1
00:00:00,000 --> 00:00:20,000
[Music]

2
00:00:20,000 --> 00:00:47,000
[Music]

3
00:00:47,000 --> 00:00:53,000
Imagine a delivery drone just flying over your neighborhood. It's carrying a package, right?

4
00:00:53,000 --> 00:00:56,000
But halfway through the route, it realizes its battery is dying.

5
00:00:56,000 --> 00:00:58,000
Right. Which usually means it failed delivery or a crash.

6
00:00:58,000 --> 00:01:04,000
Exactly. But instead of turning around or like leading for human intervention,

7
00:01:04,000 --> 00:01:09,000
it automatically locates a charging pad in a stranger's backyard, it lands,

8
00:01:09,000 --> 00:01:13,000
and then it pays that homeowner for the electricity directly from its own crypto wallet.

9
00:01:13,000 --> 00:01:18,000
Completely autonomously. Yeah. No corporate expense accounts, no human supervision at all.

10
00:01:18,000 --> 00:01:19,000
Right.

11
00:01:19,000 --> 00:01:25,000
The machine just handled the entire transaction itself. Welcome to this deep dive into the source material.

12
00:01:25,000 --> 00:01:29,000
Today, you and I are unpacking the Depend Revolution of 2026.

13
00:01:29,000 --> 00:01:30,000
It's a huge topic.

14
00:01:30,000 --> 00:01:34,000
It really is. We're looking at how decentralized physical infrastructure networks,

15
00:01:34,000 --> 00:01:42,000
that's what Depend stands for, by the way, are just fundamentally changing how we build, share, and, well, profit from real world resources.

16
00:01:42,000 --> 00:01:45,000
Things like computing power, internet, and energy.

17
00:01:45,000 --> 00:01:50,000
Yeah. And it's a complete inversion of how we normally think about the foundations of our society.

18
00:01:50,000 --> 00:01:57,000
I mean, historically building a power grid or a global data network requires billions in a front capital.

19
00:01:57,000 --> 00:01:59,000
Right. You needed massive corporations.

20
00:01:59,000 --> 00:02:07,000
Exactly. Massive corporations building colossal concrete dams or endless heavily guarded rows of servers.

21
00:02:07,000 --> 00:02:13,000
Infrastructure was basically a fortress. It was centralized, exclusive, and totally untouchable.

22
00:02:13,000 --> 00:02:16,000
But now that fortress is kind of being dismantled.

23
00:02:16,000 --> 00:02:24,000
It is. The ownership is being handed over to everyday people who deploy their own hardware and earn tokens for providing essential services to the network.

24
00:02:24,000 --> 00:02:31,000
So we're moving from infrastructure as this service, you just blindly pay for to an ecosystem you actively participate in.

25
00:02:31,000 --> 00:02:35,000
But to ground this, let's look at the actual numbers in our sources because they're wild.

26
00:02:35,000 --> 00:02:40,000
By early 2026, the depends sector hit a $10 billion market cap.

27
00:02:40,000 --> 00:02:42,000
Yeah, which is massive.

28
00:02:42,000 --> 00:02:49,000
It is. Now, I know some people listening might look at that and say, wait, didn't it peak around 19 billion back in late 2025?

29
00:02:49,000 --> 00:02:55,000
It did. Yeah. But honestly, that correction is arguably the best thing that could have happened to the sector.

30
00:02:55,000 --> 00:02:57,000
Really? How so?

31
00:02:57,000 --> 00:03:04,000
Well, the speculative hype burned off. When you look at the landscape today, that $10 billion valuation is driven by actual numbers.

32
00:03:04,000 --> 00:03:09,000
The actual measurable utility, not just traders chasing mean tokens.

33
00:03:09,000 --> 00:03:14,000
We have over 8.8 million active devices deployed. Wow. Over 8 million.

34
00:03:14,000 --> 00:03:20,000
8.8 million, yeah, across 199 countries spread across 150 distinct projects.

35
00:03:20,000 --> 00:03:23,000
Yeah. I mean, the physical footprint is just massive now.

36
00:03:23,000 --> 00:03:27,000
Okay, let's unpack this because 8.8 million devices is a huge number.

37
00:03:27,000 --> 00:03:33,000
But how do we actually go from a speculative crypto idea to a system where you can like turn an idle gaming laptop

38
00:03:33,000 --> 00:03:39,000
or a rooftop solar panel into passive income while you sleep? How is this working in the real physical world?

39
00:03:39,000 --> 00:03:43,000
What's fascinating here is that dip in is finally transitioned into being invisible infrastructure.

40
00:03:43,000 --> 00:03:46,000
Invisible, meaning what exactly?

41
00:03:46,000 --> 00:03:51,000
Meaning the end users, whether they're out developers, logistics companies, or just everyday consumers,

42
00:03:51,000 --> 00:03:59,000
they're now benefiting from cheaper, faster services without ever needing to know that a blockchain is running in the background.

43
00:03:59,000 --> 00:04:05,000
The complexity is totally hidden. Which is the golden rule of good tech, right? It should just work.

44
00:04:05,000 --> 00:04:13,000
But to understand why dip in is exploding right now, I think we have to look at the massive physical bottleneck that's just crushing the tech into something.

45
00:04:13,000 --> 00:04:17,000
Oh, absolutely. The AI compute crunch. Yeah. Artificial intelligence is just starving.

46
00:04:17,000 --> 00:04:23,000
The old centralized models just cannot feed it enough computing power. No, they can't.

47
00:04:23,000 --> 00:04:27,000
AI's growth curve has slammed into a very hard physical limit.

48
00:04:27,000 --> 00:04:35,000
The projections from our sources show that by 2028 data centers alone will consume roughly 12% of all US electricity.

49
00:04:35,000 --> 00:04:43,000
Half percent, that's insane. It's completely unsustainable. Tech giants are hitting a ceiling where they literally cannot manufacture or secure enough top to your GPUs.

50
00:04:43,000 --> 00:04:50,000
Or honestly, even secure the municipal permits to draw the necessary power from the grid to run these massive AI models.

51
00:04:50,000 --> 00:04:56,000
Right. So we have this critical compute crunch. And this is exactly where decentralized compute steps in.

52
00:04:56,000 --> 00:05:01,000
It reminds me a bit of the early days of crypto mining. Oh, yeah. The Bitcoin mining days. Yeah.

53
00:05:01,000 --> 00:05:09,000
Where people ran their computers incredibly hot to solve random arbitrary math problems just to win some digital coins.

54
00:05:09,000 --> 00:05:19,000
Decentralized compute operates on a similar distributed model. But instead of doing useless math, your hardware is actually churning through AI training data.

55
00:05:19,000 --> 00:05:23,000
Exactly. You're powering the AI revolution and you're getting paid for it.

56
00:05:23,000 --> 00:05:28,000
It makes so much sense. But what are the actual economics behind this for the companies buying the power?

57
00:05:28,000 --> 00:05:33,000
Well, the cost incentives are making enterprise adoption basically inevitable at this point.

58
00:05:33,000 --> 00:05:43,000
If an AI startup wants to rent a top tier and video H100 GPU on a legacy cloud provider like AWS, for example, they're paying about $7.90 per hour.

59
00:05:43,000 --> 00:05:45,000
Okay. Almost eight bucks an hour.

60
00:05:45,000 --> 00:05:52,000
Right. But if they source that exact same hardware through a decentralized network, it costs between $2.56 and $5.95 per hour.

61
00:05:52,000 --> 00:05:56,000
Wow. So that's what? A 45 to 60% discount?

62
00:05:56,000 --> 00:06:03,000
Exactly. And for a startup burning through millions in compute costs, that isn't just a nice little discount.

63
00:06:03,000 --> 00:06:07,000
That is the difference between survival and bankruptcy.

64
00:06:07,000 --> 00:06:11,000
And the sources show this is generating serious revenue for these networks.

65
00:06:11,000 --> 00:06:18,000
Like, render network is a great example. They started out doing decentralized 3D rendering for digital artists.

66
00:06:18,000 --> 00:06:20,000
Yeah. Visual effects. That sort of thing.

67
00:06:20,000 --> 00:06:25,000
And they realized they were sitting on a gold mine of compute and expanded into generative AI.

68
00:06:25,000 --> 00:06:28,000
Now they're pulling in around $38 million in monthly revenue.

69
00:06:28,000 --> 00:06:30,000
Monthly. That's huge.

70
00:06:30,000 --> 00:06:36,000
And then you have a cost network, which is maintaining nearly 80% utilization of their GPUs.

71
00:06:36,000 --> 00:06:41,000
The demand is just undeniable. They're operating as true digital utilities at this point.

72
00:06:41,000 --> 00:06:45,000
Oh, and hold on. Let's look at the actual mechanics of this because I have a major security question here.

73
00:06:45,000 --> 00:06:46,000
Okay. Lay it on me.

74
00:06:46,000 --> 00:06:58,000
So if I'm an AI researcher, right? And my highly complex, totally proprietary model is being processed on some teenagers gaming PC halfway across the world.

75
00:06:58,000 --> 00:07:00,000
How do I know they actually did the work?

76
00:07:00,000 --> 00:07:03,000
Ah, the classic verification problem.

77
00:07:03,000 --> 00:07:11,000
Yeah. Like, what stops them from faking the computation, sending back random garbage data and just pocketing the crypto reward?

78
00:07:11,000 --> 00:07:26,000
Well, that is literally the multi million dollar question in decentralized computing because generating cryptographic zero knowledge proofs to verify complex deep learning is often more computationally expensive than just doing the work in the first place.

79
00:07:26,000 --> 00:07:28,000
Right. It defeats the whole purpose.

80
00:07:28,000 --> 00:07:33,000
Exactly. So the sources highlight a brilliant alternative from a network called genson.

81
00:07:33,000 --> 00:07:37,000
They use something called a Verde Verification Protocol Verde Verification.

82
00:07:37,000 --> 00:07:38,000
Okay. How does that work?

83
00:07:38,000 --> 00:07:43,000
It relies on a referee delegation system. It's essentially an automated game of finding the liar.

84
00:07:43,000 --> 00:07:48,000
A refereed system. Break that down for me. I think the source is called it a bisection game.

85
00:07:48,000 --> 00:07:52,000
Are they just cutting the code in half repeatedly until they find the lie?

86
00:07:52,000 --> 00:07:56,000
Conceptually, yeah. That's exactly what they do. The system is split into three roles.

87
00:07:56,000 --> 00:08:02,000
First, you have the solvers. They do the actual computational work, but they have to lock up or stake their own tokens as collateral.

88
00:08:02,000 --> 00:08:04,000
Okay. So they have skin in the game.

89
00:08:04,000 --> 00:08:09,000
Exactly. Then you have verifiers who run random statistical checks on the work.

90
00:08:09,000 --> 00:08:11,000
And finally, you have whistleblowers.

91
00:08:11,000 --> 00:08:18,000
Okay. The network snitches basically. Yeah. If a whistleblower flags a solver for faking the work,

92
00:08:18,000 --> 00:08:24,000
the protocol initiates this bisection game. It compares the intermediate states of the computation.

93
00:08:24,000 --> 00:08:28,000
If the results match at the halfway point, they know the first half of the work is honest.

94
00:08:28,000 --> 00:08:31,000
So then they cut the second half and half and check again.

95
00:08:31,000 --> 00:08:39,000
You got it. They keep having the problem space until they pinpoint the exact single mathematical operation where the solver lied.

96
00:08:39,000 --> 00:08:46,000
That is fascinating. It's like a highly aggressive automated game of guess who that just zeroes in on the exact point of fraud.

97
00:08:46,000 --> 00:08:47,000
It really is.

98
00:08:47,000 --> 00:08:49,000
And what happens when they actually catch the lie?

99
00:08:49,000 --> 00:08:55,000
Oh, it's brutal. The bad actors stake tokens are slashed, meaning they are permanently confiscated by the network.

100
00:08:55,000 --> 00:08:59,000
And the whistleblower gets a significant cut of those seized tokens as a reward.

101
00:08:59,000 --> 00:09:06,000
Wow. So it creates this intense game theoretic environment where honesty is quite literally the only profitable strategy.

102
00:09:06,000 --> 00:09:09,000
Precisely. It's a self-policing immune system for computation.

103
00:09:09,000 --> 00:09:12,000
I love that. But here's another hurdle though.

104
00:09:12,000 --> 00:09:15,000
AI frontier models are enormous.

105
00:09:15,000 --> 00:09:21,000
Like a standard consumer gaming PC doesn't have the memory to hold a massive language model all at once, right?

106
00:09:21,000 --> 00:09:24,000
Historically, yes. That was a very hard barrier.

107
00:09:24,000 --> 00:09:29,000
You need massive centralized server racks just to fit the model into memory.

108
00:09:29,000 --> 00:09:34,000
But the source is point to a major breakthrough with the SphereRoid blockchain protocol.

109
00:09:34,000 --> 00:09:36,000
SphereRoid blockchain. What does that do?

110
00:09:36,000 --> 00:09:39,000
It fundamentally changes the math of how models are trained.

111
00:09:39,000 --> 00:09:43,000
It breaks these massive AI models down into tiny independent blocks.

112
00:09:43,000 --> 00:09:50,000
So worker node only needs to compute a microscopic traction of the model, which only requires a few gigabytes of memory.

113
00:09:50,000 --> 00:09:57,000
Oh, wow. So that unlocks the ability for standard consumer grade GPUs to actually contribute to complex training runs.

114
00:09:57,000 --> 00:10:00,000
Exactly. You don't need a hyperscale data center anymore.

115
00:10:00,000 --> 00:10:02,000
We've essentially crowdsourced the supercomputer.

116
00:10:02,000 --> 00:10:04,000
But every action has a reaction, right?

117
00:10:04,000 --> 00:10:10,000
We now have millions of everyday computers churning through complex AI models around the clock,

118
00:10:10,000 --> 00:10:14,000
which creates a massive physical byproduct, heat, and energy consumption.

119
00:10:14,000 --> 00:10:16,000
Oh, yeah. The power draws intense.

120
00:10:16,000 --> 00:10:23,000
Right. You can't just plug a million GPUs into the existing neighborhood grid without melting the transformers down the street.

121
00:10:23,000 --> 00:10:26,000
This brings us to a really critical breaking point.

122
00:10:26,000 --> 00:10:31,000
How do we power all this infrastructure without just blacking out our cities?

123
00:10:31,000 --> 00:10:35,000
I mean, centralized grids are already failing every summer as it is.

124
00:10:35,000 --> 00:10:38,000
And this is where it depends intersects directly with the energy sector.

125
00:10:38,000 --> 00:10:44,000
We are decentralizing the power grid itself through virtual power plants or VPPs.

126
00:10:44,000 --> 00:10:47,000
Virtual power plants, so software instead of smoke stacks.

127
00:10:47,000 --> 00:10:52,000
Exactly. A VPP uses software to coordinate highly distributed energy resources.

128
00:10:52,000 --> 00:10:59,000
We were talking about residential solar panels, home battery walls, and even electric vehicles sitting idle and garages.

129
00:10:59,000 --> 00:11:01,000
So it networks all of them together?

130
00:11:01,000 --> 00:11:10,000
Yeah. It networks all of these fragmented resources so they can act as one massive, highly flexible power plant that dispatches energy precisely where the grid is strained.

131
00:11:10,000 --> 00:11:15,000
I always think of this as the Airbnb of energy, but taken to its absolute logical conclusion.

132
00:11:15,000 --> 00:11:16,000
That's a great way to look at it.

133
00:11:16,000 --> 00:11:29,000
Right. Because just like Airbnb uses software to unlock empty bedrooms, smart contracts in a VPP act as this automated invisible property manager for your solar panels.

134
00:11:29,000 --> 00:11:38,000
Like if your neighbor turns on their air conditioning and the local grid strains, the smart contract instantly routes your excess stored solar power to their house.

135
00:11:38,000 --> 00:11:42,000
And it deposits fractions of a cent into your wallet all in milliseconds.

136
00:11:42,000 --> 00:11:48,000
Exactly. You transition from being just a passive consumer paying a bill to an active energy producer.

137
00:11:48,000 --> 00:11:52,000
And the real world implementation of this is already incredibly successful.

138
00:11:52,000 --> 00:11:55,000
The source is detailed a company called Fuse Energy over in the UK.

139
00:11:55,000 --> 00:11:57,000
Oh, the one founded by the former Revolute Executive.

140
00:11:57,000 --> 00:12:03,000
That's the one. By acting as a hybrid between a tokenized dip in operator and a traditional utility company,

141
00:12:03,000 --> 00:12:08,000
they reach a $400 million annual recurring revenue by the end of 2025.

142
00:12:08,000 --> 00:12:10,000
$400 million just incredible.

143
00:12:10,000 --> 00:12:12,000
And they serve over 200,000 households.

144
00:12:12,000 --> 00:12:16,000
But most importantly, they've actually reduced energy costs for those families by 10%.

145
00:12:16,000 --> 00:12:21,000
Here's where it gets really interesting though. The grid isn't just becoming decentralized.

146
00:12:21,000 --> 00:12:26,000
The way it manages itself is becoming biologically inspired.

147
00:12:26,000 --> 00:12:29,000
Ah, you're referring to the SwarmNet 5G framework.

148
00:12:29,000 --> 00:12:31,000
Yes, this blew my mind.

149
00:12:31,000 --> 00:12:38,000
Wild, isn't it? Rather than relying on a slow centralized control room to route power or detect failures,

150
00:12:38,000 --> 00:12:45,000
which can take hundreds of milliseconds, SwarmNet 5G puts the intelligence at the very edge of the network.

151
00:12:45,000 --> 00:12:48,000
The local grid nodes use what they call spiking neural networks.

152
00:12:48,000 --> 00:12:54,000
Right. And a spiking neural network physically mimics how the human brain processes information correct.

153
00:12:54,000 --> 00:12:57,000
Like it doesn't process continuous streams of empty data.

154
00:12:57,000 --> 00:13:01,000
It only fires when a specific threshold is met.

155
00:13:01,000 --> 00:13:04,000
Exactly, which saves an enormous amount of computational overhead.

156
00:13:04,000 --> 00:13:08,000
And the way these nodes coordinate across the grid is through this concept called Stigmergy,

157
00:13:08,000 --> 00:13:10,000
which is literally modeled on ant colonies.

158
00:13:10,000 --> 00:13:13,000
Yeah, it's biomimicry at its finest.

159
00:13:13,000 --> 00:13:19,000
Just as ants leave physical pheromone trails to signal the shortest path to food to the rest of the Swarm,

160
00:13:19,000 --> 00:13:27,000
these grid nodes leave digital pheromones across the network to signal congestion, power availability, or a physical fault on the line.

161
00:13:27,000 --> 00:13:29,000
So let's make that tangible for the listener.

162
00:13:29,000 --> 00:13:34,000
What does an ant colony power grid actually achieve in the real world?

163
00:13:34,000 --> 00:13:36,000
Speed and resilience, primarily.

164
00:13:36,000 --> 00:13:44,000
The data shows this decentralized Swarm intelligence can detect power grid faults in just 8.17 milliseconds.

165
00:13:44,000 --> 00:13:50,000
Yes, for comparison, traditional scatter systems take nearly 100 milliseconds.

166
00:13:50,000 --> 00:13:57,000
Wow, to put that in perspective, that means like a heavy tree branch could snap a power line during a storm, right?

167
00:13:57,000 --> 00:14:03,000
And before that severed cable even hits the ground, the Swarm intelligence has already sensed the voltage drop, isolated the broken line,

168
00:14:03,000 --> 00:14:06,000
and rerouted backup power to the neighborhood.

169
00:14:06,000 --> 00:14:10,000
Exactly. The neighborhood wouldn't even see their lights flicker. It happens that fast.

170
00:14:10,000 --> 00:14:18,000
Furthermore, by dynamically balancing the load using those digital pheromones we talked about, the system reduces overall energy waste by 43%.

171
00:14:18,000 --> 00:14:23,000
43% less waste just by letting the nodes talk to each other autonomously, that is staggering.

172
00:14:23,000 --> 00:14:29,000
So we now have decentralized computing power and a highly efficient decentralized energy grid.

173
00:14:29,000 --> 00:14:35,000
But this race is a really profound question about the economy itself. Who is actually buying these resources?

174
00:14:35,000 --> 00:14:36,000
Right, who are the customers?

175
00:14:36,000 --> 00:14:42,000
Yeah, I mean software developers and logistics companies are sure, but it's not just humans anymore.

176
00:14:42,000 --> 00:14:45,000
Not at all. We are seeing the dawn of the machine economy.

177
00:14:45,000 --> 00:14:52,000
In February, 2026, a massive shift occurred when open mind and circle launched the buy 402 protocol.

178
00:14:52,000 --> 00:14:55,000
The buy 402 protocol. What exactly did that change?

179
00:14:55,000 --> 00:15:00,000
Well, they activated a dormant internet status code error 402 payment required.

180
00:15:00,000 --> 00:15:05,000
And they used it to allow machines to make instant nano payments using USDC stablecoins.

181
00:15:05,000 --> 00:15:09,000
Okay, so this goes right back to our opening scenario with the drone. Let's call the drone bits.

182
00:15:09,000 --> 00:15:14,000
It's realized as its battery is low, finds a charging station and just pays for it.

183
00:15:14,000 --> 00:15:20,000
Because historically, an autonomous system hitting a paywall would trigger an alert for a human, right?

184
00:15:20,000 --> 00:15:27,000
A human would have to input a corporate credit card, clear an enterprise billing department, and pass a KYC know your customer check.

185
00:15:27,000 --> 00:15:30,000
Yeah, a massive administrative bottleneck.

186
00:15:30,000 --> 00:15:33,000
But the buy 402 protocol removes all of that friction.

187
00:15:33,000 --> 00:15:35,000
So the machines can just transact directly.

188
00:15:35,000 --> 00:15:40,000
If we connect this to the bigger picture, machines now have verifiable on chain identities.

189
00:15:40,000 --> 00:15:46,000
Protocols like fabric, link of physical robot, or even just a piece of software directly to a blockchain identity.

190
00:15:46,000 --> 00:15:48,000
So it's essentially an ID card for an AI.

191
00:15:48,000 --> 00:15:53,000
Right. And this allows AI agents to act entirely on their own behalf.

192
00:15:53,000 --> 00:16:02,000
A software agent can analyze a task, realize it needs more processing power, go to a depend network like IO.net, seamlessly purchased GPU time,

193
00:16:02,000 --> 00:16:07,000
run its analysis, pay the decentralized provider, and shut down.

194
00:16:07,000 --> 00:16:08,000
Zero human bottlenecks.

195
00:16:08,000 --> 00:16:12,000
It is literally giving robots their own bank account so they can run their own errands.

196
00:16:12,000 --> 00:16:13,000
That's a great way to put it.

197
00:16:13,000 --> 00:16:20,000
But physical robots like drones or autonomous vehicles, they have a massive physical constraint that software AI doesn't have.

198
00:16:20,000 --> 00:16:22,000
They run on local batteries.

199
00:16:22,000 --> 00:16:26,000
You cannot put a massive power-hungry server rack inside a robot dog.

200
00:16:26,000 --> 00:16:28,000
No, it would green the battery in five minutes.

201
00:16:28,000 --> 00:16:32,000
Right. So how are they doing the heavy lifting out in the real world?

202
00:16:32,000 --> 00:16:35,000
This is where neuromorphic computing comes into play.

203
00:16:35,000 --> 00:16:43,000
To power this physical AI out in the real world, machines are moving away from traditional GPUs entirely.

204
00:16:43,000 --> 00:16:48,000
They're using neuromorphic chips like the Intel OEE2 or the Inetera Pulsar.

205
00:16:48,000 --> 00:16:50,000
Neuromorphic, meaning brain like.

206
00:16:50,000 --> 00:16:52,000
Exactly. These aren't just software emulations.

207
00:16:52,000 --> 00:16:57,000
The physical silicon itself is designed to mimic the neural structure and synapses of the human brain.

208
00:16:57,000 --> 00:17:00,000
Because they only consume power when a spike occurs.

209
00:17:00,000 --> 00:17:03,000
Just like those spiking neural networks we discussed.

210
00:17:03,000 --> 00:17:07,000
For the grid, they operate on a tiny fraction of the wattage of traditional processors.

211
00:17:07,000 --> 00:17:11,000
Oh wow. So they are explicitly designed for edge computing in autonomous machines.

212
00:17:11,000 --> 00:17:24,000
Yes. And we are even seeing depend networks like Dynex creating decentralized marketplaces specifically to source and rent out this highly efficient low power neuromorphic compute.

213
00:17:24,000 --> 00:17:30,000
So the machines are getting smarter, they're highly energy efficient, and they have the wallets to act autonomously.

214
00:17:30,000 --> 00:17:38,000
But for a drone or a self-driving car to actually navigate the physical world, it needs incredibly accurate real-time data about its environment.

215
00:17:38,000 --> 00:17:39,000
Oh absolutely.

216
00:17:39,000 --> 00:17:44,000
I mean a satellite map from three years ago is completely useless if there's construction on your street today.

217
00:17:44,000 --> 00:17:47,000
How are these machines getting real-time environmental data?

218
00:17:47,000 --> 00:17:49,000
Through sensor and mapping depends.

219
00:17:49,000 --> 00:17:55,000
We are essentially crowdsourcing the physical mapping of the world. Hi-v-mappers are a great example of this.

220
00:17:55,000 --> 00:17:57,000
Hi-v-mapper, right. I've seen people with those cameras.

221
00:17:57,000 --> 00:18:03,000
Yeah. Users install specialized dash cams in their cars and simply do their daily commute.

222
00:18:03,000 --> 00:18:08,000
As they drive, they map the roads in high definition and the network rewards them with tokens.

223
00:18:08,000 --> 00:18:13,000
They've mapped over 710 million kilometers using this distributed model.

224
00:18:13,000 --> 00:18:17,000
710 million kilometers just by driving to work.

225
00:18:17,000 --> 00:18:24,000
Exactly. Another vital one is geodnet, which provides RTK or real-time kinematic GPS corrections.

226
00:18:24,000 --> 00:18:27,000
Right. Let's break down why RTK matters.

227
00:18:27,000 --> 00:18:33,000
If a delivery drone is flying to your house, standard GPS on your phone is accurate to within maybe what?

228
00:18:33,000 --> 00:18:34,000
Three to five meters?

229
00:18:34,000 --> 00:18:35,000
Yeah, roughly.

230
00:18:35,000 --> 00:18:40,000
So if a drone relies on standard GPS, it's going to crash into your mailbox or your living room window.

231
00:18:40,000 --> 00:18:43,000
It needs centimeter-level accuracy to land safely.

232
00:18:43,000 --> 00:18:49,000
Right. And standard satellites drift due to atmospheric interference, so they just aren't that precise.

233
00:18:49,000 --> 00:18:53,000
What geodnet does is deploy ground-based reference stations.

234
00:18:53,000 --> 00:18:56,000
Everyday people install these little base stations on their roofs.

235
00:18:56,000 --> 00:18:58,000
Okay. And what do the stations do?

236
00:18:58,000 --> 00:19:03,000
The stations calculate the exact error in the satellite signal at that specific location in moment

237
00:19:03,000 --> 00:19:05,000
and then they broadcast a real-time correction.

238
00:19:05,000 --> 00:19:12,000
Geodnet has over 20,000 of these user-owned base stations globally now, providing the centimeter-level accuracy

239
00:19:12,000 --> 00:19:15,000
that autonomous machines require to navigate safely.

240
00:19:15,000 --> 00:19:18,000
And the applications extend way beyond just delivery drones, right?

241
00:19:18,000 --> 00:19:24,000
Look at agriculture. The sources show DJI agriculture has deployed over 600,000 drones across 100 countries.

242
00:19:24,000 --> 00:19:26,000
It's totally revolutionized farming.

243
00:19:26,000 --> 00:19:34,000
Yeah, these drone swarms map agricultural territories in real-time and precision spray crops based on sensor data.

244
00:19:34,000 --> 00:19:37,000
The environmental impact is just staggering.

245
00:19:37,000 --> 00:19:43,000
They've reduced pesticide use by 35% and cut carbon emissions by 51 megatons.

246
00:19:43,000 --> 00:19:48,000
It is a massive leap in precision agriculture, completely powered by distributed hardware.

247
00:19:48,000 --> 00:19:51,000
But wait, hold on. I have to genuinely challenge the promise here for a second.

248
00:19:51,000 --> 00:19:52,000
Okay, go ahead.

249
00:19:52,000 --> 00:19:56,000
We started this deep dive talking about democratizing infrastructure.

250
00:19:56,000 --> 00:20:03,000
You know, regular people using an idle gaming PC or putting a small antenna on their roof for passive income.

251
00:20:03,000 --> 00:20:09,000
The source is also site this produce survey detailing the costs of these agricultural drone operations.

252
00:20:09,000 --> 00:20:11,000
Ah, yes, the reality check.

253
00:20:11,000 --> 00:20:18,000
Right. Setting up a professional agricultural drone crew costs an average of $145,000 just for the equipment and licensing.

254
00:20:18,000 --> 00:20:24,000
Plus, there are variable costs, wages, fuel, insurance, running around $36,000 a month.

255
00:20:24,000 --> 00:20:28,000
$145,000 setup isn't exactly a side hustle for an everyday person.

256
00:20:28,000 --> 00:20:34,000
So, are we just replacing old tech monopolies with new, highly capitalized, crypto monopolies?

257
00:20:34,000 --> 00:20:37,000
That is a very necessary reality check.

258
00:20:37,000 --> 00:20:41,000
The pin is not a magical money tree and we are seeing sector bifurcate.

259
00:20:41,000 --> 00:20:42,000
Bifurcate how?

260
00:20:42,000 --> 00:20:50,000
Well, there is a big difference between a hobbyist running a dash cam on their commute and a professional deploying physical infrastructure at scale.

261
00:20:50,000 --> 00:20:54,000
The ladder requires serious capital expenditure or catbacks.

262
00:20:54,000 --> 00:21:00,000
I mean, hardware, degrades in the rain and dirt, components break, and strict aviation regulations must be followed.

263
00:21:00,000 --> 00:21:03,000
It requires rigorous full-time management.

264
00:21:03,000 --> 00:21:08,000
So, the idea of effortless passive income is a bit of a myth at the professional scale.

265
00:21:08,000 --> 00:21:11,000
For major operations, absolutely. You can't just set it and forget it.

266
00:21:11,000 --> 00:21:16,000
However, the core democratization aspect still remains intact at the protocol level.

267
00:21:16,000 --> 00:21:18,000
How so? If it's still that expensive?

268
00:21:18,000 --> 00:21:30,000
Because $145,000 drone operation is expensive, yes, but it is still fundamentally different from a single billion dollar centralized corporation owning the entire agricultural data layer for the whole planet.

269
00:21:30,000 --> 00:21:33,000
Depend lowers the barrier to entry from billions down to thousands.

270
00:21:33,000 --> 00:21:35,000
Ah, okay, I see.

271
00:21:35,000 --> 00:21:42,000
You still have thousands of independent mid-size operators contributing to and owning a piece of the network.

272
00:21:42,000 --> 00:21:49,000
It creates a highly resilient global infrastructure that no single CEO can just shut down on a whim.

273
00:21:49,000 --> 00:21:56,000
Got it. It's distributing the power and the profits to a much wider base, even if it requires real capital and real work.

274
00:21:56,000 --> 00:21:59,000
Let's synthesize this journey for everyone listening.

275
00:21:59,000 --> 00:22:05,000
The depend revolution of 2026 is democratizing access to the fundamental building blocks of our future.

276
00:22:05,000 --> 00:22:06,000
It really is.

277
00:22:06,000 --> 00:22:16,000
You are turning an idle GPU into an AI training node, letting a smart contract turn your solar panels into a micro-power plant, or feeding real time mapping data to the network.

278
00:22:16,000 --> 00:22:21,000
You are no longer just a passive consumer paying a utility bill. You are becoming the infrastructure itself.

279
00:22:21,000 --> 00:22:25,000
And this raises an important question about the future of global commerce as a whole.

280
00:22:25,000 --> 00:22:34,000
As Depend replaces centralized middlemen with transparent machine readable protocols, we are seeing the rapid acceleration of what economists call "agentic GDP."

281
00:22:34,000 --> 00:22:36,000
"Agentic GDP." That's a great term.

282
00:22:36,000 --> 00:22:48,000
Yeah. This is the economic output generated entirely by autonomous AI agents trading with each other, negotiating contracts, and purchasing resources without any human oversight.

283
00:22:48,000 --> 00:22:57,000
To give you a sense of the scale, in the virtual ecosystem alone, a "gentic GDP" reached $477 million earlier this year.

284
00:22:57,000 --> 00:23:07,000
Now, nearly half a billion dollars of machines doing real business with other machines using physical infrastructure owned and maintained by all of us, it completely upends the economic landscape.

285
00:23:07,000 --> 00:23:08,000
You change everything.

286
00:23:08,000 --> 00:23:13,000
It does. We've spent the last hundred years building infrastructures specifically for humans, right?

287
00:23:13,000 --> 00:23:17,000
Power lines tailored to our houses, internet cables routed to our office buildings.

288
00:23:17,000 --> 00:23:21,000
But as the Depend ecosystem matures, I want to leave you with a final thought to mull over.

289
00:23:21,000 --> 00:23:23,000
Ooh, I like where this is going.

290
00:23:23,000 --> 00:23:37,000
What happens to human jobs and what happens to traditional corporate monopolies when the highest paying, most reliable and most frequent customers for our energy, our internet, and our physical data are actually autonomous machines?

291
00:23:37,000 --> 00:23:47,000
Because suddenly, that untouchable fortress of centralized infrastructure we talked about at the start, it's not just being dismantled, it's being rebuilt by us, and we are renting it out to the robots.

292
00:23:47,000 --> 00:23:51,000
It really forces us to rethink our entire place in the economic hierarchy, doesn't it?

293
00:23:51,000 --> 00:23:52,000
Yeah, it does.

294
00:23:52,000 --> 00:23:54,680
Well thanks for joining us on this Deep Dive until next time.