206
00:10:15,240 --> 00:10:17,640
It won't need to exist anymore 
because agents will do that 

207
00:10:17,640 --> 00:10:19,920
already and not charge you in 
the process. 

208
00:10:20,560 --> 00:10:23,560
And this is where the thesis 
goes from what I think is an 

209
00:10:23,560 --> 00:10:27,680
intriguing premise to offering 
some of the worst examples you 

210
00:10:27,680 --> 00:10:29,560
could offer, logically flawed 
ones. 

211
00:10:30,200 --> 00:10:32,360
For example, one of the ones 
that they point out that will be

212
00:10:32,360 --> 00:10:35,600
displaced or disintermediated is
DoorDash. 

213
00:10:35,840 --> 00:10:38,080
And this was a stock that was 
hit hard yesterday. 

214
00:10:38,280 --> 00:10:42,920
DoorDash was the poster child of
a Moat being disrupted by agents

215
00:10:42,920 --> 00:10:47,000
and their removal of friction. 
Now DoorDash stock since this 

216
00:10:47,000 --> 00:10:50,080
released is down about 6%. 
So let's go ahead and hear their

217
00:10:50,080 --> 00:10:53,120
argument of why DoorDash is 
going to be doomed when it comes

218
00:10:53,120 --> 00:10:55,760
to agents coding. 
Agents had collapsed the barrier

219
00:10:55,760 --> 00:10:57,760
to entry for launching a 
delivery app. 

220
00:10:58,160 --> 00:11:01,720
So now you can, you can code a 
delivery app very quickly. 

221
00:11:01,880 --> 00:11:04,520
They say a competent developer 
could deploy A functional 

222
00:11:04,520 --> 00:11:09,320
competitor in weeks and dozens 
did enticing drivers away from 

223
00:11:09,320 --> 00:11:12,680
DoorDash and Uber Eats by 
passing 90 to 95% of the 

224
00:11:12,680 --> 00:11:15,680
delivery fee through to the 
driver. 

225
00:11:16,000 --> 00:11:18,600
So again, this is hypothetical. 
This hasn't happened. 

226
00:11:18,920 --> 00:11:22,560
But they're saying now that vibe
coding allows you to easily code

227
00:11:22,560 --> 00:11:24,680
things like delivery 
applications, which is true. 

228
00:11:24,760 --> 00:11:27,600
Make it easier to code things. 
You can basically make a 

229
00:11:27,600 --> 00:11:33,000
competent functioning, competent
competitor in weeks to DoorDash 

230
00:11:33,000 --> 00:11:35,880
and Uber Eats. 
Then those competent competitors

231
00:11:36,440 --> 00:11:38,800
cannot have the greed that 
DoorDash has. 

232
00:11:38,800 --> 00:11:42,120
Instead they can just pass the 
savings along to the driver. 

233
00:11:42,680 --> 00:11:46,480
Multi app dashboards let gig 
workers track incoming jobs from

234
00:11:46,480 --> 00:11:50,240
20 or 30 different platforms at 
once, eliminating the lock in 

235
00:11:50,240 --> 00:11:51,960
that the incumbents depended 
upon. 

236
00:11:52,400 --> 00:11:55,760
The market fragmented overnight 
and margins compressed to nearly

237
00:11:55,760 --> 00:11:59,160
nothing. 
Agents accelerated both sides of

238
00:11:59,160 --> 00:12:01,880
the destruction. 
They enabled the competitors and

239
00:12:01,880 --> 00:12:04,600
they used them. 
The DoorDash Mote was literally 

240
00:12:04,760 --> 00:12:06,800
quote you're hungry, you're 
lazy. 

241
00:12:06,960 --> 00:12:10,080
This is the app on your home 
screen, and Agents doesn't. 

242
00:12:10,320 --> 00:12:13,280
Agents don't have a home screen.
It checks DoorDash. 

243
00:12:13,280 --> 00:12:17,160
It checks Uber Eats, the 
restaurant's own website, and 20

244
00:12:17,240 --> 00:12:20,760
new vibe coded alternatives so 
that it can pick the lowest fee 

245
00:12:20,960 --> 00:12:24,600
and the fastest delivery time. 
Habitual app loyalty, the entire

246
00:12:24,600 --> 00:12:27,320
basis of the business model, 
simply didn't exist for the 

247
00:12:27,320 --> 00:12:30,800
machine. 
This was oddly poetic as perhaps

248
00:12:30,800 --> 00:12:34,480
the only example in this entire 
saga of agents doing a favor for

249
00:12:34,480 --> 00:12:37,720
the soon to be replaced 
displaced white collar workers 

250
00:12:38,240 --> 00:12:40,200
when they ended up as delivery 
drivers. 

251
00:12:40,400 --> 00:12:41,840
So there's the argument with 
DoorDash. 

252
00:12:41,840 --> 00:12:46,280
It's basically that DoorDash and
Uber Eats contain high margins. 

253
00:12:46,280 --> 00:12:48,000
They charge a lot when you order
from them. 

254
00:12:48,400 --> 00:12:50,720
So an agent could come along and
it can shop. 

255
00:12:50,920 --> 00:12:53,680
It can shop across all the 
different food delivery 

256
00:12:53,680 --> 00:12:55,720
networks. 
And since you can vibe code a 

257
00:12:55,720 --> 00:12:57,920
new app in just a couple of 
weeks, there's going to be a lot

258
00:12:57,920 --> 00:13:01,080
of them, 20 to 30 of them since 
the agent, which is like a 

259
00:13:01,080 --> 00:13:04,480
ChatGPT or Gemini since it can 
shop for you. 

260
00:13:04,760 --> 00:13:08,240
Frictionless and it can go to 
all these different potential 

261
00:13:08,360 --> 00:13:11,560
delivery companies or just the 
the website of the restaurant 

262
00:13:11,560 --> 00:13:13,840
itself. 
It can find the lowest prices, 

263
00:13:13,840 --> 00:13:16,440
which will naturally compress 
the margins to zero. 

264
00:13:17,080 --> 00:13:19,040
And that seems good. 
At first glance, it seems a 

265
00:13:19,040 --> 00:13:21,840
little bit convincing until you 
dive in a little bit deeper and 

266
00:13:21,840 --> 00:13:23,800
highlight some of the big 
problems with this type of 

267
00:13:23,800 --> 00:13:25,760
argument. 
One of the problems with this 

268
00:13:25,760 --> 00:13:28,600
thought process is if you just 
put yourself in the customer's 

269
00:13:28,600 --> 00:13:30,560
position here. 
Let's say that you're you're 

270
00:13:30,560 --> 00:13:33,320
using Chachi BT and you're 
getting hungry. 

271
00:13:33,320 --> 00:13:34,560
So you notice that you're 
hungry. 

272
00:13:35,440 --> 00:13:37,800
In that case, you notice that 
you're hungry, but you don't 

273
00:13:37,800 --> 00:13:42,200
know what you're going to eat. 
So you could tell Chachi BT I'm 

274
00:13:42,200 --> 00:13:45,200
hungry and it can say, hey, 
you've liked these type of 

275
00:13:45,200 --> 00:13:47,120
foods, would you like to order 
them again? 

276
00:13:47,680 --> 00:13:50,280
And you go, Nah, what are some 
other foods out there? 

277
00:13:50,760 --> 00:13:53,520
And then it has to assemble and 
create a list of foods that you 

278
00:13:53,520 --> 00:13:55,720
can order. 
It has to bring in all the 

279
00:13:55,720 --> 00:13:59,320
images and then you have to say,
hey, what about you're typing 

280
00:13:59,320 --> 00:14:00,880
this in again, You're typing all
this in. 

281
00:14:01,080 --> 00:14:03,280
What about restaurants that are 
within 30 minutes? 

282
00:14:03,280 --> 00:14:05,920
Because I want to get it fast. 
Then it has to revise and give 

283
00:14:05,920 --> 00:14:08,040
you another answer. 
Restaurants within 30 minutes. 

284
00:14:08,640 --> 00:14:10,760
But then you want to see the 
reviews of all those restaurants

285
00:14:10,760 --> 00:14:12,720
to know which ones are are good 
and which ones aren't. 

286
00:14:12,720 --> 00:14:14,560
So you want 4 star reviews and 
up. 

287
00:14:14,880 --> 00:14:17,680
So you have to type that in. 
Now order them by by reviews. 

288
00:14:18,080 --> 00:14:19,760
OK, so then you're on the 
restaurant level. 

289
00:14:19,760 --> 00:14:22,240
You're looking at reviews of it.
And let's say that you finally 

290
00:14:22,240 --> 00:14:24,280
narrow down which restaurant you
want to order from. 

291
00:14:24,880 --> 00:14:27,240
OK, You say, hey, Chachi PT, I 
want to order from this 

292
00:14:27,240 --> 00:14:30,480
restaurant, but I don't know 
exactly what they have or what 

293
00:14:30,480 --> 00:14:31,960
the prices are. 
Can you tell me that? 

294
00:14:32,560 --> 00:14:36,200
And then it has to assemble a 
list of the the menu from that 

295
00:14:36,200 --> 00:14:39,160
restaurant of all the different 
items, what the feature is, what

296
00:14:39,160 --> 00:14:42,280
the daily special is. 
OK, so you, you have that 

297
00:14:42,280 --> 00:14:45,960
assembled by Chachi PT Now and 
again you're having Chachi PT 

298
00:14:45,960 --> 00:14:49,440
build this menu on the fly as 
you continually type in. 

299
00:14:50,000 --> 00:14:52,440
Now you're on like the 6th 
message now and you say, OK, I 

300
00:14:52,440 --> 00:14:54,960
want to get a #3 I want to get 
that meal. 

301
00:14:55,600 --> 00:14:57,880
But there's lots of 
customizations in every single 

302
00:14:57,880 --> 00:14:59,520
meal from every single 
restaurant. 

303
00:14:59,760 --> 00:15:02,720
For example, if you order from 
Five Guys, do you want the small

304
00:15:02,720 --> 00:15:04,760
burger or the normal one? 
Do you want the double Patty or 

305
00:15:04,760 --> 00:15:06,520
the one with cheese? 
Do you want the one with bacon? 

306
00:15:06,760 --> 00:15:08,520
Do you want it to have Mayo, 
ketchup? 

307
00:15:08,520 --> 00:15:09,680
Do you want to have onions on 
it? 

308
00:15:09,680 --> 00:15:10,920
Do you want to have Pickles on 
it? 

309
00:15:11,320 --> 00:15:14,680
There are hundreds of 
permutations and customizations 

310
00:15:14,680 --> 00:15:17,480
within a single order. 
Literally hundreds of them in 

311
00:15:17,480 --> 00:15:19,720
some cases. 
Chick-fil-A, just that 

312
00:15:19,720 --> 00:15:22,560
restaurant has hundreds of 
different customizations on a 

313
00:15:22,560 --> 00:15:25,480
single order. 
All the all the way to what type

314
00:15:25,480 --> 00:15:29,280
of sauce you want to get. 
Are you going to prompt Chachi 

315
00:15:29,280 --> 00:15:31,440
BT all the way through this 
process? 

316
00:15:31,920 --> 00:15:34,080
By the time you're done, your 
fingers, your thumbs will be 

317
00:15:34,080 --> 00:15:35,800
sore from the amount of typing 
you do. 

318
00:15:36,320 --> 00:15:39,080
You'll have to gone through so 
many different customization 

319
00:15:39,080 --> 00:15:42,800
layers and the idea that ChatGPT
will just know. 

320
00:15:42,920 --> 00:15:45,320
That's another thing that a lot 
of people argue here. 

321
00:15:45,560 --> 00:15:48,360
They'll say, well, you've talked
to ChatGPT so much, it'll know 

322
00:15:48,360 --> 00:15:50,560
you inside and out. 
It'll know you better than 

323
00:15:50,560 --> 00:15:54,560
yourself. 
I was a top 1% user of ChatGPT 

324
00:15:54,560 --> 00:15:59,560
and Gemini in 2025, top 1%. 
And I promise you, it has no 

325
00:15:59,560 --> 00:16:03,240
clue what I want to eat tonight.
It has no idea what what 

326
00:16:03,360 --> 00:16:06,440
restaurant, what menu, what 
items, because it changes all 

327
00:16:06,440 --> 00:16:08,880
the time. 
Food preferences change all the 

328
00:16:08,880 --> 00:16:10,600
time. 
And even though there may be 

329
00:16:10,600 --> 00:16:13,360
something that you order on a 
routine basis, maybe you get the

330
00:16:13,360 --> 00:16:14,920
same Starbucks drink every 
morning. 

331
00:16:15,240 --> 00:16:18,440
When it comes to food delivery, 
this is something that's hyper 

332
00:16:18,440 --> 00:16:20,400
specialized. 
Thousands of different 

333
00:16:20,400 --> 00:16:24,480
customizations and orders, and 
of course different restaurants 

334
00:16:24,480 --> 00:16:27,920
that are open different hours. 
Every single time that you would

335
00:16:27,920 --> 00:16:31,960
have to order through an agent, 
you'd have to rebuild a menu 

336
00:16:32,240 --> 00:16:35,520
prompted to death because you're
doing prompt by prompt to 

337
00:16:35,520 --> 00:16:38,640
rebuild all these things to get 
to the point of actually 

338
00:16:38,640 --> 00:16:41,640
ordering your food. 
Now when we're talking about 

339
00:16:41,640 --> 00:16:45,400
friction, in this case, does the
agent actually save you on 

340
00:16:45,400 --> 00:16:48,200
friction? 
Compare it to just opening up 

341
00:16:48,200 --> 00:16:51,040
the DoorDash app. 
The DoorDash app is already 

342
00:16:51,080 --> 00:16:53,560
hyper specialized in food 
delivery. 

343
00:16:53,960 --> 00:16:55,920
You open it up, it shows you all
the restaurants that you 

344
00:16:55,920 --> 00:16:58,240
recently ordered from, all your 
favorite meals from all of them,

345
00:16:58,480 --> 00:17:02,040
all the toggles and switches 
like 30 minute distances or 

346
00:17:02,040 --> 00:17:03,840
below. 
You have everything perfectly 

347
00:17:03,840 --> 00:17:06,119
tested in an AB tested user 
interface. 

348
00:17:06,920 --> 00:17:09,800
You don't have to prompt 
anything, but if you want to, 

349
00:17:10,240 --> 00:17:12,880
DoorDash has an AI search engine
built into it as well. 

350
00:17:12,880 --> 00:17:15,319
So you have that option 
available to you if you want. 

351
00:17:15,640 --> 00:17:18,560
But you don't have to. 
And in most cases, people won't 

352
00:17:18,680 --> 00:17:22,040
because they rather just tap a 
couple images on the screen, hit

353
00:17:22,040 --> 00:17:25,280
confirm order, then endlessly 
prompt an agent. 

354
00:17:25,280 --> 00:17:29,320
That's an open text interface. 
What you lose with agentic 

355
00:17:29,320 --> 00:17:33,440
ordering is an interface. 
You lose the ability to see the 

356
00:17:33,440 --> 00:17:36,120
food. 
People don't know exactly what 

357
00:17:36,120 --> 00:17:38,360
they want to eat. 
And any server will tell you 

358
00:17:38,360 --> 00:17:40,920
that they want to see images, 
they want to browse. 

359
00:17:40,920 --> 00:17:42,480
They want to see things that are
brought up. 

360
00:17:42,480 --> 00:17:44,360
When you're looking at one 
restaurant, maybe there's some 

361
00:17:44,360 --> 00:17:46,360
food that's similar to an A 
different 1. 

362
00:17:46,640 --> 00:17:48,720
You want to see different colors
and images. 

363
00:17:48,720 --> 00:17:51,440
That's what crates hunger. 
And that's when you say, oh, 

364
00:17:51,440 --> 00:17:55,480
that looks really appetizing and
agents don't solve that problem.

365
00:17:56,000 --> 00:18:00,000
So inadvertently by saying the 
agents are going to remove 

366
00:18:00,000 --> 00:18:03,840
friction by bypassing DoorDash 
and Uber Eats app, they're 

367
00:18:03,840 --> 00:18:06,480
actually creating more friction.
There's no way around it. 

368
00:18:06,480 --> 00:18:09,560
It is more difficult to get to 
the point of ordering through an

369
00:18:09,560 --> 00:18:12,760
agent than it is to just go to 
the app that's already hyper 

370
00:18:12,760 --> 00:18:14,760
specialized in the specific 
task. 

371
00:18:15,000 --> 00:18:18,880
A comparison that I would make 
is, you know how Amazon has that

372
00:18:18,880 --> 00:18:22,600
ability to order through Alexa? 
Like you can have the speaker 

373
00:18:22,600 --> 00:18:25,920
there and you can say, hey, 
Alexa, get me some more Tide 

374
00:18:25,920 --> 00:18:28,320
Pods. 
Everybody has the ability to do 

375
00:18:28,320 --> 00:18:30,720
that. 
You have open-ended ordering on 

376
00:18:30,720 --> 00:18:34,240
Alexa devices, but how much has 
that really taken off? 

377
00:18:34,280 --> 00:18:37,800
Amazon doesn't really give the 
numbers because it's not that 

378
00:18:37,800 --> 00:18:40,200
popular. 
Sure, there's cases that you can

379
00:18:40,200 --> 00:18:42,400
cite where people have done it 
and maybe it's actually growing 

380
00:18:42,400 --> 00:18:45,800
really fast, but in the whole, 
it's not popular at all. 

381
00:18:46,040 --> 00:18:48,400
And the reason why is because 
people don't just want to talk 

382
00:18:48,400 --> 00:18:50,600
to a, a bot. 
They don't want to talk to 

383
00:18:50,600 --> 00:18:52,560
something. 
They want to visually see 

384
00:18:52,680 --> 00:18:55,520
precisely what their order is, 
how much it costs. 

385
00:18:55,520 --> 00:18:57,560
They want to see it visually in 
front of them. 

386
00:18:57,960 --> 00:19:00,560
So most people, the vast 
majority, order through 

387
00:19:00,560 --> 00:19:03,680
Amazon.com. 
They go to the website, they put

388
00:19:03,680 --> 00:19:06,040
in their order, they see it in 
the checkout, they know exactly 

389
00:19:06,040 --> 00:19:08,440
what they're ordering and they 
complete the process. 

390
00:19:08,880 --> 00:19:11,720
They don't like ordering through
a speaker, just like most people

391
00:19:11,720 --> 00:19:15,560
won't like ordering their food 
through an AI that has endless 

392
00:19:15,560 --> 00:19:18,560
prompting, endless texting. 
And that's not the only problem 

393
00:19:18,560 --> 00:19:20,440
with this one examle, which is 
DoorDash. 

394
00:19:20,760 --> 00:19:23,600
There are more critical flaws 
with this Tye of thought rocess.

395
00:19:24,120 --> 00:19:27,280
One of them is the simple fact 
that when you're ordering food, 

396
00:19:27,440 --> 00:19:29,400
your goal is not to get the 
lowest price. 

397
00:19:29,640 --> 00:19:32,640
Now from a tech investor 
perspective, from Citronee, they

398
00:19:32,640 --> 00:19:35,560
probably look at it as the only 
thing the agent really cares 

399
00:19:35,560 --> 00:19:37,280
about is getting the lowest 
price. 

400
00:19:37,320 --> 00:19:40,600
So you put your order in for a 
McDonald's burger and fries, and

401
00:19:40,600 --> 00:19:44,080
the agent goes and finds 50 
different delivery platforms 

402
00:19:44,080 --> 00:19:47,200
that can pick up from McDonald's
and it selects the one that is 

403
00:19:47,200 --> 00:19:50,440
the lowest price. 
Maybe it's $0.50 lower than 

404
00:19:50,440 --> 00:19:52,960
DoorDash. 
So now your order, instead of 

405
00:19:52,960 --> 00:19:55,760
being routed to DoorDash, is now
routed to some third party 

406
00:19:55,760 --> 00:19:59,120
smaller company that's offering 
a cheaper take rate than 

407
00:19:59,120 --> 00:20:01,840
DoorDash by $0.50. 
You save $0.50. 

408
00:20:01,840 --> 00:20:04,000
But in the process you got your 
order routed to a different 

409
00:20:04,000 --> 00:20:06,200
company, and that company 
doesn't have as good of 

410
00:20:06,200 --> 00:20:08,280
logistics. 
It's not as saturated with 

411
00:20:08,280 --> 00:20:10,560
drivers, so it takes longer to 
get your order. 

412
00:20:10,880 --> 00:20:13,720
Your order arrives 10 minutes 
later than it normally would. 

413
00:20:13,760 --> 00:20:16,920
Your food's soggy and cold. 
All of a sudden you're 

414
00:20:16,920 --> 00:20:20,680
frustrated at the agent for 
routing your order to some third

415
00:20:20,680 --> 00:20:23,240
party delivery service that 
doesn't have its act together. 

416
00:20:23,600 --> 00:20:27,560
So you tell the agent please 
only route it to big mainstream 

417
00:20:27,880 --> 00:20:30,040
logistics companies like 
DoorDash and Uber. 

418
00:20:30,120 --> 00:20:33,240
All of a sudden the whole thesis
of fractional cost going down is

419
00:20:33,240 --> 00:20:35,800
gone and the biggest companies 
continue to succeed. 

420
00:20:36,480 --> 00:20:39,760
The truth is, people ordering 
from DoorDash and Uber do not 

421
00:20:39,760 --> 00:20:42,920
care solely about fees. 
If they cared only about the 

422
00:20:42,920 --> 00:20:45,640
fees and getting the lowest 
price, they wouldn't be ordering

423
00:20:45,640 --> 00:20:47,920
food delivery. 
What they care about is a 

424
00:20:47,920 --> 00:20:52,840
combination of price as well as 
service and quality and 

425
00:20:52,840 --> 00:20:56,040
consistency and being able to 
get refunds and being taken care

426
00:20:56,040 --> 00:20:59,880
of if there's any issues. 
And an agent destroys nearly all

427
00:20:59,880 --> 00:21:02,480
of that. 
Consider the fact that part of 

428
00:21:02,480 --> 00:21:05,080
the reason that I can get 
refunds on DoorDash, part of the

429
00:21:05,080 --> 00:21:07,240
reason they're so eager to 
refund me if I ever have a 

430
00:21:07,240 --> 00:21:10,120
problem is because DoorDash 
knows that I'm a a good 

431
00:21:10,120 --> 00:21:12,120
customer. 
I do lots of orders with them. 

432
00:21:12,120 --> 00:21:14,720
Most of them are good. 
So when something goes wrong, 

433
00:21:14,840 --> 00:21:17,520
they can look at my history, 
they can say, hey, most of the 

434
00:21:17,520 --> 00:21:20,160
orders from this guy are good. 
So when something goes wrong, we

435
00:21:20,160 --> 00:21:22,320
should give him a full refund 
because he's overall a good 

436
00:21:22,320 --> 00:21:23,960
customer. 
We want to keep him happy. 

437
00:21:24,360 --> 00:21:28,640
Having a healthy order backlog 
is important to a customer. 

438
00:21:28,880 --> 00:21:31,360
It allows you more leverage as a
customer. 

439
00:21:31,680 --> 00:21:35,480
If an agent routes your orders 
to 50 different companies, you 

440
00:21:35,480 --> 00:21:39,040
have no outstanding order 
backlog and therefore makes it 

441
00:21:39,040 --> 00:21:41,040
harder to be taken care of as a 
customer. 

442
00:21:41,120 --> 00:21:43,800
Because if you only have one 
random order from some random 

443
00:21:43,800 --> 00:21:46,440
place, then you're not going to 
be able to get a refund nearly 

444
00:21:46,440 --> 00:21:48,600
as easily as if you have 50 
orders from them. 

445
00:21:48,760 --> 00:21:51,080
This destroys the customer 
relationship with the company, 

446
00:21:51,120 --> 00:21:53,480
their order history. 
You also have the problem of 

447
00:21:53,480 --> 00:21:56,200
memberships. 
Both DoorDash and Uber Eats have

448
00:21:56,200 --> 00:21:59,080
extensive memberships. 
We have Dashpass and we have 

449
00:21:59,080 --> 00:22:01,440
Uber 1:00 which gives you lots 
of perks. 

450
00:22:01,520 --> 00:22:04,280
People that use a DoorDash 
membership are not going to want

451
00:22:04,280 --> 00:22:07,760
to order outside of DoorDash and
it will make it cheaper after 

452
00:22:07,760 --> 00:22:09,160
they already paid for that 
membership. 

453
00:22:09,400 --> 00:22:12,160
The same thing with Uber eats. 
So you have further customer 

454
00:22:12,160 --> 00:22:14,800
lock in by these memberships 
which make it even more 

455
00:22:14,800 --> 00:22:18,440
difficult to have the scenario 
where there are dozens of 

456
00:22:18,440 --> 00:22:19,960
different food delivery 
networks. 

457
00:22:20,400 --> 00:22:24,040
Why would anybody use multi app 
dashboards and different food 

458
00:22:24,040 --> 00:22:27,200
delivery networks when they have
their membership with DoorDash 

459
00:22:27,240 --> 00:22:29,840
or their Uber one pass that 
makes this one uniquely cheap 

460
00:22:30,360 --> 00:22:32,840
where they get benefits from it 
and rewards from ordering 

461
00:22:33,000 --> 00:22:35,080
routinely? 
The whole scenario here, this 

462
00:22:35,080 --> 00:22:37,560
hypothetical, is that agents are
making it so that there's way 

463
00:22:37,560 --> 00:22:39,440
less friction so that they can 
shop around. 

464
00:22:39,440 --> 00:22:41,320
They can lower the margins down 
to 0. 

465
00:22:41,680 --> 00:22:44,800
In reality, these agents would 
make it substantially more 

466
00:22:44,800 --> 00:22:48,320
friction on behalf of the user. 
Users want interfaces, they want

467
00:22:48,320 --> 00:22:50,200
routines. 
They want to be able to know and

468
00:22:50,200 --> 00:22:52,000
see what they're ordering with 
precision. 

469
00:22:52,440 --> 00:22:55,240
Ordering food is not an 
open-ended game where you just 

470
00:22:55,280 --> 00:22:57,440
want to be surprised of what 
comes to your doorstep. 

471
00:22:57,560 --> 00:22:59,920
So these are the type of 
arguments that are being given 

472
00:22:59,920 --> 00:23:01,360
here. 
And again, this one isn't just 

473
00:23:01,360 --> 00:23:03,560
some random argument. 
They highlight this as the 

474
00:23:03,560 --> 00:23:06,040
poster child, the prime 
argument. 

475
00:23:06,040 --> 00:23:08,800
This is the example of a company
that will be disintermediated, 

476
00:23:09,040 --> 00:23:11,000
but it's it couldn't be further 
from the truth. 

477
00:23:11,480 --> 00:23:13,880
DoorDash is perhaps one of the 
most insulated companies from 

478
00:23:13,880 --> 00:23:17,280
agentic technology because of 
how difficult it is to replicate

479
00:23:17,280 --> 00:23:20,200
a menu, to replicate the 
hundreds of of different 

480
00:23:20,200 --> 00:23:22,160
customizations for every single 
item. 

481
00:23:22,600 --> 00:23:25,080
Food is a very visual game. 
People want to see things when 

482
00:23:25,080 --> 00:23:28,600
they're ordering and it takes 
far more work to constantly text

483
00:23:28,600 --> 00:23:30,840
and prompt something to build it
in the way that you want. 

484
00:23:31,160 --> 00:23:35,040
Then already going to a well 
built application, already a 

485
00:23:35,040 --> 00:23:38,320
unified user interface, one 
that's familiar, one that people

486
00:23:38,320 --> 00:23:42,440
have used, one that's AB tested 
for customers inadvertently, the

487
00:23:42,440 --> 00:23:45,880
agent would be more friction and
more situations. 

488
00:23:46,280 --> 00:23:48,560
And there's far more issues in 
it than just that. 

489
00:23:49,080 --> 00:23:51,480
So when I look at these type of 
arguments, they are convincing 

490
00:23:51,480 --> 00:23:54,840
on the surface until you really 
start to dive in to the nuances 

491
00:23:55,160 --> 00:23:57,920
and that's where you get lost. 
That's where a lot of investors 

492
00:23:57,920 --> 00:24:00,840
don't make it. 
They just sell because Citronees

493
00:24:00,840 --> 00:24:02,760
said that DoorDash is going to 
be replaced. 

494
00:24:03,160 --> 00:24:06,440
The stock market goes down. 
These companies drop 6 to 7% in 

495
00:24:06,440 --> 00:24:08,600
a single day. 
Hundreds of billions of dollars 

496
00:24:08,600 --> 00:24:12,920
worth of market cap was wiped 
out on a completely fan fiction 

497
00:24:13,000 --> 00:24:15,800
false premise. 
Now the DoorDash example is bad 

498
00:24:15,960 --> 00:24:17,880
because it undermines the case 
that they're making. 

499
00:24:18,080 --> 00:24:20,600
They're saying it's the poster 
child, but in reality it's very 

500
00:24:20,600 --> 00:24:23,120
resistant. 
But the next example that they 

501
00:24:23,120 --> 00:24:26,880
name is it goes from bad to 
comical. 

502
00:24:27,320 --> 00:24:30,920
Now we're going from delivery 
food networks like Uber Eats and

503
00:24:30,920 --> 00:24:34,440
DoorDash to the payment networks
Visa and MasterCard. 

504
00:24:35,080 --> 00:24:38,360
They say once the agents 
controlled the transaction, they

505
00:24:38,360 --> 00:24:40,680
went looking for a bigger paper 
clip. 

506
00:24:40,760 --> 00:24:43,800
There's only so much price 
matching and aggregation to do. 

507
00:24:44,200 --> 00:24:47,320
The biggest way to repeatedly 
save the users money, especially

508
00:24:47,320 --> 00:24:50,200
when the agents start 
transacting themselves, was to 

509
00:24:50,200 --> 00:24:53,760
eliminate fees. 
In machine to machine commerce, 

510
00:24:53,880 --> 00:24:57,960
the 2 to 3% card interchange 
rate became an obvious target. 

511
00:24:58,400 --> 00:25:01,760
Agents went looking for faster 
and cheaper options than cards. 

512
00:25:02,280 --> 00:25:07,000
Most settled on using stable 
coins via Solona or Ethereum, 

513
00:25:07,360 --> 00:25:10,320
where settlement was near 
instant and the transaction cost

514
00:25:10,320 --> 00:25:12,360
was measured in a fraction of a 
penny. 

515
00:25:12,920 --> 00:25:14,440
And then they have a 
hypothetical. 

516
00:25:14,440 --> 00:25:16,360
Again, this is a fake headline 
of the future. 

517
00:25:16,760 --> 00:25:21,640
MasterCard Q 12027 Net revenues 
plus 6% year over year. 

518
00:25:21,880 --> 00:25:26,720
Purchase volume growth slows to 
3.4% year over year from 5.9% 

519
00:25:26,720 --> 00:25:29,400
the prior year. 
Management notes that agent LED 

520
00:25:29,400 --> 00:25:32,760
price optimization and pressure 
and discretionary categories. 

521
00:25:33,120 --> 00:25:38,440
Bloomberg, April 29th, 2027. 
It was Mastercard's Q 12027 

522
00:25:38,440 --> 00:25:40,880
report that was the point of no 
return. 

523
00:25:41,360 --> 00:25:44,880
Agentic commerce went from being
a product story to a plumbing 

524
00:25:44,880 --> 00:25:47,200
story. 
MasterCard dropped 9% the 

525
00:25:47,200 --> 00:25:50,520
following day, Visa 2, but 
paired their losses after 

526
00:25:50,520 --> 00:25:53,360
analysts pointed out at stronger
position stablecoin 

527
00:25:53,360 --> 00:25:56,960
infrastructure. 
Agentic commerce routing around 

528
00:25:56,960 --> 00:26:01,160
interchange posed a far greater 
risk to card focus banks and 

529
00:26:01,200 --> 00:26:05,280
monoline issuers who collected 
the majority of that 2 to 3% fee

530
00:26:05,280 --> 00:26:08,120
and had built an entire business
segment around rewards programs 

531
00:26:08,120 --> 00:26:11,000
funded by merchant subsidies. 
American Express was hit the 

532
00:26:11,000 --> 00:26:13,440
hardest, a combination from 
headwinds from white collar 

533
00:26:13,440 --> 00:26:16,760
workforce reductions gutting its
customer base and agent routing 

534
00:26:17,000 --> 00:26:19,480
around interchange gutting the 
revenue model. 

535
00:26:19,720 --> 00:26:23,320
Synchrony, Capital One and 
Discovery fell more than 10% the

536
00:26:23,320 --> 00:26:25,800
following weeks as well. 
Their remotes were made of 

537
00:26:25,800 --> 00:26:28,880
friction and friction was going 
to 0. 

538
00:26:29,920 --> 00:26:31,240
That's the whole thesis there, 
right? 

539
00:26:31,240 --> 00:26:36,080
Friction is going to zero. 
And this is where Citronee, with

540
00:26:36,080 --> 00:26:38,040
all due respect, and I, I really
like this. 

541
00:26:38,040 --> 00:26:40,160
I think it's an interesting 
hypothetical so I'm not trying 

542
00:26:40,160 --> 00:26:42,800
to bash them. 
With all due respect, they 

543
00:26:42,800 --> 00:26:46,080
couldn't be more wrong here. 
One of their assumptions, and 

544
00:26:46,080 --> 00:26:49,680
this is a very tech kind of nerd
focused assumption, a lot of 

545
00:26:49,680 --> 00:26:53,560
investors have this assumption 
too, is that all customers want 

546
00:26:53,560 --> 00:26:55,760
is less friction. 
That's all customers want. 

547
00:26:55,920 --> 00:26:57,200
And that's an assumption that's 
correct. 

548
00:26:57,200 --> 00:27:00,360
In some cases, customers don't 
like friction in a lot of cases.

549
00:27:01,000 --> 00:27:03,920
But in reality, there's some 
cases, some unique ones, where 

550
00:27:03,920 --> 00:27:07,200
customers do like friction, 
where they love friction. 

551
00:27:07,240 --> 00:27:11,440
They don't want friction to go 
to zero, and settlement is one 

552
00:27:11,440 --> 00:27:13,640
of them. 
Payments is one of them. 

553
00:27:14,080 --> 00:27:15,920
Customers love friction and 
payments. 

554
00:27:16,320 --> 00:27:19,520
They absolutely adore friction 
and they would hate for it to go

555
00:27:19,520 --> 00:27:21,720
to 0. 
Customers reject. 

556
00:27:21,800 --> 00:27:24,720
They continually reject 
propositions where friction goes

557
00:27:24,720 --> 00:27:27,360
to zero and payments. 
Let me give you a a couple 

558
00:27:27,360 --> 00:27:29,520
examples. 
The whole value of a credit card

559
00:27:29,560 --> 00:27:32,400
is friction. 
That's why high income people 

560
00:27:32,480 --> 00:27:35,080
use credit cards. 
That's why people like me use a 

561
00:27:35,080 --> 00:27:37,200
credit card is because of the 
additional friction. 

562
00:27:37,800 --> 00:27:41,200
Allow me to explain. 
When I swipe a credit card, I 

563
00:27:41,200 --> 00:27:43,440
don't want settlement to happen 
instantly. 

564
00:27:43,960 --> 00:27:47,000
I want it to show that I paid 
money, but I want there to be 

565
00:27:47,000 --> 00:27:49,760
friction in the process. 
I want it to be routed through 

566
00:27:49,760 --> 00:27:51,560
my bank. 
I want it to be the bank's money

567
00:27:51,560 --> 00:27:56,040
that paid that money, not mine. 
I want my money to stay put, to 

568
00:27:56,040 --> 00:27:58,560
have a delay. 
I want there to be a period of 

569
00:27:58,560 --> 00:28:02,560
review where I can look at what 
the bank purchased on my behalf.

570
00:28:02,840 --> 00:28:05,680
I can review it and make sure 
that it was charged correctly, 

571
00:28:05,800 --> 00:28:08,880
that I received the product, and
I don't need to dispute it. 

572
00:28:09,400 --> 00:28:12,280
I want to be able to review all 
of it before I actually put my 

573
00:28:12,280 --> 00:28:16,480
money into the process. 
That's friction that slows down 

574
00:28:16,480 --> 00:28:18,760
the process. 
When you're purchasing things 

575
00:28:18,760 --> 00:28:21,760
and you're transacting, you want
to slow down the process. 

576
00:28:22,200 --> 00:28:25,360
In fact, the fact that I don't 
have to pay the money right away

577
00:28:25,360 --> 00:28:27,840
to buy something means that I 
can hold it in my savings 

578
00:28:27,840 --> 00:28:31,120
account and earn 4% interest for
another two weeks or another 

579
00:28:31,120 --> 00:28:33,520
month. 
That gives me the time money 

580
00:28:33,520 --> 00:28:36,040
value. 
So by having friction in this 

581
00:28:36,040 --> 00:28:38,720
process, by having a slower 
period of me paying for 

582
00:28:38,720 --> 00:28:42,560
something that is advantageous 
and consumers want to delay 

583
00:28:42,560 --> 00:28:45,920
payment for as long as possible.
That's why you have buy now, pay

584
00:28:45,920 --> 00:28:47,400
later. 
That's why you have credit cards

585
00:28:47,600 --> 00:28:51,240
where you can buy something 
today and pay for it in a month.

586
00:28:51,720 --> 00:28:55,320
That is friction. 
Furthermore, the thought process

587
00:28:55,320 --> 00:28:58,160
of money immediately moving from
my account to someone else's 

588
00:28:58,160 --> 00:29:02,400
account the instant I hit the 
buy button is terrifying. 

589
00:29:02,800 --> 00:29:04,200
That's the last thing that I 
want to do. 

590
00:29:04,680 --> 00:29:07,320
I don't want money to 
immediately exit my checking 

591
00:29:07,320 --> 00:29:10,040
account when I buy something 
online on some random website. 

592
00:29:10,640 --> 00:29:13,000
I want to be able to have time 
to review it and make sure I got

593
00:29:13,000 --> 00:29:15,520
the thing that I ordered before 
money moves from my account. 

594
00:29:15,760 --> 00:29:19,360
And again, that's friction. 
In this specific case, I want as

595
00:29:19,360 --> 00:29:22,120
much friction as possible. 
I wish there could be more 

596
00:29:22,360 --> 00:29:25,520
guardrails and time money value 
in the process. 

597
00:29:26,160 --> 00:29:30,760
So this is a a, a premise where 
the entire premise of what 

598
00:29:30,760 --> 00:29:32,600
they're building this on is 
wrong. 

599
00:29:33,280 --> 00:29:36,400
The premise that customers in 
the financial world want 0 

600
00:29:36,400 --> 00:29:40,000
friction is entirely wrong. 
Customers don't want 0 friction 

601
00:29:40,000 --> 00:29:42,360
in the financial world, they 
want as much friction as 

602
00:29:42,360 --> 00:29:44,280
possible. 
And so when you build an 

603
00:29:44,280 --> 00:29:47,240
assumption on a wrong premise, 
every single other assumption 

604
00:29:47,240 --> 00:29:50,520
based on that is also wrong. 
Another problem with the idea 

605
00:29:50,520 --> 00:29:53,600
that agents will just wipe away 
that credit card fee and save 

606
00:29:53,600 --> 00:29:56,680
the customer money is that in 
the eyes of the customer, that 

607
00:29:56,680 --> 00:30:00,080
fee isn't real. 
Customers aren't paying that 3% 

608
00:30:00,080 --> 00:30:01,840
fee. 
Customers get credit card 

609
00:30:01,840 --> 00:30:03,800
rewards whenever they swipe 
their card. 

610
00:30:04,040 --> 00:30:05,520
The merchants are paying that 
fee. 

611
00:30:05,880 --> 00:30:07,680
The merchants have hated that 
fee forever. 

612
00:30:08,120 --> 00:30:10,680
Merchants, any merchant doesn't 
like paying a credit card fee. 

613
00:30:10,760 --> 00:30:12,360
They don't like it at all. 
They wish that they could go 

614
00:30:12,360 --> 00:30:14,120
away. 
They've tried to legislate 

615
00:30:14,120 --> 00:30:17,280
reasons for it to go away. 
The merchant is already on board

616
00:30:17,280 --> 00:30:18,560
with wanting that fee to go 
away. 

617
00:30:18,560 --> 00:30:20,720
And there's already many 
alternatives that would make 

618
00:30:20,720 --> 00:30:22,400
that fee go away, like a debit 
card. 

619
00:30:22,880 --> 00:30:25,960
But there's a reason that high 
income earners don't like using 

620
00:30:25,960 --> 00:30:29,560
debit cards because they're 
instantaneous, because they are 

621
00:30:29,560 --> 00:30:32,760
frictionless. 
We already have that option that

622
00:30:32,760 --> 00:30:36,320
brings friction to 0, which is a
debit card, and customers don't 

623
00:30:36,320 --> 00:30:38,560
like it because it has no 
friction. 

624
00:30:38,720 --> 00:30:42,040
They're hypothetical of what 
could destroy MasterCard already

625
00:30:42,040 --> 00:30:45,440
exists today, and it's not 
destroying MasterCard because 

626
00:30:45,440 --> 00:30:47,760
when you study these companies, 
you know that the high income 

627
00:30:47,760 --> 00:30:50,280
earners want friction upon their
purchases and they want the 

628
00:30:50,280 --> 00:30:53,840
rewards and they're fine if the 
merchant pays the reward. 

629
00:30:54,320 --> 00:30:57,760
See, whenever I swipe a credit 
card, I get a reward of 2 to 3%.

630
00:30:58,000 --> 00:31:01,160
I get that money back When I 
spend hundreds of thousands of 

631
00:31:01,160 --> 00:31:04,960
dollars on business expenses, on
paying AWS hosting fees or what 

632
00:31:04,960 --> 00:31:07,840
have it. 
I get lots of money as a result 

633
00:31:07,840 --> 00:31:09,640
in credit card rewards paid back
to me. 

634
00:31:10,200 --> 00:31:13,760
I have no interest in using 
stablecoin, none at all. 

635
00:31:13,960 --> 00:31:17,800
Why would I want to give up the 
3% fee so that the merchant can 

636
00:31:17,800 --> 00:31:20,160
save money? 
That's a bad proposition for the

637
00:31:20,160 --> 00:31:25,240
customer, so I would never adopt
A gentic technology that brings 

638
00:31:25,240 --> 00:31:29,760
that fee to 0, making it so that
I don't earn a reward that's not

639
00:31:29,760 --> 00:31:32,920
in my interest or any other high
income earners interest at all. 

640
00:31:33,280 --> 00:31:36,960
Any American Express user signs 
up for those cards specifically 

641
00:31:36,960 --> 00:31:40,560
for the rewards so that they can
get paid every single purchase. 

642
00:31:40,960 --> 00:31:45,440
By taking the friction to 0 and 
wiping away that fee, you also 

643
00:31:45,440 --> 00:31:48,440
wipe away the rewards. 
You can't have one without the 

644
00:31:48,440 --> 00:31:50,760
other. 
You cannot have the 2 to 3% 

645
00:31:50,760 --> 00:31:54,120
reward without the 2 to 3% fee 
from the merchant. 

646
00:31:54,520 --> 00:31:57,880
So in the process of wiping away
and bringing friction to zero, 

647
00:31:57,880 --> 00:32:01,320
you also wipe away the primary 
value proposition of the credit 

648
00:32:01,320 --> 00:32:03,440
card. 
And that just won't happen. 

649
00:32:04,240 --> 00:32:07,240
I don't know anybody. 
I don't know a single person, 

650
00:32:07,240 --> 00:32:10,800
Not one wealthy person that 
wants to do transactions through

651
00:32:10,800 --> 00:32:13,000
stablecoin. 
Not a single one. 

652
00:32:13,160 --> 00:32:17,160
So when your entire premise is 
built on a false premise that 

653
00:32:17,160 --> 00:32:20,800
customers want friction to go to
0, everything else falls apart. 

654
00:32:21,120 --> 00:32:23,160
Customers like friction in 
payments. 

655
00:32:23,560 --> 00:32:25,160
They like the time value of 
money. 

656
00:32:25,480 --> 00:32:27,800
They like the security of being 
able to buy something now and 

657
00:32:27,800 --> 00:32:30,120
pay for it later and dispute it 
in the process. 

658
00:32:30,560 --> 00:32:31,880
They like the credit card 
rewards. 

659
00:32:32,080 --> 00:32:34,040
They don't care if the 
merchant's paying them. 

660
00:32:34,200 --> 00:32:36,360
They never have. 
So all the arguments here fall 

661
00:32:36,360 --> 00:32:38,800
flat when faced with actual 
consumer behavior. 

662
00:32:39,000 --> 00:32:41,480
Now this doomsday article 
continues on, but this time it 

663
00:32:41,480 --> 00:32:44,600
expands its horizon. 
It broadens itself from 

664
00:32:44,600 --> 00:32:47,680
individual companies like a 
DoorDash and American Express or

665
00:32:47,680 --> 00:32:51,200
MasterCard. 
That this time a Daisy chain of 

666
00:32:51,200 --> 00:32:55,680
correlated bets where everything
starts to unravel 1 domino at a 

667
00:32:55,680 --> 00:32:59,080
time. 
And I will say this is one of 

668
00:32:59,080 --> 00:33:02,400
the most beautifully written and
architected bear cases. 

669
00:33:02,560 --> 00:33:06,080
It's like a dream, a beautiful 
dream of destruction. 

670
00:33:06,400 --> 00:33:08,040
Let's go ahead and just take a 
look at it. 

671
00:33:08,080 --> 00:33:10,040
I I really do respect the 
riders. 

672
00:33:10,040 --> 00:33:13,440
This is fantastic riding. 
They go over how private credit 

673
00:33:13,440 --> 00:33:18,120
had grown from under 1 trillion 
in 2015 to 2.5 trillion in 2026.

674
00:33:18,520 --> 00:33:21,040
A meaningful share of the 
capital had been deployed into 

675
00:33:21,040 --> 00:33:24,240
software and technology deals. 
Many of them leveraged buyouts 

676
00:33:24,240 --> 00:33:27,040
of SAS companies. 
So in the future SAS companies 

677
00:33:27,040 --> 00:33:31,720
are SAS companies are so toast 
in this future bear case that 

678
00:33:31,720 --> 00:33:34,920
the valuations go so low that 
private equities buying them up 

679
00:33:34,920 --> 00:33:38,080
left and right using debt and 
buying up SAS, SAS companies 

680
00:33:38,080 --> 00:33:40,360
taking them private. 
So that's the future of SAS. 

681
00:33:40,360 --> 00:33:42,280
It's just these companies that 
become worthless. 

682
00:33:42,720 --> 00:33:46,200
They're like dying burrito 
chains or you know, sandwich 

683
00:33:46,200 --> 00:33:49,960
chains, which we have a private 
equity coming in and buying up 

684
00:33:49,960 --> 00:33:52,320
left and right. 
We have a headline here again 

685
00:33:52,320 --> 00:33:55,560
from Moody's. 
This is in April of 2027. 

686
00:33:56,120 --> 00:34:00,720
Moody's downgrades 18 billion of
PE back software debt across 14 

687
00:34:00,720 --> 00:34:04,480
issuers citing secular revenue 
headwinds from AI driven 

688
00:34:04,520 --> 00:34:08,000
competitive disruption, the 
largest single sector action 

689
00:34:08,000 --> 00:34:11,960
since energy in 2015. 
This is another fictional future

690
00:34:11,960 --> 00:34:15,440
headline that's that's Moody's 
now talking about the 

691
00:34:15,440 --> 00:34:19,400
destruction of SAS companies. 
Secular revenue headwinds. 

692
00:34:20,040 --> 00:34:22,280
Everyone remembers what happened
after the downgrade. 

693
00:34:22,560 --> 00:34:25,199
Industry veterans had already 
seen the playbook. 

694
00:34:25,199 --> 00:34:27,400
Following the 2015 energy 
downgrade. 

695
00:34:27,840 --> 00:34:31,639
Software backed loans began 
defaulting in Q3 of 2027. 

696
00:34:32,360 --> 00:34:35,600
Private equity portfolio 
companies and Information 

697
00:34:35,600 --> 00:34:37,719
Services and consulting 
followed. 

698
00:34:38,199 --> 00:34:41,320
Several multi billion dollar 
leverage buyouts of well known 

699
00:34:41,320 --> 00:34:43,159
SAS companies entered 
restructuring. 

700
00:34:43,560 --> 00:34:48,159
Zendesk was the smoking gun. 
Zendesk misses the covenants as 

701
00:34:48,280 --> 00:34:52,000
AI driven customer service 
automation erodes annual 

702
00:34:52,000 --> 00:34:55,679
recurring revenue. 5 billion 
direct lending facility marked 

703
00:34:55,679 --> 00:34:58,240
to $0.58. 
And this section continues on 

704
00:34:58,240 --> 00:35:01,600
detailing the domino effect of 
when all of these SAS companies 

705
00:35:01,880 --> 00:35:04,040
eventually go to zero. 
And like the rest of this, 

706
00:35:04,040 --> 00:35:06,640
there's a number of logical 
problems with these assumptions.

707
00:35:07,160 --> 00:35:08,880
First of all, you again have 
Zendesk. 

708
00:35:08,880 --> 00:35:10,000
Here is the one that they 
highlight. 

709
00:35:10,360 --> 00:35:13,760
Now, when we look at Zendesk, 
this is also a company that's in

710
00:35:13,800 --> 00:35:15,760
many of the Fortune 500 
companies. 

711
00:35:16,280 --> 00:35:19,040
This article suggests that those
companies will simply just RIP 

712
00:35:19,040 --> 00:35:20,960
out Zendesk and replace it with 
an agent. 

713
00:35:21,520 --> 00:35:23,200
But how long would that actually
take? 

714
00:35:23,720 --> 00:35:27,560
Consider how deeply integrated 
HR software is in the processes,

715
00:35:27,840 --> 00:35:29,760
the embedded nature of these 
products. 

716
00:35:30,320 --> 00:35:36,000
Even if agentic technology was 
vastly superior to Zendesk, it 

717
00:35:36,000 --> 00:35:39,400
would take half a decade plus to
get to the point of actually 

718
00:35:39,400 --> 00:35:41,560
replacing it. 
And that's a big if. 

719
00:35:42,160 --> 00:35:44,880
But the integration cost, the 
cost of actually doing anything 

720
00:35:44,880 --> 00:35:47,520
is incredibly difficult in a 
Fortune 500 company. 

721
00:35:48,000 --> 00:35:51,600
In most cases, the solution has 
to be dramatically better to 

722
00:35:51,600 --> 00:35:54,000
even consider it. 
That's switching costs, which 

723
00:35:54,000 --> 00:35:56,080
are entirely ignored in this 
article. 

724
00:35:56,720 --> 00:35:59,760
The other thing that I I just 
look at when I'm reading this is

725
00:35:59,760 --> 00:36:02,440
that when you're reading 
articles like this, they almost 

726
00:36:02,440 --> 00:36:06,000
assume that humans are 
inflexible, that humans don't 

727
00:36:06,040 --> 00:36:09,520
adapt, that humans don't have a 
brain and can figure things out 

728
00:36:09,520 --> 00:36:12,880
when things change, which is the
complete opposite of all human 

729
00:36:12,880 --> 00:36:15,680
history. 
Humans are highly flexible and 

730
00:36:15,680 --> 00:36:18,720
adaptable. 
Humans take on new tasks all the

731
00:36:18,720 --> 00:36:21,960
time. 
They they adapt to situations in

732
00:36:21,960 --> 00:36:23,640
different disruptions all the 
time. 

733
00:36:23,640 --> 00:36:26,440
One of the big concerns here is 
that all these SAS companies go 

734
00:36:26,440 --> 00:36:29,360
to zero. 
Well, if that's the case, if 

735
00:36:29,360 --> 00:36:32,720
Zendesk and all these other SAS 
companies went to zero, that 

736
00:36:32,720 --> 00:36:35,120
means that all the other 
companies paying for them. 

737
00:36:35,560 --> 00:36:37,760
Imagine all those companies that
are paying money to all these 

738
00:36:37,760 --> 00:36:40,000
SAS companies. 
All those companies would have 

739
00:36:40,000 --> 00:36:42,720
margins go up. 
So every company that's paying 

740
00:36:42,720 --> 00:36:45,000
for a SAS company, let's say 
you're paying for Salesforce, 

741
00:36:45,000 --> 00:36:49,480
adobezendeskmonday.com, you name
it, you're paying for all these 

742
00:36:49,480 --> 00:36:52,480
different services. 
You're paying $500 a month per 

743
00:36:52,480 --> 00:36:54,960
employee for all these different
licenses and seat based 

744
00:36:54,960 --> 00:36:57,800
services. 
If you could erase all of those,

745
00:36:58,240 --> 00:37:01,560
imagine the margins of all these
companies paying for all that. 

746
00:37:02,320 --> 00:37:06,240
Every other company that's not 
SAS would have dramatically 

747
00:37:06,240 --> 00:37:09,840
higher margins as a result. 
Every other company that's not 

748
00:37:09,840 --> 00:37:14,000
SAS would have entirely bigger 
budgets to be able to hire more 

749
00:37:14,000 --> 00:37:18,080
people, to be able to expand, to
be able to do land grabs, to be 

750
00:37:18,080 --> 00:37:21,160
able to create entirely new 
products and departments and 

751
00:37:21,160 --> 00:37:23,600
developments, to compete in new 
verticals. 

752
00:37:24,160 --> 00:37:27,080
Which whenever you do that, no 
matter how technologically 

753
00:37:27,080 --> 00:37:29,200
advanced you are, it requires 
humans. 

754
00:37:29,720 --> 00:37:31,640
That's the problem with this 
assumption overall. 

755
00:37:31,920 --> 00:37:33,920
It's talking about the 
destruction of one category, 

756
00:37:33,920 --> 00:37:37,160
which is SAS, without mentioning
that when that happens, the 

757
00:37:37,160 --> 00:37:39,480
companies that are paying for 
all of that, that now have those

758
00:37:39,600 --> 00:37:42,560
services being automated by 
agents, are dramatically more 

759
00:37:42,560 --> 00:37:45,480
profitable. 
Profitable companies expand. 

760
00:37:45,520 --> 00:37:47,720
That's typically what they do 
all throughout human history. 

761
00:37:47,720 --> 00:37:50,560
They expand, they grab more 
territory, they go into 

762
00:37:50,560 --> 00:37:53,360
different categories and that 
requires more people. 

763
00:37:53,960 --> 00:37:56,840
So this is a a very rigid look 
at the future. 

764
00:37:56,960 --> 00:38:00,040
It's completely unrealistic. 
It ignores human behavior. 

765
00:38:00,400 --> 00:38:03,600
It focuses hyper focused on a 
few companies and then even 

766
00:38:03,600 --> 00:38:06,840
those companies that it hyper 
focuses on, it gives very flawed

767
00:38:06,840 --> 00:38:09,280
arguments that don't really 
represent reality. 

768
00:38:09,960 --> 00:38:12,680
Saying that people want 0 
friction in transactions with 

769
00:38:12,680 --> 00:38:15,600
MasterCard, saying that people 
want to order things through a 

770
00:38:15,600 --> 00:38:18,200
text message. 
These are assumptions on their 

771
00:38:18,200 --> 00:38:20,320
surface are just on the false 
premise. 

772
00:38:20,680 --> 00:38:23,080
Yet it's enough to move the 
markets down dramatically. 

773
00:38:23,200 --> 00:38:26,760
Now I don't mind responding to 
these type of hypotheticals and 

774
00:38:27,080 --> 00:38:29,720
and giving my defense of them, 
my rebuttal of them and explain 

775
00:38:29,720 --> 00:38:32,560
why I think a lot of this is 
completely ridiculous. 

776
00:38:32,880 --> 00:38:35,480
But I'd also ask investors, ones
that are listening to this, ones

777
00:38:35,480 --> 00:38:37,520
that are learning, investing and
trying to take themselves 

778
00:38:37,520 --> 00:38:40,520
seriously as an investor, can we
just get a grip? 

779
00:38:41,320 --> 00:38:44,120
I mean, can we? 
Can we for a minute in this 

780
00:38:44,120 --> 00:38:48,880
market just get some sense and 
have some stability and be level

781
00:38:48,880 --> 00:38:51,120
headed? 
Can we go a day without the 

782
00:38:51,120 --> 00:38:53,520
market being ripped and tugged 
and pulled in different 

783
00:38:53,520 --> 00:38:57,000
directions by a hyper viral fan 
fiction article? 

784
00:38:57,560 --> 00:39:01,560
This reads like fan fiction. 
It reads like a Marvel story. 

785
00:39:02,040 --> 00:39:05,480
So much of it is completely 
flawed, so much of it is 

786
00:39:05,480 --> 00:39:08,000
embellished. 
So much of it ignores important 

787
00:39:08,000 --> 00:39:10,520
market dynamics. 
Now, again, no offense to the 

788
00:39:10,520 --> 00:39:12,560
author. 
It's an interesting article and 

789
00:39:12,560 --> 00:39:14,640
it's well written. 
But it's also completely 

790
00:39:14,640 --> 00:39:17,760
unrealistic. 
Yet it's moving the market, just

791
00:39:17,760 --> 00:39:19,320
like many of these ones before 
them. 

792
00:39:19,760 --> 00:39:22,680
They have compelling narratives 
that sound good on the surface 

793
00:39:22,680 --> 00:39:25,640
and it's causing investors to 
question their own reality. 

794
00:39:26,440 --> 00:39:31,120
It poses a a future where AI is 
this unstoppable alien force 

795
00:39:31,120 --> 00:39:33,680
that takes over the world that 
no one can do anything about, 

796
00:39:33,680 --> 00:39:36,720
that destroys everything. 
In fact, in every time period 

797
00:39:36,720 --> 00:39:39,080
known to man, there's people 
talking about the end of the 

798
00:39:39,080 --> 00:39:42,480
world, the doomsday, the doom 
and gloom, how nothing will be 

799
00:39:42,480 --> 00:39:44,480
the same and the future will 
look like nothing you've ever 

800
00:39:44,480 --> 00:39:47,880
seen, and how the stocks you own
today are essentially worthless.

801
00:39:48,040 --> 00:39:50,760
This is what Peter Lynch 
literally talked about decades 

802
00:39:50,760 --> 00:39:53,280
ago. 
OK, now we'll get to the 

803
00:39:53,280 --> 00:39:55,360
important stuff. 
There's always something to 

804
00:39:55,360 --> 00:39:58,960
worry about. 
This is the difference. 

805
00:39:58,960 --> 00:40:00,240
This is what happens in the 
stock market. 

806
00:40:00,320 --> 00:40:03,360
Because, see, everybody's got 
the brain power to do well in 

807
00:40:03,360 --> 00:40:04,720
the stock market. 
Question is whether you have the

808
00:40:04,720 --> 00:40:06,760
stomach for it. 
That's the key organ in the 

809
00:40:06,760 --> 00:40:08,880
body. 
There's always something to 

810
00:40:08,880 --> 00:40:10,440
worry about. 
I mean, there's always something

811
00:40:10,440 --> 00:40:12,600
to worry about. 
In the 50s, it was depression 

812
00:40:12,600 --> 00:40:15,360
and nuclear war. 
The 50s was the best decade. 

813
00:40:15,360 --> 00:40:17,040
This century. 
Do you remember when oil went 

814
00:40:17,040 --> 00:40:18,840
from 4 to 40? 
You remember that period? 

815
00:40:19,320 --> 00:40:21,480
And all the countries of the 
world are going to go bankrupt. 

816
00:40:22,600 --> 00:40:24,840
And then and the big banks go 
bankrupt and we're going to have

817
00:40:24,840 --> 00:40:26,920
a Great Depression and the stock
market's going down and you're 

818
00:40:27,240 --> 00:40:28,920
going to wind up selling pencils
and apples, you know? 

819
00:40:29,200 --> 00:40:32,000
Doesn't that sound familiar? 
You're going to end up selling 

820
00:40:32,000 --> 00:40:34,320
pencils and apples. 
The whole stock market's going 

821
00:40:34,320 --> 00:40:37,440
to go bankrupt. 
This is Peter Lynch saying this 

822
00:40:37,440 --> 00:40:41,600
prediction has happened before 
in almost the same exact way. 

823
00:40:41,600 --> 00:40:45,640
The only thing that's changed is
the time history repeats and 

824
00:40:45,640 --> 00:40:47,800
we're seeing again another 
repetition of it. 

825
00:40:47,960 --> 00:40:51,920
So I'm asking investors today to
realize that right now you're in

826
00:40:51,920 --> 00:40:54,040
a time period where there's a 
lot of uncertainty, there's 

827
00:40:54,040 --> 00:40:56,520
questions in the air, and 
nothing about that is new. 

828
00:40:56,520 --> 00:40:59,520
Nothing's abnormal. 
You'll be in the same situation 

829
00:40:59,520 --> 00:41:02,080
in five years and in 10 years. 
As Peter Lynch highlights, his 

830
00:41:02,080 --> 00:41:04,520
entire investment career has 
been filled with uncertainties 

831
00:41:04,520 --> 00:41:05,760
and predictions of doom and 
gloom. 

832
00:41:06,240 --> 00:41:10,200
Whether it's oil collapsing the 
market, whether it's nuclear 

833
00:41:10,200 --> 00:41:13,600
warfare, the Great Depression or
the recession or the the even 

834
00:41:13,600 --> 00:41:17,200
greater depression, whatever it 
may be, there's always something

835
00:41:17,200 --> 00:41:18,880
to worry about in all times of 
the market. 

836
00:41:19,240 --> 00:41:21,960
So if you're investor, get a 
grip, lock in. 

837
00:41:22,400 --> 00:41:27,960
This is a time to be optimistic.
Not to be greedy, not to be 

838
00:41:27,960 --> 00:41:30,880
unprudent, but we should be 
optimistic about the future. 

839
00:41:31,000 --> 00:41:33,040
Optimism is what makes money in 
the end. 

840
00:41:33,440 --> 00:41:35,640
You can look at the biggest 
examples and although there's 

841
00:41:36,200 --> 00:41:38,920
investors that try to look smart
by always having bearish calls 

842
00:41:38,920 --> 00:41:41,360
and when the market will 
collapse and there's some of 

843
00:41:41,360 --> 00:41:43,520
them like that, you have the 
Michael Buries of the world. 

844
00:41:43,880 --> 00:41:46,200
There's always way more 
investors that have made way 

845
00:41:46,200 --> 00:41:49,480
more money through optimism of 
having realistic positive 

846
00:41:49,480 --> 00:41:52,560
assumptions of the future. 
Knowing that many of these 

847
00:41:52,560 --> 00:41:54,920
companies are going to be highly
successful, extremely 

848
00:41:54,920 --> 00:41:57,560
prosperous. 
That if money is destroyed in 

849
00:41:57,560 --> 00:42:00,800
some companies that can't 
maneuver and adapt to agentic 

850
00:42:00,800 --> 00:42:04,040
technology, many of them will. 
And the ones that go away will 

851
00:42:04,040 --> 00:42:06,920
mean higher margins for the 
existing companies that will be 

852
00:42:06,920 --> 00:42:08,960
able to expand and grow in the 
process. 

853
00:42:09,120 --> 00:42:11,320
What we have right now is 
investors selling off great 

854
00:42:11,320 --> 00:42:14,920
companies for entirely 
unrealistic presumptions of the 

855
00:42:14,920 --> 00:42:17,800
future. 
Ones that are seeded in fear and

856
00:42:17,800 --> 00:42:22,560
doubt, not realistic optimism. 
So that's my thoughts on this 

857
00:42:22,560 --> 00:42:24,280
and that's how I'm responding as
an investor. 

858
00:42:24,720 --> 00:42:27,200
I'll continue to invest in the 
companies that I think will be 

859
00:42:27,200 --> 00:42:29,080
incredibly profitable in the 
future. 

860
00:42:29,080 --> 00:42:32,120
And I think that the disruption 
risk is dramatically overstated 

861
00:42:32,120 --> 00:42:35,520
by many of these articles. 
And this certainly isn't going 

862
00:42:35,520 --> 00:42:38,520
to be the last. 
As we see more AI technology 

863
00:42:38,520 --> 00:42:41,720
emerge, more demos, more clawed 
plug insurance, we're going to 

864
00:42:41,720 --> 00:42:45,040
see more of this doom and gloom.
But that's the latest as of now.

865
00:42:45,360 --> 00:42:46,840
Now moving on, we get to some 
big news. 

866
00:42:46,840 --> 00:42:51,560
Over the past seven days, 
Netflix gave the opportunity for

867
00:42:51,560 --> 00:42:54,200
Warner Brothers to go back to 
Paramount and say, hey, look, 

868
00:42:54,400 --> 00:42:57,280
you have 7 days to revise your 
offer and give us your best and 

869
00:42:57,280 --> 00:42:59,680
final offer. 
And that has happened. 

870
00:43:00,040 --> 00:43:03,400
Warner Bros Discovery says that 
Paramount made a higher bid and 

871
00:43:03,400 --> 00:43:05,920
the board will weigh the offer 
against the Netflix deal. 

872
00:43:05,960 --> 00:43:08,520
Basically, what's going on here 
is the offer that Paramount 

873
00:43:08,520 --> 00:43:10,840
submitted as private. 
So we know that they submitted 

874
00:43:10,840 --> 00:43:12,400
an offer. 
It was higher than their 

875
00:43:12,400 --> 00:43:14,520
previous one, but the details 
are private. 

876
00:43:14,520 --> 00:43:16,600
We don't know the number. 
Everybody's speculating of what 

877
00:43:16,600 --> 00:43:18,560
it could be. 
It could be anywhere from like 

878
00:43:18,560 --> 00:43:23,320
$31.00 with better terms to $34 
on the high end. 

879
00:43:23,880 --> 00:43:26,840
And either way, this is 
Paramount's last ditch attempt 

880
00:43:27,000 --> 00:43:31,320
to steal Warner Brothers away 
from Netflix to say, hey, Warner

881
00:43:31,320 --> 00:43:33,640
Brothers, we want to go with 
Paramount over Netflix because 

882
00:43:33,640 --> 00:43:36,000
their offer's so much more 
attractive now. 

883
00:43:36,200 --> 00:43:40,200
They can't necessarily do that 
unless Netflix looks at their 

884
00:43:40,200 --> 00:43:42,720
offer and decides not to beat 
it. 

885
00:43:43,120 --> 00:43:45,600
So Netflix can look at it and 
they get the first call. 

886
00:43:45,600 --> 00:43:47,040
They can match the offer or beat
it. 

887
00:43:47,640 --> 00:43:50,560
So Netflix is going to simply 
look at it, look at the revised 

888
00:43:50,560 --> 00:43:52,760
offer and see if they want to 
adjust theirs at all or if they 

889
00:43:52,760 --> 00:43:54,400
need to do anything to beat that
offer. 

890
00:43:54,400 --> 00:43:56,840
And that's up to them. 
Netflix has all the power in the

891
00:43:56,840 --> 00:43:58,720
situation. 
I think that Netflix is going to

892
00:43:58,720 --> 00:43:59,760
look at it and say, you know 
what? 

893
00:44:00,040 --> 00:44:02,920
If we have to pay an additional 
four to six months of free cash 

894
00:44:02,920 --> 00:44:06,240
flow to buy this service, to 
make it so that we own all of 

895
00:44:06,240 --> 00:44:10,160
this for the next 20-30 years so
that we can drop on this well of

896
00:44:10,160 --> 00:44:12,840
incredibly valuable content and 
distribute it through our 

897
00:44:12,840 --> 00:44:15,400
massive distribution, that's 
probably worth it. 

898
00:44:15,800 --> 00:44:18,440
Internally, Netflix knows 
exactly what they're willing to 

899
00:44:18,440 --> 00:44:21,280
pay and what is overpriced. 
They already have that modeled 

900
00:44:21,280 --> 00:44:22,960
out. 
They already have their numbers.

901
00:44:23,280 --> 00:44:26,040
So they can look at this and 
weigh it against what is their 

902
00:44:26,040 --> 00:44:28,440
price of what they believe this 
asset is worth. 

903
00:44:28,880 --> 00:44:32,400
And as Ted Sarandos says, they 
are disciplined buyers. 

904
00:44:32,800 --> 00:44:35,080
If they don't like the deal, if 
it's simply just too much and it

905
00:44:35,080 --> 00:44:38,280
makes it not worth it, they will
walk away and Netflix will be 

906
00:44:38,280 --> 00:44:41,360
able to collect over a 2 1/2 
billion dollar fee. 

907
00:44:41,920 --> 00:44:44,360
So even if they walk away, they 
get something out of this. 

908
00:44:44,880 --> 00:44:47,680
But if I have to guess, and 
again this is hypothetical, 

909
00:44:47,680 --> 00:44:49,760
maybe they won't. 
My guess is that Paramount 

910
00:44:49,760 --> 00:44:51,920
submits a higher offer. 
I think that Netflix makes 

911
00:44:51,920 --> 00:44:55,000
theirs more attractive and locks
it down and then it's done from 

912
00:44:55,000 --> 00:44:56,720
there. 
But we'll have to wait and see. 

913
00:44:57,280 --> 00:44:59,360
Now moving on, we have another 
bit of news here. 

914
00:44:59,360 --> 00:45:03,280
This is a big headline. 
Meta and AMD agreed to AI chip 

915
00:45:03,280 --> 00:45:05,520
deal worth more than $100 
billion. 

916
00:45:06,080 --> 00:45:09,760
These deals used to be really 
cool, like we'd see these and 

917
00:45:09,760 --> 00:45:12,640
everybody would get excited. 
Oh my gosh, $100 billion deal 

918
00:45:12,640 --> 00:45:15,120
with a chip company between a 
big tech company. 

919
00:45:15,680 --> 00:45:18,080
But now that they've happened so
frequently, it just seems like 

920
00:45:18,080 --> 00:45:20,720
another day, like every other 
day there's a deal like this. 

921
00:45:20,720 --> 00:45:23,280
It seems like I'm getting a 
little bit numb to looking at 

922
00:45:23,280 --> 00:45:25,320
these deals. 
The instant reaction to this 

923
00:45:25,320 --> 00:45:27,360
deal is AMD, of course, is up 
big. 

924
00:45:27,440 --> 00:45:31,480
They got a new customer, so 
they're up nearly 10% and Meta's

925
00:45:31,480 --> 00:45:33,520
up a little bit. 
So Meta's not even being hurt by

926
00:45:33,520 --> 00:45:35,400
this. 
Most investors already know that

927
00:45:35,400 --> 00:45:38,200
Meta's they're buying big, that 
they're investing in CapEx. 

928
00:45:38,200 --> 00:45:41,320
So nothing is new about that. 
Now, I've recently invested A 

929
00:45:41,320 --> 00:45:42,880
substantial amount of money in 
Meta. 

930
00:45:42,880 --> 00:45:44,680
I've made it my third largest 
position. 

931
00:45:44,920 --> 00:45:48,720
It's about $150,000 invested in 
it and I love what the business 

932
00:45:48,720 --> 00:45:50,520
is doing. 
They're becoming one of the few 

933
00:45:50,520 --> 00:45:53,480
supercomputing powers. 
Many people are bearish on this 

934
00:45:53,480 --> 00:45:55,080
CapEx spend. 
I'm not. 

935
00:45:55,080 --> 00:45:58,480
Long term, I realize that it 
will hurt the financials in the 

936
00:45:58,480 --> 00:46:02,200
short term, it'll make it so 
that their accounting and the 

937
00:46:02,200 --> 00:46:04,680
fact that they're depreciating 
these assets over time will hurt

938
00:46:04,680 --> 00:46:07,080
their earnings. 
But I think long term, this sets

939
00:46:07,080 --> 00:46:09,720
these companies up to be so 
uniquely strong. 

940
00:46:10,000 --> 00:46:12,960
Meta will have one of the 
strongest moats in the world, 

941
00:46:13,280 --> 00:46:16,240
both a network Moat and an 
infrastructure Moat. 

942
00:46:16,520 --> 00:46:18,280
It'll be virtually 
indestructible. 

943
00:46:18,560 --> 00:46:22,480
So I view this company as 
incredibly Moat heavy, super 

944
00:46:22,480 --> 00:46:26,520
wide and deep Moat with ample 
growth opportunity priced at a 

945
00:46:26,520 --> 00:46:29,720
22 Ford PE. 
To me, that feels like a very 

946
00:46:29,920 --> 00:46:33,160
attractive value proposition and
attractive situation to invest 

947
00:46:33,160 --> 00:46:34,560
in. 
That's why I continue to feel 

948
00:46:34,560 --> 00:46:36,800
good about this one. 
Finally, we move on to the fail 

949
00:46:36,800 --> 00:46:38,320
of the week. 
In this case, it's an 

950
00:46:38,320 --> 00:46:43,520
unintentional this is an 
accidental hacking of about 7000

951
00:46:43,520 --> 00:46:45,920
robot vacuums. 
We have Sammy as Dewfall. 

952
00:46:45,960 --> 00:46:48,560
He claims that he wasn't trying 
to hack every robot vacuum in 

953
00:46:48,560 --> 00:46:51,240
the world. 
He just wanted to remote control

954
00:46:51,240 --> 00:46:56,600
his brand new DJI Romo vacuum 
with the PS-5 game PS-5 game 

955
00:46:56,600 --> 00:46:59,360
pad, he tells The Verge. 
Because it sounded fun. 

956
00:47:00,080 --> 00:47:02,760
But when his homegrown remote 
control app started talking to 

957
00:47:02,760 --> 00:47:05,800
DJI servers, it wasn't just one 
vacuum cleaner that replied 

958
00:47:06,120 --> 00:47:09,320
Roughly 7000 of them. 
All of them around the world 

959
00:47:09,320 --> 00:47:12,240
begin treating as Dewfall like 
their boss. 

960
00:47:12,440 --> 00:47:15,680
He could remotely control them 
and look and listen through 

961
00:47:15,680 --> 00:47:18,360
their live camera feeds. 
This means that this guy 

962
00:47:18,360 --> 00:47:21,640
inadvertently hacked other 
people's robot vacuums, which 

963
00:47:21,680 --> 00:47:25,480
have cameras on board, meaning 
that he hacked cameras of other 

964
00:47:25,480 --> 00:47:29,520
people's homes. 7000 people. 
He would watch them map out each

965
00:47:29,520 --> 00:47:32,520
room of a house, generating a 
complete 2D floor plan. 

966
00:47:32,800 --> 00:47:35,560
You could use any robot's IP 
address to find its rough 

967
00:47:35,560 --> 00:47:38,040
location. 
I found my device was just in 

968
00:47:38,040 --> 00:47:42,480
one of the ocean of devices. 
So it has a map here of like a 

969
00:47:42,760 --> 00:47:45,120
we have an area map of all the 
different devices that he's 

970
00:47:45,120 --> 00:47:47,400
found and here it has like a 
live camera feed. 

971
00:47:47,720 --> 00:47:50,400
Like some of these have on board
cameras so they can see objects 

972
00:47:50,400 --> 00:47:53,960
and avoid them, but it also 
inadvertently operates as a 

973
00:47:53,960 --> 00:47:57,520
camera going around your home. 
You could just drive around and 

974
00:47:57,520 --> 00:48:00,160
film people at will and they 
don't even know they're being 

975
00:48:00,160 --> 00:48:01,880
filmed. 
They might just think that their

976
00:48:01,880 --> 00:48:04,400
robot's acting a little funny. 
And this has to be one of the 

977
00:48:04,400 --> 00:48:07,160
worst hacking or personal 
invasion of people's privacies 

978
00:48:07,720 --> 00:48:10,440
possibly ever done. 
I mean, have we had a situation 

979
00:48:10,680 --> 00:48:14,960
where 7000 devices that are 
mobile and have cameras have 

980
00:48:14,960 --> 00:48:17,960
been hacked in people's homes? 
In most cases, if you hack a 

981
00:48:17,960 --> 00:48:21,480
camera in someone's home, at 
least it's just one location. 

982
00:48:22,080 --> 00:48:24,280
These ones are literally mobile.
You can drive them around. 

983
00:48:24,280 --> 00:48:27,680
So he has a camera that he can 
stare in other people's homes. 

984
00:48:27,760 --> 00:48:31,280
He says right now that DJI 
hasn't even fixed all the 

985
00:48:31,280 --> 00:48:34,240
vulnerabilities that he's found.
One of them is the ability to 

986
00:48:34,240 --> 00:48:37,560
view your own DJI Romo video 
stream without needing its 

987
00:48:37,560 --> 00:48:40,040
security pin. 
Another one is so bad, I won't 

988
00:48:40,040 --> 00:48:42,520
describe it until DJI has more 
time to fix it. 

989
00:48:42,600 --> 00:48:46,000
So DJI has a massive critical 
failure and there's security 

990
00:48:46,000 --> 00:48:48,160
here. 
And if you have one of these 

991
00:48:48,160 --> 00:48:50,920
robots, especially with the 
camera, you might want to just 

992
00:48:50,920 --> 00:48:53,800
put the robot away, take it off 
its wheels, flip it upside down,

993
00:48:53,800 --> 00:48:55,640
make sure it can't drive 
anywhere without your 

994
00:48:55,640 --> 00:48:57,760
permission. 
Because as of right now, it 

995
00:48:57,760 --> 00:49:01,520
could literally be viewed by 
anyone with a clod code plugin. 

996
00:49:01,960 --> 00:49:04,920
And until the update happens, 
until the company publicly 

997
00:49:05,120 --> 00:49:08,360
acknowledges it and it fixes all
these these security exploits, 

998
00:49:08,800 --> 00:49:10,960
it might be a concern. 
That's going to be it for this 

999
00:49:10,960 --> 00:49:12,080
time. 
Hope you enjoyed. 

1000
00:49:12,240 --> 00:49:13,040
See you in the next one.
