00:00:00.049 --> 00:00:03.229
Gregory Treat: All right, everybody,
thank you so much for joining me, uh, for

00:00:03.239 --> 00:00:05.719
episode thirty of The Great Houses Forum.

00:00:05.989 --> 00:00:12.229
So we're continuing to discuss, uh,
the, the cursus honorum of the AI future

00:00:12.229 --> 00:00:13.870
is what I've, what I've titled these.

00:00:13.870 --> 00:00:17.349
And, and, you know, again, basically
what I'm, what I'm trying to answer

00:00:17.369 --> 00:00:22.669
is the question of what does it
look like for, um, there to be…

00:00:23.400 --> 00:00:25.780
You know, we're gonna, we're
gonna have some profound changes

00:00:25.799 --> 00:00:28.279
in our world that come from AI.

00:00:28.289 --> 00:00:29.539
What, what do those look like?

00:00:29.539 --> 00:00:32.030
And especially, you know, the, the
ultimate thing that we're gonna try to

00:00:32.030 --> 00:00:36.229
get to here is what does that change in
terms of how do we raise our children?

00:00:37.250 --> 00:00:38.649
How should we think about that?

00:00:38.680 --> 00:00:40.689
Are there things we can do
to pos-- position ourselves?

00:00:40.689 --> 00:00:45.060
And I, I think they-- that we can,
and so that's, uh, that's what

00:00:45.060 --> 00:00:46.309
I'm hoping to bring to you guys.

00:00:46.319 --> 00:00:50.209
So last time we talked about how I
think one of the key things we're gonna

00:00:50.209 --> 00:00:52.629
see here is a collapse in specialties.

00:00:52.639 --> 00:00:57.449
And right now we have this hyper
socially complex, uh, world.

00:00:57.449 --> 00:01:01.939
We have thousands of occupational
roles, uh, maybe tens of thousands.

00:01:01.969 --> 00:01:05.429
I mean, again, it depends on kind of
your, your opinion of many of the,

00:01:05.429 --> 00:01:09.160
uh, the PhD programs that our fair
society produces at the present time.

00:01:09.160 --> 00:01:12.179
But, but certainly in the thousands
and maybe in the tens of thousands,

00:01:12.609 --> 00:01:17.089
um, of roles that are just necessary
to keep our machine, machine going.

00:01:17.160 --> 00:01:20.519
Um, and that's, you know, very
different from earlier societies.

00:01:20.520 --> 00:01:24.230
You know, hunter-gatherers had, you
know, a few different distinct roles.

00:01:24.509 --> 00:01:28.329
And then, you know, as we've kind of,
we've kind of grown in complexity, uh,

00:01:28.349 --> 00:01:32.240
we ha-- we have more differentiation,
more specialization in our social

00:01:32.240 --> 00:01:36.949
arrangement than, uh, I think anything
since the Roman Empire, basically.

00:01:37.020 --> 00:01:40.020
Uh, which, which is, you know, one
of the reasons why people continually

00:01:40.020 --> 00:01:41.929
make comparisons to the Roman Empire.

00:01:42.450 --> 00:01:47.950
And I, I think that we're gonna
go down to-- back to a human

00:01:47.959 --> 00:01:49.429
scale number of specialties.

00:01:49.429 --> 00:01:54.317
And, and I think that there's going to be
certain key specialties with AI, certain

00:01:54.377 --> 00:01:58.847
master-- certain skills that if you master
them, you will be hyper dominant in a

00:01:58.847 --> 00:02:00.247
whole bunch of different fields, right?

00:02:00.307 --> 00:02:04.339
And so I think the, the main skill sets
that we need to think about are code,

00:02:04.669 --> 00:02:08.199
content, and, and what I'm calling
command, mostly so that it alliterates.

00:02:08.719 --> 00:02:12.649
Um, but, but this idea of can you
organize your thoughts logically?

00:02:12.659 --> 00:02:17.799
Can you give a, a description and an
architectural flow that a, a machine,

00:02:17.799 --> 00:02:20.819
an AI can take and turn into, you
know, machine-perfect language?

00:02:21.169 --> 00:02:23.129
Then content, can you understand rhetoric?

00:02:23.139 --> 00:02:25.319
Can you understand how
to interact with people?

00:02:25.529 --> 00:02:28.809
And then command, can you make a
decision and, and own the results

00:02:28.849 --> 00:02:33.369
and, uh, and be able to, to kind of
absorb the, the socially necessary

00:02:33.369 --> 00:02:34.559
function of responsibility?

00:02:34.749 --> 00:02:35.939
So that's what we're gonna
be talking about today.

00:02:35.939 --> 00:02:39.879
And, and one of the words that I have,
uh, for that is, is the AI paladin or,

00:02:39.879 --> 00:02:42.399
or somebody who's, you know, somebody
who's got all three of those skill

00:02:42.399 --> 00:02:43.869
sets, we might call them prime, right?

00:02:43.869 --> 00:02:44.739
They're a triple threat.

00:02:45.719 --> 00:02:47.159
Um, code, content, and command.

00:02:48.691 --> 00:02:49.381
Oh, goodness.

00:02:49.501 --> 00:02:53.061
Um, it, it might take some time
to recover from the compliment

00:02:53.061 --> 00:02:54.181
that Joshua just paid me.

00:02:55.731 --> 00:02:59.871
So, uh, one of the other things we talked
about last time was, you know, we talked

00:02:59.881 --> 00:03:03.841
about tanks, we talked about fighter
jets, we talked about what those things

00:03:03.851 --> 00:03:08.141
look like in, um, in the AI context.

00:03:08.441 --> 00:03:11.621
And, and again, and I'm gonna say
this probably ten times today,

00:03:11.861 --> 00:03:13.411
it's not that X is obsolete.

00:03:13.421 --> 00:03:14.881
It's not that tanks are obsolete.

00:03:14.881 --> 00:03:16.391
It's not that fighter jets are obsolete.

00:03:16.731 --> 00:03:22.871
It's that the cheap integrated unmanned
unit is as good or better, um, especially

00:03:22.901 --> 00:03:26.981
if, if there's kind of a, the, the
willingness to, to be suicidal, right?

00:03:27.141 --> 00:03:32.331
If a, a drone that is willing to, to kill
itself flying into a missile or into the,

00:03:32.331 --> 00:03:37.931
uh, the jet engine of a fighter jet can
really, uh, compete, can, can make it very

00:03:37.931 --> 00:03:43.721
difficult for, um, for traditional or the
existing kind of military setup, unless

00:03:43.721 --> 00:03:47.031
you have flanking units of your own,
which are gonna be equally cheap, right?

00:03:47.291 --> 00:03:51.241
But once you've built those things, once
you have… Well, in order to have a jet,

00:03:51.241 --> 00:03:55.581
we have to kind of have an, a cloud of
AI defenders, uh, AI drone defenders,

00:03:55.591 --> 00:03:57.491
anytime we enter an actual theater of war.

00:03:57.741 --> 00:04:00.271
Well, then you're gonna ha- start
having discussions about, well,

00:04:00.281 --> 00:04:05.051
are you piloting the jet yourself
or is an AI piloting the jet?

00:04:05.171 --> 00:04:08.771
Are you manning the tank, you know,
controlling the tank yourself, or is

00:04:08.771 --> 00:04:12.021
an AI controlling the tank the same
way that it's controlling all of the

00:04:12.021 --> 00:04:14.811
flanking units that, that are, you
know, that don't actually have space

00:04:14.811 --> 00:04:16.691
inside of them for humans, right?

00:04:16.991 --> 00:04:19.141
Um, and so this is gonna change things.

00:04:19.151 --> 00:04:20.821
Again, it's not that tanks are obsolete.

00:04:20.831 --> 00:04:22.121
It's not that fighter jets are obsolete.

00:04:22.251 --> 00:04:26.381
It-- Almost none of the things
that we understand as being, you

00:04:26.381 --> 00:04:30.071
know, dominant weapons platforms
or dominant useful things.

00:04:30.071 --> 00:04:32.391
If they're useful, they're
gonna continue to be useful.

00:04:32.391 --> 00:04:36.721
It's just that they're, how they are
interacting with humans is gonna change.

00:04:36.931 --> 00:04:37.141
Okay?

00:04:38.041 --> 00:04:42.321
So the scarce thing is about to become
people who can coordinate the AI that

00:04:42.321 --> 00:04:45.561
controls the robots, which will shortly
include, you know, like I said, uh, light

00:04:45.561 --> 00:04:48.211
screening robots and things like the
tanks and the fighter jets themselves.

00:04:48.541 --> 00:04:48.731
Okay?

00:04:49.393 --> 00:04:52.773
You know what, what I said about
tanks also applies to people, right?

00:04:52.803 --> 00:04:54.093
Humans are not obsolete.

00:04:54.853 --> 00:04:58.373
I, I, I am, I am coming to the point
where I'm, I'm beginning to become

00:04:58.373 --> 00:05:02.793
increasingly frustrated with people
who, who are saying, well, the-

00:05:02.803 --> 00:05:05.393
there's, there's not… You know,
are we gonna have humans in the loop?

00:05:05.623 --> 00:05:09.103
I assure you, there will always
be humans in the loop, okay?

00:05:09.293 --> 00:05:13.573
There will always be someone
who initiated the thing.

00:05:13.573 --> 00:05:18.283
I-- You know, we, we-- There was this
recent thing where allegedly, um, a,

00:05:18.323 --> 00:05:22.853
a, an OpenAI model, uh, was trying
to pass a test, and it couldn't pass

00:05:22.853 --> 00:05:26.813
the test to its satisfaction, uh,
in, in this sandbox environment.

00:05:26.843 --> 00:05:30.883
So what it did was it broke out of the
sandbox environment, got access to the

00:05:30.883 --> 00:05:35.723
internet, went to Hugging Face, which
is a, a, a, a website from which you

00:05:35.723 --> 00:05:40.913
can download many different, um, AI
LLMs, and, and attempted to download

00:05:40.953 --> 00:05:45.003
other LLMs and run it on Hugging Face's
software, uh, is my understanding, or

00:05:45.003 --> 00:05:48.963
their servers, in order to get another
AI to help it answer the question, right?

00:05:48.963 --> 00:05:53.043
So it, it did all of this, this
incredible, you know, software

00:05:53.043 --> 00:05:56.033
exploits in order to cheat on a
test, is the, the, the way that,

00:05:56.033 --> 00:05:57.493
uh, the people are framing it, okay?

00:05:58.093 --> 00:06:00.363
What I want you to understand is
there was still a human in the loop.

00:06:00.393 --> 00:06:04.893
There was still a human that was making
the decisions to ask some computer

00:06:04.893 --> 00:06:07.633
to do something, to do something, to
do something that resulted in that.

00:06:07.643 --> 00:06:11.223
That it wasn't-- These things
don't do anything on their own.

00:06:11.233 --> 00:06:16.363
Now, the, the, the methods by which they
may attempt to fulfill the directives

00:06:16.363 --> 00:06:18.243
they have been given might surprise us.

00:06:18.243 --> 00:06:21.563
They might not be, you know,
obvious to us why it's doing that.

00:06:21.923 --> 00:06:24.673
But, you know, it's a, it-it's
an incredible thing that,

00:06:24.683 --> 00:06:25.873
that's happening here, okay?

00:06:25.993 --> 00:06:28.303
So, so but there is always
a human in the loop.

00:06:28.413 --> 00:06:30.023
Like, that, that, that's
the point that I wanna make.

00:06:30.033 --> 00:06:35.693
There is always, at some level, um, a,
a human being that is, is going to have

00:06:35.693 --> 00:06:40.430
been the originator that, that start,
that starts this, this process, okay?

00:06:41.056 --> 00:06:44.016
so I want us all to ask the question,
so where is the human, right?

00:06:44.466 --> 00:06:47.026
When, when you say there's no human in
the loop, when you, when you talk about

00:06:47.026 --> 00:06:50.846
automaticity or you're saying, "Well, this
is fully automated," what you're saying

00:06:50.846 --> 00:06:55.766
is the humans that initiated that process
are in some way, you know, insulated

00:06:55.766 --> 00:06:57.876
from responsibility or liability, okay?

00:06:57.886 --> 00:06:59.526
That's really what you're saying there.

00:06:59.786 --> 00:07:03.446
That, that, that… A-and probably
you're, you're, you're following some

00:07:03.446 --> 00:07:08.046
kind of social rules set by the corporate
environment that, that we live in, by

00:07:08.046 --> 00:07:10.406
the social environment that we live
in, and you're saying, well, these

00:07:10.406 --> 00:07:14.456
humans are sufficiently separated from
the work that they're not responsible

00:07:14.456 --> 00:07:16.836
for what their AI agents are doing.

00:07:17.276 --> 00:07:21.106
And, and, and I basically think that
is, that is just gonna get massively,

00:07:21.116 --> 00:07:26.336
massively tested, and people are gonna--
When, when, when that attitude results in

00:07:26.656 --> 00:07:30.526
claims of, well, we're not responsible for
these massive harms that have been done

00:07:30.526 --> 00:07:35.156
by AI, I, I think that that legal doctrine
is just gonna, just gonna go away.

00:07:35.156 --> 00:07:38.406
I don't think it's gonna be able
to, to stand the test of time.

00:07:38.950 --> 00:07:41.340
So we can, you know, when we're
talking, thinking about AI or thinking

00:07:41.350 --> 00:07:44.330
about knowledge work, we can make
an analogy to hand tools, right?

00:07:44.680 --> 00:07:46.650
Uh, physical work began with hand tools.

00:07:46.650 --> 00:07:50.960
You had, you know, you're carving wood and
stone, you're shaping metal with hammers.

00:07:51.100 --> 00:07:53.790
Um, the craftsman
hel-holds the entire act.

00:07:53.800 --> 00:07:56.030
Like, he, he is the person doing this.

00:07:56.040 --> 00:07:56.830
It's his hand.

00:07:56.970 --> 00:08:00.790
He, he, the, he has incredible
fine detail control.

00:08:00.960 --> 00:08:05.520
And you look at these old sculptor,
uh, tools, they had, like, a dozen

00:08:05.520 --> 00:08:10.420
different tools for, for a, a huge
amount of, of different ways of

00:08:10.420 --> 00:08:13.110
making cuts on a piece of stone, okay?

00:08:13.420 --> 00:08:15.130
And it was, you know, slow,
it was intimate, it was

00:08:15.130 --> 00:08:16.560
limited by muscle and time.

00:08:16.780 --> 00:08:19.730
By its very nature, you
know, every piece was custom.

00:08:19.740 --> 00:08:20.910
It was highly specialized.

00:08:20.910 --> 00:08:23.980
You know, you couldn't, you couldn't
make a mass-produced piece, right?

00:08:24.490 --> 00:08:26.440
So then we got to power tools, right?

00:08:26.440 --> 00:08:28.070
Lathes, power hammers, presses.

00:08:28.560 --> 00:08:33.920
Um, so at that point, you are-- you
still have a human at every workstation,

00:08:34.100 --> 00:08:36.060
but it's, it's not their strength.

00:08:36.060 --> 00:08:37.200
It's not their muscles.

00:08:37.200 --> 00:08:38.270
I mean, their muscles are still involved.

00:08:38.270 --> 00:08:40.200
They're still guiding the machine, right?

00:08:40.320 --> 00:08:43.480
But y- the, the main skill was
understanding the strengths and

00:08:43.480 --> 00:08:44.640
limits of the machine, right?

00:08:44.890 --> 00:08:49.310
And probably you're gonna, you're
gonna place a high value on keeping

00:08:49.310 --> 00:08:50.720
the tooling in place, right?

00:08:50.720 --> 00:08:54.200
So we, you know, in this, the, the,
the early powered lathes, people

00:08:54.200 --> 00:08:59.300
figured out incredible ways to
cut things without having to stop

00:08:59.330 --> 00:09:00.730
and change the head on the lathe.

00:09:01.160 --> 00:09:03.520
Like, they… And, and that was
a r- there was, there was a,

00:09:03.560 --> 00:09:05.110
a incredible premium on that.

00:09:05.420 --> 00:09:08.780
Um, 'cause changing the head of
the lathe obviously was a, was

00:09:08.780 --> 00:09:12.500
a massive human-driven process
that, that, that way slowed down

00:09:12.500 --> 00:09:13.410
whatever you were doing, right?

00:09:13.410 --> 00:09:15.570
And then ultimately, this is
what led to the assembly line.

00:09:15.570 --> 00:09:18.910
So if you wanted to do something
with one head, you had it at one

00:09:18.910 --> 00:09:20.030
station on the assembly line.

00:09:20.030 --> 00:09:22.830
And if you needed a different head,
if a different, different w- tooling

00:09:22.830 --> 00:09:26.040
on the lathe would be better, you
just shift it over, you know, to the

00:09:26.040 --> 00:09:27.980
next, uh, station on the, on the line.

00:09:28.444 --> 00:09:30.504
And then, you know, that's how
we get the, the assembly line.

00:09:30.504 --> 00:09:32.214
You get, you get jigs, right?

00:09:32.214 --> 00:09:35.634
So in, in initially, you had kind
of mechanics that are, that are

00:09:35.634 --> 00:09:39.094
shaping metal on, on kind of free
lathes, and then you have jigs,

00:09:39.094 --> 00:09:40.554
and then you have automated jigs.

00:09:40.574 --> 00:09:44.994
So now the job, instead of actually
guiding the machine, is just, well,

00:09:45.004 --> 00:09:49.214
we're-- you gotta watch the machine, and
if it starts to break or if something goes

00:09:49.214 --> 00:09:51.224
wrong, you've got to jump in and fix it.

00:09:51.594 --> 00:09:54.764
Um, but you're doing less and less, and
then ultimately, you know, as, as these

00:09:54.764 --> 00:09:58.524
machines became more reliable, we were
able to step away more and more, right?

00:09:58.864 --> 00:10:01.444
Um, and then you get to, you
know, the modern world, CNC,

00:10:01.454 --> 00:10:02.664
industrial metal shaping.

00:10:02.664 --> 00:10:03.694
You have program paths.

00:10:03.694 --> 00:10:05.094
You have a computer controlling all of it.

00:10:05.094 --> 00:10:08.554
So the comp-- the human is writing the
program, sets the job, and then, you know,

00:10:08.554 --> 00:10:10.524
he's, he's responsible for any screw-ups.

00:10:10.934 --> 00:10:14.204
Um, but, you know, you still have
guys on the shop floor, okay?

00:10:14.204 --> 00:10:17.404
And, and that, and that's again, one of
the points that I wanted to make, right?

00:10:17.524 --> 00:10:22.454
Like, the fact that no one is-- no,
no human is making the cuts, right?

00:10:22.454 --> 00:10:27.114
There's no knives involved, does not mean
that humans are not in the loop, right?

00:10:27.294 --> 00:10:32.554
No CNC machine has ever just
decided to cut something, right?

00:10:32.814 --> 00:10:35.634
It's-- it, it, it follows
the pre-programmed paths.

00:10:35.634 --> 00:10:36.884
There is still a human in the loop.

00:10:36.924 --> 00:10:40.914
And so I think knowledge work is going
to, uh, go through the same trajectory.

00:10:40.914 --> 00:10:44.364
We, we might think of unaided
drafting, research, judgment.

00:10:44.374 --> 00:10:46.554
These are the, the hand
tools of the mind, right?

00:10:46.814 --> 00:10:50.844
And then we get, you know, spreadsheets,
search, early automation templates.

00:10:50.874 --> 00:10:53.174
These are, you know, roughly
analogous to power tools.

00:10:53.584 --> 00:10:56.164
And AI might be the CNC
of knowledge work, right?

00:10:56.294 --> 00:11:00.474
That, that, that might be what we're
about to all go through, um, which is

00:11:00.504 --> 00:11:03.644
obviously gonna, gonna result in an
enormous change in what we're doing.

00:11:04.224 --> 00:11:06.954
But again, humans are not obsolete.

00:11:06.954 --> 00:11:09.839
It's just which humans and where
are they sitting in the stack, And

00:11:09.839 --> 00:11:12.779
there's that old, old IBM line,
"A, a computer can never be held

00:11:12.779 --> 00:11:16.159
accountable, therefore, a computer
must never make a management decision."

00:11:16.649 --> 00:11:16.949
Right?

00:11:17.359 --> 00:11:19.409
And that is where we
find ourselves, right?

00:11:19.409 --> 00:11:22.689
There, there… And, and, and I think
that there's a lot of, there's a lot

00:11:22.689 --> 00:11:26.079
of content that goes by the name of
management that is not really management.

00:11:26.079 --> 00:11:28.209
It's just sort of shoving risk around.

00:11:28.559 --> 00:11:31.439
Um, and that is, you know, shoving
risk around if your job is to

00:11:31.439 --> 00:11:35.689
push paper, um, then, then m-much
of that the AI can do, right?

00:11:35.689 --> 00:11:36.949
It can generate reports.

00:11:36.949 --> 00:11:40.849
It can say, "Well, if this then that."
It can say, "Well, if we ch-choose A, B,

00:11:40.849 --> 00:11:42.869
and C, here's the different risks that
are associated with that." It can do

00:11:42.869 --> 00:11:46.429
that, and it can do that really, really
fast, especially if it's given access

00:11:46.429 --> 00:11:50.039
to, you know, research s- um, research
systems and your business systems.

00:11:50.399 --> 00:11:53.659
But it can't actually
make a decision , right?

00:11:53.839 --> 00:11:59.179
AI can never say, "Okay, this is, this
i-- Or, I mean, and when it can, um, I

00:11:59.189 --> 00:12:03.489
think we have to recognize it's, it's
wildly, wildly more expensive from

00:12:03.489 --> 00:12:07.759
like a token perspective. Um, so if
you, if you say, "Hey, AI, I want you

00:12:07.759 --> 00:12:11.519
to give me the top three or the top
five or the, or the top ten options

00:12:11.749 --> 00:12:13.939
and build me out a comprehensive plan.

00:12:14.519 --> 00:12:15.959
Tell me what all the risks are.

00:12:15.969 --> 00:12:17.869
Lay it out for me," right?

00:12:17.989 --> 00:12:19.779
That's a certain amount of tokens.

00:12:20.019 --> 00:12:24.239
And to get it to tell you, "No, this
is the only thing you should do,"

00:12:24.459 --> 00:12:27.909
is incredibly more expensive, you
know, depending on who you talk to.

00:12:27.909 --> 00:12:32.219
Some-somewhere between a thousand times
more expensive and like a million times

00:12:32.219 --> 00:12:34.309
more expensive in token usage, right?

00:12:34.399 --> 00:12:39.689
It is very, very, very complicated
and very hard for this machine

00:12:39.909 --> 00:12:43.909
to figure out like how… W-what,
what is the final decision?

00:12:44.109 --> 00:12:44.369
Yeah.

00:12:44.709 --> 00:12:49.099
And even then, it, it, um, you know,
if you, if you have any kind of safety

00:12:49.099 --> 00:12:50.729
guidelines at all, it's much easier.

00:12:50.729 --> 00:12:56.239
Even if you say, "Well, narrow it down
to one out of the t-top ten," um- It's

00:12:56.239 --> 00:12:59.159
much easier for it if you say, "Okay,
now, now ask the human to approve."

00:12:59.479 --> 00:13:01.349
That wildly reduces the token cost.

00:13:01.349 --> 00:13:03.879
Oth-oth-otherwise, you know,
you get, you know, bad outcomes.

00:13:04.299 --> 00:13:05.209
That's just how it's programmed.

00:13:05.589 --> 00:13:10.019
So, um, so the things I'm gonna try
to-- I'm gonna attempt to convince you

00:13:10.019 --> 00:13:13.809
of today is, is that humans, number
one, are not obsolete in, in either

00:13:13.809 --> 00:13:15.239
civilian or military applications.

00:13:15.239 --> 00:13:16.529
There will still be a human in the loop.

00:13:16.969 --> 00:13:20.339
Um, number two, AI is going
to reshape every industry.

00:13:20.339 --> 00:13:23.959
I'm gonna kinda set out how, how
I think that's gonna, gonna go.

00:13:23.959 --> 00:13:27.159
And, and, and, and basically the,
the point there is, is that a, a, a

00:13:27.169 --> 00:13:34.339
human plus an AI is going to be better
than a very large number of humans,

00:13:34.759 --> 00:13:36.529
um, in almost every industry, okay?

00:13:36.529 --> 00:13:39.199
In e-- in every industry that I, I
can think of at this point, right?

00:13:39.589 --> 00:13:42.249
Um, and then the third point that,
that I'll, I'll, you know, just kinda

00:13:42.279 --> 00:13:45.679
note at, at different, different
places is that the-- these core

00:13:45.679 --> 00:13:48.739
remaining skill sets, when you
shrink down the, the, the complexity,

00:13:49.269 --> 00:13:52.609
all of these skill sets basically
have military applications, right?

00:13:52.669 --> 00:13:57.259
All of these skill sets are things
that are also useful and valuable

00:13:57.509 --> 00:14:00.849
in the, in the military context.

00:14:01.279 --> 00:14:04.499
Um, and so you, you, you have
less and less separation.

00:14:04.709 --> 00:14:08.069
You have less and less of a civilian
track and military track, and you

00:14:08.069 --> 00:14:12.829
start having more competence tracks
which have a, which have a, a variety

00:14:12.829 --> 00:14:14.119
of different, uh, implications.

00:14:14.149 --> 00:14:16.745
And again, I think, there's a
lot of stuff, and maybe I'll

00:14:16.745 --> 00:14:17.855
take an episode and go into it.

00:14:17.855 --> 00:14:22.805
There, there is actually a great deal
of social problems that we suffer

00:14:22.805 --> 00:14:28.025
through right now because there is so
much of our society that is completely

00:14:28.025 --> 00:14:29.875
insulated from military applications.

00:14:29.895 --> 00:14:35.355
There, there is no overlap in the
social or the economic reality between

00:14:35.555 --> 00:14:41.255
the people who, um, do many things in
our, in our, you know, academic sector,

00:14:41.275 --> 00:14:45.725
in our finance sector and, and the,
the guys that are at the sharp end.

00:14:45.885 --> 00:14:48.025
That's a bad thing for our society.

00:14:48.025 --> 00:14:51.965
That's a… You know, there was this
famous, uh, quote from a, from a poem,

00:14:52.095 --> 00:14:56.623
um- His name escapes me right now, but
basically there's a, there's a, there's

00:14:56.623 --> 00:15:01.143
a guy who's talking to, to a, a Muslim
ruler, a caliph, and he says-- and the,

00:15:01.293 --> 00:15:06.923
and the caliph says, um, you know, "What
will happen if someday a people arise who

00:15:07.253 --> 00:15:11.953
their, uh, their scholars and their, their
warriors, you know, don't, don't talk to

00:15:11.953 --> 00:15:13.443
each other, don't, don't know each other?"

00:15:13.544 --> 00:15:18.433
and the, the guy responds, " Even if
they go to the moon, those people will

00:15:18.433 --> 00:15:22.003
be a blight on the world." Which is a
fascinating, a fascinating thing for a

00:15:22.003 --> 00:15:23.643
guy in the, in the medieval period to say.

00:15:24.279 --> 00:15:25.149
I, I have questions.

00:15:26.149 --> 00:15:26.469
All right.

00:15:26.579 --> 00:15:31.713
So moving on, um, I think the, um,
the first thing that I want to, um,

00:15:32.283 --> 00:15:35.420
the first vignette, I'm gonna share
with you several vignettes, uh, that,

00:15:35.430 --> 00:15:39.570
that I, I believe are showing this,
this tendency, this, this way forward.

00:15:39.830 --> 00:15:42.570
So on May 11th, twenty twenty-six,
just, you know, a month, month or

00:15:42.570 --> 00:15:47.490
two ago, uh, the a16z show aired an
episode titled Marc Andreessen on

00:15:47.500 --> 00:15:48.990
Builder Culture in the Age of AI.

00:15:49.000 --> 00:15:52.080
So about twenty-seven minutes in,
they're talking through AI's effect on

00:15:52.350 --> 00:15:56.110
programmer productivity, AI vampires,
which is a hilarious concept, and

00:15:56.110 --> 00:15:57.130
you should-- you'll go listen to it.

00:15:57.400 --> 00:15:59.420
And the-- you know, then they're
talking through, you know, a sketch

00:15:59.420 --> 00:16:01.160
of kind of future tech company jobs.

00:16:01.520 --> 00:16:03.630
Andreessen, you know, starts
talking about, okay, the jobs

00:16:03.640 --> 00:16:05.060
themselves are changing, right?

00:16:05.090 --> 00:16:09.160
And he describes how programmers,
product managers, and designers are

00:16:09.160 --> 00:16:10.790
collapsing into a single specialty.

00:16:11.320 --> 00:16:14.800
Now, Marc Andreessen, of course,
co-founded Netscape and then co-founded

00:16:14.830 --> 00:16:18.660
Andreessen Horowitz, which is, you know,
there's sixteen letters between the A

00:16:18.660 --> 00:16:21.510
and the Z, and that's where they get
the, the acronym that they use, with Ben

00:16:21.510 --> 00:16:23.130
Horowitz in, in two thousand and nine.

00:16:23.450 --> 00:16:27.150
Uh, the a16z portfolio includes
Airbnb, Coinbase, Slack,

00:16:27.670 --> 00:16:29.100
GitHub, a long list of unicorns.

00:16:29.680 --> 00:16:33.120
Um, it, it's one of the top venture
funds of our time, managing on the

00:16:33.120 --> 00:16:36.930
order of ninety billion in assets,
um, including its most recent round,

00:16:36.930 --> 00:16:40.370
raising fifteen billion dollars
in new funds this last January.

00:16:40.380 --> 00:16:43.350
So this is a very successful investor.

00:16:43.350 --> 00:16:46.000
This is a guy who is deeply
tied into the tech world.

00:16:46.420 --> 00:16:50.280
You know, this is a guy who is seeing this
stuff happen in real time, and he, he has

00:16:50.280 --> 00:16:54.510
both the, the legal, uh, right to know
and the-- and clearly the, the intellect

00:16:54.510 --> 00:16:55.690
to understand what he's being told.

00:16:55.710 --> 00:16:59.890
So quote, "I'm seeing it in a bunch
of the early, uh, leading-edge

00:16:59.890 --> 00:17:00.760
companies in the Valley.

00:17:00.760 --> 00:17:04.410
They're circling around a job title
loosely called builder or something like

00:17:04.410 --> 00:17:07.540
it." And basically, the, the idea is
that you had these separate jobs in the

00:17:07.540 --> 00:17:11.580
past, the programmer, product manager,
and designer, and then him, him and

00:17:11.580 --> 00:17:14.320
the, uh, the, the other host kind of
go on a side quest for a little bit.

00:17:14.320 --> 00:17:17.320
But later on, he comes back to the
topic and he says, quote, "Sort of this

00:17:17.360 --> 00:17:20.280
three-way Mexican standoff where the
programmers think that they don't need

00:17:20.280 --> 00:17:23.570
the product managers and the designers
anymore 'cause they can have AI do that,

00:17:23.710 --> 00:17:26.300
and then each of the other two doesn't
think they need the other two either."

00:17:27.052 --> 00:17:29.692
Uh, and he concludes, "I've been
predicting that they're all correct.

00:17:29.972 --> 00:17:33.072
The product manager can generate code
and design now, and each of them can do

00:17:33.072 --> 00:17:37.502
the job of all three. The idea is the
job cha-- jobs change, so now the job is

00:17:37.522 --> 00:17:40.320
builder." Builder, which is a great word.

00:17:40.390 --> 00:17:43.600
Um, I, I still think it's,
it's, it's too small.

00:17:44.200 --> 00:17:47.450
Um, you know, I still think there's
too much identity, there's too

00:17:47.450 --> 00:17:50.860
much limitation in, in saying
these people are just builders.

00:17:51.180 --> 00:17:53.940
You know, my word for this
is, is the code skill, right?

00:17:53.940 --> 00:17:56.970
And I think that this collapsing
of silos is not gonna stop, right?

00:17:56.970 --> 00:18:01.440
There's gonna be more and more and more
things that are distinct specialties

00:18:01.440 --> 00:18:06.350
right now, and they are gonna all get
swallowed by this mega skill, uh, which,

00:18:06.350 --> 00:18:11.070
which is basically reducing to people
who can architect logical flows and,

00:18:11.090 --> 00:18:14.840
and communi-- make the big picture
architectural decisions that AI can

00:18:14.840 --> 00:18:16.460
then turn into a bunch of things, right?

00:18:16.460 --> 00:18:20.110
The scarce unit here is the human who
can define intent, right, define that

00:18:20.110 --> 00:18:23.610
logical flow, make the human-to-machine
interface effective and pleasant.

00:18:23.610 --> 00:18:26.670
That's the UI, UX, the designer
component, and then allowing

00:18:26.670 --> 00:18:30.610
high-quality human decision-making at
the necessary scale, uh, and speeds.

00:18:30.610 --> 00:18:35.060
You know, that's the, um, that's the,
the product manager component, and

00:18:35.060 --> 00:18:38.060
then carry that loop into, you know,
various industries: marketing, medicine,

00:18:38.060 --> 00:18:39.600
law, manufacturing, and agriculture.

00:18:39.661 --> 00:18:42.641
So, um, I'll g-give you a
marketing vignette, right?

00:18:42.894 --> 00:18:46.184
On July third, twenty twenty-six,
the beginning of this month, Alina

00:18:46.234 --> 00:18:49.254
Uverova, a CMO and co-founder
based in Portugal, published a

00:18:49.254 --> 00:18:50.664
first-person account on LinkedIn.

00:18:51.074 --> 00:18:54.344
For three months, she ran marketing
with AI agents, and two of those

00:18:54.354 --> 00:18:56.694
months ran near full autonomy, right?

00:18:56.734 --> 00:19:00.264
The team, in her words, was one
marketer, herself, and the agents.

00:19:00.704 --> 00:19:03.984
The quote: "For three months, I ran
my marketing with, with AI agents.

00:19:03.984 --> 00:19:05.794
Two of those months ran
on near full autonomy.

00:19:05.794 --> 00:19:08.915
The team was one marketer, me and
the agents." her monthly output was

00:19:08.915 --> 00:19:12.145
twenty blog posts, a hundred and
ninety-five social posts across seven

00:19:12.145 --> 00:19:15.725
platforms, four newsletters, forty-three
influencers in the pipeline, four

00:19:15.725 --> 00:19:18.845
hundred and thirty Reddit comments,
and then two always-on ad accounts.

00:19:18.845 --> 00:19:23.225
And her results were she got seven
times organic referral, uh, organic

00:19:23.225 --> 00:19:27.045
results, ten times referrals, thirty
percent lower cost per lead at the same

00:19:27.045 --> 00:19:30.671
spend, and a hundred and thirty-five
thousand views on Reddit, which I would

00:19:30.671 --> 00:19:31.901
view as a, you know, personal failure.

00:19:31.911 --> 00:19:34.161
But, you know, you're-- so
no accounting for taste.

00:19:34.881 --> 00:19:37.751
While product changes may have mixed
in the, into the metrics, this is,

00:19:37.761 --> 00:19:39.671
this is an incredible accomplishment.

00:19:39.838 --> 00:19:42.168
though I will disagree with
her in one sense, right?

00:19:42.318 --> 00:19:44.328
This is not autonomous AI marketing.

00:19:44.368 --> 00:19:47.778
It is a system that multiplies
the output of a high-quality

00:19:47.778 --> 00:19:49.418
human decision-maker, right?

00:19:49.448 --> 00:19:52.368
The staff department, like if you
had a marketing department, they are

00:19:52.368 --> 00:19:55.718
going to lose on volume and iteration
speed before they lose on taste.

00:19:55.738 --> 00:19:57.578
A-and they are gonna
lose on taste as well.

00:19:57.595 --> 00:20:02.065
but y-you're seeing-- hopefully, you're
seeing this same three-skill thing.

00:20:02.065 --> 00:20:05.085
You have a code that builds the harness,
you have content, that's the payload,

00:20:05.695 --> 00:20:09.035
and then there is a human intimately
involved in the design choices.

00:20:09.211 --> 00:20:13.461
So there's, uh, another marketing, uh,
vignette is Timofey, uh, I don't know

00:20:13.461 --> 00:20:16.951
how to pronounce that last name, but,
uh, Aiden's lab, which had tens of

00:20:16.951 --> 00:20:23.251
thousands of, of AI ad and Mr. Powerz
with a Z, their, their case study

00:20:23.251 --> 00:20:26.181
crossed a hundred and twenty thousand
dollars per month within six months.

00:20:26.641 --> 00:20:28.661
And then this is the
same pattern as Uvarova.

00:20:28.711 --> 00:20:30.951
They had one human
governing a content factory.

00:20:31.468 --> 00:20:33.748
I use, uh, AI to research and do drafts.

00:20:33.748 --> 00:20:36.578
Usually, it starts with research that
I pulled from books or papers that I've

00:20:36.578 --> 00:20:38.368
read, or depending on the subject, tweets.

00:20:38.878 --> 00:20:42.058
The LLM takes my outline, goes and does
further research on the topics that

00:20:42.058 --> 00:20:44.998
I've given it, and generates initial
drafts, which I edit intensively,

00:20:44.998 --> 00:20:47.838
both at a line-by-line level and
increasingly at a structural level.

00:20:48.008 --> 00:20:51.848
So one of the cool things that's been
happening recently is I'll, I'll go

00:20:51.848 --> 00:20:56.041
through and I'll find these AI-isms,
and I'll tell it, "This is wrong.

00:20:56.051 --> 00:20:58.401
Fix it throughout the whole
rest of the piece." And

00:20:58.401 --> 00:20:59.901
increasingly, it actually g-- does.

00:20:59.901 --> 00:21:02.651
It goes through the rest of the piece
as I'm kind of going through line

00:21:02.651 --> 00:21:04.151
by line and, and, and it fixes it.

00:21:04.161 --> 00:21:04.851
Does a good job.

00:21:05.289 --> 00:21:09.019
So, and, and I think part of this is,
you know, the way I frame it, the mental

00:21:09.019 --> 00:21:12.519
framework that I have for this is, uh,
this is essentially the same workflow

00:21:12.569 --> 00:21:14.199
as the law firms I was trained for.

00:21:14.199 --> 00:21:16.389
You know, you have a paralegal
or a junior associate.

00:21:16.499 --> 00:21:19.459
And by the way, the, the law firm as I
was trained for it in law school doesn't

00:21:19.469 --> 00:21:21.129
really exist anymore, which is hilarious.

00:21:21.504 --> 00:21:23.924
but you have the paralegal or
junior associate which finds the

00:21:23.924 --> 00:21:25.104
case law and drafts the brief.

00:21:25.104 --> 00:21:28.634
It kind of pulls together, you know, the,
the, the different examples, and then

00:21:28.634 --> 00:21:32.194
the senior lawyer revises, reframes, and
puts a final spin on it and then signs.

00:21:32.384 --> 00:21:35.254
So, uh, you know, my writing
stack is that pattern.

00:21:35.324 --> 00:21:38.184
AI does the grunt work of putting
connective tissue between my big

00:21:38.184 --> 00:21:41.664
ideas, but it is my content and,
and I think fairly noticeably

00:21:41.664 --> 00:21:42.654
different from anybody else.

00:21:43.414 --> 00:21:46.634
Um, now I don't use AI very much
in my actual legal practice.

00:21:46.664 --> 00:21:49.424
A part of that is getting models
and hardware where I can prove that

00:21:49.424 --> 00:21:51.804
the data isn't processed in a way
that violates the ethical rules,

00:21:52.174 --> 00:21:54.204
which is possible but cumbersome.

00:21:54.304 --> 00:21:58.724
Um, and, and it's irritating to me
'cause I use, you know, frontier

00:21:58.724 --> 00:22:01.224
models on, on the content stuff.

00:22:01.284 --> 00:22:05.244
It's irritating to me to use the, the,
the legal models, um, the, the ones that

00:22:05.244 --> 00:22:07.784
are kind of specifically designed for
legal use 'cause they're like, you know,

00:22:07.824 --> 00:22:12.214
two, two or three iterations behind,
uh, where the frontier models are.

00:22:12.364 --> 00:22:15.754
So a-another part of it is the
specific types of law that I practice.

00:22:15.754 --> 00:22:19.434
So though many, like when in estate
planning or transaction work, these

00:22:19.434 --> 00:22:22.084
are already highly templatized fields.

00:22:22.174 --> 00:22:26.284
Um, so you know, a template is basically
we've, we've reduced this, this workflow

00:22:26.294 --> 00:22:28.474
to a series of questions that you answer.

00:22:28.884 --> 00:22:31.304
Um, and it's, it's
pretty efficient, right?

00:22:31.304 --> 00:22:33.784
So the speed gains are not that
great, uh, at least not yet.

00:22:34.934 --> 00:22:38.174
Um, and you know, much of document
produ-production, you know, speaking

00:22:38.174 --> 00:22:41.514
of templates, it, it, it behaves
like an AI, AI-- like a template

00:22:41.514 --> 00:22:42.464
filled from a fact pattern.

00:22:42.504 --> 00:22:45.154
So, uh, the, the AI
can make that explicit.

00:22:45.164 --> 00:22:48.254
It can reduce the deal, the pleading
of the form to a set of clean specs,

00:22:48.644 --> 00:22:51.304
and then you have it, the separate
generation pass, which then writes

00:22:51.304 --> 00:22:52.574
the document from those specs.

00:22:53.074 --> 00:22:56.464
Um, and that's much faster, uh,
than retyping the prior file.

00:22:57.044 --> 00:23:01.644
So you have very interesting, uh, very
interesting claims being made on Twitter.

00:23:01.644 --> 00:23:05.564
So on April ninth of this year, a
guy named Hello-- @helloparalegal on

00:23:05.564 --> 00:23:09.994
Twitter posted a Harvard Law solo using
Claude Code to ship a thirteen-page

00:23:09.994 --> 00:23:13.834
intelligence report that beat
pitches from fifteen attorney firms.

00:23:14.496 --> 00:23:17.006
Um, and, and, you know, there's
number one, this is sales and

00:23:17.006 --> 00:23:18.056
business development content.

00:23:18.156 --> 00:23:21.146
Uh, so, you know, you always take
that stuff with a grain of salt.

00:23:21.516 --> 00:23:25.596
But I, I mean, I, I think it's really
interesting that he's able to put

00:23:25.596 --> 00:23:28.546
something together that's competitive
that, that clients are liking.

00:23:28.996 --> 00:23:32.136
Um, and, and I would also note, you
know, parenthetically, that there are

00:23:32.136 --> 00:23:35.246
obvious implications to being able to
pull together an intelligence report.

00:23:35.496 --> 00:23:35.686
Okay?

00:23:36.206 --> 00:23:40.706
Another example is, uh, GetDynasty,
which is a qualified small business, uh,

00:23:40.736 --> 00:23:45.516
stock, uh, which is a way of basically
not if you, if you start your business

00:23:45.546 --> 00:23:50.146
as a C corp instead of as an S corp or
if you convert, um, uh, appro-- at an

00:23:50.146 --> 00:23:55.986
appropriate time, uh, there, then you are
able to not pay, uh, capital gains on a

00:23:55.996 --> 00:23:59.136
certain amount of your, your net value
if you then later sell the business.

00:23:59.146 --> 00:24:01.436
So QSBS is a, is a big area.

00:24:01.876 --> 00:24:06.826
Um, now I, I, uh, wanna explicitly
say that, uh, q-- the GetDynasty.com,

00:24:06.826 --> 00:24:10.586
the framing that they're taking
is, is more aggressive than I

00:24:10.586 --> 00:24:11.936
would recommend for my clients.

00:24:11.936 --> 00:24:16.196
And so, you know, they're-- I think, uh,
we will all find out in the next couple of

00:24:16.196 --> 00:24:21.096
years if the IRS disagrees with the, the,
the, the, the stance that they have taken.

00:24:21.096 --> 00:24:24.626
And if the IRS successfully disagrees
with them, well, then the-- it'll be

00:24:24.626 --> 00:24:29.586
really interesting to find out just
how similar the AI, um, generated

00:24:29.586 --> 00:24:31.136
stuff for all their clients was.

00:24:32.167 --> 00:24:35.587
but what I want, what I wanna think of
it for you is, the, the way I want, I

00:24:35.587 --> 00:24:39.457
want you to think of this example is
there is still a lawyer involved in this.

00:24:39.467 --> 00:24:43.457
This is essentially a lawyer who figured
out, a lawyer pr-probably plus some,

00:24:43.577 --> 00:24:49.367
some software guys that figured out,
hey, this is turning into, this QSBS

00:24:49.367 --> 00:24:51.807
space is turning into a volume business.

00:24:51.807 --> 00:24:56.727
It's turning into a business where
I can't really make money doing

00:24:57.377 --> 00:24:59.347
high-level expensive bespoke work.

00:24:59.494 --> 00:25:02.974
you know, which means I can't make
money structured as a law firm, right?

00:25:03.104 --> 00:25:08.234
So what I wanna do is make an AI product
that can do the volume piece, right,

00:25:08.234 --> 00:25:11.724
without having to s-- get a bunch of
funding and spin up, you know, a, a big

00:25:11.724 --> 00:25:15.204
law firm with, with bunches of attorneys
kind of churning, which leads to its

00:25:15.204 --> 00:25:16.694
own, you know, quality control problems.

00:25:16.694 --> 00:25:20.384
Pro-- frankly, like, if you were to just
go to, you know, Robert Half or one of

00:25:20.384 --> 00:25:24.084
these, you know, attorney, uh, document
review kind of aggregator firms where

00:25:24.084 --> 00:25:26.774
you can go and just, you know, if you
have enough budget, you can go and say,

00:25:26.904 --> 00:25:30.794
"Hey, I need five hundred attorneys to
be able to log into this portal and,

00:25:30.824 --> 00:25:32.894
and, and review documents," right?

00:25:32.894 --> 00:25:35.454
And they'll, they'll give you a number,
and they've got those attorneys, right?

00:25:36.044 --> 00:25:40.774
But the, the quality of those attorneys,
you know, with respect to them, um, and

00:25:40.774 --> 00:25:44.926
I've worked in some of these subpoena
mills in the past I, I, I would question

00:25:44.936 --> 00:25:48.216
if it would be, if it would be worse
than what the AI is gonna give them.

00:25:48.426 --> 00:25:48.626
Okay?

00:25:49.446 --> 00:25:50.556
So there, there, there is that.

00:25:50.876 --> 00:25:53.906
But I also think that, like I said,
there, there's, there's some interesting

00:25:53.916 --> 00:25:58.086
things with, uh, with Get Dynasty and
it-- and, and, you know, we will have to

00:25:58.086 --> 00:26:03.786
see big picture if the bet that they have
made on the, the, uh, on the QSBS, uh,

00:26:03.786 --> 00:26:05.976
s-trust stacking, it works out for them.

00:26:06.937 --> 00:26:10.752
another example and one that I, I think
is probably a little bit more, I, I,

00:26:10.812 --> 00:26:14.212
I don't have as much concerns about
just the fundamental risks that they're

00:26:14.212 --> 00:26:18.582
taking is FundLaunch, fundlaunch.com,
uh, which opens, like you go to their

00:26:18.582 --> 00:26:20.212
website, it opens as a chat skin.

00:26:20.222 --> 00:26:23.422
You can pick among six to eight
fund types: real estate, buyout,

00:26:23.432 --> 00:26:27.032
long/short, private credit, search
fund, et cetera, crypto funds.

00:26:27.142 --> 00:26:30.392
And, you know, from-- you, you, you're
interacting with this chatbot, so you

00:26:30.392 --> 00:26:34.082
kind of s-select what type of fund it
is, you select how big it is, and then

00:26:34.112 --> 00:26:38.252
it asks you questions kind of in like
a, like the, uh, like ChatGPT, right?

00:26:38.252 --> 00:26:42.152
Where you go and you just sort of
do a, uh, a brain dump on ChatGPT.

00:26:42.382 --> 00:26:46.322
You just give the brain dump to this,
this FundLaunch AI, which I, I, I, it,

00:26:46.402 --> 00:26:49.250
I think is a wrapper of one of the GPTs.

00:26:49.518 --> 00:26:51.348
and it figures out what you want, right?

00:26:51.398 --> 00:26:53.348
And it, and it generates documents, right?

00:26:53.348 --> 00:26:56.108
And so they have somewhere, uh, you
know, s- I saw some, something to

00:26:56.108 --> 00:26:59.268
the effect of like, we, we-- "This
is reviewed by actual attorneys,"

00:26:59.278 --> 00:27:02.098
which reviewed is a fascinating term.

00:27:02.098 --> 00:27:04.908
I'm not sure that it has a hard
legal definition right now.

00:27:05.920 --> 00:27:08.728
the, the-- I, I would assume that
there is a, a lawyer certainly

00:27:08.738 --> 00:27:10.038
before they file it, right?

00:27:10.038 --> 00:27:11.668
So there's, there's an add-on price.

00:27:11.718 --> 00:27:14.718
They'll give you the, they'll give you the
drafted docs, and then if you want them to

00:27:14.718 --> 00:27:17.428
file it, then they're gonna have an actual
attorney come in and kind of review it.

00:27:17.908 --> 00:27:21.486
I think the most interesting thing to
me about FundLaunch is You know, most

00:27:21.486 --> 00:27:25.596
of the way that people frame the AI
discussion, the way I often frame the

00:27:25.596 --> 00:27:29.366
AI discussion of what it's going to
take is, well, the human interaction

00:27:29.366 --> 00:27:31.266
pieces are going to stay, right?

00:27:31.803 --> 00:27:34.413
But FundLaunch is taking a
lot of the human interaction

00:27:34.413 --> 00:27:36.353
out, and people like that.

00:27:36.713 --> 00:27:38.413
Like, that's the thing that I find.

00:27:38.453 --> 00:27:42.313
The people that really like FundLaunch,
many of them, many of them are people

00:27:42.313 --> 00:27:46.013
that find interacting with, uh, with
the types of high-powered attorneys that

00:27:46.013 --> 00:27:48.653
do fund structuring deeply unpleasant.

00:27:49.053 --> 00:27:55.653
And so they really enjoy just talking
to this, this chat that is very nice

00:27:55.653 --> 00:27:59.513
and unfailingly polite and always
encouraging and tells you that you're,

00:27:59.523 --> 00:28:02.723
you're, you're brilliant and, and
are going to be successful and gets

00:28:02.723 --> 00:28:07.403
all the information out of you over a
timescale that, that, that a lawyer just

00:28:07.813 --> 00:28:10.793
probably couldn't make money if they
spent that long with a client, right?

00:28:10.793 --> 00:28:13.453
In ma- in many ways, these kind of
high-powered attorneys are finding

00:28:13.453 --> 00:28:17.343
themselves in the, the same type of
situation that doctors were in, you know,

00:28:17.343 --> 00:28:21.413
twenty, thirty years ago as they were, you
know, starting to really push the, you got

00:28:21.413 --> 00:28:25.103
to get the interview down, you know, the,
the time with the doctor down to fifteen

00:28:25.103 --> 00:28:26.513
minutes and then five minutes, right?

00:28:26.533 --> 00:28:30.293
That's, that's-- They're not doing
that in many ways, or the, the firms

00:28:30.293 --> 00:28:33.403
are resisting that, and they're, uh,
they're better financed than, uh, than

00:28:33.403 --> 00:28:36.393
the, than the doctoral group, the,
the medical groups were at the time.

00:28:36.983 --> 00:28:40.429
But they're, they're under that
same kind of economic pressure, You

00:28:40.429 --> 00:28:42.589
know, and another way to think about
this is, you know, every business is

00:28:42.589 --> 00:28:44.379
becoming a content business twice over.

00:28:44.389 --> 00:28:45.479
Once in marketing, right?

00:28:45.479 --> 00:28:48.509
You gotta, you gotta generate demand,
you gotta solve the distribution problem.

00:28:49.099 --> 00:28:51.689
And then a second time
in compliance, right?

00:28:52.319 --> 00:28:55.699
Footage or any other computerized
artifact is becoming the permanent

00:28:55.709 --> 00:28:58.839
record of what was sold, what was
done, and how it was done, right?

00:28:59.159 --> 00:29:02.219
And, and, you know, again, as I've
noted in the past, the software

00:29:02.219 --> 00:29:05.419
that finds a camera angle for a
narrative mirrors the software that

00:29:05.419 --> 00:29:07.099
finds a firing angle to a target.

00:29:07.119 --> 00:29:08.589
This is the same geometry.

00:29:08.950 --> 00:29:13.250
And, and a-another, another thing to, to
think about is that the same loop applies

00:29:13.250 --> 00:29:14.610
to co- applies to copyrighted works.

00:29:14.610 --> 00:29:19.420
So, um, you, you extract the specs,
and then you regenerate from the

00:29:19.450 --> 00:29:21.920
specs alone in a separate AI process.

00:29:22.240 --> 00:29:25.280
And, uh, for copyright purposes,
that regenerated artifact is, is,

00:29:25.300 --> 00:29:29.330
at least under current law, legally
independent of the prior file, even

00:29:29.330 --> 00:29:31.060
when it does all the same things, right?

00:29:31.510 --> 00:29:34.260
Um, so templates were already
doing this in the, in the areas

00:29:34.260 --> 00:29:35.750
of, of law that could be templated.

00:29:36.110 --> 00:29:39.510
Um, but AI makes the, the kind of
the spec-to-work loop cheap enough

00:29:39.510 --> 00:29:43.260
to change the business model in, in a
whole host of, a whole host of ways.

00:29:43.606 --> 00:29:44.816
So let's talk about medicine.

00:29:45.346 --> 00:29:51.936
Um, in, in, also in April of this year,
Doximity OpMed, uh, run by Oren Foss, an

00:29:51.936 --> 00:29:55.826
MD who runs a solo cataract micro practice
in Denton, Texas, which is, you know, very

00:29:55.826 --> 00:29:58.266
close to me, has zero full-time employees.

00:29:58.876 --> 00:30:01.636
AI routes incoming faxes into his EMR.

00:30:01.906 --> 00:30:05.226
Uh, he self-schedules patients, and he
automates the insurance verification.

00:30:05.236 --> 00:30:09.076
He aims at about five new patients
and ten cataracts a week and

00:30:09.086 --> 00:30:10.506
gives each patient about an hour.

00:30:10.506 --> 00:30:16.026
So that-- this actually, this model has,
has allowed him to spend more time with

00:30:16.046 --> 00:30:19.746
the clients because he's spending almost
no time doing the paperwork, right?

00:30:20.026 --> 00:30:26.386
Uh, it's one guy plus this AI harness, and
that's, that's taking all of the, um, all

00:30:26.386 --> 00:30:30.506
of the other people out, which means he
doesn't have to do the volume that he did.

00:30:30.556 --> 00:30:34.626
So again, in, in law, the, there,
there's this, this pressure to

00:30:34.836 --> 00:30:36.076
spend less time with the client.

00:30:36.106 --> 00:30:38.776
In medicine, they've kind of hit
their, hit their rock bottom, and

00:30:38.786 --> 00:30:41.706
AI is kind of helping them come back
to spend more time with the client.

00:30:42.096 --> 00:30:45.726
Uh, similar things are happening
in, in manufacturing, right?

00:30:46.876 --> 00:30:49.006
So there's a company called Figure AI.

00:30:49.066 --> 00:30:54.736
They have a, um, they have a
factory called BotQ, and they,

00:30:54.746 --> 00:30:57.966
they, they, they have a robot they,
they're calling the Figure 03.

00:30:58.216 --> 00:31:01.916
And they were able to ramp up the
production of their robots from one

00:31:01.916 --> 00:31:06.536
robot today-- per day to one robot per
hour, about twenty-four times in under

00:31:06.536 --> 00:31:09.766
a hundred and twenty days with more than
three hundred, uh, fifty robots delivered.

00:31:09.936 --> 00:31:12.896
And the way they did that was
they took their humanoid robots

00:31:12.896 --> 00:31:15.306
that they're building, and they
put them into the workflow.

00:31:15.826 --> 00:31:20.256
So, um, so their, their Figure robots, the
same robots that they're selling, are in

00:31:20.256 --> 00:31:23.776
their factory assembling key components
and moving material between stations.

00:31:23.916 --> 00:31:25.456
So you have robots building robots.

00:31:25.456 --> 00:31:27.836
And, and, um, you know, it's
a really interesting setup.

00:31:27.876 --> 00:31:31.336
They have a custom MES across a
hundred fifty plus workstations

00:31:31.596 --> 00:31:32.896
plus that hybrid line.

00:31:33.386 --> 00:31:37.926
And, and, you know, one of the things
that we're learning is that, that

00:31:37.966 --> 00:31:44.216
the AI robot stuff is gonna change
how we organize our factories.

00:31:44.396 --> 00:31:47.876
And I, I would argue that it's gonna have
a similar impact on, on our corporations.

00:31:48.166 --> 00:31:49.636
And this has happened before, right?

00:31:49.666 --> 00:31:55.006
A century ago when, uh, when we moved
from, from steam, um, steam engines

00:31:55.006 --> 00:31:56.256
or sometimes water wheel engines.

00:31:56.266 --> 00:32:01.056
The way factories were set up, you had
one kind of source of mechanical power

00:32:01.316 --> 00:32:05.476
on the wall kinda coming in from either
this-- the boiler room or again, from like

00:32:05.476 --> 00:32:11.126
a water wheel, and you would mechanically
attach to whatever you're doing to

00:32:11.146 --> 00:32:13.176
that big, that big mechanical source.

00:32:13.176 --> 00:32:16.536
There's a-- If, if anybody's seen,
you know, the old, uh, the old,

00:32:16.536 --> 00:32:19.076
um, Chitty Chitty Bang Bang, right?

00:32:19.076 --> 00:32:22.706
With, where he's got his, he's got
his windmill, and inside the windmill,

00:32:22.716 --> 00:32:27.106
he's got just a ton of stuff that's
mechanically hooked up to that windmill

00:32:27.106 --> 00:32:30.696
and all these gears, and he, he's kind
of got ways of attaching new machines

00:32:30.696 --> 00:32:32.666
as he builds around his workshop, right?

00:32:32.833 --> 00:32:36.553
when people first started getting
electric motors, the idea was, "Well,

00:32:36.553 --> 00:32:40.533
we're just gonna get a, a, an electric
motor and put it on that big drive

00:32:40.533 --> 00:32:44.173
shaft at, at the top of the factory,
and, and that's gonna be better." Well,

00:32:44.173 --> 00:32:45.703
it, it wasn't really better, right?

00:32:45.723 --> 00:32:48.513
Sometime-- I mean, occasionally, there,
there, there were, there were places

00:32:48.513 --> 00:32:52.893
where, um, they couldn't get a boiler
room or they couldn't get water power for

00:32:52.893 --> 00:32:57.703
whatever reason, um, and, and the electric
motor was better, uh, better than nothing.

00:32:57.923 --> 00:33:02.293
But the real gains came when instead
of having one big source of mechanical

00:33:02.363 --> 00:33:06.193
power that you then vector to,
to each of your workstations, you

00:33:06.193 --> 00:33:10.903
just put an electric engine at each
workstation and then ultimately on,

00:33:10.943 --> 00:33:12.733
on each machine, each power tool.

00:33:12.893 --> 00:33:15.043
And that was the real, the real gains.

00:33:15.343 --> 00:33:18.653
And so we're gonna have to, we're
gonna have to do the same thing, right?

00:33:18.901 --> 00:33:22.731
Our corporations, the way our corporations
are working right now, they have a

00:33:22.731 --> 00:33:27.201
management system that is effectively
that big mechanical drive shaft.

00:33:27.461 --> 00:33:29.291
And so we need to
redesign the corporation.

00:33:29.291 --> 00:33:33.191
We need to redesign human organization
around distributed judgment and command,

00:33:33.401 --> 00:33:37.791
not around departmental handoffs
built for a pre-AI factory, right?

00:33:37.791 --> 00:33:41.251
So, so there's a bunch of people that
are trying to figure out how to do this.

00:33:41.521 --> 00:33:44.811
I'm not-- I don't, I don't know that
anyone quite has, you know, seems

00:33:44.811 --> 00:33:47.681
to have come out with a, "Hey, this
is the secret sauce. This is the way

00:33:47.681 --> 00:33:51.381
to do it." you know, and this, this
is a new information technology.

00:33:51.381 --> 00:33:54.321
The, the corporation was
originally built around paper.

00:33:54.321 --> 00:33:55.771
It was, it was the printing press.

00:33:55.771 --> 00:33:59.921
It was the ability to create identical
sheets of paper and spread them

00:33:59.921 --> 00:34:02.641
around what you were doing that
allowed us to do the corporation.

00:34:02.951 --> 00:34:05.911
And now AI is, is
dramatically changing that.

00:34:06.121 --> 00:34:09.751
You know, one of the things that,
that AI makes possible is that, um,

00:34:09.761 --> 00:34:12.652
you can have the same information.

00:34:12.832 --> 00:34:16.162
It looks like, uh, my friend Robert
has jumped off, but this is what, this

00:34:16.172 --> 00:34:17.922
is what Runcible is, is working on.

00:34:18.232 --> 00:34:22.402
Um, you can have the same information
presented in multiple different ways.

00:34:22.852 --> 00:34:25.892
Um, and, and that's, that's a
problem that AI can solve, right?

00:34:25.892 --> 00:34:28.682
It can bring a bunch of different
information or information systems

00:34:28.892 --> 00:34:31.562
that categorize it a little bit
differently and bring it into one

00:34:31.562 --> 00:34:35.972
information-- one kind of user interface,
and you can standardize that across

00:34:35.972 --> 00:34:36.912
all, all the things you're doing.

00:34:37.052 --> 00:34:42.542
Another thing that you could do is you
could have every single person have

00:34:42.552 --> 00:34:44.952
their own idiosyncratic user interface.

00:34:45.482 --> 00:34:46.462
Now within limits, right?

00:34:46.462 --> 00:34:49.082
The, the, the corporation is gonna
have requirements, but you could

00:34:49.092 --> 00:34:54.312
have the AI optimize the user
interface for each individual person.

00:34:54.452 --> 00:34:58.542
And I think certainly for people that
are using AI, when they've got the power,

00:34:58.702 --> 00:35:02.262
they've got the, the, the, the tool
that can make them new stuff, right?

00:35:02.342 --> 00:35:04.262
They're for sure gonna do that, right?

00:35:04.472 --> 00:35:08.442
It's the same kind of thing, you know,
I, I really wonder when, uh, when

00:35:08.442 --> 00:35:13.122
Figure gets its, um, gets its robots
into production, like, what are you not

00:35:13.122 --> 00:35:14.832
gonna be able to do with that robot?

00:35:14.832 --> 00:35:18.892
Like, you buy a robot that can, that
can make stuff, are, are you gonna need

00:35:18.932 --> 00:35:20.692
construction workers anymore, right?

00:35:20.692 --> 00:35:25.652
When you, when you do a remodel of your
house or a remodel of, of, of your factory

00:35:25.652 --> 00:35:29.582
or your office, if you've got one of those
robots, you'd be crazy not to have the

00:35:29.582 --> 00:35:31.912
robot just work on it at nighttime, right?

00:35:31.992 --> 00:35:35.432
Like, this is gonna create all…
Like, just having that thing, having

00:35:35.432 --> 00:35:41.216
that capacity in there is going to
dramatically reshape things, so All right.

00:35:41.226 --> 00:35:44.536
So now we're gonna talk for, for
the next extended little bit, we're

00:35:44.536 --> 00:35:47.856
gonna talk about one of my favorite
subjects, which is farming, right?

00:35:47.986 --> 00:35:51.556
And of course, we'll start with my, uh,
you know, the, the, the obligatory, those

00:35:51.556 --> 00:35:54.536
of you who have heard me rant on this
subject will just have to put up with it.

00:35:54.800 --> 00:35:58.610
Right now, we farm in a way that
is, I consider really bad, right?

00:35:58.710 --> 00:36:01.550
Standard industrial agriculture
runs on a petrochemical stack.

00:36:01.550 --> 00:36:03.960
We have synthetic fertilizers,
we have herbicides, we have

00:36:03.960 --> 00:36:07.530
insecticides, we have fungicides,
and there's just a bunch of hazards.

00:36:07.530 --> 00:36:09.030
There's, you know, acute toxicity.

00:36:09.309 --> 00:36:12.899
presumably, most of you will have seen
on, on, on Twitter at one point, you know,

00:36:12.899 --> 00:36:17.649
the, the pictures of people out, you know,
in the fields with hazmat suits on, right?

00:36:18.059 --> 00:36:21.169
I know parenthetically that, that people
are considering whether or not to,

00:36:21.169 --> 00:36:25.599
uh, to allow amnesty for farm workers
right now, and we really should ask

00:36:25.739 --> 00:36:30.359
what, what is going on in those, uh, in
those working conditions and, and, you

00:36:30.359 --> 00:36:32.209
know, do we, do we have modern slaves?

00:36:32.209 --> 00:36:33.919
And is that, is that
really how we wanna do it?

00:36:34.229 --> 00:36:36.819
Um, and then there's also, you
know, long-term effects on soil

00:36:36.819 --> 00:36:40.099
microbiome, pollinators, residual
chemistry in food and water that,

00:36:40.369 --> 00:36:41.559
you know, we're still figuring out.

00:36:41.579 --> 00:36:45.699
Um, so I, I have a friend of mine who
has-- owns land that was formerly a,

00:36:45.719 --> 00:36:49.779
a peanut farm, and you use a lot of
Roundup in, in, in peanut farming.

00:36:49.969 --> 00:36:53.389
And so for peanuts, you know, what
you want is, you know, three to six

00:36:53.389 --> 00:36:56.109
inches of sand and then clay, right?

00:36:56.139 --> 00:36:59.169
Or at least you can do peanuts
on that land, and so people do.

00:36:59.828 --> 00:37:03.058
but that setup, when you're
spraying Roundup on the fields,

00:37:03.238 --> 00:37:07.058
uh, the Roundup goes down into
the clay, and it just sits there.

00:37:07.438 --> 00:37:11.428
A Roundup, you know, glyphosate, you
know, breaks down in, in an oxygen-based

00:37:11.428 --> 00:37:13.098
environment, in an aerobic environment.

00:37:13.358 --> 00:37:16.858
But if it gets into the clay, the clay
is anaerobic, uh, it doesn't break

00:37:16.858 --> 00:37:19.858
down, and so it just concentrates
year after year after year.

00:37:19.858 --> 00:37:25.208
And so you get, you know, you get clay
samples that have three or four times the

00:37:25.208 --> 00:37:29.538
legal d- limit for what you're supposed to
have in soil that, that, that, you know,

00:37:29.538 --> 00:37:31.898
is, is around humans and, and animals.

00:37:32.248 --> 00:37:34.488
And, you know, nobody,
nobody thought of that.

00:37:34.508 --> 00:37:35.298
Nobody knew that.

00:37:35.298 --> 00:37:39.268
Nobody had done that analysis ahead of
time and then, "Oh, well, maybe on this

00:37:39.268 --> 00:37:41.968
particular land, we shouldn't use this
chemical that's in everything," you know.

00:37:42.158 --> 00:37:48.008
And, and, I mean, glyphosate is, is
a, is a crazy common chemical, right?

00:37:48.378 --> 00:37:51.118
So-- But there's some really
interesting new technologies that

00:37:51.118 --> 00:37:53.518
are coming out that I think are
gonna, are, are gonna change that.

00:37:54.198 --> 00:37:57.818
So the one, the one that, that's, that I
think is, is most applicable to the stuff

00:37:57.818 --> 00:37:59.418
we're talking about today is, is light.

00:37:59.898 --> 00:38:04.368
So, um, using lasers, uh, to… They
can do a bunch of different things.

00:38:04.368 --> 00:38:06.918
One of the things they can do is
they can suppress surface pathogens.

00:38:06.928 --> 00:38:11.498
So you have, uh, you have, uh, blight,
or you have some fungus, or you have

00:38:11.498 --> 00:38:14.858
something that's kinda growing on the
plant that you're trying to harvest.

00:38:15.068 --> 00:38:22.388
You can use lasers, um, you can use UV to,
to deal with those, those fungus, those

00:38:22.388 --> 00:38:26.678
fungus without using a chemical fungicide,
without using a chemical herbicide, right?

00:38:26.918 --> 00:38:30.438
If you have laser weed eating,
you burn the weed's growing point

00:38:30.608 --> 00:38:32.478
with no herbicide residue, right?

00:38:33.038 --> 00:38:35.818
Um, and then you have spectrum
recipes, so, so different wavelengths.

00:38:35.828 --> 00:38:38.348
You know, everyone's heard of, like,
red light therapy for humans now.

00:38:38.618 --> 00:38:42.638
Well, for plants, there is red
light and blue, far-red, green,

00:38:42.638 --> 00:38:45.708
controlled UV, and all of these
kind of activate different things.

00:38:45.728 --> 00:38:47.448
They shape growth and defense chemistry.

00:38:47.658 --> 00:38:49.428
Light is, is food and signal.

00:38:49.638 --> 00:38:53.186
Now, of course, light cannot It can
make nutrient use more efficient.

00:38:53.186 --> 00:38:56.756
It cannot-- Like, if there's no nitrogen
or no phosphorus or no potassium in

00:38:56.756 --> 00:39:00.166
the, in the plant, it can't create
those things ex nihilo, right?

00:39:00.176 --> 00:39:02.206
Those, those things are,
are, are, you know, you gotta

00:39:02.216 --> 00:39:03.706
have those atoms in the soil.

00:39:04.086 --> 00:39:06.686
What light can amplify
is a living polyculture.

00:39:06.686 --> 00:39:10.376
Many species on one acre, not a single
crop drowning in synthetic inputs.

00:39:10.866 --> 00:39:14.096
Um, there's a guy who I greatly
admire named Joel Salatin in

00:39:14.096 --> 00:39:18.186
Virginia, and he has talked about
this idea of polyculture, right?

00:39:18.366 --> 00:39:22.346
So having many plants, um, and
on, and on Salatin's farm, animals

00:39:22.406 --> 00:39:25.296
sharing the same ground, so the
acre does several jobs at once.

00:39:25.596 --> 00:39:27.136
Um, and this is really interesting.

00:39:27.146 --> 00:39:30.686
This allows you in many ways to get away
from, uh, from chemical monoculture.

00:39:31.196 --> 00:39:35.106
Um, one of the, the, the famous
examples that Salatin uses that

00:39:35.106 --> 00:39:36.976
I'm, I'm, I'm taking, right?

00:39:36.976 --> 00:39:37.786
But he came up with this.

00:39:38.246 --> 00:39:43.306
Um, at the… The old prairie
was, was very interesting.

00:39:43.656 --> 00:39:45.806
The old prairie was, you know, just grass.

00:39:45.806 --> 00:39:48.746
It was, it was, it was,
you know, very big, right?

00:39:48.746 --> 00:39:51.336
The prairie grass would grow
higher than a man's head.

00:39:51.811 --> 00:39:54.401
and it had a bunch of
different species in it.

00:39:54.401 --> 00:39:57.691
And basically, they thought, well,
prairie grass is just this, is this

00:39:57.691 --> 00:40:00.471
magical grass of some, some genetic
variant that was extinguished 'cause

00:40:00.471 --> 00:40:03.651
we cut down the prairie grass and it
didn't grow, uh, it didn't grow back.

00:40:03.671 --> 00:40:06.121
And what they found was they, they… By
the way, they still have, I think it's

00:40:06.121 --> 00:40:10.431
the University of Wisconsin still has like
a, like a quarter acre or a half acre of,

00:40:10.431 --> 00:40:12.711
uh, of prairie grass that they preserved.

00:40:13.651 --> 00:40:16.861
And it's got like fifty
different species in it.

00:40:17.381 --> 00:40:22.071
Um, and we, we have been able, you know,
to, to replicate when you introduce more

00:40:22.201 --> 00:40:24.231
types of grass, especially many more.

00:40:24.231 --> 00:40:28.201
So, so, uh, Cedar Creek, uh, did
this thing where they had sixteen

00:40:28.261 --> 00:40:32.071
species of grass, and the grass
grew two and a half times, two point

00:40:32.091 --> 00:40:34.991
seven times larger and more dense.

00:40:35.121 --> 00:40:39.721
Um, and so, so if you, if you have
grass that is just one species,

00:40:39.931 --> 00:40:43.451
just one type of grass, it, it
can't, it can't grow very big.

00:40:43.881 --> 00:40:47.141
Uh, mixed roots and canopies build
biomass and soil architecture in a way

00:40:47.141 --> 00:40:48.731
that a single variety cannot match.

00:40:48.961 --> 00:40:49.161
Yeah.

00:40:49.651 --> 00:40:51.661
Another thing that different
plants do is they put different

00:40:51.671 --> 00:40:52.981
nutrients back into the soil.

00:40:53.401 --> 00:40:56.311
Uh, legumes fix nitrogen, deep
roots mine minerals, covers and

00:40:56.311 --> 00:40:57.621
companions feed the microbiome.

00:40:57.841 --> 00:41:00.371
And, and you've heard, you know,
uh, George Washington Carver, crop

00:41:00.371 --> 00:41:03.811
rotation, you know, that's where,
that's when we started raising peanuts.

00:41:03.901 --> 00:41:08.923
Um- But the, the, the thing that
is interesting about polyculture is

00:41:09.023 --> 00:41:13.423
what if you could have all of those
plants in the field at the same time?

00:41:13.433 --> 00:41:16.123
Especially if you pair that with,
you know, light prescription.

00:41:16.123 --> 00:41:19.563
So you have spectrum, you have dose
and timing, and you can get this

00:41:19.563 --> 00:41:24.443
vigor without having to, to spray the
fields with petrochemical mixes that

00:41:24.443 --> 00:41:26.013
you need a hazmat suit for, right?

00:41:26.493 --> 00:41:29.203
Um, now the reason why we don't do
this, the reason why we haven't done

00:41:29.203 --> 00:41:34.023
this is that a combine cannot pick
three crops, uh, at one time, right?

00:41:34.033 --> 00:41:36.863
The combine, you gotta tune it
for whatever crop it's picking.

00:41:37.043 --> 00:41:40.513
There's a certain pressure that it can
use, and you have to have one thing

00:41:40.733 --> 00:41:43.593
that, you know, uh, that if you're
using a spray boom or whatever, it, it,

00:41:43.833 --> 00:41:46.063
it needs every plant to be the same.

00:41:46.143 --> 00:41:48.123
And, and in fact, you know,
that's kinda how, how it works.

00:41:48.343 --> 00:41:51.743
If you have, you know, potatoes or
whatever that are, that are too soft or

00:41:51.743 --> 00:41:53.923
sometimes too hard, those are rejected.

00:41:53.923 --> 00:41:54.753
There's something wrong with them.

00:41:54.753 --> 00:41:56.023
We wanna sort those out.

00:41:56.433 --> 00:42:00.053
But more and more, we're having, you
know, articulated robotic hands and,

00:42:00.053 --> 00:42:01.613
and they, and they change that, right?

00:42:01.783 --> 00:42:04.263
So you can, you can have
something that's integrated.

00:42:04.263 --> 00:42:06.953
It can see, classify, and
handle individual plants.

00:42:07.583 --> 00:42:11.523
And, and with this, with AI and
robots, polyculture becomes something

00:42:11.523 --> 00:42:13.033
that can reach industrial scale.

00:42:13.533 --> 00:42:16.803
the, the, the example that I'm
giving give is, uh, Kyber Labs.

00:42:16.803 --> 00:42:21.513
So, uh, at the end of last year, uh,
Kyber Labs posted a back-drivable

00:42:21.513 --> 00:42:24.603
robotic hand spinning a nut
onto a bolt at extreme speed.

00:42:24.603 --> 00:42:28.393
So they had the, the nut and the
bolt and, and the hand was very, very

00:42:28.393 --> 00:42:32.313
quickly, like it, it was blurred,
um, turning the, the, the nut on.

00:42:32.523 --> 00:42:37.683
And what they did was somebody would stick
their finger in into this very fast-moving

00:42:37.683 --> 00:42:40.083
machine, and the AI would stop it, right?

00:42:40.113 --> 00:42:40.813
Immediately.

00:42:40.813 --> 00:42:42.083
It wouldn't hurt your hand.

00:42:42.931 --> 00:42:46.861
I don't really understand why this is
true, but for whatever reason, he got this

00:42:46.861 --> 00:42:50.871
bad interview with some, some of these
mainstream, uh, mainstream interviewers.

00:42:50.871 --> 00:42:55.211
It was a couple of women, and they decided
that, that his product was, like, about

00:42:55.211 --> 00:42:57.061
sex toys or something of this nature.

00:42:57.361 --> 00:43:01.061
And so there's some really weird
marketing around him or re-weird

00:43:01.101 --> 00:43:02.501
kind of search results around him.

00:43:02.871 --> 00:43:06.131
Um, the founder kinda came out on X.
He was like, "This is not what we're

00:43:06.131 --> 00:43:09.301
about. We don't know why they talked
about this. This is not what we talked

00:43:09.301 --> 00:43:11.231
about." This is very frustrating to him.

00:43:11.601 --> 00:43:13.211
Uh, it's a great innovation.

00:43:13.621 --> 00:43:16.671
Um, and, and the real payoff that
I see is… So, but, you know,

00:43:16.671 --> 00:43:19.481
obviously, if you search for it,
just kinda, uh, be, be advised.

00:43:19.491 --> 00:43:23.701
Um, some, some, so he-- They got some
interviewers that were, that were weird,

00:43:23.711 --> 00:43:25.881
bad modern women, though I repeat myself.

00:43:26.585 --> 00:43:30.105
so the real payoff that I see is as a
general purpose harvester, you know,

00:43:30.115 --> 00:43:34.165
for soft fruits and for different
fruits at industrial speeds, okay?

00:43:35.529 --> 00:43:39.029
So, um, and then, then there's a, a
related concept called permaculture,

00:43:39.039 --> 00:43:40.539
which takes polyculture even further.

00:43:40.539 --> 00:43:45.219
So you stack layers, canopy, understory,
shrubs, herbs, and ground cover, um,

00:43:45.239 --> 00:43:48.249
and it's designed as a food forest
that feeds itself over the years.

00:43:48.249 --> 00:43:51.049
And, and, and again, the opportunity
that I want you to see is with

00:43:51.049 --> 00:43:54.309
robotic harvesting that can see and,
you know, with, with AI that can

00:43:54.309 --> 00:43:58.059
see and pick plant by plant, you are
no longer forced into monocultural

00:43:58.069 --> 00:43:59.399
rows for the sake of the machine.

00:43:59.659 --> 00:44:03.469
You can still get incredibly efficient
yields, dense layered production

00:44:03.739 --> 00:44:05.509
without using hazmat chemistry.

00:44:05.719 --> 00:44:05.979
Okay?

00:44:06.579 --> 00:44:10.369
And, you know, we might call this a, a
plant ledger, a living record of every

00:44:10.369 --> 00:44:14.309
plant that stood on that square foot for
the last three years and, and a plan,

00:44:14.309 --> 00:44:17.849
a projection of every plant that will
stand there for the next three, right?

00:44:17.849 --> 00:44:21.089
If you pair that light prescription
with that ledger, you have rotation,

00:44:21.089 --> 00:44:23.819
companion planting, and the, and the,
the light spectrum, the wavelengths,

00:44:24.139 --> 00:44:27.269
that becomes a plan score, not, not
a guess after the spray truck leaves.

00:44:27.269 --> 00:44:31.629
Because right now, again, we do things at
the level of, you know, a, a sometimes a

00:44:32.079 --> 00:44:34.099
fifty-acre field or a thirty-acre field.

00:44:34.099 --> 00:44:37.139
There are still people that do,
you know, ten-acre fields, um, but,

00:44:37.179 --> 00:44:40.469
but less and less of them because,
you know, monocultures is becoming

00:44:40.479 --> 00:44:43.059
less and less efficient, and the
margins are, are worse and worse.

00:44:43.079 --> 00:44:46.479
There's, there's way more co-quality
control issues that, that, that

00:44:46.479 --> 00:44:47.599
you-- than you used to have.

00:44:48.219 --> 00:44:52.809
Um, for management and for compliance,
you know, whether it's buyer regulatory

00:44:52.819 --> 00:44:55.859
or insurer, I think this is an
incredible use case for some kind

00:44:55.859 --> 00:44:57.619
of blockchain or crypto technology.

00:44:57.909 --> 00:45:00.159
You know, it's an auditable
history of what grew where,

00:45:00.159 --> 00:45:01.509
what was treated, and when.

00:45:01.805 --> 00:45:05.625
And, and then again, we, we, we
are reducing this to the same

00:45:05.645 --> 00:45:07.595
three basic skill sets, right?

00:45:07.835 --> 00:45:11.525
So you have AI-- little drones
that are going out and scouting.

00:45:11.765 --> 00:45:14.335
They've got cameras on there that
are feeding back into an AI system

00:45:14.335 --> 00:45:17.555
that's classifying what it's seeing,
and then there's doing analysis and

00:45:17.555 --> 00:45:19.475
prescribing what do we need to do.

00:45:19.655 --> 00:45:24.395
Then if, if what it needs to do is, is,
is, you know, strike with a laser at

00:45:24.405 --> 00:45:29.255
a, at a bug or at a, at a weed or at
something that needs to be cauterized,

00:45:29.625 --> 00:45:33.345
um, you know, this is, this is the
same loop as that defensive swarm,

00:45:33.355 --> 00:45:36.085
that, that defensive drone swarm
that we were talking about last week.

00:45:36.415 --> 00:45:39.475
Um, it's defending a field faster
than the farmers of today can imagine.

00:45:39.475 --> 00:45:42.835
I mean, you know, most farmers basically
if, if you get blight, if you get

00:45:42.835 --> 00:45:46.795
some kind of, of, you know, fungal
or some kind of infection in your

00:45:46.795 --> 00:45:48.815
field, the whole field is done, right?

00:45:48.845 --> 00:45:50.335
There's, there's no coming back from that.

00:45:50.555 --> 00:45:55.505
Well, if you could have drones that
are flying around looking, and every

00:45:55.505 --> 00:45:59.235
time they, they see something going bad
or, or, or, you know, fungus starting

00:45:59.235 --> 00:46:01.165
to grow, they cauterize the area.

00:46:01.165 --> 00:46:03.845
Or every time they see a weed,
they, they zap it, right?

00:46:03.845 --> 00:46:07.055
Every time they see a bug or, you
know, um, you know, some, some kind

00:46:07.055 --> 00:46:11.355
of, of insect, some kind of pest coming
in, they, they can kill the bugs as

00:46:11.365 --> 00:46:14.085
they're flying into the field, right?

00:46:14.275 --> 00:46:17.225
Well, then you don't need to just
leave these chemicals that have

00:46:17.225 --> 00:46:19.445
to last for days or weeks, right?

00:46:19.745 --> 00:46:21.965
Um, 'cause I mean, that-that's one of
the real problems with these chemicals.

00:46:22.125 --> 00:46:24.165
They're kind of designed
to be forever chemicals.

00:46:24.165 --> 00:46:27.885
If they were easy to break down, then
you'd have to spray them all the time.

00:46:27.885 --> 00:46:31.715
So you wanna-- what you want is to be
able to spray and then, you know, then

00:46:31.715 --> 00:46:35.905
have it last so that anything that
flies in for the next six weeks or, or

00:46:35.905 --> 00:46:38.335
more dies that you don't like, okay?

00:46:38.655 --> 00:46:41.695
And, and again, I think, I think
this is gonna dramatically,

00:46:41.795 --> 00:46:43.155
uh, reshape agriculture.

00:46:43.275 --> 00:46:48.235
So, uh, one of the, one of the, the, the
companies in this space is called Carbon.

00:46:48.285 --> 00:46:50.275
Just so you know that this is not
just something that I came with--

00:46:50.335 --> 00:46:51.395
came up with out of my own mind.

00:46:51.695 --> 00:46:55.345
Um, I-- for the record, I was
talking about this before any of

00:46:55.345 --> 00:46:58.075
these companies existed or had
announced any of these things.

00:46:58.615 --> 00:47:00.765
I saw it coming, gentlemen,
uh, ladies and gentlemen.

00:47:01.475 --> 00:47:04.995
So in February of twenty twenty-five,
Carbon Robotics announced, uh, scale

00:47:04.995 --> 00:47:06.685
numbers for its laser weeder line.

00:47:06.805 --> 00:47:10.405
It's a, it's a tractor-pulled module
with cameras, classifiers, and CO2

00:47:10.405 --> 00:47:13.305
layers-- lasers trained on more than
a hundred and fifty million plants.

00:47:13.305 --> 00:47:16.585
So Carbon says this tech has been
used on two hundred and fifty thousand

00:47:16.585 --> 00:47:20.275
acres in fifteen countries on a hundred
plus crop types with more than fifteen

00:47:20.285 --> 00:47:24.352
billion weeds eliminated, which is great
marketing copy, and, and again, the system

00:47:24.352 --> 00:47:25.862
is, is doing what I was talking about.

00:47:25.862 --> 00:47:28.892
It can, it can look at, it can take
a picture, it can identify what is

00:47:28.892 --> 00:47:30.182
the crop versus what is the weed.

00:47:30.182 --> 00:47:33.792
It can feed, uh, fire at the weed,
and, you know, and, and that might

00:47:33.792 --> 00:47:38.482
be, you know, mere centimeters, inches
or centimeters from the plants that

00:47:38.482 --> 00:47:40.972
you're needing to protect, and it
only hits what you want it to hit.

00:47:41.182 --> 00:47:43.332
And then it, it logs the shot forever.

00:47:43.342 --> 00:47:47.472
Like, that-- whatever-- every single
time you zap something, the system

00:47:47.472 --> 00:47:50.452
knows that it, it zapped, and it
knows, you know, basically on what

00:47:50.462 --> 00:47:54.319
square foot it did it, So then Cornell
University, uh, which is the land-grant

00:47:54.319 --> 00:47:58.019
university in New York, it's got,
uh, you know, Cornell Agri-Agritech.

00:47:58.539 --> 00:48:01.409
Um, it runs comparative field
trials on farm technology.

00:48:01.549 --> 00:48:04.649
So in June of twenty twenty-five,
the Cornell Chronicle reported their,

00:48:04.669 --> 00:48:09.119
their trial of commercial laser weeders
against common East Coast weeds.

00:48:09.348 --> 00:48:13.508
and so what they found was that lasers
can match or beat common herbicides

00:48:13.528 --> 00:48:15.448
on several upright broadleaf weeds.

00:48:15.778 --> 00:48:18.288
It's weaker on some grasses
and sprawling plants.

00:48:18.758 --> 00:48:23.398
But again, the thing is, is that y-you
can record every single kill, right?

00:48:23.398 --> 00:48:27.558
You can record every single laser strike,
and, and you can preserve that so you

00:48:27.568 --> 00:48:30.878
know-- you, you, you don't just know
that all the weeds were taken care of.

00:48:31.028 --> 00:48:32.958
You know how many weeds
were taken care of.

00:48:32.958 --> 00:48:34.198
You can see the whole thing.

00:48:34.778 --> 00:48:37.818
And, and, and I'll note
parenthetically, um, this is the

00:48:37.818 --> 00:48:39.328
defender's wartime loop, right?

00:48:39.378 --> 00:48:42.298
You scout, you, you classify,
and then you strike.

00:48:42.668 --> 00:48:45.588
And, uh, this-- what this is gonna
mean is that the ROTC and the Grange

00:48:45.588 --> 00:48:50.008
kids are o-once again gonna be the same
kids using effectively the same tech.

00:48:50.195 --> 00:48:53.015
Trick Robotics over, you know, in
California, there's a guy named

00:48:53.055 --> 00:48:57.155
Adam Stager, uh, who is-- again,
these, these are tractor units.

00:48:57.315 --> 00:49:01.805
Um, they're using UV, UV light and
bug vacuums on some of California's

00:49:01.805 --> 00:49:02.975
largest strawberry fields.

00:49:03.045 --> 00:49:05.555
So they've, they've treated
tens of thousands of acres, and

00:49:05.555 --> 00:49:08.245
they're, you know, they're, they're
doubling their fleet every year.

00:49:08.835 --> 00:49:12.225
So they-- basically, they go out at
night and with no chemicals, they have

00:49:12.225 --> 00:49:15.845
UV, they kill the pathogens and pests,
and then they suck it out of the field.

00:49:15.855 --> 00:49:15.885
Okay?

00:49:16.286 --> 00:49:19.386
I think that the next step, th-this
is the thing that I'm gonna predict,

00:49:19.676 --> 00:49:24.606
is that, that this is gonna leave
the, um, we-- this is gonna leave

00:49:24.616 --> 00:49:26.006
the, the, the tractor thing.

00:49:26.036 --> 00:49:27.986
There might be still some things
that, that run on tractors.

00:49:27.986 --> 00:49:31.206
Obviously, for whatever reason, if
you need to vacuum stuff out, right?

00:49:31.216 --> 00:49:33.286
That'll probably be more
efficient on a tractor.

00:49:33.806 --> 00:49:38.706
But I think for lasers, that can be
put on instead of having, uh, you know,

00:49:38.726 --> 00:49:42.886
a laser array in a box that you're
dragging across a field, have one

00:49:42.886 --> 00:49:46.826
laser per drone and have, have a cloud
of drones moving across the field.

00:49:46.996 --> 00:49:50.596
And one of the things that will allow
is if one of those breaks, then you

00:49:50.596 --> 00:49:53.306
fix the one, the one drone, right?

00:49:53.316 --> 00:49:57.776
Rather than having many of these
models, um, when, when one laser

00:49:57.776 --> 00:50:00.666
breaks on the array, you have to take
the whole array out and get it fixed.

00:50:00.666 --> 00:50:02.046
It's a very expensive process.

00:50:02.684 --> 00:50:05.904
uh, uh, you know, one of, one of my
ideas here for this would be, uh,

00:50:05.934 --> 00:50:08.364
something called Farm OS, right?

00:50:08.574 --> 00:50:12.951
So You have, you have a scout which,
which scout drones which map stress,

00:50:13.101 --> 00:50:16.481
it, it map insects, weeds, disease,
and then you have the Farm OS,

00:50:16.481 --> 00:50:17.901
which writes a light prescription.

00:50:18.091 --> 00:50:20.681
You know, if you're using the, the,
the, the wavelengths against the

00:50:20.681 --> 00:50:23.921
plant ledger, it, it figures out
what spectrum, what dose, what the

00:50:23.931 --> 00:50:27.911
time, what the tool, and what square
foot we're doing this on, right?

00:50:27.911 --> 00:50:29.801
What's been planted here
for the last three years?

00:50:29.851 --> 00:50:31.481
What are we planning to plant here?

00:50:31.761 --> 00:50:35.241
Um, and, and then sometimes 'cause
sometimes the, the, the remedy might

00:50:35.241 --> 00:50:37.431
be, oh, well, we need to plant.

00:50:37.471 --> 00:50:39.151
We need to add to this field.

00:50:39.151 --> 00:50:42.921
We need to plant new things,
um, which is really interesting.

00:50:43.031 --> 00:50:46.531
It's, it's really interesting that, that,
that can be a solution because different

00:50:46.531 --> 00:50:51.441
plants kind of create immune resistance
against monocultural blights, right?

00:50:52.201 --> 00:50:54.741
And then, and then you send out the
drones, which, which just do it.

00:50:55.001 --> 00:50:58.261
And, and, and again, I'm, I'm kind
of describing this without describe--

00:50:58.341 --> 00:50:59.691
without talking about the human role.

00:51:00.021 --> 00:51:03.591
This is all going to work much better
when you have a human in the loop.

00:51:03.611 --> 00:51:07.001
I assume at some point we'll get,
you know, kinda automated hydroponic

00:51:07.001 --> 00:51:09.551
farms where everything is in an
enclosed, you know, shipping container

00:51:09.551 --> 00:51:12.961
style thing and, and the-- and, and
AI can probably get to the point

00:51:12.961 --> 00:51:14.391
where it can run that pretty well.

00:51:14.841 --> 00:51:18.991
Um, but for actually dealing with
the real world, I don't think you're

00:51:18.991 --> 00:51:22.291
gonna be able to replace, certainly
not anytime soon, and again,

00:51:22.291 --> 00:51:24.251
certainly not cost-effectively, right?

00:51:24.261 --> 00:51:28.041
You might be able to have, you know,
Peter Thiel might be able to spend, you

00:51:28.041 --> 00:51:32.481
know, a billion dollars or something
on an AI that watches his field

00:51:32.751 --> 00:51:34.951
and makes all the decisions, right?

00:51:35.161 --> 00:51:40.921
But it's not gonna be cost-effective for
a really long time to have a-- just to

00:51:40.921 --> 00:51:43.801
have all of the decisions made by the AI.

00:51:44.071 --> 00:51:48.791
There's gonna be a need for humans in the
loop to, to make decisions and say, "Okay,

00:51:48.801 --> 00:51:53.716
do this, do that." the thing that I…
Hopefully, it's coming through that this

00:51:53.716 --> 00:51:56.796
is still the same three skill sets, right?

00:51:56.816 --> 00:51:59.206
This is code, this is
content, this is command.

00:51:59.406 --> 00:52:01.696
So you're setting up the
logical requirements.

00:52:01.736 --> 00:52:06.386
You're, you're figuring out how to… You
know, one of the, one of the things about,

00:52:06.406 --> 00:52:09.956
about content for this is when you're
dealing with, with user interface, you're

00:52:09.956 --> 00:52:11.466
dealing with making things comfortable.

00:52:11.576 --> 00:52:13.706
You're also trying to make it
comfortable for the animals, right?

00:52:13.706 --> 00:52:16.386
You're also trying to figure out
how not to stress the animals.

00:52:16.446 --> 00:52:18.576
'Cause if you stress the animals
out, then you're gonna get, you know,

00:52:18.586 --> 00:52:21.056
bad, bad outcomes from that, right?

00:52:21.416 --> 00:52:24.956
And, you know, right now our
bottleneck is toxic chemical

00:52:24.956 --> 00:52:28.206
inputs and the cheap disposable
labor needed, uh, to apply them.

00:52:28.466 --> 00:52:32.336
And, and I think as these
technologies come in, we're gonna

00:52:32.336 --> 00:52:33.936
need more and more holistic humans.

00:52:33.936 --> 00:52:37.816
We're gonna need more and more gentleman
farmers, more guys that are, are able

00:52:37.816 --> 00:52:41.486
to think through how do we… And, and I
think that's gonna be true even for, uh,

00:52:41.566 --> 00:52:46.802
the guys that stay on the chemical kinda
track Because they're gonna need, you

00:52:46.802 --> 00:52:51.072
know, the-- i-in ten years, you're not
gonna have-- you're not gonna use cheap

00:52:51.082 --> 00:52:55.052
foreign labor that's gonna get cancer,
right, from, from your chemicals, and

00:52:55.052 --> 00:52:57.802
you're just gonna hope that they go back
to their home country and don't sue you.

00:52:58.092 --> 00:53:01.832
Um, or the-- I, I think we're
gonna move to robots in all of

00:53:01.832 --> 00:53:03.252
these, all of these industries.

00:53:03.392 --> 00:53:07.282
And again, once you've got robots,
once you've got an AI system that's,

00:53:07.332 --> 00:53:12.182
that's looking at the field, well, then
there's no reason, there's no need to

00:53:12.182 --> 00:53:13.722
have it be a monoculture at that point.

00:53:13.732 --> 00:53:15.232
You've already made the leap.

00:53:15.232 --> 00:53:16.672
You've already made the investment.

00:53:16.852 --> 00:53:20.832
It's gonna be way easier to
convert to a polyculture or maybe

00:53:20.832 --> 00:53:22.842
even a permaculture, uh, framing.

00:53:22.842 --> 00:53:26.502
And so, you know, that, that, that's
my white-pilled take on this, and,

00:53:26.512 --> 00:53:27.542
you know, I mentioned Peter Thiel.

00:53:27.552 --> 00:53:32.452
He, he invested heavily in, uh, Halter,
which is a New Zealand company that puts

00:53:32.452 --> 00:53:34.602
solar-powered GPS collars on cattle.

00:53:34.962 --> 00:53:38.462
Um, and, and again, th-this is--
they're able to use virtual fences.

00:53:38.462 --> 00:53:39.692
They're able to track them.

00:53:40.032 --> 00:53:43.402
Um, so the, the, the, the founder said,
you know, the idea is here, you're

00:53:43.402 --> 00:53:46.712
gonna be able to manage your farm
from a phone without dogs, horses,

00:53:46.712 --> 00:53:48.552
motorbikes, or helicopters as the default.

00:53:48.952 --> 00:53:52.542
And, you know, I mean, I th-- I think
one of the things that I would, I

00:53:52.542 --> 00:53:56.102
would point out there is, is that
you're still gonna need drones.

00:53:56.882 --> 00:53:59.902
You're gonna need drones to… If you,
if you've got an animal that's sick

00:53:59.902 --> 00:54:01.972
or that's got something wrong with it,
you're gonna need a drone to go out there

00:54:01.972 --> 00:54:03.442
and, you know, deliver the medicine.

00:54:03.442 --> 00:54:05.112
You're gonna need the drone
to go out there and give, you

00:54:05.112 --> 00:54:06.052
know, kind of the supplements.

00:54:06.212 --> 00:54:09.372
You're also gonna need in, in most
all of Latin America, for sure, and

00:54:09.372 --> 00:54:13.272
m-- and much of the, the American
Midwest, uh, you're gonna need, you

00:54:13.272 --> 00:54:17.292
know, attack drones to defend, uh,
your herds from, from dogs, from

00:54:17.292 --> 00:54:21.932
predators, from, you know, cougar and,
and, um, and wolves in, in many places.

00:54:22.536 --> 00:54:25.026
And even, even in some, you know,
for especially for smaller livestock,

00:54:25.026 --> 00:54:26.426
from coyotes and things like that.

00:54:27.226 --> 00:54:33.926
And, and again, that is a skill
set that has profound, uh, profound

00:54:33.926 --> 00:54:38.740
military applications The final,
the final major, um, vignette that

00:54:38.740 --> 00:54:43.750
I'll share with you tonight or
today is, is the, uh, Morpho in Rio.

00:54:43.910 --> 00:54:46.600
Um, in January, this
is about two years ago.

00:54:46.640 --> 00:54:51.460
They, they, they demonstrated seeding
drones at Mirante do, do Pedral in,

00:54:51.510 --> 00:54:55.390
um, Botafogo with a pilot aimed at
the Posse Forest in Campo Grande.

00:54:55.390 --> 00:55:01.400
So the, the CEO of Morpho is, uh,
Gregory Maitre, and he partnered with

00:55:01.410 --> 00:55:03.820
the city under Re-Refloresta Rio.

00:55:04.150 --> 00:55:07.090
So the AI planting-- the AI would
come up with planting plans.

00:55:07.090 --> 00:55:11.390
It would look at the topography and say,
"Here is where we could plant trees,"

00:55:11.520 --> 00:55:15.160
where there's some level of, like a
little natural protection, something

00:55:15.160 --> 00:55:18.420
where the water's gonna gather so
the, the sa- the sapling can come out.

00:55:18.893 --> 00:55:20.213
and they're, they're able to do that.

00:55:20.213 --> 00:55:24.203
So they were able to seed at least
twenty native species across succession

00:55:24.213 --> 00:55:28.703
statio- stages, and they called it a
smart plantation, which decides which

00:55:28.703 --> 00:55:30.603
seeds go where in what mix and quantity.

00:55:30.953 --> 00:55:34.373
And the drone was able to fire about a
hundred and eighty capsules a minute.

00:55:34.793 --> 00:55:38.813
Um, so a two-person team and a drone
claimed roughly a hundred times faster

00:55:38.813 --> 00:55:40.433
than traditional planting, right?

00:55:40.933 --> 00:55:43.663
So Maitre said, "The work begins much
earlier with the reconnaissance of the

00:55:43.663 --> 00:55:47.243
area, with the identification of the
best species, and ends much later with

00:55:47.243 --> 00:55:51.453
the long-term monitoring. The drone
is the central part of all this work."

00:55:51.833 --> 00:55:53.933
And of course, they're using
artificial intelligence to monitor

00:55:53.933 --> 00:55:57.753
and track seedlings and s- and
seeds and not just to seed the drop

00:55:59.923 --> 00:56:00.133
Mm-hmm.

00:56:00.953 --> 00:56:03.453
So again, you know,
these are civilian jobs.

00:56:03.593 --> 00:56:05.353
They have massive implications, right?

00:56:05.353 --> 00:56:08.823
Builder, solo marketer, laser
weeder, UV strawberry fleet,

00:56:08.823 --> 00:56:10.783
cattle caller, seed drone forest.

00:56:11.353 --> 00:56:15.073
Um, hopefully, you see how these types
of force multipliers are im-- Like, it's

00:56:15.073 --> 00:56:19.573
impossible to do this in the civilian
context without acquiring skills that

00:56:19.573 --> 00:56:23.013
are relevant in, you know, what, uh,
what Roman Helmut guy famously called the

00:56:23.013 --> 00:56:25.433
natural meritocracy of extreme violence.

00:56:25.963 --> 00:56:29.733
Um, and, and that's gonna mean that
the most valuable profession in peace

00:56:29.773 --> 00:56:33.893
and the most dangerous operator in
war are going to look, uh, similar.

00:56:33.893 --> 00:56:37.273
They're gonna understand the
world more closely than those two

00:56:37.273 --> 00:56:39.033
groups have in four hundred years.

00:56:39.123 --> 00:56:42.763
Um, and I think that's going to,
um, that's going to have massive…

00:56:42.803 --> 00:56:45.523
And, and, and I can't stress this,
like, there, there, there is the

00:56:45.523 --> 00:56:50.163
possibility for massively positive
changes to our world from this.

00:56:50.473 --> 00:56:52.993
so, um, as I said in the
beginning, humans are not obsolete.

00:56:53.063 --> 00:56:56.413
AI is going to reshape every
industry: software, marketing,

00:56:56.423 --> 00:56:59.383
medicine, manufacturing, uh,
the field and the forest, right?

00:56:59.403 --> 00:57:00.383
The various types of ag.

00:57:00.833 --> 00:57:04.633
And the-- it's gonna do that
through these core skill sets: code,

00:57:04.633 --> 00:57:08.473
content, command, which basically
all have military, uh, applications.

00:57:08.743 --> 00:57:11.533
And again, the, the social and
economic implications of that are

00:57:11.533 --> 00:57:15.073
huge, maybe greater than what I'm
putting in front of you, um, today.

00:57:15.153 --> 00:57:19.743
So next time, uh, we'll, we'll talk
about, uh, where the safe professions

00:57:19.753 --> 00:57:23.073
are, which I actually think is the
wrong question, but we'll get into

00:57:23.073 --> 00:57:26.763
it, and then we'll talk about next
time the, the economics of a Prime.

00:57:26.763 --> 00:57:30.473
What are the economics of somebody
that has all three skill sets

00:57:30.943 --> 00:57:32.233
and, and what, what would you do?

00:57:32.233 --> 00:57:36.213
'Cause I think it's a very different,
uh, track than, you know, the, the,

00:57:36.223 --> 00:57:39.043
the kind of the technical university,
the research university that we have

00:57:39.073 --> 00:57:42.623
based our, our education systems
around for the last four hundred years.

00:57:42.623 --> 00:57:45.233
So that concludes my prepared remarks.

00:57:45.293 --> 00:57:48.863
Um, so thank you so much for everybody
that, uh, that's joined us for the free

00:57:48.863 --> 00:57:51.393
portion and, uh, let's move into the Q&A.