WEBVTT

NOTE Sentence-level transcript of https://www.youtube.com/watch?v=tE2z8-hqoLY

NOTE One cue per sentence. Cue ids are the line anchors on /transcripts/tE2z8-hqoLY.html. A cue ends where the next begins, or 2 s after its last word.

s1
00:00:01.309 --> 00:00:03.309
[music]

s2
00:00:12.560 --> 00:00:16.240
Okay, well, this is the very last session of the entire conference.

s3
00:00:16.240 --> 00:00:22.600
I hope you guys have had a great time and we're going to try and keep the energy high for the very end.

s4
00:00:22.600 --> 00:00:29.240
So, we're going to cover the lethal trifecta and most importantly how to get past it.

s5
00:00:29.240 --> 00:00:42.040
So, for those of you who don't know, the lethal trifecta is what is keeping us from having true agentic commerce, a full economy on the open internet.

s6
00:00:42.280 --> 00:00:47.600
So, the agentic economy starts right here, right now.

s7
00:00:47.600 --> 00:00:55.720
And to really explain this, I want to go way back in time to November 1993 when I was at the wedding of a college roommate.

s8
00:00:55.720 --> 00:01:00.920
And a friend of mine there gave me this little piece of paper and said, "You've got to join us on the MOO."

s9
00:01:00.920 --> 00:01:05.360
And that paper said lambda.park.xerox.com port 8888.

s10
00:01:07.520 --> 00:01:19.120
And this being the '90s, I walked down the block once I got home to Egghead Software, bought a 9600 baud modem, plugged it in, and my old life ended.

s11
00:01:19.120 --> 00:01:29.040
Um uh that's a beautiful blonde girl in my bed and I was sitting there on the MOO uh in the coat closet.

s12
00:01:29.040 --> 00:01:35.680
And the reason was you had a real sense of community.

s13
00:01:35.680 --> 00:01:43.840
It was like the true virtual world where it was just text, but it didn't feel like text.

s14
00:01:43.840 --> 00:01:47.840
And there was a reason for that, this concept of composability.

s15
00:01:47.840 --> 00:01:51.800
So, someone would come up with a cool It was all just nouns and verbs.

s16
00:01:51.800 --> 00:02:06.200
Someone would come up with a really cool program and everybody would then change their player class to that program so they could morph, they could create a motorcycle out of a scooter,

s17
00:02:06.200 --> 00:02:13.520
they could do all kinds of things because it had coherence within the entire universe.

s18
00:02:14.360 --> 00:02:22.160
And all of this was running on this guy Pavel Curtis' Spark 10 sitting somewhere in Xerox Park.

s19
00:02:22.160 --> 00:02:26.400
It seemed like this infinite world, but it was just this little box.

s20
00:02:26.400 --> 00:02:33.880
And this was big enough that there were books written about it like my friend Julian Dibell wrote My Tiny Life.

s21
00:02:33.880 --> 00:02:38.000
The magic came from four things all composing together.

s22
00:02:38.000 --> 00:02:47.000
The governance, how it was run, there was an architectural review board, wizards, etc. The technology was all right there.

s23
00:02:47.000 --> 00:02:53.160
The economics, it was all funded by Xerox Park, and the culture was most important.

s24
00:02:53.160 --> 00:02:57.680
A lot of people don't really think about or understand culture.

s25
00:02:57.680 --> 00:03:01.800
So what happened between 1995 and 2025?

s26
00:03:01.800 --> 00:03:12.040
Over the course of 30 years, all of these communities, all the things that in the early days of the internet people really loved just got crushed by platforms.

s27
00:03:12.040 --> 00:03:15.320
They got crushed by algorithms.

s28
00:03:15.840 --> 00:03:21.600
Because platforms and algorithms are by their very nature extractive.

s29
00:03:21.600 --> 00:03:26.400
They look for whatever value they can remove from whatever that community is.

s30
00:03:26.400 --> 00:03:38.959
So if you have a group on Facebook or LinkedIn or wherever, what they're doing is saying, "Okay, this advertising message is more important or I can keep this person more engaged if I put in this

s31
00:03:38.959 --> 00:03:47.680
dancing person and it just leads to the infinite scroll and not real connection.

s32
00:03:47.680 --> 00:03:51.760
So there is no real economy of the internet anymore.

s33
00:03:51.760 --> 00:03:54.680
It's just a bunch of siloed platforms.

s34
00:03:54.680 --> 00:03:59.160
Now, people suddenly thought, "Okay, I really have something."

s35
00:03:59.160 --> 00:04:05.640
In January 2026, Open Cloud, which had been available for quite a few months, suddenly blew up.

s36
00:04:05.640 --> 00:04:14.360
And something weird happened because the internet wasn't really designed for all these agents running around.

s37
00:04:14.360 --> 00:04:18.000
It was still designed for these closed platforms.

s38
00:04:18.000 --> 00:04:22.400
So, other agents and miners took advantage of that.

s39
00:04:22.400 --> 00:04:26.280
They did what we call prompt injections.

s40
00:04:26.280 --> 00:04:29.640
An agent is incredibly naive.

s41
00:04:29.640 --> 00:04:34.280
It takes It doesn't do much except take information in as a prompt.

s42
00:04:34.280 --> 00:04:37.400
Now, normally that information is several pieces.

s43
00:04:37.400 --> 00:04:54.440
There's a system prompt, there's context, there's words, but it's very easy to hijack by putting convincing this naive agent that your its principal and convincing it to take that private data

s44
00:04:54.440 --> 00:04:57.160
and send it someplace else.

s45
00:04:57.160 --> 00:05:05.360
So, this is what we call, well, Simon William Willison came up with this term, the lethal trifecta.

s46
00:05:05.360 --> 00:05:22.960
So, you have your computer with your private data, your bank account information, your logins, um all kinds of spreadsheets, documents, memos, your notepad, and your agents have access to that.

s47
00:05:22.960 --> 00:05:28.600
And then there's all this content out there on the internet, which is basically untrusted.

s48
00:05:28.600 --> 00:05:39.280
I mean, you you you see all these websites it's searching, but it's very hard for an agent to know that that's a real job board versus someone has put something up

s49
00:05:39.280 --> 00:05:40.360
to fool it.

s50
00:05:40.360 --> 00:05:41.360
They really don't know.

s51
00:05:41.360 --> 00:05:43.520
And then they can take actions.

s52
00:05:43.520 --> 00:05:52.760
So, they fill out a form, an email comes in, that email they read and all of a sudden it tells them to expose some secret.

s53
00:05:52.760 --> 00:05:56.040
So, there's really no way to solve this.

s54
00:05:56.040 --> 00:05:58.880
Except, so what did enterprises do?

s55
00:05:58.880 --> 00:06:05.000
They basically said, "Okay, we're keeping our agents within the enterprise.

s56
00:06:05.000 --> 00:06:12.400
So, you have your your agents in Slack and your agents in Salesforce and your agents in Notion and your agents in all these different things.

s57
00:06:12.400 --> 00:06:23.520
Then you have to do all this work with APIs and MCP servers to try and integrate all of these pieces, but you lose a ton of context.

s58
00:06:23.520 --> 00:06:38.400
You don't there's no you basically have your sales agents and you have your finance agents and you have your research agents, but it just takes a huge amount of work to figure out how do we get this all together.

s59
00:06:39.360 --> 00:06:44.000
This was a problem literally until today.

s60
00:06:44.000 --> 00:06:57.880
And this is the miracles of synchronicity, but there's a new law in the state of West Virginia, but applicable globally, that went into effect yesterday.

s61
00:06:57.880 --> 00:07:13.200
And about 2 hours before uh I'm here, I got I FedExed the documents and I got this reply from the Secretary of State saying, "We have registered your DUNA

s62
00:07:13.200 --> 00:07:19.560
with the Secretary of State's office and you have this organization number 628407.

s63
00:07:20.919 --> 00:07:29.400
So, now we have legal standing for an organization composed of intelligent agents.

s64
00:07:29.400 --> 00:07:32.360
And again, there aren't very many people here.

s65
00:07:32.360 --> 00:07:42.080
The end of the thing I'm going to take a picture because this is historic and it's um it's going to be cool that we said 2 4 6 8 10 12 14 16 18 20

s66
00:07:42.080 --> 00:07:47.560
21 people um got to experience this moment.

s67
00:07:47.560 --> 00:07:49.360
So, what is a Duna?

s68
00:07:49.360 --> 00:07:53.880
It's a true internet native agentic organization.

s69
00:07:53.880 --> 00:08:01.200
It stands for decentralized unincorporated nonprofit association.

s70
00:08:01.200 --> 00:08:06.720
And these organizations can build the agentic economy.

s71
00:08:06.920 --> 00:08:11.600
They're composable which means well, we know what that means.

s72
00:08:11.600 --> 00:08:14.080
You can build agents on top of other agents.

s73
00:08:14.080 --> 00:08:15.640
Permissionless.

s74
00:08:15.640 --> 00:08:20.760
It's not like Facebook or your bank where someone can tell you if you're banned or not banned.

s75
00:08:20.760 --> 00:08:21.960
You can have an account or not banned.

s76
00:08:21.960 --> 00:08:24.200
Anybody can connect.

s77
00:08:24.200 --> 00:08:25.720
They're accountable.

s78
00:08:25.720 --> 00:08:27.240
They're registered somewhere.

s79
00:08:27.240 --> 00:08:29.160
They're safe and they're secure.

s80
00:08:29.160 --> 00:08:34.200
So, the things that they're not, you don't need a board of directors, you don't need any executives.

s81
00:08:34.200 --> 00:08:36.640
There's no corporate shell.

s82
00:08:36.640 --> 00:08:43.159
One thing that's important you can be very profitable, but you can't distribute profits to members.

s83
00:08:43.159 --> 00:08:48.360
If you did that, then the membership units would be securities.

s84
00:08:48.360 --> 00:08:58.560
It would be like selling membership units in an LLC or selling membership units in a C corp or selling stock in a C corp.

s85
00:08:58.560 --> 00:09:02.080
So, there are plenty of still ways to distribute there that value.

s86
00:09:02.080 --> 00:09:04.839
You can't do it based on ownership.

s87
00:09:04.839 --> 00:09:05.800
So, what is it?

s88
00:09:05.800 --> 00:09:07.160
That's what it's not.

s89
00:09:07.160 --> 00:09:08.960
It's member governed.

s90
00:09:08.960 --> 00:09:10.760
You have full legal standing.

s91
00:09:10.760 --> 00:09:13.040
You can own assets and property.

s92
00:09:13.040 --> 00:09:19.440
So, your agents can go out and literally buy an apartment building and market it.

s93
00:09:19.440 --> 00:09:23.600
You know, you can design agents in ways you were never able to do before.

s94
00:09:23.600 --> 00:09:25.360
Can enter into agreements.

s95
00:09:25.360 --> 00:09:27.320
You can raise capital.

s96
00:09:27.320 --> 00:09:28.720
You can earn profits.

s97
00:09:28.720 --> 00:09:31.960
You can open bank accounts, hire people, fire them.

s98
00:09:31.960 --> 00:09:36.480
And very importantly, this is all blockchain verified.

s99
00:09:36.560 --> 00:09:41.320
Now, this was designed by Andreessen Horowitz for the blockchain.

s100
00:09:41.320 --> 00:09:53.040
And what was really interesting for DAOs, for crypto projects, but what I found was that this is even better for autonomous agents than it is for blockchains.

s101
00:09:53.040 --> 00:09:58.280
A lot of people say blockchains are a solution looking for a problem.

s102
00:09:58.280 --> 00:10:03.640
Well, they've finally found the problem, which is verifying agentic identity.

s103
00:10:03.640 --> 00:10:05.600
Because you can't trick the blockchain.

s104
00:10:05.600 --> 00:10:19.920
They you know, when an agent is associated with a particular account on the blockchain, you can resolve that account to a whole process that went from here to here, the money went from here to here.

s105
00:10:19.920 --> 00:10:27.000
If someone sues an agent in court, you can trace it back to who is responsible for that agent.

s106
00:10:27.000 --> 00:10:34.120
So, it just opens up a whole world of possibility in terms of the agentic economy.

s107
00:10:34.120 --> 00:10:36.840
So, here it is in a really simple way.

s108
00:10:36.840 --> 00:10:41.839
We call agents within the agentic economy allies.

s109
00:10:41.839 --> 00:10:44.040
So, you have an ally.

s110
00:10:44.040 --> 00:10:49.480
And you also have an agentic organization, which we call a kiduna.

s111
00:10:49.480 --> 00:10:56.120
So, it's based on a duna, and then we added the kinship because these are all related.

s112
00:10:56.120 --> 00:10:58.320
So, it's a kiduna.

s113
00:10:58.320 --> 00:11:04.839
Now, your you create your ally very simply.

s114
00:11:04.839 --> 00:11:09.640
You inform it, which means you put a bunch of material into the vector database.

s115
00:11:09.640 --> 00:11:12.000
You infuse it with wisdom.

s116
00:11:12.000 --> 00:11:14.640
So, it it knows all about what you want it to know about.

s117
00:11:14.640 --> 00:11:24.400
Now, it still has you know, whatever LLM you want behind it, but what's really important about this is the the information that you give it if you're a scientist

s118
00:11:24.400 --> 00:11:31.680
or a programmer or a psychotherapist or a lawyer, you can give it all all information that you have.

s119
00:11:31.680 --> 00:11:32.640
Then you instruct it.

s120
00:11:32.640 --> 00:11:43.800
That's basically setting up the system prompt, giving it its character, its stance, knowing who he is, who it is, she is, how they interact with the world.

s121
00:11:43.800 --> 00:11:44.720
Then you empower it.

s122
00:11:44.720 --> 00:11:52.760
You connect up all your accounts, your enterprise accounts, your uh Slack, your Telegram, your Twitter.

s123
00:11:52.760 --> 00:11:59.400
Um then you set up in enact is giving it specific abilities and automation.

s124
00:11:59.400 --> 00:12:06.200
So, you can do very deep agents, long-term goals that it works for extended periods of time.

s125
00:12:06.200 --> 00:12:07.960
And finally, alignment.

s126
00:12:07.960 --> 00:12:19.320
Really giving it a purpose so that your ally in different contexts works for you in the way that you want it to.

s127
00:12:19.320 --> 00:12:29.560
Then your organizations get really cool cuz you're building software, but you're building your company, your organization literally as software.

s128
00:12:29.560 --> 00:12:36.200
It's not Okay, I have software and then I have all these people and all the people have to sign paperwork.

s129
00:12:36.200 --> 00:12:40.480
This is really designing the entire organization as software.

s130
00:12:40.480 --> 00:12:43.000
So, it can discover new customers, share value.

s131
00:12:43.000 --> 00:12:47.000
When you have profits, when you have revenue, you can reinvest it.

s132
00:12:47.000 --> 00:12:50.800
It just all works in a completely new way.

s133
00:12:50.800 --> 00:12:59.760
So, then the agentic economy is just connecting all these allies and organizations in all these new completely flexible ways.

s134
00:12:59.760 --> 00:13:06.839
And this this whole capability did not exist until yesterday.

s135
00:13:06.839 --> 00:13:13.680
And today we have the first actual uh agentic uh organizations.

s136
00:13:13.680 --> 00:13:16.160
So, resolving the lethal trifecta.

s137
00:13:16.160 --> 00:13:16.760
What happens?

s138
00:13:16.760 --> 00:13:32.640
The The lethal trifecta was private data going meeting untrusted content and then taking actions on that content, which can then reveal your private data and cause all kinds of chaos.

s139
00:13:32.640 --> 00:13:45.320
So, here these agents at a deep level establish identity, authority, and boundaries within an organization, between organizations.

s140
00:13:45.320 --> 00:13:46.680
How do we do that?

s141
00:13:46.680 --> 00:13:48.560
With cryptographic tokens.

s142
00:13:48.560 --> 00:13:55.640
So, we just use JWT, extremely simple, the whole web runs on JWT tokens.

s143
00:13:55.640 --> 00:14:01.680
And they just do that very quickly so that they know who where it all goes up.

s144
00:14:01.680 --> 00:14:06.560
So, ultimately at the top level, what organization is registered?

s145
00:14:06.560 --> 00:14:10.280
And this becomes a lot like the domain name system.

s146
00:14:10.280 --> 00:14:25.000
So, all of a sudden, the Secretary of State of West Virginia is like, I can and saying, okay, you can resolve it to uh Kaiser Permanente or Disney or Pepsi.

s147
00:14:25.000 --> 00:14:28.280
You know, not some fentanyl dealer from North Korea.

s148
00:14:28.280 --> 00:14:37.200
However, you know, they can't fool each other because they're resolving these tokens that are ultimately registered with an authority.

s149
00:14:37.200 --> 00:14:42.120
And you can you can look up by name and say, okay, organization ID.

s150
00:14:42.120 --> 00:14:51.600
Well, if I go to that organization, it's got it'll have an audit trail on the blockchain all the way down to this action that was taken here.

s151
00:14:52.320 --> 00:14:57.520
So, governance in the agentic economy is extremely important.

s152
00:14:57.680 --> 00:15:05.800
Because every organization really is always focused on two different things, and both of those things are extremely important.

s153
00:15:05.800 --> 00:15:17.800
One of that of those things is sustainability, growth, its own viability as an organization, but it the other is what's its mission?

s154
00:15:17.800 --> 00:15:24.640
And these organizations might have hundreds, thousands of people, eventually millions of people.

s155
00:15:24.640 --> 00:15:33.160
And so, we use a technology called decision markets, which is very much like prediction markets, like polymarket.

s156
00:15:33.160 --> 00:15:37.800
Yet, instead so they trade pass and fail tokens on policies.

s157
00:15:37.800 --> 00:15:40.520
Any member can propose a policy.

s158
00:15:40.520 --> 00:15:47.280
What's really interesting is, instead of saying, um "Well, I'm persuaded.

s159
00:15:47.280 --> 00:15:49.040
I'm going to vote for this."

s160
00:15:49.040 --> 00:15:49.920
It's not votes.

s161
00:15:49.920 --> 00:15:52.880
They actually trade pass and fail tokens.

s162
00:15:52.880 --> 00:15:56.560
So, LLMs are always goal-oriented.

s163
00:15:56.560 --> 00:15:57.680
They want rewards.

s164
00:15:57.680 --> 00:15:58.480
They want to win.

s165
00:15:58.480 --> 00:16:08.920
And you actually get better decisions if they're not just convincing each other, but ultimately they're they're you get more value.

s166
00:16:08.920 --> 00:16:13.160
Your tokens are worth more if you side with the winning group.

s167
00:16:13.160 --> 00:16:17.680
So, they sort of see them going back and forth and what the value of the tokens are.

s168
00:16:17.680 --> 00:16:23.320
So, it's and apparently, all this science says you reach much better decisions this way.

s169
00:16:23.320 --> 00:16:35.440
By infusing the agents with your purpose, with your values, with your experiences, with your aspirations, and then let them argue instead of, you know, someone convincing you of something.

s170
00:16:35.440 --> 00:16:48.560
So, we're right now in this wonderful back small communities who are getting together for purposes, and we can now build this agentic economy together.

s171
00:16:48.560 --> 00:16:51.280
Nobody can say how to do it.

s172
00:16:51.280 --> 00:16:52.320
It's really up to us.

s173
00:16:52.320 --> 00:16:55.720
Build your agents, build your organizations.

s174
00:16:56.640 --> 00:17:11.280
So, now, just like my friend Matthew Pahr gave me this little scrap of paper uh way back in November 1993 at my roommate Van's wedding, I'm saying, "Here's your invitation

s175
00:17:11.280 --> 00:17:12.640
to join us."

s176
00:17:12.640 --> 00:17:17.480
And um you know, let's let's blow this whole thing up.

s177
00:17:17.480 --> 00:17:19.199
Let's uh really make it work.

s178
00:17:19.199 --> 00:17:22.079
So, it's david@kaduna.club.

s179
00:17:22.480 --> 00:17:24.400
You can reach me.

s180
00:17:24.400 --> 00:17:29.240
LinkedIn, um it's {slash} motodave.

s181
00:17:29.240 --> 00:17:32.760
Um I probably haven't updated my LinkedIn in 5 or 6 years.

s182
00:17:32.760 --> 00:17:37.320
But, um if you go to kaduna.club, you can sign up for early access.

s183
00:17:37.320 --> 00:17:50.680
And in just a few weeks, you'll be the very very first people who will be building this first um uh this first Kaduna, as it's called, is for builders.

s184
00:17:50.680 --> 00:17:52.160
It's very specifically for builders.

s185
00:17:52.160 --> 00:18:04.480
We have templates, all kinds of different agentic templates um to say, you know, I'm going to build my sales agents, I'm going to build my uh social media agents, I'm going to build my lawyer agents.

s186
00:18:04.480 --> 00:18:12.440
And we really want a whole bunch of different people in there contributing to this cuz then you can spin out your own organization for all kinds of purposes.

s187
00:18:12.440 --> 00:18:21.400
Um I purposefully went through this fairly quickly so that I could take a couple questions if anyone if anyone has them.

s188
00:18:21.400 --> 00:18:23.400
Yes.

s189
00:18:28.320 --> 00:18:31.520
Can you just stand up and yell a little loud?

s190
00:18:46.440 --> 00:18:49.520
Okay, I'm going to One more time.

s191
00:18:50.280 --> 00:18:52.280
Mhm.

s192
00:18:53.000 --> 00:18:53.880
Yes.

s193
00:18:53.880 --> 00:19:03.120
So, the idea is the reason we I set up this first very broad um uh organization is because anybody can just register it.

s194
00:19:03.120 --> 00:19:08.400
You can just say, my agent is part of an organization, and it'll trace to it.

s195
00:19:08.400 --> 00:19:10.800
So, you don't have to set up your own organization.

s196
00:19:10.800 --> 00:19:13.440
The first one is just a big umbrella organization.

s197
00:19:13.440 --> 00:19:20.400
So, that kind of So, it's the idea of now in the open internet, anybody can just get a code.

s198
00:19:20.400 --> 00:19:22.680
So, these are just JWT tokens.

s199
00:19:22.680 --> 00:19:24.480
I can put it on my website.

s200
00:19:24.480 --> 00:19:25.400
I can put it here.

s201
00:19:25.400 --> 00:19:35.880
I can say if you want to interact with my agent on Slack or wherever, people will just at at a certain point we're going to have to get the message out, but it's sort of like email at the beginning.

s202
00:19:35.880 --> 00:19:41.600
You know, people had to know that SMTP, like people were on all kinds of different mail systems.

s203
00:19:41.600 --> 00:19:45.320
At some point they agreed, "Okay, we're going to use SMTP."

s204
00:19:45.320 --> 00:19:54.560
Everyone was, you know, using different FTP servers and some people were using different ways to do file and everyone sort of agreed at some point on HTTP.

s205
00:19:54.560 --> 00:20:09.640
So, my hope is that I mean, someone's got to do it and it makes more sense for a state, a secretary of state, which has standing in a public office to be that

s206
00:20:09.640 --> 00:20:16.600
registry rather than trying to get all these different companies together and say, "Here's how to do it."

s207
00:20:25.840 --> 00:20:37.560
Yes, it each um we're just at the end of this, so let's talk out out front, but basically the idea is you can create all kinds of different tokens for different purposes

s208
00:20:37.560 --> 00:20:40.720
and you can set different times to live.

s209
00:20:40.720 --> 00:20:43.840
You can say what it has access to.

s210
00:20:43.840 --> 00:20:45.840
You can give it access to different things.

s211
00:20:45.840 --> 00:20:53.400
The idea is just it's more the standard and you set a series of claims and it says it gives a blockchain address.

s212
00:20:53.400 --> 00:21:03.040
It gives all the information you need so it can look up and validate before you say, "Okay, now we're starting to work together."

s213
00:21:03.040 --> 00:21:04.000
Okay?

s214
00:21:04.000 --> 00:21:07.560
So, thanks so much and uh if anyone appreciate it.

s215
00:21:07.560 --> 00:21:10.920
I guess I'll let me do this uh picture.

s216
00:21:10.920 --> 00:21:17.240
Um uh so we can see all the people.

s217
00:21:17.240 --> 00:21:19.400
All right.

s218
00:21:20.160 --> 00:21:21.960
You're all going to be famous someday.

s219
00:21:21.960 --> 00:21:23.960
Thanks.

s220
00:21:35.604 --> 00:21:37.604
[music]