WEBVTT
1
00:00:00.140 --> 00:00:02.580
I'm Ashton Addison from the Crypto Coin Show. And today
2
00:00:02.660 --> 00:00:06.019
on Blockchain Interviews, we have Yanev Tal, who co-founded The Graph,
3
00:00:06.099 --> 00:00:10.119
the protocol that made blockchain data open and queryable that
4
00:00:10.199 --> 00:00:13.419
thousands of applications depend on. He stepped away from that
5
00:00:13.779 --> 00:00:16.960
to quietly build something new. And that project is GEO,
6
00:00:17.320 --> 00:00:21.160
an open knowledge network. And today they're launching their first product.
7
00:00:21.769 --> 00:00:25.269
Excited to dive into this with GEO Debates. Yanev, thank
8
00:00:25.309 --> 00:00:26.210
you for taking the time.
9
00:00:26.230 --> 00:00:29.050
Thanks, Ashton. It's great to be with you.
10
00:00:30.030 --> 00:00:33.729
Likewise, and I can imagine it's launch day with years
11
00:00:33.770 --> 00:00:36.189
in the making, and we're speaking on the day of.
12
00:00:36.289 --> 00:00:38.450
You must be super busy, and there's a lot to
13
00:00:38.490 --> 00:00:42.859
catch up on. But this is about as new as
14
00:00:42.899 --> 00:00:45.719
it gets, so excited to dive into this. And maybe
15
00:00:45.740 --> 00:00:47.560
we can start with a little bit on the backstory.
16
00:00:48.340 --> 00:00:52.700
The graph is being used by thousands of companies, infrastructures,
17
00:00:52.740 --> 00:00:55.270
and developers, and now you're moving over to more of
18
00:00:55.289 --> 00:00:58.490
a consumer-facing application. Maybe you can start with that story
19
00:00:58.630 --> 00:01:02.039
on you built something incredible and now you yourself have
20
00:01:02.100 --> 00:01:04.019
pivoted to more of the consumer facing apps.
21
00:01:05.200 --> 00:01:09.540
Yeah. So founded the graph back in 2017 and we
22
00:01:09.780 --> 00:01:13.340
were really one of the first infrastructure protocols. I mean,
23
00:01:13.359 --> 00:01:17.180
we kind of defined that category and we really believed
24
00:01:17.299 --> 00:01:20.560
in the vision of web three, right? That that the
25
00:01:20.599 --> 00:01:23.040
web was too concentrated into the hands of a few,
26
00:01:23.060 --> 00:01:25.579
you know, big tech companies and that we need to
27
00:01:25.620 --> 00:01:29.519
make things open, composable, interoperable and, Give more power to
28
00:01:29.599 --> 00:01:32.629
individuals so that you're not locked in. And I still
29
00:01:32.670 --> 00:01:35.250
believe that every bit as much today as I did
30
00:01:35.290 --> 00:01:37.930
back then. And so, you know, the graph was really
31
00:01:37.969 --> 00:01:41.549
successful at an infrastructure level. It is really serving like,
32
00:01:41.870 --> 00:01:44.209
you know, the things that people were building on blockchains.
33
00:01:44.980 --> 00:01:46.799
And a lot of that has been DeFi. You know,
34
00:01:46.819 --> 00:01:49.579
it's a limited set of kind of applications. And for me,
35
00:01:49.719 --> 00:01:50.900
you know, I was looking at this and I'm like,
36
00:01:51.359 --> 00:01:55.019
clearly like Web3 still does not exist in the way
37
00:01:55.079 --> 00:01:58.329
that we always saw it. And, you know, for me,
38
00:01:58.430 --> 00:02:02.010
it's such an important thing to work on because, you know,
39
00:02:02.750 --> 00:02:05.689
we do need to give people more power, not have
40
00:02:05.730 --> 00:02:11.060
it so concentrated. And in today's information environment, it just
41
00:02:11.069 --> 00:02:13.930
feels like it's so hard to know what to trust. Um,
42
00:02:14.289 --> 00:02:19.139
and we need better verifiability and better coordination systems. And so,
43
00:02:19.159 --> 00:02:22.280
you know, to me, the only way to really push
44
00:02:22.300 --> 00:02:25.080
things forward, you know, because I, I didn't see it happening.
45
00:02:25.139 --> 00:02:27.860
I didn't see it working. It was like, we have
46
00:02:27.879 --> 00:02:29.719
to go out and we have to do this ourselves.
47
00:02:31.180 --> 00:02:32.259
Yeah, I agree.
48
00:02:33.000 --> 00:02:36.969
And I love that, that vision because I felt the
49
00:02:37.009 --> 00:02:40.870
same way in 2017, we founded a consumer facing application
50
00:02:41.050 --> 00:02:45.729
for blockchain based ticketing. And it's like there's so many
51
00:02:45.830 --> 00:02:49.189
use cases. Of course, the financial industry, decentralized finance is
52
00:02:49.229 --> 00:02:52.169
super important. And that's an early use case. But blockchain
53
00:02:52.189 --> 00:02:55.870
is going to be touching everywhere where outside of finance,
54
00:02:56.210 --> 00:03:01.349
all of the industries that have data or transactions, entertainment, social.
55
00:03:01.509 --> 00:03:05.229
And I think that the social part especially needs some
56
00:03:05.250 --> 00:03:07.490
catching up on. We need to update and bring the
57
00:03:07.530 --> 00:03:08.370
power back to the people.
58
00:03:09.409 --> 00:03:12.449
A hundred percent. Couldn't agree more. Yeah.
59
00:03:13.330 --> 00:03:18.669
When you pivoted from the graph, did you self-fund this
60
00:03:18.729 --> 00:03:21.650
whole thing? And you've been working heads down sort of
61
00:03:21.969 --> 00:03:23.629
in the bear market. I think we're coming back now.
62
00:03:23.969 --> 00:03:26.389
But throughout the bear market, sort of heads down, just
63
00:03:26.430 --> 00:03:27.490
building this the entire time?
64
00:03:28.689 --> 00:03:32.219
That's right. Really have not taken a break, have not
65
00:03:32.240 --> 00:03:37.000
taken a beat. It's been straight from that in 2022
66
00:03:37.000 --> 00:03:42.620
to self-funding GEO. And so essentially had to start over
67
00:03:42.689 --> 00:03:46.800
with the team, you know, put a team together in place. Fortunately,
68
00:03:46.819 --> 00:03:48.199
you know, there were a lot of people that really,
69
00:03:48.219 --> 00:03:51.580
you know, believed in the vision of Web3 and, you know, what,
70
00:03:51.979 --> 00:03:55.860
you know, I've been trying to build for the last,
71
00:03:55.979 --> 00:03:59.740
I mean, while. So we've got a really great team.
72
00:04:00.000 --> 00:04:02.620
And yeah, it's taken four years. Now, we have actually
73
00:04:02.680 --> 00:04:06.759
released little things along the way. But what we're building
74
00:04:07.080 --> 00:04:11.120
is really a core internet protocol. So you think of
75
00:04:11.159 --> 00:04:14.969
something like HTTP. And so imagine both trying to do
76
00:04:15.310 --> 00:04:18.569
the protocol and a consumer app at the same time.
77
00:04:19.629 --> 00:04:22.310
That's a really tough thing to do because they're almost
78
00:04:22.430 --> 00:04:27.670
diametrically opposed. A protocol's job is to be stable and
79
00:04:27.790 --> 00:04:32.120
perfect and never change. And then a consumer app's job
80
00:04:32.259 --> 00:04:35.600
is to iterate and evolve and meet consumers where they are.
81
00:04:35.639 --> 00:04:39.420
So doing both of those together has been quite the challenge.
82
00:04:40.220 --> 00:04:44.480
And because we're trying to build a foundation for Web3,
83
00:04:45.639 --> 00:04:48.939
you can't scale the thing until it's really ready. And
84
00:04:48.959 --> 00:04:51.399
you don't really know when it's ready until you've just
85
00:04:51.490 --> 00:04:54.889
tried a lot of different things. And so that's really
86
00:04:55.009 --> 00:04:59.939
been the process over the last four years of... Lots
87
00:04:59.980 --> 00:05:03.899
of different use cases, lots of different things that we're
88
00:05:03.939 --> 00:05:07.939
trying to really get the full landscape of what does
89
00:05:08.000 --> 00:05:15.980
it take to build open, composable, interoperable applications that are verifiable.
90
00:05:16.579 --> 00:05:19.620
And there are just so many different layers to that
91
00:05:20.360 --> 00:05:22.540
to finally get to a polished consumer app.
92
00:05:23.740 --> 00:05:24.180
Definitely.
93
00:05:24.220 --> 00:05:27.779
And I'd love to dive into the application more. And
94
00:05:28.199 --> 00:05:30.360
we can get into the protocol later, which is super important.
95
00:05:30.759 --> 00:05:35.019
Bitcoin has been a revolutionary protocol and it would be
96
00:05:35.060 --> 00:05:38.019
great to live up to its standards. But on the
97
00:05:38.300 --> 00:05:42.019
consumer app, on the social side, maybe we can paint
98
00:05:42.040 --> 00:05:46.000
a picture for the problem that Jio is actually solving.
99
00:05:46.259 --> 00:05:49.430
And maybe I just pretense this by saying this is
100
00:05:49.459 --> 00:05:52.889
different than other social applications. But over the years, we've
101
00:05:52.930 --> 00:05:57.660
seen people try to build decentralized Twitter, you know, alternatives
102
00:05:57.740 --> 00:06:01.220
to Web3, to Web2 social media that everyone in the
103
00:06:01.240 --> 00:06:01.740
world is on.
104
00:06:02.019 --> 00:06:04.459
And we can't seem to get it right, or at
105
00:06:04.500 --> 00:06:05.699
least there's no network effect.
106
00:06:05.750 --> 00:06:08.970
So people stay stuck on these apps that we don't
107
00:06:09.009 --> 00:06:11.170
want to use, you know, and also they're consuming our
108
00:06:11.310 --> 00:06:13.389
time and our life. And it's like, we need to
109
00:06:13.430 --> 00:06:16.089
get away from this, but we need to find what's
110
00:06:16.129 --> 00:06:20.079
that magic key that allows us to move into Web3
111
00:06:20.120 --> 00:06:20.839
on the social side.
112
00:06:21.379 --> 00:06:26.569
Yeah, great question. So Geo's a network for organizing knowledge
113
00:06:27.050 --> 00:06:30.449
that we can trust. And you can actually build any
114
00:06:30.490 --> 00:06:34.250
type of application on top of that foundation once you
115
00:06:34.300 --> 00:06:38.319
structure the knowledge in an open, composable, verifiable way. And
116
00:06:38.439 --> 00:06:41.560
so the key insight that I think a lot of
117
00:06:41.600 --> 00:06:44.680
people missed, and it took us a while, but we
118
00:06:44.699 --> 00:06:48.019
always kind of had an inkling, is the trick to
119
00:06:48.060 --> 00:06:51.850
getting Web3 right is in how you structure the information.
120
00:06:52.810 --> 00:06:55.410
And the way that you structure the information is in
121
00:06:55.449 --> 00:06:58.550
a graph, right? Not just any kind of graph, but
122
00:06:59.050 --> 00:07:03.209
a knowledge graph. And so knowledge graphs are the most
123
00:07:03.290 --> 00:07:07.129
flexible way of representing information where you have entities and
124
00:07:07.189 --> 00:07:11.709
relations between entities. And these relationships can be like free form.
125
00:07:12.149 --> 00:07:13.990
And so you don't have to know upfront how everything
126
00:07:14.050 --> 00:07:17.410
is connected, but over time you can create those connections
127
00:07:17.509 --> 00:07:20.500
and that provides better context and understanding. And it's something
128
00:07:20.540 --> 00:07:24.399
that can grow and grow and grow. And the original
129
00:07:24.519 --> 00:07:28.720
architects of the web actually understood that. So Tim Berners-Lee
130
00:07:29.199 --> 00:07:31.610
and those guys were actually trying to build something called
131
00:07:31.670 --> 00:07:35.990
the semantic web, which was this exact idea. But there
132
00:07:36.009 --> 00:07:37.649
were a bunch of things that were missing back then
133
00:07:37.769 --> 00:07:40.490
and some issues in their original implementation that made it
134
00:07:40.529 --> 00:07:43.589
like never really get traction, never really work. But they
135
00:07:43.610 --> 00:07:46.449
understood like this is what the web was supposed to be.
136
00:07:47.279 --> 00:07:50.199
And now we have AI, now we have, you know,
137
00:07:50.300 --> 00:07:53.959
crypto open decentralized networks. And those are actually both of
138
00:07:54.019 --> 00:07:56.399
the things that we were missing that we needed to
139
00:07:56.560 --> 00:08:01.410
actually build this open composable graph of knowledge and information.
140
00:08:01.910 --> 00:08:04.029
And so in, in, in web three, you know, we
141
00:08:04.069 --> 00:08:09.149
saw experiments like far caster, like lens, and, you know,
142
00:08:09.230 --> 00:08:11.449
many of them, they didn't have the success that we
143
00:08:11.529 --> 00:08:15.560
all would have wanted. And I think that, you know,
144
00:08:15.639 --> 00:08:17.920
one of the core reasons was, first of all, they
145
00:08:17.959 --> 00:08:20.740
didn't understand that key insight. So, you know, they were
146
00:08:20.800 --> 00:08:24.490
still building like specific applications, right? They were building like
147
00:08:24.579 --> 00:08:28.370
the social app, right? Where it's just posts and maybe
148
00:08:28.389 --> 00:08:32.070
you can like respond to posts. But that's an application
149
00:08:32.110 --> 00:08:37.649
specific data model versus having an open generalized, you know,
150
00:08:38.169 --> 00:08:42.070
connected graph of any type of thing, any type of application, right?
151
00:08:42.110 --> 00:08:45.070
That's what really creates the network effects. You need that
152
00:08:45.269 --> 00:08:49.179
open graph model. And then, you know, you can build
153
00:08:49.240 --> 00:08:51.740
all kinds of applications and those applications, they need to
154
00:08:51.779 --> 00:08:55.240
be differentiated, right? They can't just be like another Twitter
155
00:08:55.279 --> 00:08:57.620
clone or it's going to be really difficult to compete
156
00:08:57.639 --> 00:09:01.129
against Twitter. So you need the open graph, you know,
157
00:09:01.169 --> 00:09:04.129
knowledge graph data model. And then you need to build
158
00:09:04.490 --> 00:09:06.750
new experiences. And if you do that well, we can
159
00:09:06.789 --> 00:09:10.090
really make this Web3 thing happen. Definitely.
160
00:09:10.850 --> 00:09:11.909
That sounds really exciting.
161
00:09:12.009 --> 00:09:16.580
And I feel like your experience building the graph has
162
00:09:16.700 --> 00:09:23.220
influenced this in building the right architecture for a consumer-facing application,
163
00:09:23.519 --> 00:09:26.240
but also the introduction of AI. And more people are
164
00:09:26.259 --> 00:09:31.340
now connecting using MCPs and APIs to all of their applications.
165
00:09:31.399 --> 00:09:34.419
Everything's sort of interconnected. And this feels like the perfect
166
00:09:34.440 --> 00:09:37.379
time to integrate both of those things into like a
167
00:09:37.419 --> 00:09:38.200
social application.
168
00:09:38.879 --> 00:09:41.039
A hundred percent. So, you know, I think for a moment,
169
00:09:41.419 --> 00:09:44.240
you know, the crypto Web3 people were probably looking at AI,
170
00:09:44.539 --> 00:09:46.820
you know, and say like, damn, you know, those guys
171
00:09:46.840 --> 00:09:49.000
are moving really fast. There's all kinds of interesting things
172
00:09:49.039 --> 00:09:52.269
happening over there. But the reality is I think AI
173
00:09:52.429 --> 00:09:55.070
and crypto are really going to go hand in hand,
174
00:09:55.990 --> 00:09:59.090
and Web3 specifically. And in a lot of ways, the
175
00:09:59.149 --> 00:10:03.049
acceleration of AI, well, not only enables Web3 in a
176
00:10:03.100 --> 00:10:07.379
way that was difficult to do before, but it's really
177
00:10:07.419 --> 00:10:11.980
going to accelerate it because what AI is doing is–
178
00:10:12.759 --> 00:10:16.710
it's giving people these really powerful tools where suddenly they
179
00:10:16.769 --> 00:10:20.029
want to have more control because they have these powerful
180
00:10:20.049 --> 00:10:22.570
tools and they don't want to be limited. So suddenly
181
00:10:22.649 --> 00:10:26.360
now you want your agent to have access, right? To,
182
00:10:27.080 --> 00:10:29.419
all of your data, all of your information, and you
183
00:10:29.460 --> 00:10:31.720
want to have full control. You don't want it to
184
00:10:31.759 --> 00:10:36.750
be locked with random rules and siloed and controlled by
185
00:10:37.120 --> 00:10:38.990
all of these different companies. I mean, you can connect
186
00:10:39.000 --> 00:10:41.190
to them with APIs and MCPs, but it's not the same.
187
00:10:41.830 --> 00:10:44.370
If you have all your data in one place, organize
188
00:10:44.389 --> 00:10:45.870
the way that you want, you can update it, you
189
00:10:45.889 --> 00:10:49.250
have no limits, you're just going to be way more effective.
190
00:10:49.990 --> 00:10:52.769
And so that's what people want. We're also seeing people
191
00:10:52.809 --> 00:10:56.629
like Vibecode Apps And that's so powerful. So, you know,
192
00:10:56.750 --> 00:11:00.500
AI is accelerating this transition away from is like one
193
00:11:00.539 --> 00:11:05.399
size fits all centralized solutions. Now, you know, people want
194
00:11:05.440 --> 00:11:08.440
to do things themselves and they have that power. But
195
00:11:08.600 --> 00:11:12.100
we still need to solve some of the core primitives. Right.
196
00:11:12.139 --> 00:11:15.120
That really enables that. And that's where Web3 comes in. Right.
197
00:11:15.139 --> 00:11:19.029
That's the composability, the interoperability. OK, how do we structure
198
00:11:19.090 --> 00:11:23.220
this stuff? So that you can actually, you know, have
199
00:11:23.379 --> 00:11:30.179
composability and interoperability across applications. And then once you do that, then,
200
00:11:30.210 --> 00:11:33.210
you know, how do you actually, you know, build up
201
00:11:33.350 --> 00:11:37.250
trust in a distributed setting? Right. Because when you have,
202
00:11:37.990 --> 00:11:41.649
you know, these centralized platforms, then everybody can just like
203
00:11:41.769 --> 00:11:44.169
trust those platforms. Right. Twitter is going to take care
204
00:11:44.210 --> 00:11:45.950
of that. Google is going to take care of that.
205
00:11:46.350 --> 00:11:49.070
And then we can trust the platforms and that's it. Well, now, suddenly,
206
00:11:49.529 --> 00:11:51.870
if I'm dealing with a million people that I don't know,
207
00:11:51.889 --> 00:11:55.340
you know, we're just, you know, directly. How do we
208
00:11:55.549 --> 00:11:59.059
build up that trust? So that's exactly what Web3 was
209
00:11:59.100 --> 00:12:01.980
meant to address. And so these two technologies are just
210
00:12:02.000 --> 00:12:03.879
going to supercharge each other. And I think we're going
211
00:12:03.899 --> 00:12:06.980
to see a lot of exciting things. Definitely. I agree.
212
00:12:07.000 --> 00:12:10.440
I'm excited to see the crypto part and the verification
213
00:12:10.460 --> 00:12:15.860
and underlying protocols support that authenticity in social. I feel
214
00:12:16.120 --> 00:12:21.009
like right now in the social platforms, there's two things
215
00:12:21.570 --> 00:12:24.490
that are making it crazy. The first is that everyone's
216
00:12:24.529 --> 00:12:28.610
making AI content, AI videos, and it's getting to the
217
00:12:28.629 --> 00:12:31.649
point where you can't really tell what's real and what's
218
00:12:31.700 --> 00:12:32.100
not real.
219
00:12:32.460 --> 00:12:35.279
And secondly, the company, they don't really.
220
00:12:37.149 --> 00:12:40.230
Don't put enough disclaimers to let people know like, yeah,
221
00:12:40.250 --> 00:12:43.929
you know, people are posting AI or fake news and
222
00:12:43.980 --> 00:12:46.820
it's like, or music and people listening to this music
223
00:12:46.860 --> 00:12:48.919
all day, they don't even know that it's not created
224
00:12:48.960 --> 00:12:49.460
by humans.
225
00:12:50.360 --> 00:12:52.000
And it doesn't really say anywhere.
226
00:12:53.360 --> 00:12:56.879
Yeah. Yeah, that's right. So I think verification, you know,
227
00:12:57.080 --> 00:12:59.860
building a web of trust, right? All of these things,
228
00:13:00.259 --> 00:13:02.779
super important to do right now. Definitely.
229
00:13:03.200 --> 00:13:08.690
And now further on growing, Geodebates, on the debate side,
230
00:13:09.129 --> 00:13:11.929
as the first product in Geo.
231
00:13:11.990 --> 00:13:13.009
Talk to me about that.
232
00:13:13.129 --> 00:13:16.759
There's a lot of debates on the internet, on Facebook comments,
233
00:13:16.840 --> 00:13:19.200
and people going back and forth, and on X especially,
234
00:13:20.000 --> 00:13:22.700
but none of it is really, I don't know if
235
00:13:22.720 --> 00:13:25.620
we're really getting the amount of value out of it
236
00:13:25.679 --> 00:13:26.570
that we should be.
237
00:13:27.870 --> 00:13:31.809
Yeah, so today we launched Geodebates, and it's a new
238
00:13:32.090 --> 00:13:36.730
social app for short-form video debates. And what we want
239
00:13:36.769 --> 00:13:40.590
to do is help elevate the quality of discourse online
240
00:13:41.370 --> 00:13:44.669
and remind people that we're all just people and actually
241
00:13:44.889 --> 00:13:50.629
we should be exploring ideas. And that by the process
242
00:13:50.669 --> 00:13:53.250
of debate, we can learn more about each other. We
243
00:13:53.289 --> 00:13:55.549
can learn more about what's happening in the world and
244
00:13:55.620 --> 00:13:59.940
hopefully get a little bit closer to truth. So we
245
00:13:59.960 --> 00:14:03.529
just launched this today. And we're really excited to see
246
00:14:04.090 --> 00:14:07.049
the types of conversations that people can have on this
247
00:14:07.110 --> 00:14:11.850
new platform. We want to be having discussions about the
248
00:14:11.889 --> 00:14:15.409
biggest issues every day. And hopefully people will be able
249
00:14:15.429 --> 00:14:20.629
to find high-quality conversations and discussions. And then what we
250
00:14:20.669 --> 00:14:23.429
want to do is help structure that in a way
251
00:14:23.899 --> 00:14:27.820
That's better than exists on any of the social platforms today. Right.
252
00:14:27.840 --> 00:14:30.139
Cause you know, most of what we see on social,
253
00:14:30.580 --> 00:14:33.399
it's really just feeds, right? It's a single feed and
254
00:14:33.419 --> 00:14:36.169
it's whoever shouts the loudest, you know, goes the farthest
255
00:14:36.690 --> 00:14:38.649
and you can end up in some, you know, comment,
256
00:14:38.730 --> 00:14:41.289
you know, maybe like rabbit hole thread, but then like that,
257
00:14:41.429 --> 00:14:44.970
that context is really lost. And really what we need
258
00:14:45.009 --> 00:14:47.929
is a way to grow knowledge over time so that
259
00:14:47.980 --> 00:14:50.500
we can move, you know, further and further and build
260
00:14:50.539 --> 00:14:54.389
better and better alignment. So by adding structure to, to
261
00:14:54.429 --> 00:14:58.070
the conversations. We think that's going to enable, you know, well,
262
00:14:58.129 --> 00:15:02.090
first better conversations, but also, you know, us to, to really,
263
00:15:02.330 --> 00:15:04.769
you know, hopefully build some alignment over time.
264
00:15:06.639 --> 00:15:09.500
Yeah, it definitely can get lost and you go down
265
00:15:09.519 --> 00:15:12.480
the rabbit hole and there's never really a, you know,
266
00:15:12.559 --> 00:15:16.500
who won this debate. So with geo debates, how do you,
267
00:15:17.220 --> 00:15:20.120
as a, as a consumer of it, maybe you're not
268
00:15:20.179 --> 00:15:21.159
actually producing the content.
269
00:15:21.879 --> 00:15:23.519
Do you comment to take a side?
270
00:15:23.539 --> 00:15:24.179
Do you vote?
271
00:15:24.299 --> 00:15:27.759
How exactly does that work to contribute to the conversation?
272
00:15:28.820 --> 00:15:32.139
Yeah, a lot of it is really around voting. And
273
00:15:32.240 --> 00:15:36.100
so if you go to Geodebates, you'll see an explore
274
00:15:36.159 --> 00:15:39.899
page where you can see recent debates. We also have
275
00:15:40.799 --> 00:15:43.610
other pieces of content that's part of this knowledge graph.
276
00:15:43.629 --> 00:15:45.789
So you might see some news stories or things that
277
00:15:45.830 --> 00:15:49.809
are happening and individual claims. And you can take a
278
00:15:49.830 --> 00:15:53.730
position on these claims. So, you know, whatever the topic is,
279
00:15:53.750 --> 00:15:55.389
you know, it could be, you know, the war in
280
00:15:55.409 --> 00:16:02.059
the Middle East or in Europe or, you know, local politics, health, AI, right?
281
00:16:02.080 --> 00:16:05.159
There's all of these, you know, topics that are controversial
282
00:16:05.179 --> 00:16:07.240
and people are trying to figure out, like, what are
283
00:16:07.259 --> 00:16:10.129
the facts and, like, what should we do about these
284
00:16:10.169 --> 00:16:13.690
different situations? And so that gets broken down into these
285
00:16:13.789 --> 00:16:18.460
individual claims that you can vote on. And then people
286
00:16:18.500 --> 00:16:21.299
can choose to debate each other on a claim. And
287
00:16:21.340 --> 00:16:23.639
that kicks you into this kind of video, you know,
288
00:16:23.740 --> 00:16:27.090
live debate experience with somebody that you get matched with.
289
00:16:28.250 --> 00:16:31.529
And so you can view those debates. And if you
290
00:16:31.549 --> 00:16:33.450
click on one of those debates, you get this like
291
00:16:33.529 --> 00:16:36.970
TikTok style interface. So you can just keep swiping to
292
00:16:37.009 --> 00:16:40.110
see the conversations that people are having. And if you're interested,
293
00:16:40.149 --> 00:16:41.370
you can jump into one yourself.
294
00:16:43.029 --> 00:16:47.779
That's very interesting. And I'm curious about, When these claims
295
00:16:47.820 --> 00:16:53.429
come up, I feel like if you have better information,
296
00:16:53.470 --> 00:16:56.440
you know something is true. It's great to vote on it,
297
00:16:56.529 --> 00:16:59.950
but there's also, at least with the crypto niche, people
298
00:16:59.970 --> 00:17:01.309
are putting their money where their mouth is.
299
00:17:01.889 --> 00:17:02.830
So is that.
300
00:17:02.809 --> 00:17:06.960
Something that is involved at this stage or at a
301
00:17:07.000 --> 00:17:11.680
later stage where people can actually bet or back their
302
00:17:11.740 --> 00:17:14.480
claims if they believe it to be right to put
303
00:17:14.519 --> 00:17:14.839
money up?
304
00:17:16.130 --> 00:17:21.869
So to me, I think that we should separate matters
305
00:17:21.910 --> 00:17:26.710
of fact and opinion from, for example, let's say predictions.
306
00:17:27.549 --> 00:17:31.950
So prediction markets have obviously been really big over the
307
00:17:31.970 --> 00:17:36.390
last few years, super interesting and exciting. And that is
308
00:17:36.450 --> 00:17:38.900
a place where I think making bets is makes a
309
00:17:38.920 --> 00:17:42.440
lot of sense because, you know, we're talking about something
310
00:17:42.460 --> 00:17:44.940
that nobody can know, right? A prediction is something in
311
00:17:44.980 --> 00:17:48.650
the future and nobody knows for sure. So, you know,
312
00:17:49.029 --> 00:17:51.549
we might as well bet on it. And then at
313
00:17:51.589 --> 00:17:53.710
some point in the future, we'll find out whether or
314
00:17:53.730 --> 00:17:56.509
not that thing was true. It should be easy enough
315
00:17:56.529 --> 00:17:59.710
for us to agree so we can settle the bet.
316
00:18:00.089 --> 00:18:03.130
Perfect use case for betting. Now, matters of fact and
317
00:18:03.210 --> 00:18:07.170
matters of, let's say, taste or preference, I don't think
318
00:18:07.210 --> 00:18:09.880
they work that way. And I don't think you want
319
00:18:09.900 --> 00:18:13.319
to like corrupt that process. Right. Imagine if like in
320
00:18:13.339 --> 00:18:15.759
the scientific process, you know, you could just spend a
321
00:18:15.779 --> 00:18:19.569
little more money and then, you know, your theorem is true. Surprise.
322
00:18:19.670 --> 00:18:22.529
You know, like I don't think that's how truth works.
323
00:18:22.650 --> 00:18:26.069
I don't think that's like how our political system should work. Right.
324
00:18:26.089 --> 00:18:30.960
That's called corruption. when money impacts the political process. So
325
00:18:31.589 --> 00:18:36.589
I think we need better systems of coordination to better
326
00:18:36.630 --> 00:18:40.890
discover truth, to get people aligned. Now, I do think
327
00:18:40.990 --> 00:18:46.230
that money plays a role and that's for incentives because
328
00:18:46.279 --> 00:18:49.359
there is work to be done to organize knowledge and
329
00:18:49.380 --> 00:18:52.819
it benefits a large number of people to have better,
330
00:18:53.380 --> 00:18:57.980
more organized, more trustworthy information is something that benefits a
331
00:18:58.079 --> 00:19:02.119
very large number of people. And so I think it's
332
00:19:02.180 --> 00:19:07.369
worth incentivizing that work, but I would keep those things distinct.
333
00:19:08.170 --> 00:19:10.450
Yeah, that's a great perspective.
334
00:19:11.029 --> 00:19:15.849
And with the debates, right now with AI, it seems
335
00:19:15.910 --> 00:19:18.049
like you ask a question, you get an answer, and
336
00:19:18.089 --> 00:19:20.490
it's like, well, that must be the truth, or at
337
00:19:20.539 --> 00:19:21.670
least I'm trusting AI.
338
00:19:21.980 --> 00:19:23.799
This is the single source of truth that this is
339
00:19:23.819 --> 00:19:24.460
the correct answer.
340
00:19:24.930 --> 00:19:26.670
And maybe that's not always the case that people are
341
00:19:26.690 --> 00:19:30.039
talking about hallucinations or, you know, it just agreeing with
342
00:19:30.079 --> 00:19:30.700
whatever you say.
343
00:19:30.720 --> 00:19:32.700
It doesn't, most of the time it doesn't disagree. If
344
00:19:32.740 --> 00:19:35.339
you say, well, isn't it this, which is wrong. And
345
00:19:35.359 --> 00:19:36.900
I'll say, yeah, yeah. Just to agree with you.
346
00:19:38.759 --> 00:19:43.589
Is that something that won't happen in, in geo debates that,
347
00:19:43.910 --> 00:19:47.049
you know, the hallucinations of AI or sort of agreeing
348
00:19:47.059 --> 00:19:50.920
uh with your personal perspective uh can you avoid.
349
00:19:50.740 --> 00:19:55.109
That yeah so you know i think the hallucinations are
350
00:19:55.140 --> 00:19:58.029
becoming less of a problem you know the models are
351
00:19:58.089 --> 00:20:02.960
getting like quite good at the like intelligence Right. But
352
00:20:03.460 --> 00:20:06.480
you still end up with the problem of, you know,
353
00:20:06.619 --> 00:20:10.190
what is the AI answering the question against? Right. What
354
00:20:10.250 --> 00:20:13.690
is that ground truth? And for that, we still don't
355
00:20:13.730 --> 00:20:16.549
have an answer. Right. So, you know, if you ask
356
00:20:16.569 --> 00:20:18.720
a question, you know, it'll use the tools that are
357
00:20:18.759 --> 00:20:21.940
available to it. Right. It might do some some web search.
358
00:20:22.359 --> 00:20:24.000
It might, you know, look a few things up and
359
00:20:24.039 --> 00:20:26.730
it's like, OK, here's what I found. Here's your answer. Right.
360
00:20:26.750 --> 00:20:28.829
But did it look at the right sources? Were those
361
00:20:28.890 --> 00:20:32.450
sources correct? You know, according to who? So, you know,
362
00:20:32.670 --> 00:20:35.650
you kind of have that problem of ground truth. And
363
00:20:36.250 --> 00:20:38.049
and I think we still don't have a good solution
364
00:20:38.089 --> 00:20:40.930
to that. So that's really is, you know, what Gio
365
00:20:41.150 --> 00:20:44.069
aims to address. You know, we believe that you get
366
00:20:44.569 --> 00:20:48.359
to the truth by having an open system that lots
367
00:20:48.430 --> 00:20:51.759
of people can participate. Right. Where, you know, things are
368
00:20:51.880 --> 00:20:54.380
open to debate and there's challenge and there's all of
369
00:20:54.420 --> 00:20:58.529
this kind of. you know, process and you're compounding and
370
00:20:58.549 --> 00:21:01.369
you're adding context and more people and more knowledge and
371
00:21:01.410 --> 00:21:06.670
more information over time. And through that process, over time,
372
00:21:06.690 --> 00:21:09.609
you can get closer to truth. And so if it's open,
373
00:21:09.710 --> 00:21:12.910
anyone can participate, you can build reputation, and you have
374
00:21:12.970 --> 00:21:16.430
really powerful tools, then, you know, over time, we can
375
00:21:16.529 --> 00:21:20.490
all build up this, you know, verifiable knowledge graph that
376
00:21:20.519 --> 00:21:23.680
becomes more and more, you know, trustworthy and dependable over time.
377
00:21:25.049 --> 00:21:25.450
Definitely.
378
00:21:25.730 --> 00:21:29.650
And when you do search AI, it goes into the
379
00:21:29.670 --> 00:21:32.369
web and it looks what people were debating on Reddit
380
00:21:32.490 --> 00:21:35.059
as an example. And, you know, people can be lying
381
00:21:35.099 --> 00:21:37.140
there and it could pick that up and be like, OK, well,
382
00:21:37.259 --> 00:21:39.480
that had some upvotes, but that doesn't mean it's the truth.
383
00:21:40.440 --> 00:21:41.930
And I guess that's where geo debates comes in.
384
00:21:42.240 --> 00:21:47.079
Right. Exactly. Now, I think another element that we're seeing
385
00:21:47.119 --> 00:21:50.140
these days is, you know, you hear this about this
386
00:21:50.160 --> 00:21:53.799
thing called the dead Internet theory. And I think that's
387
00:21:53.839 --> 00:21:56.960
really accelerated. You know, the fact is, like, you know,
388
00:21:57.460 --> 00:21:59.940
the common threads aren't exactly the same as they used
389
00:21:59.980 --> 00:22:03.980
to be. Right. And maybe some of that is a slopification. Right.
390
00:22:04.019 --> 00:22:06.400
Now you're like, OK, is this even a person? Why
391
00:22:06.519 --> 00:22:11.339
am I wasting my time? But before it was the centralization. Right.
392
00:22:11.420 --> 00:22:15.009
It was like, you know. this is just owned by Facebook. Why, like,
393
00:22:15.049 --> 00:22:18.670
why am I giving Facebook my time? You know, like, and,
394
00:22:18.849 --> 00:22:21.450
and so like, I think there was this time, you know,
395
00:22:21.529 --> 00:22:25.539
in the early internet where people were willing to spend
396
00:22:25.630 --> 00:22:30.400
their time and their creative energy having discussions and like contributing, right.
397
00:22:31.359 --> 00:22:34.660
To this open thing. And the less it became open,
398
00:22:34.779 --> 00:22:37.619
the less people really wanted to like give their creative energy.
399
00:22:38.160 --> 00:22:41.390
And so I think that like, you know, by creating,
400
00:22:41.410 --> 00:22:45.960
you know, an open composable web, it'll be worth people's
401
00:22:46.019 --> 00:22:50.140
time because they know that they can continue to use
402
00:22:50.660 --> 00:22:53.559
anything that they create, right? Any time that I spend
403
00:22:54.000 --> 00:22:58.259
to contribute my knowledge, to contribute my thoughts, to organize whatever,
404
00:22:58.859 --> 00:23:01.000
I'll be able to keep using that. And it, right,
405
00:23:01.059 --> 00:23:06.009
like it is improving my own, like, you know, workflows
406
00:23:06.069 --> 00:23:08.109
or knowledge because, you know, I have access to that.
407
00:23:08.130 --> 00:23:11.650
I'm not giving it away. But Why not make it open?
408
00:23:11.670 --> 00:23:13.930
And then I can also build reputation, right? Other people
409
00:23:13.950 --> 00:23:16.250
can benefit from it too. And so we get back
410
00:23:16.309 --> 00:23:18.029
to this open web.
411
00:23:19.750 --> 00:23:22.970
Talk to me about launch day moving forward in the
412
00:23:23.019 --> 00:23:27.440
first quarter. How can you and the team make Geodebates
413
00:23:28.140 --> 00:23:32.599
grow rapidly? Is it more content, more claims, more people
414
00:23:32.640 --> 00:23:35.720
that want to be content creators and actually put up videos?
415
00:23:36.180 --> 00:23:37.210
How can you scale it?
416
00:23:38.299 --> 00:23:42.240
Yeah. So, I mean, if your audience that's watching, if
417
00:23:42.480 --> 00:23:45.549
we have any content creators in the crowd, I mean,
418
00:23:45.589 --> 00:23:48.349
we would love to see you on Geodebates. I'd love
419
00:23:48.390 --> 00:23:50.650
to get you up there and whatever you're interested in,
420
00:23:51.009 --> 00:23:53.869
you can participate in a debate and we'd love to
421
00:23:53.910 --> 00:23:56.369
see you there. And for everyone else to just check
422
00:23:56.430 --> 00:24:00.750
out the conversations that are happening and to vote. I
423
00:24:00.829 --> 00:24:03.869
think we want to have an impact on the world
424
00:24:03.910 --> 00:24:05.349
we live in, right? I think a lot of people
425
00:24:05.390 --> 00:24:08.369
have this feeling that like, You know, maybe there's less
426
00:24:08.430 --> 00:24:12.009
and less, you know, ways to, you know, exert your
427
00:24:12.109 --> 00:24:14.869
agency to have an influence. Maybe the system's corrupt, like
428
00:24:14.890 --> 00:24:18.529
whatever it is. we got to change that around, right?
429
00:24:19.769 --> 00:24:21.849
We need people to believe in the future and to
430
00:24:21.910 --> 00:24:24.970
know that they can make a change. And, and I
431
00:24:25.009 --> 00:24:27.450
think the open web is going to make that possible. So,
432
00:24:27.849 --> 00:24:30.019
you know, what, what can we do? I mean, we
433
00:24:30.039 --> 00:24:33.859
want to find interesting people in every field. And, you know,
434
00:24:33.920 --> 00:24:35.640
I think what we do is, I mean, it's, you know,
435
00:24:35.660 --> 00:24:38.390
we create a platform. So it's, It's a way for
436
00:24:38.470 --> 00:24:41.109
people that have knowledge or interest in any type of
437
00:24:41.150 --> 00:24:43.710
thing to be able to express their opinions, to meet
438
00:24:43.750 --> 00:24:47.470
other interesting people. So there's a great community aspect, and
439
00:24:47.490 --> 00:24:51.660
we're giving people powerful tools. And my hope is that
440
00:24:51.720 --> 00:24:55.359
we're going to see really great coverage in a lot
441
00:24:55.400 --> 00:24:58.220
of the areas that are really kind of impacting people.
442
00:24:58.799 --> 00:25:02.900
And I'd love to help bring focus to some of
443
00:25:02.920 --> 00:25:06.710
the most important debates or topics and just get more
444
00:25:06.730 --> 00:25:08.920
people involved. Definitely.
445
00:25:08.960 --> 00:25:11.019
I would love to see that too. I'll do my
446
00:25:11.059 --> 00:25:13.940
best to be one of those content creators. What's the
447
00:25:13.980 --> 00:25:15.920
best way to access the, you said it's on a
448
00:25:15.940 --> 00:25:18.700
web app. How do you access it and how can
449
00:25:18.740 --> 00:25:21.299
you make it convenient for non-technical people?
450
00:25:22.259 --> 00:25:25.759
Yeah. So geo debates is available today. You can just
451
00:25:25.779 --> 00:25:29.680
go to geo browser.io and hit explore. And right away,
452
00:25:29.700 --> 00:25:31.420
you're going to start seeing, you know, some of the
453
00:25:31.460 --> 00:25:34.299
top debates that are coming out and you know, more
454
00:25:34.319 --> 00:25:38.750
debates every day. Um, So that's available now. And, you know,
455
00:25:38.789 --> 00:25:41.029
we really made it so that you don't have to
456
00:25:41.069 --> 00:25:44.410
be technical to use it at all. It is, you know,
457
00:25:44.470 --> 00:25:50.670
a crypto application. It's on the Ethereum blockchain on a
458
00:25:50.690 --> 00:25:55.210
layer three. And so it's all open and composable and verifiable. But,
459
00:25:55.319 --> 00:25:58.000
you know, the users don't really have to know that, right?
460
00:25:58.279 --> 00:26:02.299
They're creating an account using Privy. It's all abstracted. And,
461
00:26:02.799 --> 00:26:04.839
you know, it's just like any social app. You upvote,
462
00:26:04.880 --> 00:26:08.019
you downvote. It's super easy to use. Awesome.
463
00:26:08.119 --> 00:26:08.680
Sounds great.
464
00:26:08.779 --> 00:26:12.380
And I'm always interested in the technical background. But as
465
00:26:12.440 --> 00:26:15.950
we know, for blockchain to be adopted in the mainstream,
466
00:26:16.430 --> 00:26:18.450
the end users can't know they're using blockchain.
467
00:26:19.210 --> 00:26:19.710
Exactly.
468
00:26:21.289 --> 00:26:24.230
So all the best with this. I'm looking forward to
469
00:26:24.670 --> 00:26:27.890
creating some content, doing some votes. We'll leave a link
470
00:26:27.990 --> 00:26:32.019
to Geodebates in the show notes. Would love to follow up,
471
00:26:32.700 --> 00:26:34.220
if not before the end of this year.
472
00:26:34.670 --> 00:26:35.950
To see the.
473
00:26:35.890 --> 00:26:38.329
Progress and wishing you and the team all the best.
474
00:26:38.390 --> 00:26:39.819
And let's follow up in the near future.
475
00:26:40.660 --> 00:26:43.039
Thank you so much, Afton. Yeah, I really enjoyed the conversation.
476
00:26:43.079 --> 00:26:45.839
I'm looking forward to chatting more. Hopefully seeing you over
477
00:26:45.900 --> 00:26:46.779
in the Geoverse.
1
00:00:00.140 --> 00:00:02.580
I'm Ashton Addison from the Crypto Coin Show. And today
2
00:00:02.660 --> 00:00:06.019
on Blockchain Interviews, we have Yanev Tal, who co-founded The Graph,
3
00:00:06.099 --> 00:00:10.119
the protocol that made blockchain data open and queryable that
4
00:00:10.199 --> 00:00:13.419
thousands of applications depend on. He stepped away from that
5
00:00:13.779 --> 00:00:16.960
to quietly build something new. And that project is GEO,
6
00:00:17.320 --> 00:00:21.160
an open knowledge network. And today they're launching their first product.
7
00:00:21.769 --> 00:00:25.269
Excited to dive into this with GEO Debates. Yanev, thank
8
00:00:25.309 --> 00:00:26.210
you for taking the time.
9
00:00:26.230 --> 00:00:29.050
Thanks, Ashton. It's great to be with you.
10
00:00:30.030 --> 00:00:33.729
Likewise, and I can imagine it's launch day with years
11
00:00:33.770 --> 00:00:36.189
in the making, and we're speaking on the day of.
12
00:00:36.289 --> 00:00:38.450
You must be super busy, and there's a lot to
13
00:00:38.490 --> 00:00:42.859
catch up on. But this is about as new as
14
00:00:42.899 --> 00:00:45.719
it gets, so excited to dive into this. And maybe
15
00:00:45.740 --> 00:00:47.560
we can start with a little bit on the backstory.
16
00:00:48.340 --> 00:00:52.700
The graph is being used by thousands of companies, infrastructures,
17
00:00:52.740 --> 00:00:55.270
and developers, and now you're moving over to more of
18
00:00:55.289 --> 00:00:58.490
a consumer-facing application. Maybe you can start with that story
19
00:00:58.630 --> 00:01:02.039
on you built something incredible and now you yourself have
20
00:01:02.100 --> 00:01:04.019
pivoted to more of the consumer facing apps.
21
00:01:05.200 --> 00:01:09.540
Yeah. So founded the graph back in 2017 and we
22
00:01:09.780 --> 00:01:13.340
were really one of the first infrastructure protocols. I mean,
23
00:01:13.359 --> 00:01:17.180
we kind of defined that category and we really believed
24
00:01:17.299 --> 00:01:20.560
in the vision of web three, right? That that the
25
00:01:20.599 --> 00:01:23.040
web was too concentrated into the hands of a few,
26
00:01:23.060 --> 00:01:25.579
you know, big tech companies and that we need to
27
00:01:25.620 --> 00:01:29.519
make things open, composable, interoperable and, Give more power to
28
00:01:29.599 --> 00:01:32.629
individuals so that you're not locked in. And I still
29
00:01:32.670 --> 00:01:35.250
believe that every bit as much today as I did
30
00:01:35.290 --> 00:01:37.930
back then. And so, you know, the graph was really
31
00:01:37.969 --> 00:01:41.549
successful at an infrastructure level. It is really serving like,
32
00:01:41.870 --> 00:01:44.209
you know, the things that people were building on blockchains.
33
00:01:44.980 --> 00:01:46.799
And a lot of that has been DeFi. You know,
34
00:01:46.819 --> 00:01:49.579
it's a limited set of kind of applications. And for me,
35
00:01:49.719 --> 00:01:50.900
you know, I was looking at this and I'm like,
36
00:01:51.359 --> 00:01:55.019
clearly like Web3 still does not exist in the way
37
00:01:55.079 --> 00:01:58.329
that we always saw it. And, you know, for me,
38
00:01:58.430 --> 00:02:02.010
it's such an important thing to work on because, you know,
39
00:02:02.750 --> 00:02:05.689
we do need to give people more power, not have
40
00:02:05.730 --> 00:02:11.060
it so concentrated. And in today's information environment, it just
41
00:02:11.069 --> 00:02:13.930
feels like it's so hard to know what to trust. Um,
42
00:02:14.289 --> 00:02:19.139
and we need better verifiability and better coordination systems. And so,
43
00:02:19.159 --> 00:02:22.280
you know, to me, the only way to really push
44
00:02:22.300 --> 00:02:25.080
things forward, you know, because I, I didn't see it happening.
45
00:02:25.139 --> 00:02:27.860
I didn't see it working. It was like, we have
46
00:02:27.879 --> 00:02:29.719
to go out and we have to do this ourselves.
47
00:02:31.180 --> 00:02:32.259
Yeah, I agree.
48
00:02:33.000 --> 00:02:36.969
And I love that, that vision because I felt the
49
00:02:37.009 --> 00:02:40.870
same way in 2017, we founded a consumer facing application
50
00:02:41.050 --> 00:02:45.729
for blockchain based ticketing. And it's like there's so many
51
00:02:45.830 --> 00:02:49.189
use cases. Of course, the financial industry, decentralized finance is
52
00:02:49.229 --> 00:02:52.169
super important. And that's an early use case. But blockchain
53
00:02:52.189 --> 00:02:55.870
is going to be touching everywhere where outside of finance,
54
00:02:56.210 --> 00:03:01.349
all of the industries that have data or transactions, entertainment, social.
55
00:03:01.509 --> 00:03:05.229
And I think that the social part especially needs some
56
00:03:05.250 --> 00:03:07.490
catching up on. We need to update and bring the
57
00:03:07.530 --> 00:03:08.370
power back to the people.
58
00:03:09.409 --> 00:03:12.449
A hundred percent. Couldn't agree more. Yeah.
59
00:03:13.330 --> 00:03:18.669
When you pivoted from the graph, did you self-fund this
60
00:03:18.729 --> 00:03:21.650
whole thing? And you've been working heads down sort of
61
00:03:21.969 --> 00:03:23.629
in the bear market. I think we're coming back now.
62
00:03:23.969 --> 00:03:26.389
But throughout the bear market, sort of heads down, just
63
00:03:26.430 --> 00:03:27.490
building this the entire time?
64
00:03:28.689 --> 00:03:32.219
That's right. Really have not taken a break, have not
65
00:03:32.240 --> 00:03:37.000
taken a beat. It's been straight from that in 2022
66
00:03:37.000 --> 00:03:42.620
to self-funding GEO. And so essentially had to start over
67
00:03:42.689 --> 00:03:46.800
with the team, you know, put a team together in place. Fortunately,
68
00:03:46.819 --> 00:03:48.199
you know, there were a lot of people that really,
69
00:03:48.219 --> 00:03:51.580
you know, believed in the vision of Web3 and, you know, what,
70
00:03:51.979 --> 00:03:55.860
you know, I've been trying to build for the last,
71
00:03:55.979 --> 00:03:59.740
I mean, while. So we've got a really great team.
72
00:04:00.000 --> 00:04:02.620
And yeah, it's taken four years. Now, we have actually
73
00:04:02.680 --> 00:04:06.759
released little things along the way. But what we're building
74
00:04:07.080 --> 00:04:11.120
is really a core internet protocol. So you think of
75
00:04:11.159 --> 00:04:14.969
something like HTTP. And so imagine both trying to do
76
00:04:15.310 --> 00:04:18.569
the protocol and a consumer app at the same time.
77
00:04:19.629 --> 00:04:22.310
That's a really tough thing to do because they're almost
78
00:04:22.430 --> 00:04:27.670
diametrically opposed. A protocol's job is to be stable and
79
00:04:27.790 --> 00:04:32.120
perfect and never change. And then a consumer app's job
80
00:04:32.259 --> 00:04:35.600
is to iterate and evolve and meet consumers where they are.
81
00:04:35.639 --> 00:04:39.420
So doing both of those together has been quite the challenge.
82
00:04:40.220 --> 00:04:44.480
And because we're trying to build a foundation for Web3,
83
00:04:45.639 --> 00:04:48.939
you can't scale the thing until it's really ready. And
84
00:04:48.959 --> 00:04:51.399
you don't really know when it's ready until you've just
85
00:04:51.490 --> 00:04:54.889
tried a lot of different things. And so that's really
86
00:04:55.009 --> 00:04:59.939
been the process over the last four years of... Lots
87
00:04:59.980 --> 00:05:03.899
of different use cases, lots of different things that we're
88
00:05:03.939 --> 00:05:07.939
trying to really get the full landscape of what does
89
00:05:08.000 --> 00:05:15.980
it take to build open, composable, interoperable applications that are verifiable.
90
00:05:16.579 --> 00:05:19.620
And there are just so many different layers to that
91
00:05:20.360 --> 00:05:22.540
to finally get to a polished consumer app.
92
00:05:23.740 --> 00:05:24.180
Definitely.
93
00:05:24.220 --> 00:05:27.779
And I'd love to dive into the application more. And
94
00:05:28.199 --> 00:05:30.360
we can get into the protocol later, which is super important.
95
00:05:30.759 --> 00:05:35.019
Bitcoin has been a revolutionary protocol and it would be
96
00:05:35.060 --> 00:05:38.019
great to live up to its standards. But on the
97
00:05:38.300 --> 00:05:42.019
consumer app, on the social side, maybe we can paint
98
00:05:42.040 --> 00:05:46.000
a picture for the problem that Jio is actually solving.
99
00:05:46.259 --> 00:05:49.430
And maybe I just pretense this by saying this is
100
00:05:49.459 --> 00:05:52.889
different than other social applications. But over the years, we've
101
00:05:52.930 --> 00:05:57.660
seen people try to build decentralized Twitter, you know, alternatives
102
00:05:57.740 --> 00:06:01.220
to Web3, to Web2 social media that everyone in the
103
00:06:01.240 --> 00:06:01.740
world is on.
104
00:06:02.019 --> 00:06:04.459
And we can't seem to get it right, or at
105
00:06:04.500 --> 00:06:05.699
least there's no network effect.
106
00:06:05.750 --> 00:06:08.970
So people stay stuck on these apps that we don't
107
00:06:09.009 --> 00:06:11.170
want to use, you know, and also they're consuming our
108
00:06:11.310 --> 00:06:13.389
time and our life. And it's like, we need to
109
00:06:13.430 --> 00:06:16.089
get away from this, but we need to find what's
110
00:06:16.129 --> 00:06:20.079
that magic key that allows us to move into Web3
111
00:06:20.120 --> 00:06:20.839
on the social side.
112
00:06:21.379 --> 00:06:26.569
Yeah, great question. So Geo's a network for organizing knowledge
113
00:06:27.050 --> 00:06:30.449
that we can trust. And you can actually build any
114
00:06:30.490 --> 00:06:34.250
type of application on top of that foundation once you
115
00:06:34.300 --> 00:06:38.319
structure the knowledge in an open, composable, verifiable way. And
116
00:06:38.439 --> 00:06:41.560
so the key insight that I think a lot of
117
00:06:41.600 --> 00:06:44.680
people missed, and it took us a while, but we
118
00:06:44.699 --> 00:06:48.019
always kind of had an inkling, is the trick to
119
00:06:48.060 --> 00:06:51.850
getting Web3 right is in how you structure the information.
120
00:06:52.810 --> 00:06:55.410
And the way that you structure the information is in
121
00:06:55.449 --> 00:06:58.550
a graph, right? Not just any kind of graph, but
122
00:06:59.050 --> 00:07:03.209
a knowledge graph. And so knowledge graphs are the most
123
00:07:03.290 --> 00:07:07.129
flexible way of representing information where you have entities and
124
00:07:07.189 --> 00:07:11.709
relations between entities. And these relationships can be like free form.
125
00:07:12.149 --> 00:07:13.990
And so you don't have to know upfront how everything
126
00:07:14.050 --> 00:07:17.410
is connected, but over time you can create those connections
127
00:07:17.509 --> 00:07:20.500
and that provides better context and understanding. And it's something
128
00:07:20.540 --> 00:07:24.399
that can grow and grow and grow. And the original
129
00:07:24.519 --> 00:07:28.720
architects of the web actually understood that. So Tim Berners-Lee
130
00:07:29.199 --> 00:07:31.610
and those guys were actually trying to build something called
131
00:07:31.670 --> 00:07:35.990
the semantic web, which was this exact idea. But there
132
00:07:36.009 --> 00:07:37.649
were a bunch of things that were missing back then
133
00:07:37.769 --> 00:07:40.490
and some issues in their original implementation that made it
134
00:07:40.529 --> 00:07:43.589
like never really get traction, never really work. But they
135
00:07:43.610 --> 00:07:46.449
understood like this is what the web was supposed to be.
136
00:07:47.279 --> 00:07:50.199
And now we have AI, now we have, you know,
137
00:07:50.300 --> 00:07:53.959
crypto open decentralized networks. And those are actually both of
138
00:07:54.019 --> 00:07:56.399
the things that we were missing that we needed to
139
00:07:56.560 --> 00:08:01.410
actually build this open composable graph of knowledge and information.
140
00:08:01.910 --> 00:08:04.029
And so in, in, in web three, you know, we
141
00:08:04.069 --> 00:08:09.149
saw experiments like far caster, like lens, and, you know,
142
00:08:09.230 --> 00:08:11.449
many of them, they didn't have the success that we
143
00:08:11.529 --> 00:08:15.560
all would have wanted. And I think that, you know,
144
00:08:15.639 --> 00:08:17.920
one of the core reasons was, first of all, they
145
00:08:17.959 --> 00:08:20.740
didn't understand that key insight. So, you know, they were
146
00:08:20.800 --> 00:08:24.490
still building like specific applications, right? They were building like
147
00:08:24.579 --> 00:08:28.370
the social app, right? Where it's just posts and maybe
148
00:08:28.389 --> 00:08:32.070
you can like respond to posts. But that's an application
149
00:08:32.110 --> 00:08:37.649
specific data model versus having an open generalized, you know,
150
00:08:38.169 --> 00:08:42.070
connected graph of any type of thing, any type of application, right?
151
00:08:42.110 --> 00:08:45.070
That's what really creates the network effects. You need that
152
00:08:45.269 --> 00:08:49.179
open graph model. And then, you know, you can build
153
00:08:49.240 --> 00:08:51.740
all kinds of applications and those applications, they need to
154
00:08:51.779 --> 00:08:55.240
be differentiated, right? They can't just be like another Twitter
155
00:08:55.279 --> 00:08:57.620
clone or it's going to be really difficult to compete
156
00:08:57.639 --> 00:09:01.129
against Twitter. So you need the open graph, you know,
157
00:09:01.169 --> 00:09:04.129
knowledge graph data model. And then you need to build
158
00:09:04.490 --> 00:09:06.750
new experiences. And if you do that well, we can
159
00:09:06.789 --> 00:09:10.090
really make this Web3 thing happen. Definitely.
160
00:09:10.850 --> 00:09:11.909
That sounds really exciting.
161
00:09:12.009 --> 00:09:16.580
And I feel like your experience building the graph has
162
00:09:16.700 --> 00:09:23.220
influenced this in building the right architecture for a consumer-facing application,
163
00:09:23.519 --> 00:09:26.240
but also the introduction of AI. And more people are
164
00:09:26.259 --> 00:09:31.340
now connecting using MCPs and APIs to all of their applications.
165
00:09:31.399 --> 00:09:34.419
Everything's sort of interconnected. And this feels like the perfect
166
00:09:34.440 --> 00:09:37.379
time to integrate both of those things into like a
167
00:09:37.419 --> 00:09:38.200
social application.
168
00:09:38.879 --> 00:09:41.039
A hundred percent. So, you know, I think for a moment,
169
00:09:41.419 --> 00:09:44.240
you know, the crypto Web3 people were probably looking at AI,
170
00:09:44.539 --> 00:09:46.820
you know, and say like, damn, you know, those guys
171
00:09:46.840 --> 00:09:49.000
are moving really fast. There's all kinds of interesting things
172
00:09:49.039 --> 00:09:52.269
happening over there. But the reality is I think AI
173
00:09:52.429 --> 00:09:55.070
and crypto are really going to go hand in hand,
174
00:09:55.990 --> 00:09:59.090
and Web3 specifically. And in a lot of ways, the
175
00:09:59.149 --> 00:10:03.049
acceleration of AI, well, not only enables Web3 in a
176
00:10:03.100 --> 00:10:07.379
way that was difficult to do before, but it's really
177
00:10:07.419 --> 00:10:11.980
going to accelerate it because what AI is doing is–
178
00:10:12.759 --> 00:10:16.710
it's giving people these really powerful tools where suddenly they
179
00:10:16.769 --> 00:10:20.029
want to have more control because they have these powerful
180
00:10:20.049 --> 00:10:22.570
tools and they don't want to be limited. So suddenly
181
00:10:22.649 --> 00:10:26.360
now you want your agent to have access, right? To,
182
00:10:27.080 --> 00:10:29.419
all of your data, all of your information, and you
183
00:10:29.460 --> 00:10:31.720
want to have full control. You don't want it to
184
00:10:31.759 --> 00:10:36.750
be locked with random rules and siloed and controlled by
185
00:10:37.120 --> 00:10:38.990
all of these different companies. I mean, you can connect
186
00:10:39.000 --> 00:10:41.190
to them with APIs and MCPs, but it's not the same.
187
00:10:41.830 --> 00:10:44.370
If you have all your data in one place, organize
188
00:10:44.389 --> 00:10:45.870
the way that you want, you can update it, you
189
00:10:45.889 --> 00:10:49.250
have no limits, you're just going to be way more effective.
190
00:10:49.990 --> 00:10:52.769
And so that's what people want. We're also seeing people
191
00:10:52.809 --> 00:10:56.629
like Vibecode Apps And that's so powerful. So, you know,
192
00:10:56.750 --> 00:11:00.500
AI is accelerating this transition away from is like one
193
00:11:00.539 --> 00:11:05.399
size fits all centralized solutions. Now, you know, people want
194
00:11:05.440 --> 00:11:08.440
to do things themselves and they have that power. But
195
00:11:08.600 --> 00:11:12.100
we still need to solve some of the core primitives. Right.
196
00:11:12.139 --> 00:11:15.120
That really enables that. And that's where Web3 comes in. Right.
197
00:11:15.139 --> 00:11:19.029
That's the composability, the interoperability. OK, how do we structure
198
00:11:19.090 --> 00:11:23.220
this stuff? So that you can actually, you know, have
199
00:11:23.379 --> 00:11:30.179
composability and interoperability across applications. And then once you do that, then,
200
00:11:30.210 --> 00:11:33.210
you know, how do you actually, you know, build up
201
00:11:33.350 --> 00:11:37.250
trust in a distributed setting? Right. Because when you have,
202
00:11:37.990 --> 00:11:41.649
you know, these centralized platforms, then everybody can just like
203
00:11:41.769 --> 00:11:44.169
trust those platforms. Right. Twitter is going to take care
204
00:11:44.210 --> 00:11:45.950
of that. Google is going to take care of that.
205
00:11:46.350 --> 00:11:49.070
And then we can trust the platforms and that's it. Well, now, suddenly,
206
00:11:49.529 --> 00:11:51.870
if I'm dealing with a million people that I don't know,
207
00:11:51.889 --> 00:11:55.340
you know, we're just, you know, directly. How do we
208
00:11:55.549 --> 00:11:59.059
build up that trust? So that's exactly what Web3 was
209
00:11:59.100 --> 00:12:01.980
meant to address. And so these two technologies are just
210
00:12:02.000 --> 00:12:03.879
going to supercharge each other. And I think we're going
211
00:12:03.899 --> 00:12:06.980
to see a lot of exciting things. Definitely. I agree.
212
00:12:07.000 --> 00:12:10.440
I'm excited to see the crypto part and the verification
213
00:12:10.460 --> 00:12:15.860
and underlying protocols support that authenticity in social. I feel
214
00:12:16.120 --> 00:12:21.009
like right now in the social platforms, there's two things
215
00:12:21.570 --> 00:12:24.490
that are making it crazy. The first is that everyone's
216
00:12:24.529 --> 00:12:28.610
making AI content, AI videos, and it's getting to the
217
00:12:28.629 --> 00:12:31.649
point where you can't really tell what's real and what's
218
00:12:31.700 --> 00:12:32.100
not real.
219
00:12:32.460 --> 00:12:35.279
And secondly, the company, they don't really.
220
00:12:37.149 --> 00:12:40.230
Don't put enough disclaimers to let people know like, yeah,
221
00:12:40.250 --> 00:12:43.929
you know, people are posting AI or fake news and
222
00:12:43.980 --> 00:12:46.820
it's like, or music and people listening to this music
223
00:12:46.860 --> 00:12:48.919
all day, they don't even know that it's not created
224
00:12:48.960 --> 00:12:49.460
by humans.
225
00:12:50.360 --> 00:12:52.000
And it doesn't really say anywhere.
226
00:12:53.360 --> 00:12:56.879
Yeah. Yeah, that's right. So I think verification, you know,
227
00:12:57.080 --> 00:12:59.860
building a web of trust, right? All of these things,
228
00:13:00.259 --> 00:13:02.779
super important to do right now. Definitely.
229
00:13:03.200 --> 00:13:08.690
And now further on growing, Geodebates, on the debate side,
230
00:13:09.129 --> 00:13:11.929
as the first product in Geo.
231
00:13:11.990 --> 00:13:13.009
Talk to me about that.
232
00:13:13.129 --> 00:13:16.759
There's a lot of debates on the internet, on Facebook comments,
233
00:13:16.840 --> 00:13:19.200
and people going back and forth, and on X especially,
234
00:13:20.000 --> 00:13:22.700
but none of it is really, I don't know if
235
00:13:22.720 --> 00:13:25.620
we're really getting the amount of value out of it
236
00:13:25.679 --> 00:13:26.570
that we should be.
237
00:13:27.870 --> 00:13:31.809
Yeah, so today we launched Geodebates, and it's a new
238
00:13:32.090 --> 00:13:36.730
social app for short-form video debates. And what we want
239
00:13:36.769 --> 00:13:40.590
to do is help elevate the quality of discourse online
240
00:13:41.370 --> 00:13:44.669
and remind people that we're all just people and actually
241
00:13:44.889 --> 00:13:50.629
we should be exploring ideas. And that by the process
242
00:13:50.669 --> 00:13:53.250
of debate, we can learn more about each other. We
243
00:13:53.289 --> 00:13:55.549
can learn more about what's happening in the world and
244
00:13:55.620 --> 00:13:59.940
hopefully get a little bit closer to truth. So we
245
00:13:59.960 --> 00:14:03.529
just launched this today. And we're really excited to see
246
00:14:04.090 --> 00:14:07.049
the types of conversations that people can have on this
247
00:14:07.110 --> 00:14:11.850
new platform. We want to be having discussions about the
248
00:14:11.889 --> 00:14:15.409
biggest issues every day. And hopefully people will be able
249
00:14:15.429 --> 00:14:20.629
to find high-quality conversations and discussions. And then what we
250
00:14:20.669 --> 00:14:23.429
want to do is help structure that in a way
251
00:14:23.899 --> 00:14:27.820
That's better than exists on any of the social platforms today. Right.
252
00:14:27.840 --> 00:14:30.139
Cause you know, most of what we see on social,
253
00:14:30.580 --> 00:14:33.399
it's really just feeds, right? It's a single feed and
254
00:14:33.419 --> 00:14:36.169
it's whoever shouts the loudest, you know, goes the farthest
255
00:14:36.690 --> 00:14:38.649
and you can end up in some, you know, comment,
256
00:14:38.730 --> 00:14:41.289
you know, maybe like rabbit hole thread, but then like that,
257
00:14:41.429 --> 00:14:44.970
that context is really lost. And really what we need
258
00:14:45.009 --> 00:14:47.929
is a way to grow knowledge over time so that
259
00:14:47.980 --> 00:14:50.500
we can move, you know, further and further and build
260
00:14:50.539 --> 00:14:54.389
better and better alignment. So by adding structure to, to
261
00:14:54.429 --> 00:14:58.070
the conversations. We think that's going to enable, you know, well,
262
00:14:58.129 --> 00:15:02.090
first better conversations, but also, you know, us to, to really,
263
00:15:02.330 --> 00:15:04.769
you know, hopefully build some alignment over time.
264
00:15:06.639 --> 00:15:09.500
Yeah, it definitely can get lost and you go down
265
00:15:09.519 --> 00:15:12.480
the rabbit hole and there's never really a, you know,
266
00:15:12.559 --> 00:15:16.500
who won this debate. So with geo debates, how do you,
267
00:15:17.220 --> 00:15:20.120
as a, as a consumer of it, maybe you're not
268
00:15:20.179 --> 00:15:21.159
actually producing the content.
269
00:15:21.879 --> 00:15:23.519
Do you comment to take a side?
270
00:15:23.539 --> 00:15:24.179
Do you vote?
271
00:15:24.299 --> 00:15:27.759
How exactly does that work to contribute to the conversation?
272
00:15:28.820 --> 00:15:32.139
Yeah, a lot of it is really around voting. And
273
00:15:32.240 --> 00:15:36.100
so if you go to Geodebates, you'll see an explore
274
00:15:36.159 --> 00:15:39.899
page where you can see recent debates. We also have
275
00:15:40.799 --> 00:15:43.610
other pieces of content that's part of this knowledge graph.
276
00:15:43.629 --> 00:15:45.789
So you might see some news stories or things that
277
00:15:45.830 --> 00:15:49.809
are happening and individual claims. And you can take a
278
00:15:49.830 --> 00:15:53.730
position on these claims. So, you know, whatever the topic is,
279
00:15:53.750 --> 00:15:55.389
you know, it could be, you know, the war in
280
00:15:55.409 --> 00:16:02.059
the Middle East or in Europe or, you know, local politics, health, AI, right?
281
00:16:02.080 --> 00:16:05.159
There's all of these, you know, topics that are controversial
282
00:16:05.179 --> 00:16:07.240
and people are trying to figure out, like, what are
283
00:16:07.259 --> 00:16:10.129
the facts and, like, what should we do about these
284
00:16:10.169 --> 00:16:13.690
different situations? And so that gets broken down into these
285
00:16:13.789 --> 00:16:18.460
individual claims that you can vote on. And then people
286
00:16:18.500 --> 00:16:21.299
can choose to debate each other on a claim. And
287
00:16:21.340 --> 00:16:23.639
that kicks you into this kind of video, you know,
288
00:16:23.740 --> 00:16:27.090
live debate experience with somebody that you get matched with.
289
00:16:28.250 --> 00:16:31.529
And so you can view those debates. And if you
290
00:16:31.549 --> 00:16:33.450
click on one of those debates, you get this like
291
00:16:33.529 --> 00:16:36.970
TikTok style interface. So you can just keep swiping to
292
00:16:37.009 --> 00:16:40.110
see the conversations that people are having. And if you're interested,
293
00:16:40.149 --> 00:16:41.370
you can jump into one yourself.
294
00:16:43.029 --> 00:16:47.779
That's very interesting. And I'm curious about, When these claims
295
00:16:47.820 --> 00:16:53.429
come up, I feel like if you have better information,
296
00:16:53.470 --> 00:16:56.440
you know something is true. It's great to vote on it,
297
00:16:56.529 --> 00:16:59.950
but there's also, at least with the crypto niche, people
298
00:16:59.970 --> 00:17:01.309
are putting their money where their mouth is.
299
00:17:01.889 --> 00:17:02.830
So is that.
300
00:17:02.809 --> 00:17:06.960
Something that is involved at this stage or at a
301
00:17:07.000 --> 00:17:11.680
later stage where people can actually bet or back their
302
00:17:11.740 --> 00:17:14.480
claims if they believe it to be right to put
303
00:17:14.519 --> 00:17:14.839
money up?
304
00:17:16.130 --> 00:17:21.869
So to me, I think that we should separate matters
305
00:17:21.910 --> 00:17:26.710
of fact and opinion from, for example, let's say predictions.
306
00:17:27.549 --> 00:17:31.950
So prediction markets have obviously been really big over the
307
00:17:31.970 --> 00:17:36.390
last few years, super interesting and exciting. And that is
308
00:17:36.450 --> 00:17:38.900
a place where I think making bets is makes a
309
00:17:38.920 --> 00:17:42.440
lot of sense because, you know, we're talking about something
310
00:17:42.460 --> 00:17:44.940
that nobody can know, right? A prediction is something in
311
00:17:44.980 --> 00:17:48.650
the future and nobody knows for sure. So, you know,
312
00:17:49.029 --> 00:17:51.549
we might as well bet on it. And then at
313
00:17:51.589 --> 00:17:53.710
some point in the future, we'll find out whether or
314
00:17:53.730 --> 00:17:56.509
not that thing was true. It should be easy enough
315
00:17:56.529 --> 00:17:59.710
for us to agree so we can settle the bet.
316
00:18:00.089 --> 00:18:03.130
Perfect use case for betting. Now, matters of fact and
317
00:18:03.210 --> 00:18:07.170
matters of, let's say, taste or preference, I don't think
318
00:18:07.210 --> 00:18:09.880
they work that way. And I don't think you want
319
00:18:09.900 --> 00:18:13.319
to like corrupt that process. Right. Imagine if like in
320
00:18:13.339 --> 00:18:15.759
the scientific process, you know, you could just spend a
321
00:18:15.779 --> 00:18:19.569
little more money and then, you know, your theorem is true. Surprise.
322
00:18:19.670 --> 00:18:22.529
You know, like I don't think that's how truth works.
323
00:18:22.650 --> 00:18:26.069
I don't think that's like how our political system should work. Right.
324
00:18:26.089 --> 00:18:30.960
That's called corruption. when money impacts the political process. So
325
00:18:31.589 --> 00:18:36.589
I think we need better systems of coordination to better
326
00:18:36.630 --> 00:18:40.890
discover truth, to get people aligned. Now, I do think
327
00:18:40.990 --> 00:18:46.230
that money plays a role and that's for incentives because
328
00:18:46.279 --> 00:18:49.359
there is work to be done to organize knowledge and
329
00:18:49.380 --> 00:18:52.819
it benefits a large number of people to have better,
330
00:18:53.380 --> 00:18:57.980
more organized, more trustworthy information is something that benefits a
331
00:18:58.079 --> 00:19:02.119
very large number of people. And so I think it's
332
00:19:02.180 --> 00:19:07.369
worth incentivizing that work, but I would keep those things distinct.
333
00:19:08.170 --> 00:19:10.450
Yeah, that's a great perspective.
334
00:19:11.029 --> 00:19:15.849
And with the debates, right now with AI, it seems
335
00:19:15.910 --> 00:19:18.049
like you ask a question, you get an answer, and
336
00:19:18.089 --> 00:19:20.490
it's like, well, that must be the truth, or at
337
00:19:20.539 --> 00:19:21.670
least I'm trusting AI.
338
00:19:21.980 --> 00:19:23.799
This is the single source of truth that this is
339
00:19:23.819 --> 00:19:24.460
the correct answer.
340
00:19:24.930 --> 00:19:26.670
And maybe that's not always the case that people are
341
00:19:26.690 --> 00:19:30.039
talking about hallucinations or, you know, it just agreeing with
342
00:19:30.079 --> 00:19:30.700
whatever you say.
343
00:19:30.720 --> 00:19:32.700
It doesn't, most of the time it doesn't disagree. If
344
00:19:32.740 --> 00:19:35.339
you say, well, isn't it this, which is wrong. And
345
00:19:35.359 --> 00:19:36.900
I'll say, yeah, yeah. Just to agree with you.
346
00:19:38.759 --> 00:19:43.589
Is that something that won't happen in, in geo debates that,
347
00:19:43.910 --> 00:19:47.049
you know, the hallucinations of AI or sort of agreeing
348
00:19:47.059 --> 00:19:50.920
uh with your personal perspective uh can you avoid.
349
00:19:50.740 --> 00:19:55.109
That yeah so you know i think the hallucinations are
350
00:19:55.140 --> 00:19:58.029
becoming less of a problem you know the models are
351
00:19:58.089 --> 00:20:02.960
getting like quite good at the like intelligence Right. But
352
00:20:03.460 --> 00:20:06.480
you still end up with the problem of, you know,
353
00:20:06.619 --> 00:20:10.190
what is the AI answering the question against? Right. What
354
00:20:10.250 --> 00:20:13.690
is that ground truth? And for that, we still don't
355
00:20:13.730 --> 00:20:16.549
have an answer. Right. So, you know, if you ask
356
00:20:16.569 --> 00:20:18.720
a question, you know, it'll use the tools that are
357
00:20:18.759 --> 00:20:21.940
available to it. Right. It might do some some web search.
358
00:20:22.359 --> 00:20:24.000
It might, you know, look a few things up and
359
00:20:24.039 --> 00:20:26.730
it's like, OK, here's what I found. Here's your answer. Right.
360
00:20:26.750 --> 00:20:28.829
But did it look at the right sources? Were those
361
00:20:28.890 --> 00:20:32.450
sources correct? You know, according to who? So, you know,
362
00:20:32.670 --> 00:20:35.650
you kind of have that problem of ground truth. And
363
00:20:36.250 --> 00:20:38.049
and I think we still don't have a good solution
364
00:20:38.089 --> 00:20:40.930
to that. So that's really is, you know, what Gio
365
00:20:41.150 --> 00:20:44.069
aims to address. You know, we believe that you get
366
00:20:44.569 --> 00:20:48.359
to the truth by having an open system that lots
367
00:20:48.430 --> 00:20:51.759
of people can participate. Right. Where, you know, things are
368
00:20:51.880 --> 00:20:54.380
open to debate and there's challenge and there's all of
369
00:20:54.420 --> 00:20:58.529
this kind of. you know, process and you're compounding and
370
00:20:58.549 --> 00:21:01.369
you're adding context and more people and more knowledge and
371
00:21:01.410 --> 00:21:06.670
more information over time. And through that process, over time,
372
00:21:06.690 --> 00:21:09.609
you can get closer to truth. And so if it's open,
373
00:21:09.710 --> 00:21:12.910
anyone can participate, you can build reputation, and you have
374
00:21:12.970 --> 00:21:16.430
really powerful tools, then, you know, over time, we can
375
00:21:16.529 --> 00:21:20.490
all build up this, you know, verifiable knowledge graph that
376
00:21:20.519 --> 00:21:23.680
becomes more and more, you know, trustworthy and dependable over time.
377
00:21:25.049 --> 00:21:25.450
Definitely.
378
00:21:25.730 --> 00:21:29.650
And when you do search AI, it goes into the
379
00:21:29.670 --> 00:21:32.369
web and it looks what people were debating on Reddit
380
00:21:32.490 --> 00:21:35.059
as an example. And, you know, people can be lying
381
00:21:35.099 --> 00:21:37.140
there and it could pick that up and be like, OK, well,
382
00:21:37.259 --> 00:21:39.480
that had some upvotes, but that doesn't mean it's the truth.
383
00:21:40.440 --> 00:21:41.930
And I guess that's where geo debates comes in.
384
00:21:42.240 --> 00:21:47.079
Right. Exactly. Now, I think another element that we're seeing
385
00:21:47.119 --> 00:21:50.140
these days is, you know, you hear this about this
386
00:21:50.160 --> 00:21:53.799
thing called the dead Internet theory. And I think that's
387
00:21:53.839 --> 00:21:56.960
really accelerated. You know, the fact is, like, you know,
388
00:21:57.460 --> 00:21:59.940
the common threads aren't exactly the same as they used
389
00:21:59.980 --> 00:22:03.980
to be. Right. And maybe some of that is a slopification. Right.
390
00:22:04.019 --> 00:22:06.400
Now you're like, OK, is this even a person? Why
391
00:22:06.519 --> 00:22:11.339
am I wasting my time? But before it was the centralization. Right.
392
00:22:11.420 --> 00:22:15.009
It was like, you know. this is just owned by Facebook. Why, like,
393
00:22:15.049 --> 00:22:18.670
why am I giving Facebook my time? You know, like, and,
394
00:22:18.849 --> 00:22:21.450
and so like, I think there was this time, you know,
395
00:22:21.529 --> 00:22:25.539
in the early internet where people were willing to spend
396
00:22:25.630 --> 00:22:30.400
their time and their creative energy having discussions and like contributing, right.
397
00:22:31.359 --> 00:22:34.660
To this open thing. And the less it became open,
398
00:22:34.779 --> 00:22:37.619
the less people really wanted to like give their creative energy.
399
00:22:38.160 --> 00:22:41.390
And so I think that like, you know, by creating,
400
00:22:41.410 --> 00:22:45.960
you know, an open composable web, it'll be worth people's
401
00:22:46.019 --> 00:22:50.140
time because they know that they can continue to use
402
00:22:50.660 --> 00:22:53.559
anything that they create, right? Any time that I spend
403
00:22:54.000 --> 00:22:58.259
to contribute my knowledge, to contribute my thoughts, to organize whatever,
404
00:22:58.859 --> 00:23:01.000
I'll be able to keep using that. And it, right,
405
00:23:01.059 --> 00:23:06.009
like it is improving my own, like, you know, workflows
406
00:23:06.069 --> 00:23:08.109
or knowledge because, you know, I have access to that.
407
00:23:08.130 --> 00:23:11.650
I'm not giving it away. But Why not make it open?
408
00:23:11.670 --> 00:23:13.930
And then I can also build reputation, right? Other people
409
00:23:13.950 --> 00:23:16.250
can benefit from it too. And so we get back
410
00:23:16.309 --> 00:23:18.029
to this open web.
411
00:23:19.750 --> 00:23:22.970
Talk to me about launch day moving forward in the
412
00:23:23.019 --> 00:23:27.440
first quarter. How can you and the team make Geodebates
413
00:23:28.140 --> 00:23:32.599
grow rapidly? Is it more content, more claims, more people
414
00:23:32.640 --> 00:23:35.720
that want to be content creators and actually put up videos?
415
00:23:36.180 --> 00:23:37.210
How can you scale it?
416
00:23:38.299 --> 00:23:42.240
Yeah. So, I mean, if your audience that's watching, if
417
00:23:42.480 --> 00:23:45.549
we have any content creators in the crowd, I mean,
418
00:23:45.589 --> 00:23:48.349
we would love to see you on Geodebates. I'd love
419
00:23:48.390 --> 00:23:50.650
to get you up there and whatever you're interested in,
420
00:23:51.009 --> 00:23:53.869
you can participate in a debate and we'd love to
421
00:23:53.910 --> 00:23:56.369
see you there. And for everyone else to just check
422
00:23:56.430 --> 00:24:00.750
out the conversations that are happening and to vote. I
423
00:24:00.829 --> 00:24:03.869
think we want to have an impact on the world
424
00:24:03.910 --> 00:24:05.349
we live in, right? I think a lot of people
425
00:24:05.390 --> 00:24:08.369
have this feeling that like, You know, maybe there's less
426
00:24:08.430 --> 00:24:12.009
and less, you know, ways to, you know, exert your
427
00:24:12.109 --> 00:24:14.869
agency to have an influence. Maybe the system's corrupt, like
428
00:24:14.890 --> 00:24:18.529
whatever it is. we got to change that around, right?
429
00:24:19.769 --> 00:24:21.849
We need people to believe in the future and to
430
00:24:21.910 --> 00:24:24.970
know that they can make a change. And, and I
431
00:24:25.009 --> 00:24:27.450
think the open web is going to make that possible. So,
432
00:24:27.849 --> 00:24:30.019
you know, what, what can we do? I mean, we
433
00:24:30.039 --> 00:24:33.859
want to find interesting people in every field. And, you know,
434
00:24:33.920 --> 00:24:35.640
I think what we do is, I mean, it's, you know,
435
00:24:35.660 --> 00:24:38.390
we create a platform. So it's, It's a way for
436
00:24:38.470 --> 00:24:41.109
people that have knowledge or interest in any type of
437
00:24:41.150 --> 00:24:43.710
thing to be able to express their opinions, to meet
438
00:24:43.750 --> 00:24:47.470
other interesting people. So there's a great community aspect, and
439
00:24:47.490 --> 00:24:51.660
we're giving people powerful tools. And my hope is that
440
00:24:51.720 --> 00:24:55.359
we're going to see really great coverage in a lot
441
00:24:55.400 --> 00:24:58.220
of the areas that are really kind of impacting people.
442
00:24:58.799 --> 00:25:02.900
And I'd love to help bring focus to some of
443
00:25:02.920 --> 00:25:06.710
the most important debates or topics and just get more
444
00:25:06.730 --> 00:25:08.920
people involved. Definitely.
445
00:25:08.960 --> 00:25:11.019
I would love to see that too. I'll do my
446
00:25:11.059 --> 00:25:13.940
best to be one of those content creators. What's the
447
00:25:13.980 --> 00:25:15.920
best way to access the, you said it's on a
448
00:25:15.940 --> 00:25:18.700
web app. How do you access it and how can
449
00:25:18.740 --> 00:25:21.299
you make it convenient for non-technical people?
450
00:25:22.259 --> 00:25:25.759
Yeah. So geo debates is available today. You can just
451
00:25:25.779 --> 00:25:29.680
go to geo browser.io and hit explore. And right away,
452
00:25:29.700 --> 00:25:31.420
you're going to start seeing, you know, some of the
453
00:25:31.460 --> 00:25:34.299
top debates that are coming out and you know, more
454
00:25:34.319 --> 00:25:38.750
debates every day. Um, So that's available now. And, you know,
455
00:25:38.789 --> 00:25:41.029
we really made it so that you don't have to
456
00:25:41.069 --> 00:25:44.410
be technical to use it at all. It is, you know,
457
00:25:44.470 --> 00:25:50.670
a crypto application. It's on the Ethereum blockchain on a
458
00:25:50.690 --> 00:25:55.210
layer three. And so it's all open and composable and verifiable. But,
459
00:25:55.319 --> 00:25:58.000
you know, the users don't really have to know that, right?
460
00:25:58.279 --> 00:26:02.299
They're creating an account using Privy. It's all abstracted. And,
461
00:26:02.799 --> 00:26:04.839
you know, it's just like any social app. You upvote,
462
00:26:04.880 --> 00:26:08.019
you downvote. It's super easy to use. Awesome.
463
00:26:08.119 --> 00:26:08.680
Sounds great.
464
00:26:08.779 --> 00:26:12.380
And I'm always interested in the technical background. But as
465
00:26:12.440 --> 00:26:15.950
we know, for blockchain to be adopted in the mainstream,
466
00:26:16.430 --> 00:26:18.450
the end users can't know they're using blockchain.
467
00:26:19.210 --> 00:26:19.710
Exactly.
468
00:26:21.289 --> 00:26:24.230
So all the best with this. I'm looking forward to
469
00:26:24.670 --> 00:26:27.890
creating some content, doing some votes. We'll leave a link
470
00:26:27.990 --> 00:26:32.019
to Geodebates in the show notes. Would love to follow up,
471
00:26:32.700 --> 00:26:34.220
if not before the end of this year.
472
00:26:34.670 --> 00:26:35.950
To see the.
473
00:26:35.890 --> 00:26:38.329
Progress and wishing you and the team all the best.
474
00:26:38.390 --> 00:26:39.819
And let's follow up in the near future.
475
00:26:40.660 --> 00:26:43.039
Thank you so much, Afton. Yeah, I really enjoyed the conversation.
476
00:26:43.079 --> 00:26:45.839
I'm looking forward to chatting more. Hopefully seeing you over
477
00:26:45.900 --> 00:26:46.779
in the Geoverse.
