1
00:00:00,150 --> 00:00:03,660
- So I want introduce you
to the next presentation.

2
00:00:03,660 --> 00:00:06,243
It's called OpenCola,
The AntiSocial Network.

3
00:00:07,290 --> 00:00:09,060
Please give a big hand to John, here.

4
00:00:09,060 --> 00:00:12,643
His first time DEFCON and I
wanna make him feel welcome.

5
00:00:12,643 --> 00:00:15,087
(attendees applauding)

6
00:00:15,087 --> 00:00:16,590
And hopefully you'll learn something

7
00:00:16,590 --> 00:00:17,423
and you'll enjoy this.

8
00:00:17,423 --> 00:00:18,256
Thanks.

9
00:00:18,256 --> 00:00:19,980
- Great.
- Here's John.

10
00:00:19,980 --> 00:00:22,170
- All right, thank you very much.

11
00:00:22,170 --> 00:00:24,030
Hi there, thank you
for coming to the talk.

12
00:00:24,030 --> 00:00:26,550
I was supposed to do
this with Oxblood Ruffin.

13
00:00:26,550 --> 00:00:27,750
Unfortunately, he could not travel.

14
00:00:27,750 --> 00:00:30,200
So for better or for worse,
you're stuck with me.

15
00:00:31,170 --> 00:00:33,030
So I thought I would start.

16
00:00:33,030 --> 00:00:34,710
The talk today is mainly on an alternative

17
00:00:34,710 --> 00:00:36,000
to existing social media,

18
00:00:36,000 --> 00:00:37,800
but I thought I'd start
with introducing myself

19
00:00:37,800 --> 00:00:38,850
'cause it's pretty relevant

20
00:00:38,850 --> 00:00:40,800
to the whole path that brought me here.

21
00:00:44,012 --> 00:00:46,710
I was born and raised in Toronto.

22
00:00:46,710 --> 00:00:49,380
I got my first computer
when I was 10 years old.

23
00:00:49,380 --> 00:00:50,490
I got a Commodore VIC-20,

24
00:00:50,490 --> 00:00:53,850
which really sparked my
interest in computing.

25
00:00:53,850 --> 00:00:57,090
A couple years later, I got a Commodore 64

26
00:00:57,090 --> 00:00:59,460
and then got heavily into the BBS world,

27
00:00:59,460 --> 00:01:02,433
which was a big precursor to the internet.

28
00:01:03,360 --> 00:01:06,030
It was a world where people
were donating their time

29
00:01:06,030 --> 00:01:08,430
and their money and whatnot
to build communities,

30
00:01:08,430 --> 00:01:11,670
which is really what I would
love to see more of today.

31
00:01:11,670 --> 00:01:14,190
But I spent a bunch of time
hacking around in that world

32
00:01:14,190 --> 00:01:15,720
with supporting BBSs and so on,

33
00:01:15,720 --> 00:01:17,670
and so that's really where I came from.

34
00:01:19,470 --> 00:01:22,560
My first job out of university

35
00:01:22,560 --> 00:01:24,090
was at the original openCola,

36
00:01:24,090 --> 00:01:25,590
which maybe some of you remember.

37
00:01:25,590 --> 00:01:27,390
It was 20 years ago though.

38
00:01:27,390 --> 00:01:29,400
This was a company that was
founded by Corey Doctorow

39
00:01:29,400 --> 00:01:31,020
and a couple of other folks.

40
00:01:31,020 --> 00:01:34,700
It was meant to be a company that...

41
00:01:34,700 --> 00:01:36,120
It was an open source company

42
00:01:36,120 --> 00:01:37,980
that was supposed to facilitate sharing

43
00:01:37,980 --> 00:01:39,570
and building communities and so on.

44
00:01:39,570 --> 00:01:42,030
By the time I got there, Corey had left,

45
00:01:42,030 --> 00:01:43,880
and it became a closed source company

46
00:01:44,910 --> 00:01:46,230
and it was very, very different,

47
00:01:46,230 --> 00:01:48,480
but I was there building
all the search technology.

48
00:01:48,480 --> 00:01:49,830
So here is a little example

49
00:01:49,830 --> 00:01:52,200
of what it looked like at the time.

50
00:01:52,200 --> 00:01:55,200
It was a Windows XP application.

51
00:01:55,200 --> 00:01:57,330
At the top left, there's a search box.

52
00:01:57,330 --> 00:01:58,800
You enter a query.

53
00:01:58,800 --> 00:02:00,930
At the time, Google
was not super dominant.

54
00:02:00,930 --> 00:02:02,040
It was just getting there.

55
00:02:02,040 --> 00:02:04,350
And so, when you wanted to search the web,

56
00:02:04,350 --> 00:02:05,880
you had to search multiple sites

57
00:02:05,880 --> 00:02:07,710
just to find good results.

58
00:02:07,710 --> 00:02:09,480
So this would do that
for you automatically.

59
00:02:09,480 --> 00:02:11,460
It would search across a
various number of sites.

60
00:02:11,460 --> 00:02:12,660
It would give you a list of results.

61
00:02:12,660 --> 00:02:14,550
You can see in the top pane.

62
00:02:14,550 --> 00:02:15,870
You can scroll through those results.

63
00:02:15,870 --> 00:02:17,220
If you found ones that you liked,

64
00:02:17,220 --> 00:02:18,420
you'd mark them as relevant.

65
00:02:18,420 --> 00:02:19,740
It would rescore the folder,

66
00:02:19,740 --> 00:02:21,080
everything in it for you.

67
00:02:21,080 --> 00:02:23,100
It would help you draw attention to things

68
00:02:23,100 --> 00:02:24,990
that would be useful for you,

69
00:02:24,990 --> 00:02:26,570
and then it would generate new queries.

70
00:02:26,570 --> 00:02:28,290
It would do some light semantic analysis

71
00:02:28,290 --> 00:02:29,550
and generate new queries,

72
00:02:29,550 --> 00:02:32,040
and continually add
information to that folder.

73
00:02:32,040 --> 00:02:33,720
And so, you could have
as many of these folders

74
00:02:33,720 --> 00:02:34,593
as you wanted.

75
00:02:36,450 --> 00:02:38,280
You were also connected to other peers

76
00:02:38,280 --> 00:02:39,450
and so when peers would go off

77
00:02:39,450 --> 00:02:41,490
and find interesting things and save them,

78
00:02:41,490 --> 00:02:42,990
you could leverage all of that.

79
00:02:42,990 --> 00:02:43,980
Your queries would go to them

80
00:02:43,980 --> 00:02:45,780
and you'd get their results as well.

81
00:02:47,250 --> 00:02:48,960
So that was a ton of fun for me.

82
00:02:48,960 --> 00:02:50,410
One of my favorite jobs ever.

83
00:02:51,390 --> 00:02:52,980
So what's happened since?

84
00:02:52,980 --> 00:02:55,350
Well for myself, I bounced around.

85
00:02:55,350 --> 00:02:56,250
I did a couple startups

86
00:02:56,250 --> 00:02:58,530
with two of the other
founders of openCola.

87
00:02:58,530 --> 00:03:00,270
I eventually ended up at Microsoft.

88
00:03:00,270 --> 00:03:01,950
I worked in identity for a bit,

89
00:03:01,950 --> 00:03:03,660
learned a bit about security.

90
00:03:03,660 --> 00:03:06,630
I then worked on a knowledge web team,

91
00:03:06,630 --> 00:03:09,120
and learned a ton about semantic web

92
00:03:09,120 --> 00:03:11,250
and recommendation systems.

93
00:03:11,250 --> 00:03:12,270
And then after that,

94
00:03:12,270 --> 00:03:14,340
I worked at Netflix for 10
years where I bounced around

95
00:03:14,340 --> 00:03:16,980
a lot of the different
personalization teams.

96
00:03:16,980 --> 00:03:18,270
Building all of the algorithms

97
00:03:18,270 --> 00:03:19,740
that produced the Netflix experience.

98
00:03:19,740 --> 00:03:22,350
The homepage, the ordering
of videos, and so on.

99
00:03:22,350 --> 00:03:23,183
And at the end,

100
00:03:23,183 --> 00:03:24,570
I was actually running
the whole organization.

101
00:03:24,570 --> 00:03:27,180
So I have a lot of experience
with recommender systems

102
00:03:27,180 --> 00:03:28,350
and being in a big company,

103
00:03:28,350 --> 00:03:29,640
trying to balance user needs

104
00:03:29,640 --> 00:03:31,983
and the company's needs as well.

105
00:03:33,540 --> 00:03:35,280
Obviously, in that time, a
lot has changed in the world.

106
00:03:35,280 --> 00:03:37,170
Google is the dominant search engine.

107
00:03:37,170 --> 00:03:40,650
We've gone through Friendster, MySpace,

108
00:03:40,650 --> 00:03:42,330
Facebook, Twitter, Instagram.

109
00:03:42,330 --> 00:03:44,130
All of these things have happened,

110
00:03:44,130 --> 00:03:48,930
and along with them have come
some pretty big tech problems.

111
00:03:48,930 --> 00:03:51,297
We've seen a large amount
of information siloing.

112
00:03:51,297 --> 00:03:55,650
And so, all these sites benefit
by keeping you as members

113
00:03:55,650 --> 00:03:57,240
and so they don't really
want to unlock your data

114
00:03:57,240 --> 00:03:59,160
and let you use it in multiple places.

115
00:03:59,160 --> 00:04:00,750
So if I accumulate things on Facebook

116
00:04:00,750 --> 00:04:02,220
and accumulate things on Twitter,

117
00:04:02,220 --> 00:04:03,330
they're stuck in those silos.

118
00:04:03,330 --> 00:04:04,950
I can't really do anything with that data.

119
00:04:04,950 --> 00:04:06,870
I have to use those sites,

120
00:04:06,870 --> 00:04:08,780
and that serves these companies.

121
00:04:08,780 --> 00:04:11,553
It doesn't serve me or any
other end users super well.

122
00:04:13,080 --> 00:04:13,913
Secondly.

123
00:04:15,510 --> 00:04:18,150
The majority of these
companies are profit companies

124
00:04:18,150 --> 00:04:20,550
that really make their
money through advertising,

125
00:04:20,550 --> 00:04:21,810
and that means they make more money

126
00:04:21,810 --> 00:04:22,920
if you stay on their site.

127
00:04:22,920 --> 00:04:26,670
And so, the algorithms are
really tuned to grab you in

128
00:04:26,670 --> 00:04:29,580
and keep you on their
site as much as possible.

129
00:04:29,580 --> 00:04:31,920
And again, this isn't necessarily
in your best interest.

130
00:04:31,920 --> 00:04:33,450
You're being shown things
that keep you there.

131
00:04:33,450 --> 00:04:34,680
They may not be things that, actually,

132
00:04:34,680 --> 00:04:36,080
are the best things for you.

133
00:04:37,650 --> 00:04:40,860
There has been endless
amounts of privacy abuse.

134
00:04:40,860 --> 00:04:43,590
I think the favorite story
that I've linked to here

135
00:04:43,590 --> 00:04:45,270
is a John Oliver clip that's talking

136
00:04:45,270 --> 00:04:48,180
about the amount of data that
data brokers have access to,

137
00:04:48,180 --> 00:04:50,100
and that you can purchase
with a credit card.

138
00:04:50,100 --> 00:04:51,690
And in many cases, this data

139
00:04:51,690 --> 00:04:55,230
is data that would require the
government a warrant to get.

140
00:04:55,230 --> 00:04:57,090
And sometimes, they even
just go and buy the data

141
00:04:57,090 --> 00:04:58,560
from the brokers if
they can't get warrants.

142
00:04:58,560 --> 00:05:01,083
So it's a pretty crazy situation.

143
00:05:02,400 --> 00:05:03,570
We've also seen a huge amount

144
00:05:03,570 --> 00:05:05,670
of corporate concentration
of wealth and power.

145
00:05:05,670 --> 00:05:09,570
There are very few companies at the top

146
00:05:09,570 --> 00:05:11,120
that are making a lot of money.

147
00:05:13,260 --> 00:05:15,480
Because of that, they can use
their power and their wealth

148
00:05:15,480 --> 00:05:17,910
to continue to be anti-competitive

149
00:05:17,910 --> 00:05:20,190
and to continue their advantage.

150
00:05:20,190 --> 00:05:21,780
There's a great post link here

151
00:05:21,780 --> 00:05:24,990
to the antitrust case against Facebook.

152
00:05:24,990 --> 00:05:27,450
Which reminded me, it's about
how Facebook's evolution,

153
00:05:27,450 --> 00:05:28,530
that I had totally forgotten about.

154
00:05:28,530 --> 00:05:29,790
They used to in the early days,

155
00:05:29,790 --> 00:05:33,870
be really advocates for user
privacy and for the end user.

156
00:05:33,870 --> 00:05:35,190
And they got to a point,

157
00:05:35,190 --> 00:05:38,310
that as pretty much any
for-profit company's gonna get to,

158
00:05:38,310 --> 00:05:41,460
where they have to maximize
profit and in order to do so,

159
00:05:41,460 --> 00:05:43,803
they had to stop
representing the end user.

160
00:05:45,210 --> 00:05:46,410
Misinformation, for me,

161
00:05:46,410 --> 00:05:48,390
it's been huge since 2016.

162
00:05:48,390 --> 00:05:50,400
Just trying to think
about ways, potentially,

163
00:05:50,400 --> 00:05:52,050
to solve these problems,

164
00:05:52,050 --> 00:05:56,670
and putting out information
that is not accurate

165
00:05:56,670 --> 00:05:58,530
and distorts people's perceptions

166
00:05:58,530 --> 00:05:59,790
serves lots of different reasons.

167
00:05:59,790 --> 00:06:03,450
It could be to distort a political opinion

168
00:06:03,450 --> 00:06:06,227
and it could also be
simply to try and make ads.

169
00:06:06,227 --> 00:06:07,980
Things that people put in information

170
00:06:07,980 --> 00:06:09,240
that attracts attention.

171
00:06:09,240 --> 00:06:10,083
You click on it.

172
00:06:11,040 --> 00:06:12,483
That helps people sell ads.

173
00:06:13,950 --> 00:06:16,320
Related polarization.

174
00:06:16,320 --> 00:06:17,280
Strong emotion.

175
00:06:17,280 --> 00:06:19,290
It gets a lot of people's attention.

176
00:06:19,290 --> 00:06:21,060
Particularly, negative emotion.

177
00:06:21,060 --> 00:06:23,280
And so the algorithms
that these companies use

178
00:06:23,280 --> 00:06:24,810
to figure out how to keep your attention,

179
00:06:24,810 --> 00:06:28,260
they're really amplifying a
lot of the negative emotion

180
00:06:28,260 --> 00:06:29,850
which leads to polarization.

181
00:06:29,850 --> 00:06:31,500
It leads to us feeling
that we're further apart

182
00:06:31,500 --> 00:06:32,850
than maybe we actually are.

183
00:06:34,170 --> 00:06:35,550
There's stifled innovation.

184
00:06:35,550 --> 00:06:38,220
These companies will build what
only makes profit for them.

185
00:06:38,220 --> 00:06:39,900
So there may be lots of great features

186
00:06:39,900 --> 00:06:41,370
that would be super innovative

187
00:06:41,370 --> 00:06:42,750
but you aren't gonna see them

188
00:06:42,750 --> 00:06:45,870
unless the company can
make some money from it.

189
00:06:45,870 --> 00:06:47,580
And then lastly,

190
00:06:47,580 --> 00:06:50,490
in order to manipulate networks
and manipulate discourse,

191
00:06:50,490 --> 00:06:52,740
there's been a huge rise in bots.

192
00:06:52,740 --> 00:06:54,930
So automated systems
that pretend to be people

193
00:06:54,930 --> 00:06:56,910
that will manipulate
the internet structure

194
00:06:56,910 --> 00:06:58,350
to try and trick Google,

195
00:06:58,350 --> 00:07:00,600
or manipulate discourse on
these social media platforms

196
00:07:00,600 --> 00:07:01,830
in order to try and get more eyes

197
00:07:01,830 --> 00:07:04,403
on the types of stories that
these folks are interested in.

198
00:07:05,400 --> 00:07:08,790
So if we step back though,
and really ask ourselves,

199
00:07:08,790 --> 00:07:09,990
what are these sites offering?

200
00:07:09,990 --> 00:07:12,780
It's actually not all that much

201
00:07:12,780 --> 00:07:15,840
and it's actually not
all that hard to build.

202
00:07:15,840 --> 00:07:18,840
Up here, I've got arguably the three,

203
00:07:18,840 --> 00:07:20,670
sort of, top social media sites today.

204
00:07:20,670 --> 00:07:23,100
Facebook, Twitter, and Instagram.

205
00:07:23,100 --> 00:07:24,930
Facebook, it's really just a feed.

206
00:07:24,930 --> 00:07:28,440
Every post in it has just
an image, some description.

207
00:07:28,440 --> 00:07:30,540
It's a bit of a structured item.

208
00:07:30,540 --> 00:07:32,280
And then you can do a
few things with that.

209
00:07:32,280 --> 00:07:36,150
Like, you can like, you
can share, and so on.

210
00:07:36,150 --> 00:07:37,250
You can add a comment.

211
00:07:38,100 --> 00:07:39,660
On Twitter, it's pretty
much the same thing.

212
00:07:39,660 --> 00:07:42,450
It's just an image with some text on it

213
00:07:42,450 --> 00:07:47,070
and you can comment,
retweet, like, or share.

214
00:07:47,070 --> 00:07:48,970
Instagram, pretty much the same thing.

215
00:07:50,400 --> 00:07:51,960
So if you take a step
back and you look at it.

216
00:07:51,960 --> 00:07:53,730
Well, what are you really
getting as an end user?

217
00:07:53,730 --> 00:07:56,250
You're getting the ability
to indirectly interact

218
00:07:56,250 --> 00:07:58,170
with your friends, your
family, and your community,

219
00:07:58,170 --> 00:08:00,600
and you can perform a
fairly small set of actions.

220
00:08:00,600 --> 00:08:03,360
Across all the sites
that I'm thinking about,

221
00:08:03,360 --> 00:08:07,380
it's really just post, like,
rate, comment, tag, and share.

222
00:08:07,380 --> 00:08:09,960
And I'd include even things
like Amazon and YouTube,

223
00:08:09,960 --> 00:08:12,060
which may not be considered
directly social media

224
00:08:12,060 --> 00:08:15,870
but they're places that
fit this pattern very well.

225
00:08:15,870 --> 00:08:17,730
And what you get back is a feed

226
00:08:17,730 --> 00:08:20,587
but in order to get access to all of this,

227
00:08:20,587 --> 00:08:22,500
you're really building...

228
00:08:22,500 --> 00:08:24,660
We're really allowing all of
these other problems to exist.

229
00:08:24,660 --> 00:08:26,913
You also get all the
aforementioned problems.

230
00:08:29,400 --> 00:08:30,600
So...

231
00:08:30,600 --> 00:08:34,830
I found a great quote in the
antitrust case against Facebook

232
00:08:34,830 --> 00:08:36,300
that really just asked the question

233
00:08:36,300 --> 00:08:38,640
of why is it that in a
country that's so paranoid

234
00:08:38,640 --> 00:08:41,250
about government or
corporate surveillance.

235
00:08:41,250 --> 00:08:43,863
Why is this really the
only option that we have?

236
00:08:44,910 --> 00:08:47,340
And I would say it's probably
because people that go

237
00:08:47,340 --> 00:08:49,020
and want to build these type of companies

238
00:08:49,020 --> 00:08:50,100
have two things in mind.

239
00:08:50,100 --> 00:08:51,060
They either wanna make money

240
00:08:51,060 --> 00:08:52,590
or they want to control what's going on,

241
00:08:52,590 --> 00:08:54,030
and I think that's sort of very different

242
00:08:54,030 --> 00:08:55,950
than the types of companies
that have been built up

243
00:08:55,950 --> 00:08:58,230
to give us these products today.

244
00:08:58,230 --> 00:08:59,700
So for me, fundamentally, the question

245
00:08:59,700 --> 00:09:03,240
is "How can we have a more
trustworthy internet?"

246
00:09:03,240 --> 00:09:04,950
And I do not think that profit seekers

247
00:09:04,950 --> 00:09:07,380
can really offer a solution to this.

248
00:09:07,380 --> 00:09:09,660
For example, there are certain
things they just don't do.

249
00:09:09,660 --> 00:09:10,890
They did not and could not

250
00:09:10,890 --> 00:09:13,800
really have produced the
internet in the first place.

251
00:09:13,800 --> 00:09:15,840
It was too risky, too long run.

252
00:09:15,840 --> 00:09:18,750
All of that was based off
of public funding to start

253
00:09:18,750 --> 00:09:19,890
and then, at some point,

254
00:09:19,890 --> 00:09:22,200
was taken over by private companies.

255
00:09:22,200 --> 00:09:24,210
But they did create all
the problems we've seen.

256
00:09:24,210 --> 00:09:26,310
So I don't really feel
like I would trust them

257
00:09:26,310 --> 00:09:28,020
to solve these problems.

258
00:09:28,020 --> 00:09:29,040
So I think we need to start

259
00:09:29,040 --> 00:09:30,360
with just a different incentive model

260
00:09:30,360 --> 00:09:31,320
which is rather than thinking

261
00:09:31,320 --> 00:09:33,420
about how to make money
solving the problem.

262
00:09:33,420 --> 00:09:34,800
This is a problem we're solving,

263
00:09:34,800 --> 00:09:37,260
so let's just figure out how
we should solve this problem

264
00:09:37,260 --> 00:09:38,463
from principles first.

265
00:09:40,530 --> 00:09:43,200
So in terms of trust, then.

266
00:09:43,200 --> 00:09:44,910
I've enumerated, at least what I think,

267
00:09:44,910 --> 00:09:46,773
is a minimum set of principles.

268
00:09:48,000 --> 00:09:50,610
Any system that's built to
try and solve these problems,

269
00:09:50,610 --> 00:09:52,170
I believe, needs to be transparent.

270
00:09:52,170 --> 00:09:53,490
You need to know what it's doing

271
00:09:53,490 --> 00:09:54,720
and why it's doing what it's doing.

272
00:09:54,720 --> 00:09:57,930
So you can choose to use
it based on what it's doing

273
00:09:57,930 --> 00:09:59,523
and you can detect misuse.

274
00:10:01,740 --> 00:10:04,260
The system also needs to
provide for accountability

275
00:10:04,260 --> 00:10:06,180
if somebody abuses the system.

276
00:10:06,180 --> 00:10:07,590
There has to be some mechanism

277
00:10:07,590 --> 00:10:09,600
to punish that person or that entity

278
00:10:09,600 --> 00:10:11,673
in order to disincentivize that misuse.

279
00:10:14,550 --> 00:10:16,200
As well, you need some personal control.

280
00:10:16,200 --> 00:10:17,910
Your information, your data.

281
00:10:17,910 --> 00:10:19,980
You should know what's
happening to that data.

282
00:10:19,980 --> 00:10:21,240
You should know where it's going,

283
00:10:21,240 --> 00:10:22,800
who's seeing it and how it's being used,

284
00:10:22,800 --> 00:10:25,170
whether it's for your benefit or not.

285
00:10:25,170 --> 00:10:27,450
And then lastly, but obviously not least,

286
00:10:27,450 --> 00:10:29,070
you need some degree of security.

287
00:10:29,070 --> 00:10:30,060
You shouldn't be able to pretend

288
00:10:30,060 --> 00:10:31,590
to be somebody that
you're not in this system,

289
00:10:31,590 --> 00:10:33,210
and you shouldn't be able
to access information

290
00:10:33,210 --> 00:10:34,803
in an unauthorized fashion.

291
00:10:36,990 --> 00:10:40,350
So if we look at these principles, though,

292
00:10:40,350 --> 00:10:42,270
a closed source for-profit model

293
00:10:42,270 --> 00:10:46,740
generally will minimize each
one of these principles.

294
00:10:46,740 --> 00:10:48,270
In the security sense, you know,

295
00:10:48,270 --> 00:10:50,070
you might be able to argue
that some of these companies

296
00:10:50,070 --> 00:10:51,873
do a good job of securing your data.

297
00:10:52,800 --> 00:10:53,730
Maybe, from hackers.

298
00:10:53,730 --> 00:10:55,050
Although, you're probably
gonna know a lot more

299
00:10:55,050 --> 00:10:56,310
about this than I do.

300
00:10:56,310 --> 00:10:58,800
In the very least, they're
just gonna sell your data

301
00:10:58,800 --> 00:10:59,730
if it's profitable for them.

302
00:10:59,730 --> 00:11:02,340
So that's clearly not
a trustworthy behavior.

303
00:11:02,340 --> 00:11:05,430
So from my perspective, this
really strongly suggests

304
00:11:05,430 --> 00:11:07,953
a free open source
distributed software model.

305
00:11:10,440 --> 00:11:13,410
Before I dive into how it
all works under the hood.

306
00:11:13,410 --> 00:11:14,850
I'll give you a little demo to show you

307
00:11:14,850 --> 00:11:16,887
what has been built so far in OpenCola.

308
00:11:16,887 --> 00:11:19,620
And so, this is really
just a re-envisioning

309
00:11:19,620 --> 00:11:22,470
of what was originally in the product

310
00:11:22,470 --> 00:11:25,020
and the ideas that came
from the original openCola.

311
00:11:25,020 --> 00:11:26,820
It's gonna be slightly different.

312
00:11:26,820 --> 00:11:28,500
It's very similar to the
existing social media

313
00:11:28,500 --> 00:11:29,460
but there are a couple things in it

314
00:11:29,460 --> 00:11:31,260
that we do a bit differently,

315
00:11:31,260 --> 00:11:32,093
but just keep in mind

316
00:11:32,093 --> 00:11:33,600
that this is just a reference application.

317
00:11:33,600 --> 00:11:35,010
Fundamentally, this is a platform

318
00:11:35,010 --> 00:11:36,630
and the idea of it being open source

319
00:11:36,630 --> 00:11:39,263
is that you can build whatever
you want on top of this.

320
00:11:42,060 --> 00:11:44,010
So I'll switch over to the demo,

321
00:11:44,010 --> 00:11:45,123
which is visible.

322
00:11:47,280 --> 00:11:48,390
Okay.

323
00:11:48,390 --> 00:11:52,020
So let's look at setting
up the application.

324
00:11:52,020 --> 00:11:55,080
I've skipped over just the install.

325
00:11:55,080 --> 00:11:57,720
There's a server that you can run.

326
00:11:57,720 --> 00:12:00,450
What we'll see is the use
of a browser extension.

327
00:12:00,450 --> 00:12:03,120
I've skipped over the install,
just for the sake of time,

328
00:12:03,120 --> 00:12:06,900
but I'll show you how to set
it up once it's been installed.

329
00:12:06,900 --> 00:12:08,610
With your browser extension.

330
00:12:08,610 --> 00:12:11,250
Basically, it allows you
to take a few actions.

331
00:12:11,250 --> 00:12:12,630
Again, this could be extendable,

332
00:12:12,630 --> 00:12:15,090
but I've just got a
small set here for now.

333
00:12:15,090 --> 00:12:16,440
The little green bubble on the end

334
00:12:16,440 --> 00:12:17,610
is just a status indicator.

335
00:12:17,610 --> 00:12:18,810
Green means everything's okay.

336
00:12:18,810 --> 00:12:20,010
Yellow means it's doing something,

337
00:12:20,010 --> 00:12:22,650
and red means something bad happened.

338
00:12:22,650 --> 00:12:23,700
There's a little disc icon

339
00:12:23,700 --> 00:12:26,430
that allows you to save
the site that you're on.

340
00:12:26,430 --> 00:12:28,080
There is a thumbs up icon

341
00:12:28,080 --> 00:12:29,580
that allows you to like
the site you're on.

342
00:12:29,580 --> 00:12:31,200
And then there's a trust action,

343
00:12:31,200 --> 00:12:32,640
which I'll talk a lot more about later

344
00:12:32,640 --> 00:12:35,880
and is something I think
novel for social media.

345
00:12:35,880 --> 00:12:37,560
And the last thing is
there's a little search icon

346
00:12:37,560 --> 00:12:38,460
that lets you go to your feed

347
00:12:38,460 --> 00:12:41,063
and then you can search things
that you've interacted with.

348
00:12:44,280 --> 00:12:46,440
So let's just go over
to the feed, right now.

349
00:12:46,440 --> 00:12:47,980
Unsurprisingly, it's empty

350
00:12:48,930 --> 00:12:51,270
and we will go into the peer view.

351
00:12:51,270 --> 00:12:54,180
At this point, you're the
only person in the system.

352
00:12:54,180 --> 00:12:57,180
So this peer is you and you
can go and set up your name

353
00:12:57,180 --> 00:12:58,500
and an image for yourself.

354
00:12:58,500 --> 00:13:00,000
So I'm gonna pretend this is Nola

355
00:13:00,000 --> 00:13:02,010
who is the original openCola mascot.

356
00:13:02,010 --> 00:13:02,973
Put an image in.

357
00:13:04,410 --> 00:13:06,780
And then, there are three other fields

358
00:13:06,780 --> 00:13:08,190
that are set for you automatically.

359
00:13:08,190 --> 00:13:11,550
You get a globally unique OpenCola ID.

360
00:13:11,550 --> 00:13:12,990
You get a public key which is used

361
00:13:12,990 --> 00:13:16,290
to secure your e-communication

362
00:13:16,290 --> 00:13:17,940
and make sure that you can validate

363
00:13:17,940 --> 00:13:19,680
that it came from the
person you thought it did.

364
00:13:19,680 --> 00:13:20,820
And then, there's an address.

365
00:13:20,820 --> 00:13:22,380
Right now, this is an HTTP address

366
00:13:22,380 --> 00:13:24,060
'cause it's a local machine.

367
00:13:24,060 --> 00:13:26,110
I'll explain a bit more about that later.

368
00:13:32,430 --> 00:13:33,810
Okay, so now we're set up.

369
00:13:33,810 --> 00:13:36,410
Let's go and take a look
at what posting looks like.

370
00:13:37,800 --> 00:13:41,160
So in order to do some posting.

371
00:13:41,160 --> 00:13:42,030
When you're browsing the web,

372
00:13:42,030 --> 00:13:43,530
you can take some action on the pages.

373
00:13:43,530 --> 00:13:45,510
This is where we're slightly different.

374
00:13:45,510 --> 00:13:46,500
I'll show you in a second

375
00:13:46,500 --> 00:13:48,540
how you can create more organic posts.

376
00:13:48,540 --> 00:13:50,460
But the posts that we're looking at here,

377
00:13:50,460 --> 00:13:51,990
everything is very information-centric.

378
00:13:51,990 --> 00:13:53,910
So when people interact
with the same resource,

379
00:13:53,910 --> 00:13:55,020
all of that gets combined.

380
00:13:55,020 --> 00:13:56,550
So you can see what
everybody in your network

381
00:13:56,550 --> 00:14:00,120
is thinking and talking about
on that particular item.

382
00:14:00,120 --> 00:14:01,920
So we can start here and we can go and say

383
00:14:01,920 --> 00:14:04,170
that we like the DEFCON page,

384
00:14:04,170 --> 00:14:05,520
and then we can go and say

385
00:14:05,520 --> 00:14:08,343
that we want to save a Wikipedia page.

386
00:14:09,870 --> 00:14:11,640
If we go back to our feed,

387
00:14:11,640 --> 00:14:14,130
notice that now we have those two items

388
00:14:14,130 --> 00:14:16,080
that we interacted with
are sitting in our feed.

389
00:14:16,080 --> 00:14:19,020
This feed is just reverse
chronologically ordered.

390
00:14:19,020 --> 00:14:20,820
It's done some convenient extraction

391
00:14:20,820 --> 00:14:23,670
of a text description
and an image for you.

392
00:14:23,670 --> 00:14:26,490
Under each post, it
indicates who posted it

393
00:14:26,490 --> 00:14:28,590
and then there's actions
that you can take,

394
00:14:28,590 --> 00:14:30,780
and also that you can see all
the activity of your peers.

395
00:14:30,780 --> 00:14:33,030
So you can resave stuff.

396
00:14:33,030 --> 00:14:34,980
You can resave and see activity of saves.

397
00:14:34,980 --> 00:14:37,140
You can like, you can tag,

398
00:14:37,140 --> 00:14:39,640
you can comment, and you
can edit the post itself.

399
00:14:40,740 --> 00:14:42,750
One of the nice things that
it does for you, as well,

400
00:14:42,750 --> 00:14:45,180
is that when you actually
interact through the web browser,

401
00:14:45,180 --> 00:14:46,950
it's actually taking that whole page

402
00:14:46,950 --> 00:14:49,500
and it's storing it as
a single page archive.

403
00:14:49,500 --> 00:14:51,810
So not only can you just click on the link

404
00:14:51,810 --> 00:14:53,370
and see the current page,
but it'll show you...

405
00:14:53,370 --> 00:14:54,210
What I'll show you right now

406
00:14:54,210 --> 00:14:56,100
is that you can click on the archive link.

407
00:14:56,100 --> 00:14:56,933
And if you notice,

408
00:14:56,933 --> 00:14:58,800
this is being served
off your local computer.

409
00:14:58,800 --> 00:15:01,500
So you can go back and see
exactly what you interacted with

410
00:15:01,500 --> 00:15:03,600
and you can do that without being tracked.

411
00:15:12,270 --> 00:15:15,360
Okay, and so this is the more
traditional social media post.

412
00:15:15,360 --> 00:15:17,730
This is a bit of an export form right now.

413
00:15:17,730 --> 00:15:18,990
I'm gonna make it a bit simpler.

414
00:15:18,990 --> 00:15:21,840
You can edit any of the fields
that go into the template

415
00:15:21,840 --> 00:15:23,880
that gets rendered in the feed.

416
00:15:23,880 --> 00:15:25,290
Normally in social media,

417
00:15:25,290 --> 00:15:26,123
you're often just wanting

418
00:15:26,123 --> 00:15:27,510
to type something in the description.

419
00:15:27,510 --> 00:15:29,670
So if you've got something
that you wanna share,

420
00:15:29,670 --> 00:15:31,530
you can just put it in the description box

421
00:15:31,530 --> 00:15:33,000
and when you save that,

422
00:15:33,000 --> 00:15:35,370
it just creates a relatively empty entry

423
00:15:35,370 --> 00:15:36,600
with your image in it

424
00:15:36,600 --> 00:15:38,850
and that's sort of the
more traditional path.

425
00:15:42,690 --> 00:15:45,660
So it's great to have your
own personal repository.

426
00:15:45,660 --> 00:15:47,820
I find it super useful just
to keep track of things.

427
00:15:47,820 --> 00:15:50,220
But of course, you wanna be
able to connect to your friends

428
00:15:50,220 --> 00:15:52,380
and leverage all the things
that they're doing as well.

429
00:15:52,380 --> 00:15:55,890
So we'll take a look at how
you connect two peers together.

430
00:15:55,890 --> 00:15:57,903
First, we need a peer.

431
00:15:58,830 --> 00:16:01,440
So luckily I have one pre-created here.

432
00:16:01,440 --> 00:16:03,000
Clippy,

433
00:16:03,000 --> 00:16:06,150
and Clippy already has
some existing posts.

434
00:16:06,150 --> 00:16:08,430
And these two, Clippy
and Nola want to connect,

435
00:16:08,430 --> 00:16:11,280
and so what they each do is
they go into their peer view.

436
00:16:13,710 --> 00:16:15,960
Basically, there's a
token that's given to them

437
00:16:15,960 --> 00:16:18,393
that they basically
exchange with each other.

438
00:16:22,080 --> 00:16:23,280
And so I'll just mention

439
00:16:23,280 --> 00:16:25,410
that this adds a bit of
friction to connecting

440
00:16:25,410 --> 00:16:27,480
that doesn't exist in
existing social media.

441
00:16:27,480 --> 00:16:28,313
And as you'll see you later,

442
00:16:28,313 --> 00:16:29,460
this is actually pretty valuable.

443
00:16:29,460 --> 00:16:32,190
It actually provides
some great functionality

444
00:16:32,190 --> 00:16:33,600
on the other side.

445
00:16:33,600 --> 00:16:37,680
You do this token exchange out
of band of the application.

446
00:16:37,680 --> 00:16:39,120
You can use whatever
you're comfortable with.

447
00:16:39,120 --> 00:16:40,530
It's not hyper critical.

448
00:16:40,530 --> 00:16:41,820
But if you want to be super secure,

449
00:16:41,820 --> 00:16:42,930
you could use something like Signal

450
00:16:42,930 --> 00:16:44,430
to send that token to a friend.

451
00:16:44,430 --> 00:16:45,570
But the point is that you only connect

452
00:16:45,570 --> 00:16:46,950
to people that you actually know.

453
00:16:46,950 --> 00:16:47,783
That you verified.

454
00:16:47,783 --> 00:16:48,616
You know this person.

455
00:16:48,616 --> 00:16:50,260
So you've talk to them out of band.

456
00:16:51,660 --> 00:16:53,040
Once you add them, if you want to,

457
00:16:53,040 --> 00:16:55,500
you could edit their name or their image

458
00:16:55,500 --> 00:16:58,300
and see that person how you
want to see it on your side.

459
00:16:59,580 --> 00:17:00,990
And now, you're connected.

460
00:17:00,990 --> 00:17:03,450
And if we go back and look
at the individual feeds,

461
00:17:03,450 --> 00:17:05,500
we'll see that they are now synchronized.

462
00:17:08,550 --> 00:17:11,280
And on Clippy's side, Clippy
can go then and interact

463
00:17:11,280 --> 00:17:13,380
with the posts that were posted by Nola.

464
00:17:13,380 --> 00:17:14,523
Add in a comment.

465
00:17:17,820 --> 00:17:19,800
Maybe like a post.

466
00:17:19,800 --> 00:17:23,790
And if we go back over to
Nola's feed and refresh,

467
00:17:23,790 --> 00:17:27,150
notice all of those changes
took effect immediately.

468
00:17:27,150 --> 00:17:30,660
And again, it's reverse
chronologically sorted.

469
00:17:30,660 --> 00:17:32,560
The most recent activity's at the top.

470
00:17:36,990 --> 00:17:38,640
So now you're connected.

471
00:17:38,640 --> 00:17:40,320
You've got all your data.

472
00:17:40,320 --> 00:17:42,330
You probably also wanna be
able to search that data.

473
00:17:42,330 --> 00:17:43,410
That's something that's
somewhat different.

474
00:17:43,410 --> 00:17:45,060
Again, we're a bit more content-centric.

475
00:17:45,060 --> 00:17:46,740
So the idea is that you
share this information

476
00:17:46,740 --> 00:17:47,573
and then it should be really easy

477
00:17:47,573 --> 00:17:48,600
for you to find afterwards.

478
00:17:48,600 --> 00:17:50,820
You've got a single repository

479
00:17:50,820 --> 00:17:51,750
of the things that you care about

480
00:17:51,750 --> 00:17:53,950
across all the sites
that you interact with.

481
00:17:54,870 --> 00:17:56,460
So searching is pretty straightforward.

482
00:17:56,460 --> 00:17:58,650
Just enter your query in the query box.

483
00:17:58,650 --> 00:17:59,970
Here we entered writing.

484
00:17:59,970 --> 00:18:03,660
What we got back was the one
writing result that made sense.

485
00:18:03,660 --> 00:18:05,460
Notice that this looks
exactly like the feed,

486
00:18:05,460 --> 00:18:07,980
and the feed and the search
results are totally unified.

487
00:18:07,980 --> 00:18:10,050
Everything you can do in
one, you can do in the other.

488
00:18:10,050 --> 00:18:11,300
So it's super convenient.

489
00:18:12,240 --> 00:18:14,310
So you can go down and say,
edit the search result.

490
00:18:14,310 --> 00:18:16,680
If you didn't like the
description that it had extracted,

491
00:18:16,680 --> 00:18:17,550
you can edit it.

492
00:18:17,550 --> 00:18:19,830
This changes it for you and
anybody that you share it with

493
00:18:19,830 --> 00:18:21,360
but anybody who'd already has saved it,

494
00:18:21,360 --> 00:18:23,070
it won't change it for them.

495
00:18:23,070 --> 00:18:24,450
And then, you can like the item.

496
00:18:24,450 --> 00:18:25,653
You can add a tag on.

497
00:18:30,930 --> 00:18:32,160
And if you go back to the main feed,

498
00:18:32,160 --> 00:18:34,980
once again, everything has been updated

499
00:18:34,980 --> 00:18:36,390
and you see that at the top of your feed.

500
00:18:36,390 --> 00:18:39,093
And so would Nola, on her side.

501
00:18:41,100 --> 00:18:43,620
We didn't specifically
build a mobile version

502
00:18:43,620 --> 00:18:45,150
but just to show you what
it looks like right now.

503
00:18:45,150 --> 00:18:48,180
It's actually pretty functional
and it looks pretty good.

504
00:18:48,180 --> 00:18:49,230
The one challenge on mobile

505
00:18:49,230 --> 00:18:51,180
is that we don't have the plugin

506
00:18:51,180 --> 00:18:52,140
and so we had to figure out how

507
00:18:52,140 --> 00:18:54,153
to make it easy to add content in.

508
00:18:55,170 --> 00:18:56,550
That's actually pretty straightforward.

509
00:18:56,550 --> 00:18:58,740
You just use the share
functionality in your browser.

510
00:18:58,740 --> 00:19:00,360
You copy the link.

511
00:19:00,360 --> 00:19:02,490
You post it into the description.

512
00:19:02,490 --> 00:19:04,140
And if it's the only thing
you've put in your feed,

513
00:19:04,140 --> 00:19:05,610
you've posted the description,

514
00:19:05,610 --> 00:19:08,310
it's gonna do the same
parsing that it did before

515
00:19:08,310 --> 00:19:12,270
and show you a result that the description

516
00:19:12,270 --> 00:19:13,920
and the title parsed out for you.

517
00:19:18,090 --> 00:19:21,630
Okay, and then the last
thing I'm gonna show,

518
00:19:21,630 --> 00:19:23,220
talk about, is just the persistent web.

519
00:19:23,220 --> 00:19:25,560
And so this is just to
sort of underscore the fact

520
00:19:25,560 --> 00:19:27,150
that you're not just bookmarking things.

521
00:19:27,150 --> 00:19:29,550
You're also maintaining an archive

522
00:19:29,550 --> 00:19:31,530
of the things that you've interacted with.

523
00:19:31,530 --> 00:19:33,300
So I was showing this to a friend

524
00:19:33,300 --> 00:19:35,700
and I just typed in the search Corey,

525
00:19:35,700 --> 00:19:38,670
and up came four results.

526
00:19:38,670 --> 00:19:40,860
Three of them looked like they
made a ton of sense to me.

527
00:19:40,860 --> 00:19:42,720
They either were articles written by Corey

528
00:19:42,720 --> 00:19:44,190
or referred to him.

529
00:19:44,190 --> 00:19:46,170
I had recently read a book
called The Dawn of Everything

530
00:19:46,170 --> 00:19:47,040
and was a bit confused.

531
00:19:47,040 --> 00:19:48,660
I didn't know why that would pop up.

532
00:19:48,660 --> 00:19:50,880
So I went over to the Wikipedia page.

533
00:19:50,880 --> 00:19:52,500
Searched for it.

534
00:19:52,500 --> 00:19:54,120
The name wasn't there.

535
00:19:54,120 --> 00:19:55,567
So I went back and said,

536
00:19:55,567 --> 00:19:57,150
"Okay, well, what was actually indexed?"

537
00:19:57,150 --> 00:19:59,250
So I went into the
archive version instead,

538
00:19:59,250 --> 00:20:00,723
and I did a search for Corey.

539
00:20:01,710 --> 00:20:04,440
And lo and behold, the original version

540
00:20:04,440 --> 00:20:07,290
that I interacted with actually
directly mentioned Corey.

541
00:20:08,280 --> 00:20:11,370
If you're interacting with
pages that are changing,

542
00:20:11,370 --> 00:20:12,660
you have a record of the things

543
00:20:12,660 --> 00:20:14,760
at the time that you interacted with them.

544
00:20:16,410 --> 00:20:17,243
Okay.

545
00:20:18,540 --> 00:20:23,540
Let me bounce back to the slides.

546
00:20:26,070 --> 00:20:26,903
Okay.

547
00:20:28,500 --> 00:20:32,130
So I thought I'd just give a
sort of a brief explanation

548
00:20:32,130 --> 00:20:33,840
as to how this was all put together.

549
00:20:33,840 --> 00:20:35,970
It's actually pretty simple.

550
00:20:35,970 --> 00:20:39,270
What you just saw was the
browser extension in the web UI.

551
00:20:39,270 --> 00:20:42,780
There is a main API
that's running the server

552
00:20:42,780 --> 00:20:44,057
and that was written in Kotlin.

553
00:20:44,057 --> 00:20:47,853
Kotlin is sort of a more
functional and fun version of Java.

554
00:20:48,780 --> 00:20:50,820
You can use all of the
existing open source libraries

555
00:20:50,820 --> 00:20:52,140
with it and it just seemed like a,

556
00:20:52,140 --> 00:20:54,480
sort of, open source friendly choice.

557
00:20:54,480 --> 00:20:56,700
But the meat of OpenCola

558
00:20:56,700 --> 00:20:59,520
is actually all of these
components at the bottom.

559
00:20:59,520 --> 00:21:01,800
So the first one is the entity store.

560
00:21:01,800 --> 00:21:04,890
This is where all the metadata is stored

561
00:21:04,890 --> 00:21:06,510
about the posts that
you're interacting with.

562
00:21:06,510 --> 00:21:09,240
So it has all the links to
the images, the descriptions,

563
00:21:09,240 --> 00:21:11,550
and everything that
you've interacted with.

564
00:21:11,550 --> 00:21:13,500
We have a SQLite version of this

565
00:21:13,500 --> 00:21:14,820
and a Postgres version of this,

566
00:21:14,820 --> 00:21:16,920
so you can run it in process as a library

567
00:21:16,920 --> 00:21:17,753
or if you choose,

568
00:21:17,753 --> 00:21:20,640
you could host your entity
store somewhere else

569
00:21:20,640 --> 00:21:22,323
or at least scale it differently.

570
00:21:23,160 --> 00:21:24,030
There's a file store

571
00:21:24,030 --> 00:21:26,130
which stores all of
the archive information

572
00:21:26,130 --> 00:21:28,110
and it will store everything.

573
00:21:28,110 --> 00:21:30,300
That if you're sharing larger
sets of files or whatever,

574
00:21:30,300 --> 00:21:32,793
it will store any actual binary files.

575
00:21:34,080 --> 00:21:35,880
It is a hash-based store.

576
00:21:35,880 --> 00:21:37,653
So it's globally addressable.

577
00:21:38,790 --> 00:21:41,760
It stores things locally on your machine

578
00:21:41,760 --> 00:21:43,740
but it is a very simple interface.

579
00:21:43,740 --> 00:21:45,630
And you can imagine sharing one,

580
00:21:45,630 --> 00:21:47,520
and putting one up
remotely if you wanted to.

581
00:21:47,520 --> 00:21:50,370
So that you could be a bit
more efficient with it.

582
00:21:50,370 --> 00:21:52,650
The search index is just
built based off of Lucene

583
00:21:52,650 --> 00:21:54,150
which is an open source search engine.

584
00:21:54,150 --> 00:21:55,920
It's super scalable.

585
00:21:55,920 --> 00:21:58,740
There's a Lucene, one that
users direct to the library.

586
00:21:58,740 --> 00:22:00,570
So you can run it as a single process.

587
00:22:00,570 --> 00:22:02,310
But again, you can run it in Solr as well

588
00:22:02,310 --> 00:22:03,810
which is a wrap around Lucene

589
00:22:03,810 --> 00:22:07,050
and can scale to fairly
massive situations,

590
00:22:07,050 --> 00:22:09,420
and you can run it
remotely if you want to.

591
00:22:09,420 --> 00:22:11,160
And then the last piece
is the network provider

592
00:22:11,160 --> 00:22:13,890
and this allows the nodes to communicate

593
00:22:13,890 --> 00:22:15,270
with each other.

594
00:22:15,270 --> 00:22:18,210
Currently we are actually
communicating over HTTP

595
00:22:18,210 --> 00:22:20,490
but we're using ZeroTier

596
00:22:20,490 --> 00:22:22,980
which is an open source VPN solution

597
00:22:22,980 --> 00:22:25,470
that allows us to make
it look like remote peers

598
00:22:25,470 --> 00:22:26,520
are on your own network,

599
00:22:26,520 --> 00:22:29,190
which is convenient for
directly communicating

600
00:22:29,190 --> 00:22:31,350
and being able to view
other people's feeds.

601
00:22:31,350 --> 00:22:32,940
We're definitely looking
for better ways to do it.

602
00:22:32,940 --> 00:22:34,980
We're stretching what it was meant to do.

603
00:22:34,980 --> 00:22:38,010
And we want really secure
connections between people

604
00:22:38,010 --> 00:22:40,170
and it's not really meant for setting up

605
00:22:40,170 --> 00:22:42,030
as many VPNs as you have friends.

606
00:22:42,030 --> 00:22:44,333
So we're looking for a
different way to do that.

607
00:22:46,134 --> 00:22:47,520
At the core of all this, though.

608
00:22:47,520 --> 00:22:49,140
Really the core of OpenCola

609
00:22:49,140 --> 00:22:51,000
is just this data representation.

610
00:22:51,000 --> 00:22:55,200
In the database, everything
is stored as a fact.

611
00:22:55,200 --> 00:22:56,700
These are not informational facts.

612
00:22:56,700 --> 00:22:58,080
These are database facts.

613
00:22:58,080 --> 00:23:01,350
And so in a post, the
description has a fact.

614
00:23:01,350 --> 00:23:02,853
The title has a fact.

615
00:23:04,170 --> 00:23:06,030
Basically, every fact is just a two pull.

616
00:23:06,030 --> 00:23:07,800
It has these components.

617
00:23:07,800 --> 00:23:08,910
It has an authority ID

618
00:23:08,910 --> 00:23:11,460
which is essentially who
is asserting the facts.

619
00:23:11,460 --> 00:23:13,200
That would be my user ID.

620
00:23:13,200 --> 00:23:14,640
There is an entity ID

621
00:23:14,640 --> 00:23:15,900
which is what you're talking about.

622
00:23:15,900 --> 00:23:17,280
It could be a resource or a file

623
00:23:17,280 --> 00:23:19,470
or another person in the network.

624
00:23:19,470 --> 00:23:20,370
There is an attribute

625
00:23:20,370 --> 00:23:22,800
which is what property of that
entity you're talking about.

626
00:23:22,800 --> 00:23:23,850
The value.

627
00:23:23,850 --> 00:23:26,070
There's an operation about
that indicates whether

628
00:23:26,070 --> 00:23:27,900
or not you're asserting
or retracting the facts.

629
00:23:27,900 --> 00:23:29,670
And then, there's a transaction ID.

630
00:23:29,670 --> 00:23:31,680
So these facts get grouped.

631
00:23:31,680 --> 00:23:34,380
When you take an action, that
generates a set of facts.

632
00:23:34,380 --> 00:23:36,270
They get grouped into a transaction.

633
00:23:36,270 --> 00:23:37,560
That transaction is signed

634
00:23:37,560 --> 00:23:40,260
and that's the unit that's
communicated between nodes.

635
00:23:43,350 --> 00:23:45,480
This is really extendable.

636
00:23:45,480 --> 00:23:46,920
This is the type of thing
that you probably wouldn't

637
00:23:46,920 --> 00:23:48,510
wanna run in a big enterprise database.

638
00:23:48,510 --> 00:23:50,460
But for the user scale
that we're talking about,

639
00:23:50,460 --> 00:23:52,680
this is a really great choice.

640
00:23:52,680 --> 00:23:54,960
And then another just fun
fact about this database

641
00:23:54,960 --> 00:23:57,330
is because of the way that it models data,

642
00:23:57,330 --> 00:23:59,130
it's actually always accreting data.

643
00:23:59,130 --> 00:24:00,330
It's not deleting old data.

644
00:24:00,330 --> 00:24:01,560
And so if you wanted to go back

645
00:24:01,560 --> 00:24:03,630
and look at what your feed
looked like five years ago,

646
00:24:03,630 --> 00:24:05,670
you could actually just generate a query.

647
00:24:05,670 --> 00:24:07,530
Just slightly modify the
query to get the feed

648
00:24:07,530 --> 00:24:09,230
and it would produce that for you.

649
00:24:10,470 --> 00:24:12,330
Okay great, so you've seen the demo.

650
00:24:12,330 --> 00:24:15,090
You've seen a bit of how
the architecture looks like.

651
00:24:15,090 --> 00:24:16,410
So how does this help?

652
00:24:16,410 --> 00:24:18,630
I listed a bunch of
problems at the beginning.

653
00:24:18,630 --> 00:24:21,150
Well, in terms of the information silos.

654
00:24:21,150 --> 00:24:22,320
Your data is your data.

655
00:24:22,320 --> 00:24:24,870
It goes wherever you want it to go.

656
00:24:24,870 --> 00:24:26,220
You're sharing it with your friends.

657
00:24:26,220 --> 00:24:27,780
In terms of whoever you connect with,

658
00:24:27,780 --> 00:24:30,150
it's automatically
replicated to your friends.

659
00:24:30,150 --> 00:24:31,710
Your friends can see your data right now

660
00:24:31,710 --> 00:24:32,610
on social media sites.

661
00:24:32,610 --> 00:24:34,140
So that's nothing new.

662
00:24:34,140 --> 00:24:35,700
There's some nice features of that though.

663
00:24:35,700 --> 00:24:36,780
If you lose all your data,

664
00:24:36,780 --> 00:24:37,807
you can go back to your friends and say,

665
00:24:37,807 --> 00:24:39,990
"Hey, can you repopulate
my data store for me?"

666
00:24:39,990 --> 00:24:41,850
So you've got a backup of your data.

667
00:24:41,850 --> 00:24:43,560
But the point is that
you can take that data

668
00:24:43,560 --> 00:24:44,400
and put it wherever you want.

669
00:24:44,400 --> 00:24:45,510
And we imagine situations

670
00:24:45,510 --> 00:24:46,590
where you might actually want to share it

671
00:24:46,590 --> 00:24:48,570
with applications that you trust,

672
00:24:48,570 --> 00:24:50,610
and you can take that
trust away if you want to,

673
00:24:50,610 --> 00:24:53,373
but to get larger features in the future.

674
00:24:54,270 --> 00:24:56,820
In terms of the attention economy.

675
00:24:56,820 --> 00:24:58,740
There's no advertising in the system.

676
00:24:58,740 --> 00:25:00,870
It's possible that your friend
may try to advertise to you,

677
00:25:00,870 --> 00:25:04,140
but there's no advertising
component to this whatsoever

678
00:25:04,140 --> 00:25:07,740
and so there's really no
incentive to keep you in the site.

679
00:25:07,740 --> 00:25:09,090
And from my perspective,

680
00:25:09,090 --> 00:25:10,320
we're not doing anything to bug you.

681
00:25:10,320 --> 00:25:13,110
We're not SMSing you all the time

682
00:25:13,110 --> 00:25:14,310
to tell you if you've got new content.

683
00:25:14,310 --> 00:25:15,930
It's up to you if you want
to come and use the site,

684
00:25:15,930 --> 00:25:16,763
if it's useful.

685
00:25:18,630 --> 00:25:20,550
In terms of privacy abuse.

686
00:25:20,550 --> 00:25:21,390
Your data is yours

687
00:25:21,390 --> 00:25:23,880
and you should share it
with people that you trust,

688
00:25:23,880 --> 00:25:26,430
and you now are actually
in the power to do that.

689
00:25:26,430 --> 00:25:28,260
You can take it, you can revoke it.

690
00:25:28,260 --> 00:25:32,790
I think there's an opportunity
potentially here at the point

691
00:25:32,790 --> 00:25:34,680
that there are potentially
applications out there

692
00:25:34,680 --> 00:25:37,743
that want to use your data and
aggregate it in larger sets,

693
00:25:39,390 --> 00:25:41,490
you could potentially think about creating

694
00:25:41,490 --> 00:25:43,950
end service agreements
versus end user agreements.

695
00:25:43,950 --> 00:25:45,357
Meaning that, since I have all my data

696
00:25:45,357 --> 00:25:47,957
and it's valuable and it's
across every single site.

697
00:25:49,456 --> 00:25:52,410
If a service wants to use my data,

698
00:25:52,410 --> 00:25:54,660
you could conceivably create
some sort of licensing

699
00:25:54,660 --> 00:25:56,700
such that you say you're
only allowed to use it

700
00:25:56,700 --> 00:25:58,110
for these reasons,

701
00:25:58,110 --> 00:26:01,020
which is very different
than the world is today.

702
00:26:01,020 --> 00:26:03,540
So it's just something
that we'd like to explore.

703
00:26:03,540 --> 00:26:05,850
In terms of concentration
of wealth and power.

704
00:26:05,850 --> 00:26:08,790
This just provides you with
an opportunity to move away.

705
00:26:08,790 --> 00:26:11,430
So to the degree that people do move away,

706
00:26:11,430 --> 00:26:14,760
you take away the big company's power.

707
00:26:14,760 --> 00:26:16,800
And I'll talk a bit later
about how we're trying

708
00:26:16,800 --> 00:26:17,880
to make that easier to do

709
00:26:17,880 --> 00:26:18,997
'cause of course the big question is,

710
00:26:18,997 --> 00:26:20,427
"Well, will people do this?"

711
00:26:22,980 --> 00:26:24,510
In terms of misinformation.

712
00:26:24,510 --> 00:26:26,940
Every piece of data that
comes to you in the network

713
00:26:26,940 --> 00:26:28,290
comes from somebody.

714
00:26:28,290 --> 00:26:29,340
It's not random.

715
00:26:29,340 --> 00:26:31,560
It comes from somebody.

716
00:26:31,560 --> 00:26:34,920
If you see an article, say,
from the Washington Herald.

717
00:26:34,920 --> 00:26:36,480
Let's pretend that's not a real site

718
00:26:36,480 --> 00:26:38,520
and it's saying this incredible story

719
00:26:38,520 --> 00:26:39,427
and you go and look at it and say,

720
00:26:39,427 --> 00:26:40,680
"Well, wait a sec, this comes

721
00:26:40,680 --> 00:26:43,470
from some weird bot
farm that I'm aware of."

722
00:26:43,470 --> 00:26:44,730
You can mark that story.

723
00:26:44,730 --> 00:26:46,830
The point of that trust
button in the toolbar

724
00:26:46,830 --> 00:26:49,470
was that you can mark a
story as untrustworthy.

725
00:26:49,470 --> 00:26:51,210
That signals to your friends

726
00:26:51,210 --> 00:26:53,400
that if they ever come across that,

727
00:26:53,400 --> 00:26:55,590
they probably shouldn't take it seriously.

728
00:26:55,590 --> 00:26:58,950
And it can also be used to
penalize throughout the network

729
00:26:58,950 --> 00:27:01,110
at wherever that information came from.

730
00:27:01,110 --> 00:27:02,940
And obviously, you can
also just turn off people

731
00:27:02,940 --> 00:27:04,470
if they abuse your trust too much.

732
00:27:04,470 --> 00:27:05,550
But the point is that once you

733
00:27:05,550 --> 00:27:07,230
get all of this trust data in the network,

734
00:27:07,230 --> 00:27:08,280
you potentially can actually

735
00:27:08,280 --> 00:27:10,170
build some interesting algorithms

736
00:27:10,170 --> 00:27:13,473
that help you to be able
to sort of navigate things

737
00:27:13,473 --> 00:27:15,123
that you haven't seen before.

738
00:27:17,010 --> 00:27:18,570
So anyway.

739
00:27:18,570 --> 00:27:21,060
To be clear, people can
still post fake news

740
00:27:21,060 --> 00:27:22,860
but you are a moderator yourself.

741
00:27:22,860 --> 00:27:25,660
You can control whether or
not it replicates beyond you.

742
00:27:26,723 --> 00:27:28,410
In terms of polarization.

743
00:27:28,410 --> 00:27:30,990
There's less incentive to
share negative emotion.

744
00:27:30,990 --> 00:27:32,880
I definitely wouldn't claim
that people won't do it,

745
00:27:32,880 --> 00:27:35,703
but it's definitely not being
amplified by advertising.

746
00:27:37,410 --> 00:27:39,870
Don't need to say much
about stifled innovation.

747
00:27:39,870 --> 00:27:42,270
This will be an open source project soon.

748
00:27:42,270 --> 00:27:45,150
I'm just not quite at the
point of releasing everything.

749
00:27:45,150 --> 00:27:46,200
Since it will be open source,

750
00:27:46,200 --> 00:27:48,420
anybody can do whatever they want with it

751
00:27:48,420 --> 00:27:51,693
and that means you can build
whatever feature you want.

752
00:27:52,860 --> 00:27:54,840
And then in terms of bots,

753
00:27:54,840 --> 00:27:58,260
bots exist because they're
super easy to create and scale.

754
00:27:58,260 --> 00:28:00,120
You can go and create a
bunch of Facebook users.

755
00:28:00,120 --> 00:28:02,720
You can create fake links
that try and trick Google.

756
00:28:03,960 --> 00:28:05,250
And it's really cheap to do that,

757
00:28:05,250 --> 00:28:08,940
and it's an arms race to try
and figure those things out.

758
00:28:08,940 --> 00:28:10,590
There are definitely cases, especially,

759
00:28:10,590 --> 00:28:12,990
we know now with the potential
acquisition of Twitter

760
00:28:12,990 --> 00:28:14,160
where it's like there are lots of bots

761
00:28:14,160 --> 00:28:15,810
that are actually affecting
the social discourse

762
00:28:15,810 --> 00:28:18,360
and actually distorting conversation.

763
00:28:18,360 --> 00:28:19,920
But in this world to get in the network,

764
00:28:19,920 --> 00:28:21,820
you actually have to exchange a token.

765
00:28:22,800 --> 00:28:25,170
I think it would be practically impossible

766
00:28:25,170 --> 00:28:27,660
for somebody to automatically
trick enough people

767
00:28:27,660 --> 00:28:29,670
to actually have large
scale effect on the network.

768
00:28:29,670 --> 00:28:32,310
And if it does happen,
it really is your fault.

769
00:28:32,310 --> 00:28:33,750
It's not some company's fault

770
00:28:33,750 --> 00:28:35,520
who's not spending enough
time on the problem.

771
00:28:35,520 --> 00:28:38,580
It's because you let it
happen in your network.

772
00:28:38,580 --> 00:28:40,020
And then there's an
easy way to turn it off

773
00:28:40,020 --> 00:28:41,120
once you do find them.

774
00:28:43,421 --> 00:28:45,127
A question I often get about it is,

775
00:28:45,127 --> 00:28:47,127
"Well, what about information bubbles?"

776
00:28:48,330 --> 00:28:50,760
I don't know that I have a
really great answer for that.

777
00:28:50,760 --> 00:28:52,980
I don't think we directly solve this.

778
00:28:52,980 --> 00:28:55,170
I think it might be reasonable to say that

779
00:28:55,170 --> 00:28:56,610
in a world that's less polarized

780
00:28:56,610 --> 00:28:57,443
where you have more control,

781
00:28:57,443 --> 00:28:59,040
you might be more accepting of opinions

782
00:28:59,040 --> 00:29:00,603
outside your own perspective.

783
00:29:01,440 --> 00:29:02,340
I think the challenge, though,

784
00:29:02,340 --> 00:29:05,220
with information bubbles
is that they're natural.

785
00:29:05,220 --> 00:29:06,240
They've existed forever.

786
00:29:06,240 --> 00:29:07,590
I just don't think we've been

787
00:29:07,590 --> 00:29:09,870
as aware of them as we are right now,

788
00:29:09,870 --> 00:29:11,640
and I think it's really a social problem

789
00:29:11,640 --> 00:29:14,190
of people really desiring to understand.

790
00:29:14,190 --> 00:29:16,560
Build empathy and understand
people not like them,

791
00:29:16,560 --> 00:29:18,630
versus a technical problem.

792
00:29:18,630 --> 00:29:20,220
And I don't believe that any algorithm

793
00:29:20,220 --> 00:29:22,770
is ever gonna solve this
problem, personally.

794
00:29:22,770 --> 00:29:23,700
So I think that's something

795
00:29:23,700 --> 00:29:25,350
that's a bit out of bounds
of what we're doing.

796
00:29:25,350 --> 00:29:27,450
However, when we do get to the point

797
00:29:27,450 --> 00:29:29,010
where we're thinking about algorithms.

798
00:29:29,010 --> 00:29:33,210
You know, if people are willing
to share with algorithms.

799
00:29:33,210 --> 00:29:35,730
Like, help them discover new information?

800
00:29:35,730 --> 00:29:39,660
Any algorithm, you can
basically tune it in such a way.

801
00:29:39,660 --> 00:29:41,160
Sorry, you can tune it on a spectrum

802
00:29:41,160 --> 00:29:42,750
between exploring and exploiting.

803
00:29:42,750 --> 00:29:44,580
Where exploring is just
randomly giving data

804
00:29:44,580 --> 00:29:46,440
to see what people react to

805
00:29:46,440 --> 00:29:48,457
and exploiting is when the algorithm says,

806
00:29:48,457 --> 00:29:51,060
"I'm gonna give you exactly
what we think you want

807
00:29:51,060 --> 00:29:52,920
that's gonna actually get an action."

808
00:29:52,920 --> 00:29:55,720
And you can relatively easily
tune between these things.

809
00:29:56,970 --> 00:29:58,770
What I would love to see
is a case where the user

810
00:29:58,770 --> 00:30:01,170
could decide the level of explore
or exploit that they want,

811
00:30:01,170 --> 00:30:02,820
which would allow them to
experiment with their bubbles.

812
00:30:02,820 --> 00:30:05,370
And for me, I look at
something like YouTube.

813
00:30:05,370 --> 00:30:06,750
I get stuck in bubbles all the time.

814
00:30:06,750 --> 00:30:08,707
I see a friend's YouTube
feed and I'm like,

815
00:30:08,707 --> 00:30:11,340
"Why am I not seeing any of this stuff?"

816
00:30:11,340 --> 00:30:13,230
and I'd love to be able
to have control over that

817
00:30:13,230 --> 00:30:15,090
in the algorithms that I interact with.

818
00:30:15,090 --> 00:30:17,073
So we do have some thoughts on that.

819
00:30:18,780 --> 00:30:20,550
So what are we thinking about?

820
00:30:20,550 --> 00:30:23,250
Well, we're still pretty
early on in what we're doing.

821
00:30:25,050 --> 00:30:27,930
The visibility is, right now,
just restricted to your peers,

822
00:30:27,930 --> 00:30:29,430
which actually may be a good thing.

823
00:30:29,430 --> 00:30:30,990
We don't know what it's
gonna feel like for people

824
00:30:30,990 --> 00:30:32,970
once they've been using it for a while.

825
00:30:32,970 --> 00:30:34,110
But definitely, at some point,

826
00:30:34,110 --> 00:30:36,630
we wanna think about
introducing peers and so on,

827
00:30:36,630 --> 00:30:38,460
and maybe being able to
see a bit further out.

828
00:30:38,460 --> 00:30:40,830
As long as people are willing
to share their information,

829
00:30:40,830 --> 00:30:42,430
maybe more than one degree away.

830
00:30:44,100 --> 00:30:46,290
One of the big things
is how do we help people

831
00:30:46,290 --> 00:30:48,360
get off existing social media.

832
00:30:48,360 --> 00:30:51,870
There are two ideas that
we're thinking about.

833
00:30:51,870 --> 00:30:55,000
One is that because we are
not actually an application

834
00:30:56,070 --> 00:30:57,360
in the way that say Facebook is,

835
00:30:57,360 --> 00:31:00,270
we're running in the browser an extension.

836
00:31:00,270 --> 00:31:03,090
That extension can know
when you're on Facebook

837
00:31:03,090 --> 00:31:05,250
and it can know when you're
doing things on Facebook.

838
00:31:05,250 --> 00:31:09,150
So if you decide that you like
a post or you save something,

839
00:31:09,150 --> 00:31:11,940
that information can be
replicated in your own data store

840
00:31:11,940 --> 00:31:14,730
which means you can sort of
like get the value in OpenCola

841
00:31:14,730 --> 00:31:17,493
and still live on regular
social media if you want to.

842
00:31:18,450 --> 00:31:21,180
Furthermore with privacy
law, you have the right

843
00:31:21,180 --> 00:31:24,000
to request whatever data
these sites have on you.

844
00:31:24,000 --> 00:31:26,550
And so we are thinking
about writing bulk importers

845
00:31:26,550 --> 00:31:27,877
that allow you to go and say,

846
00:31:27,877 --> 00:31:28,770
"Give me all my data.

847
00:31:28,770 --> 00:31:30,060
I wanna pull this data back in."

848
00:31:30,060 --> 00:31:31,500
So that allows you to move your data over.

849
00:31:31,500 --> 00:31:34,620
And then the real question is
how do you get people over,

850
00:31:34,620 --> 00:31:36,990
and I think we can
potentially slowly map people

851
00:31:36,990 --> 00:31:39,120
between these worlds as people move over.

852
00:31:39,120 --> 00:31:40,440
But I think we've got a couple of ideas

853
00:31:40,440 --> 00:31:41,970
about how to make it a bit easier

854
00:31:41,970 --> 00:31:43,420
than it has been in the past.

855
00:31:44,880 --> 00:31:47,340
We're also thinking about identity.

856
00:31:47,340 --> 00:31:50,880
I think one of my big criticisms of, say,

857
00:31:50,880 --> 00:31:52,560
something like Facebook
is that I post something

858
00:31:52,560 --> 00:31:53,610
that is about my family

859
00:31:53,610 --> 00:31:56,190
and then every random person
I know, sort of, sees it,

860
00:31:56,190 --> 00:31:58,050
and I'm kind of like, that's
not really what I want.

861
00:31:58,050 --> 00:32:00,180
I want to be able to
create my own communities.

862
00:32:00,180 --> 00:32:01,560
I want my family community.

863
00:32:01,560 --> 00:32:02,880
I want my tech community.

864
00:32:02,880 --> 00:32:04,380
I want my music community.

865
00:32:04,380 --> 00:32:06,570
And I wanna be able to control
where that information goes,

866
00:32:06,570 --> 00:32:07,800
and so I think that's something

867
00:32:07,800 --> 00:32:09,467
that is definitely not
in the best interest

868
00:32:09,467 --> 00:32:11,250
of the existing social media companies.

869
00:32:11,250 --> 00:32:12,870
But I think there's actually
a reasonable solution

870
00:32:12,870 --> 00:32:14,730
where you can have an integrated feed

871
00:32:14,730 --> 00:32:15,563
but when you post things,

872
00:32:15,563 --> 00:32:17,340
you choose what identity it goes to.

873
00:32:17,340 --> 00:32:20,070
You can copy things between
those identities and so on,

874
00:32:20,070 --> 00:32:22,853
but I think that would be
improvement over what we have now.

875
00:32:24,810 --> 00:32:26,280
We're looking at building
a better mobile app,

876
00:32:26,280 --> 00:32:27,900
maybe, that allows you
to bump with somebody

877
00:32:27,900 --> 00:32:29,730
to be able to exchange
your contact details

878
00:32:29,730 --> 00:32:33,633
and allow a bit more rich
functionality on a mobile device.

879
00:32:34,770 --> 00:32:35,760
I think one of the things

880
00:32:35,760 --> 00:32:37,680
that I would be super excited about

881
00:32:37,680 --> 00:32:39,660
would be if we do get to a point

882
00:32:39,660 --> 00:32:41,070
where there are lots of people

883
00:32:41,070 --> 00:32:42,960
and they're willing to share their data

884
00:32:42,960 --> 00:32:44,430
with trusted applications.

885
00:32:44,430 --> 00:32:46,470
I imagine maybe an academic institution,

886
00:32:46,470 --> 00:32:48,090
and saying, "Hey, you
know what we'll do for you

887
00:32:48,090 --> 00:32:50,100
is if you give us your data,

888
00:32:50,100 --> 00:32:53,910
we can do some research and
understand social structure

889
00:32:53,910 --> 00:32:55,080
and opinion and that kind of stuff.

890
00:32:55,080 --> 00:32:58,260
And in return, we'll
host algorithms for you.

891
00:32:58,260 --> 00:33:00,600
So that you can get
broader algorithmic results

892
00:33:00,600 --> 00:33:02,580
to help you find new information."

893
00:33:02,580 --> 00:33:04,890
And in that world, because your data...

894
00:33:04,890 --> 00:33:06,510
that OpenCola has access to data

895
00:33:06,510 --> 00:33:08,370
across all the systems that you're on.

896
00:33:08,370 --> 00:33:09,690
It's got a much richer set of data

897
00:33:09,690 --> 00:33:11,403
than any one particular site does.

898
00:33:13,470 --> 00:33:15,570
Since all that data is
shared with this system

899
00:33:15,570 --> 00:33:17,250
and it even knows things
that you've done on YouTube,

900
00:33:17,250 --> 00:33:19,020
you could potentially
build a YouTube recommender

901
00:33:19,020 --> 00:33:22,230
that did not even involve YouTube.

902
00:33:22,230 --> 00:33:24,570
And so that algorithm could
produce different results

903
00:33:24,570 --> 00:33:26,760
for you that were
optimized for your purpose

904
00:33:26,760 --> 00:33:29,610
versus YouTube's purpose.

905
00:33:29,610 --> 00:33:32,490
And I think similarly, you
could apply this idea to Amazon.

906
00:33:32,490 --> 00:33:33,590
When I go onto Amazon,

907
00:33:34,470 --> 00:33:35,640
I look at all the star ratings.

908
00:33:35,640 --> 00:33:37,080
I have no idea who those people are.

909
00:33:37,080 --> 00:33:38,760
I don't really know if they're bots.

910
00:33:38,760 --> 00:33:40,440
I don't know if they're like me.

911
00:33:40,440 --> 00:33:41,850
But you could potentially use this data

912
00:33:41,850 --> 00:33:44,850
to help recommend even products for you,

913
00:33:44,850 --> 00:33:46,800
which sometimes is a good thing.

914
00:33:46,800 --> 00:33:49,410
And they could probably
do a more authentic job

915
00:33:49,410 --> 00:33:50,580
of doing that than Amazon would,

916
00:33:50,580 --> 00:33:52,863
in terms of trying to
maximize their revenue.

917
00:33:54,870 --> 00:33:56,670
We're also thinking in terms of the notion

918
00:33:56,670 --> 00:33:58,500
between publishers and consumers.

919
00:33:58,500 --> 00:34:00,360
So obviously in the
social media world today,

920
00:34:00,360 --> 00:34:01,193
there are a lot of people

921
00:34:01,193 --> 00:34:02,760
that just publish out data for people

922
00:34:02,760 --> 00:34:04,650
and so it's relatively
straightforward for us

923
00:34:04,650 --> 00:34:07,740
to allow people to create
a one way connection.

924
00:34:07,740 --> 00:34:10,020
We just haven't done this yet.

925
00:34:10,020 --> 00:34:11,460
I think it'd be super interesting

926
00:34:11,460 --> 00:34:16,410
if we could even get newspapers
to be interested in this,

927
00:34:16,410 --> 00:34:17,820
and if they were signing the information

928
00:34:17,820 --> 00:34:19,950
that they were propagating
through the network.

929
00:34:19,950 --> 00:34:21,240
The more people you got doing that,

930
00:34:21,240 --> 00:34:23,400
the harder it would be to
inject fake information

931
00:34:23,400 --> 00:34:25,353
from unknown sources into the system.

932
00:34:27,420 --> 00:34:28,920
I'll skip over the
leveraging of the network

933
00:34:28,920 --> 00:34:30,030
to enable trust algorithms.

934
00:34:30,030 --> 00:34:31,880
I think I've sort of touched on that.

935
00:34:34,152 --> 00:34:37,350
The other thing is that this
is really just a platform

936
00:34:37,350 --> 00:34:38,183
for doing these things.

937
00:34:38,183 --> 00:34:40,140
Everything I've shown you
is meant to be pluggable

938
00:34:40,140 --> 00:34:42,480
and everything I've shown you
is meant to be extendable.

939
00:34:42,480 --> 00:34:44,790
And I think on top of what's been built,

940
00:34:44,790 --> 00:34:45,870
it would be fairly straightforward

941
00:34:45,870 --> 00:34:49,170
to build an application ecosystem

942
00:34:49,170 --> 00:34:50,820
and even a recommender rig ecosystem.

943
00:34:50,820 --> 00:34:54,630
All the data is standardized
and the APIs are standardized.

944
00:34:54,630 --> 00:34:56,370
Anybody could make what
they want out of this.

945
00:34:56,370 --> 00:34:57,203
It could be different.

946
00:34:57,203 --> 00:34:58,980
Somebody could make it
look exactly like Facebook

947
00:34:58,980 --> 00:35:00,690
if they really wanted to,

948
00:35:00,690 --> 00:35:02,670
but having some basic
tooling at the bottom

949
00:35:02,670 --> 00:35:04,950
to enable that would be great.

950
00:35:04,950 --> 00:35:07,263
The last thing is search filters.

951
00:35:08,370 --> 00:35:11,670
So right now there's
probably many different ways

952
00:35:11,670 --> 00:35:13,080
you wanna look at your data,

953
00:35:13,080 --> 00:35:15,660
and so enabling rich search filters

954
00:35:15,660 --> 00:35:17,910
where you can actually specify people

955
00:35:17,910 --> 00:35:20,047
and tags and keywords and say,

956
00:35:20,047 --> 00:35:22,080
"This is my name filter that does this,"

957
00:35:22,080 --> 00:35:24,330
allows you to build your
own feeds basically.

958
00:35:26,110 --> 00:35:27,210
Okay, so now what?

959
00:35:27,210 --> 00:35:30,090
You've seen sort of
where we're at right now.

960
00:35:30,090 --> 00:35:31,650
Just to reiterate,

961
00:35:31,650 --> 00:35:33,420
we believe this is problems we're solving

962
00:35:33,420 --> 00:35:35,430
whether or not we make any money.

963
00:35:35,430 --> 00:35:37,050
We're not looking for VC funding.

964
00:35:37,050 --> 00:35:39,630
We are a nonprofit company.

965
00:35:39,630 --> 00:35:42,210
We wanna work with other
smart, cool people.

966
00:35:42,210 --> 00:35:46,680
We're looking for debate,
collaboration, and other ideas.

967
00:35:46,680 --> 00:35:47,880
We're specifically interested

968
00:35:47,880 --> 00:35:51,030
in alternatives to the network providers.

969
00:35:51,030 --> 00:35:54,090
Tor is probably the next
thing I wanna take a look at

970
00:35:54,090 --> 00:35:55,980
but there may be other options,

971
00:35:55,980 --> 00:35:57,420
that just allow two nodes

972
00:35:57,420 --> 00:35:59,250
that are sort of arbitrarily
placed on the internet

973
00:35:59,250 --> 00:36:00,840
to be able to talk to each other.

974
00:36:00,840 --> 00:36:04,053
And so, just want to explore
and have some options there.

975
00:36:05,220 --> 00:36:07,230
Peer-2-peer systems also
have some challenges

976
00:36:07,230 --> 00:36:09,270
with the fact that nodes
are, you know, transient.

977
00:36:09,270 --> 00:36:10,140
They're not always online.

978
00:36:10,140 --> 00:36:10,973
And so the question is,

979
00:36:10,973 --> 00:36:13,350
how do you make create
a seamless experience

980
00:36:13,350 --> 00:36:16,560
like you get with sort of
centralized control right now?

981
00:36:16,560 --> 00:36:18,060
We do have, as I mentioned,

982
00:36:18,060 --> 00:36:19,830
all the data is replicated
amongst your peers.

983
00:36:19,830 --> 00:36:22,080
An easy solution is
potentially to delegate

984
00:36:22,080 --> 00:36:23,460
to your peers when you're offline,

985
00:36:23,460 --> 00:36:25,260
and that can allow you to get updated

986
00:36:25,260 --> 00:36:26,640
even if you're offline,

987
00:36:26,640 --> 00:36:28,020
but I think there's other cases.

988
00:36:28,020 --> 00:36:29,670
Things that would be interesting too

989
00:36:29,670 --> 00:36:31,920
like creating dead drops
to pass information around

990
00:36:31,920 --> 00:36:34,260
or just having some other storage

991
00:36:34,260 --> 00:36:36,390
that allows for information to persist

992
00:36:36,390 --> 00:36:37,890
outside of your actual system.

993
00:36:39,480 --> 00:36:41,130
So that's all I had to talk about today.

994
00:36:41,130 --> 00:36:43,260
Hopefully you enjoyed it.

995
00:36:43,260 --> 00:36:48,260
If you want to ask any questions
or if you have any ideas,

996
00:36:48,360 --> 00:36:51,270
I'll be out front or you can reach me

997
00:36:51,270 --> 00:36:54,030
at john.midgley@opencola.io.

998
00:36:54,030 --> 00:36:54,903
Thank you.

999
00:36:54,903 --> 00:36:58,153
(attendees applauding)