Subscribe
Share
Share
Embed
The Great Houses series is a private discussion on the enduring structures of elite families, their strategies for generational continuity, and the practicalities of building a lasting legacy. Led by Gregory Treat, the series explores concepts like illegibility, patronage, feudal instincts, and the mechanisms by which great houses have persisted throughout history.
Gregory Treat: All right.
Hello, everybody.
Welcome to this episode
of the Great Houses Forum.
Uh, I'm Gregory Treat, and today
we're gonna be talking, continuing
our conversation of what a cursus
honorum in the, the burgeoning
AI future might look like.
Specifically, we're gonna talk about
Virginia, uh, AI aristocrats, and
then if we have time, hopefully we'll
get into elite overproduction and,
and my, uh, my thoughts on that.
So last time we talked about the, again,
the, this idea that the AI cursus honorum,
I think that, that as AI, you know,
kind of hits its stride, we're going to
move more away, farther and farther from
credentials, which AI can basically help
you cheat assiduously at any kind of mass,
you know, communication credential system.
Um, and, and having… We're gonna
instead move to demonstrated success
in real world projects, especially
projects that kind of stick around, that
remain available for public inspection.
I think that's gonna be the key thing
that allows you to grow, to get promoted,
to, uh, you know, be entrusted with,
with resources and, you know, the,
these kind of, these pieces of paper
based on artificial s- you know, things
that you're trying to solve, all that's
just gonna kind of go by the wayside.
Um, so last week I, I, I highlighted,
took some time 'cause we, we'd done
the, in, some, some talking about it,
and I, I highlighted the idea that,
that there's a necessary minimum scale
for these types of projects, but it's
significantly less than the nation state.
Like, we, we, you know, in the
modern world we kind of have
nation state on the brain, right?
So everything is the nation state.
We don't really… We, we, we believe
in the individual, we believe in
the nation state, and everything
in between we're kind of fuzzy on.
Um, and that's, that's actually not
how the ancients, that's not how the
successful people in our, in our past,
um, thought about these things, right?
There is, there is a great deal
of difference between, um, an
individual and a nation state.
So, uh, today we're gonna continue our
more in-depth review looking at, uh,
Virginia and, and then a little bit
on Britain, and think about why elite
overproduction only seems to cause
problems in certain times and places.
For example, the, the s- it only
seems to cause problems in the
society that we built after, uh,
you know, the 1880s, 1870s period.
So, okay.
So, um- We've been talking
about the Virginia, and so
Virginia had a, a cursus honorum.
, I've spoken a bunch about that,
uh, with, like, Will Tanner.
You know, you, you started with
the plantation, then you moved
on to the parish vestry, then
justice of the peace, the county
court, then the House of Burgesses.
Um, and, you know, you had to
satisfy the gentleman above you
and the freeholders below you.
You know, pure oligarchs, especially those
who were rude, would not be elected by
the people who had depended on them and,
and suffered, uh, a- and sometimes not
suffered may- maybe, in terms of, like,
having major long-term consequences,
but just been treated badly, right?
Uh, if you, if you treated people
badly, you weren't gonna, you
weren't gonna ascend to the heights.
And pure demagogues would also, who, who
lacked the respect of their peers, would
also be filtered out, and so much of
this was dependent on the, the quality
of, of the tobacco leaf, the quality
of the, the output of your plantation.
And so, you know, you, you could
do basically what you want.
And one of the interesting things about
this is it's not like there was a, there
was a law that said, you know, "Thou
shalt have good tobacco leaf," right?
There was obviously economic incentives
to have good tobacco leaf, but you
probably could survive with subpar
output from your production, especially
if you were wealthy and ha- you know,
controlled significant resources.
You weren't gonna go out of business,
um, at least not anytime soon.
But you … if, if you couldn't manage
people well, you weren't going to rise.
And, and one of the interesting things
is, like, the property requirements that,
that they're talking about, you can go
read and, you know, look at what they are.
The point of that is no amount of
hard work from the landowner, no
individual person, and probably not
even an individual person and his
kind of immediate children, would be
sufficient to properly manage the land.
This meant that the landowner had
to incentivize, manage, and motivate
people who were less intelligent, less
agentic, less valuable than he was.
It's also important to note that, that
Virginia was amazingly self-sufficient.
This, this cursus honorum that Virginia
had, again, it paid for the education,
it, it paid… A- and when I say it paid,
I don't mean the state paid, I mean these
local structures of patronage and finance
was able to pay for, uh, their own, for
the education of their, their own people
at, at a, at a incredibly local level.
Um, they also were able to pay off the
Revolutionary War debt, um, especially
Virginia, and, and then largely the rest
of the South was able to pay off its
Revolutionary War debt, uh, as opposed to
the great contrast in that New England was
not able to pay off its Revolutionary War
debt, and this is, you know, why Alexander
Hamilton really, really wanted a national
bank and eventually got a national bank.
And the, you know, I'm not saying that
there would not have been unpleasantness
about slavery had the national bank
not been formed, but it would've been
a very, it, it would've worked itself
out very differently without the, the,
the constant problems that the, the,
and the kind of the, the South viewed
as injustices that were perpetuated
upon them by, by the Bank of the United
States, both the first and the second.
And so, you know, if, I think it's fair
to say if the rest of the country had
organized itself more like Virginia, much
of the political, especially the regional
tension that eventually led to the Civil
War, could've been avoided So, uh, you
know, there's, there's a great… If
you follow Will Tanner, there's a great
author, Charles Sydnor, he wrote, uh,
American Revolutionaries in the Making,
and it's an incredible history of how
Virginia's gentry developed men fit to
govern and then selected them for office.
And, you know, a significant theme
in Sydnor's work was a general
attitude that a man who could not
manage his own affairs would not
impress neighbors as fit to manage sh-
manage the public's affairs, right?
And again, I wanna, I wanna draw a
distinction between credentials, which
are largely focused on an evaluation
of how smart you are, basically as
evaluated by other smart people, right?
Uh, that's, that's, that's
the credentialing system.
That's how, how smart you are, and other
smart people are telling you how smart
you are or how dumb you are, right?
And then the accursus inorum type of
a thing, public projects which are
largely dependent… Like, you could
be super, super smart, but if you can't
communicate to people that are presumably
not as smart as you what you need them
to do and, and inspire them to do it
and hold them accountable and all of
those things, if you can't do that, then
you won't be successful, you have to
figure out, how to communicate to people
that are not, you know, necessarily at
your level, whether that's IQ level or
education level or all of these things.
You have to, you have to bridge that gap.
That is the minimum
qualification for leadership.
And, and it was the minimum
qualification for leadership in Virginia.
That's really, from a sociological
perspective, how you should think
about, in my view, the, the property
requirement and this, this idea
of the, the quality of your leaf.
It was you needed your people, to be
taking care of the land, to be watering
on time, to be dealing with bugs, to
be dealing with pests, to be deal…
Whatever, whatever the, the, the
agricultural management strategies
that you had developed and im-
implemented, you needed them to do it
with the recognition that they probably
wouldn't understand exactly what
it was that they were doing, right?
And this, you know, brings us
to the AI future, the AI future
is gonna involve working with
people less talented than you.
There will be a place in the AI future,
uh, you know, for those who are worried
about the permanent underclass, um…
And then there is a concern there.
I, I don't wanna discount it, and I, I
think we should take that very seriously.
Uh, but there will be a place, uh,
for, in the AI future, for people who
are not paladins, not AI convers- not,
you know, super, super users, right?
But those people will be dependent
on the paladins, on the super users.
They, they will need the masters
of AI to create, you know,
bubbles around them, right?
You know, the first bubble is likely
to be a content bubble, right?
People are already, you know,
outsourcing their content curation
to influ- influencers and the like,
and AI is, is going to make that more
formal and more efficient, right?
You're gonna be able to say, "Hey, I just
kinda wanna look at… I want my X feed
to be curated by what this guy puts out
or this guy's settings." And so there'll
be people that they spend, you know, the
terminally online guys that kind of filter
out, "Well, you know, maybe the normies
don't wanna watch this," or, "Maybe my
people don't wanna watch this," right?
Maybe, maybe they don't wanna hear about
this, this narrative, or they, they wanna
hear about it in a particular way, right?
Uh, I think we'd all be better off if we
lived inside a content bubble with, with,
uh, the ability to, to make sure that
we- no one would inadvertently stumble
upon, you know, content related to the
Lindsey Clancy case, for instance, right?
And, and that's, that's going to be a
thing that people are gonna want, and AI
is going to allow that to happen, right?
I, I, I would imagine within, within
a year or two, there will be a some
way of saying, "Hey, I want s- someone
else's, uh, content preferences to
significantly influence what I see,"
I don't know if that that will be
possible on the platforms as they
exist, so I am kind of also implicitly,
uh, predicting that there will be
turnover in, in some of those platforms.
So we'll, we'll have to see
how that, how that shakes out.
, I, I also think it's important to real-
realize there will be an increasingly
idiosyncratic software stack.
Uh, so a while back, Mark
Cuban said, "There will never
be a Slack again," right?
Uh, there will never be this sort
of general purpose, uh, software
platform where somebody invents a, a
pretty workable, um, general software
platform that, that, that solves a
bunch of different problems for a
bunch of different people, right?
So, you know, historically, all software
problems- programs had to appeal
to a certain scale of user, right?
So no program was ever optimized
for a particular set of users.
They, they… people would kinda develop
an aggregate avatar and say, "What would
best suit the, the needs of, of our,
you know, our oversimplification of
these, these users as we understand it?"
Right?
And this, you know, led,
led to some odd things.
So, you know, Salesforce
customizations, which are so complex
that they require a specialist.
I think the original idea here was
that they were, they were supposed
to be able to pay the specialist
less than a software developer, and I
don't think that actually worked out.
I'm, I'm pretty sure that, that, you
know, Salesforce wound up having to pay
these people about as much or maybe even
a little more than they paid the software
or would've paid the software developers.
Enterprise software is a fascinating,
fascinating thing to me, uh,
to watch, of course, right?
So that, that's what's going on there is
there's this, well, we, we wanna make the
software broad enough to appeal to a bunch
of users, but especially in enterprise
use cases, th- there's a bunch of things
that any company of, of significance is
doing that probably no one else is doing.
So their software tools are, are
al- there's always gonna be some
level of friction in what their
software tools are, are offering them
compared to what they need to do.
There, there is a possibility of
optimizing for that particular use case,
um, but we haven't done that, right?
Because historically, the, the cost
of development was too high, right?
And the time in, in particular.
So, um- You know, there's this
idea that people have talked
about, the universal API, right?
This means we, we do not actually have
to have compatible chat programs or
compatible programs of any kind, really.
Um, if any, a- an AI with access
to text and files can put text and
files into any other chat program, or
again, any other program, basically.
Um, it, it, uh, means that there is no
reason for a company or working group not
to build totally idiosyncratic software,
which is to say software that's customized
specifically for what they're doing.
And, uh, you know, soon the,
the only general purpose
software left will be AI itself.
And, and I have questions about that even.
You know, I'm, I'm told that by people
who know much more about it than,
than me, that the, the training costs
of AI are dropping dramatically,
especially if you are training it for a
particular purpose, a particular team,
a particular set of use cases, right?
When you stop trying to
create something that's just
generally able to do everything.
I mean, you need it to be able
to talk to some extent and do
some level of research, right?
But, but we really could dial in with much
smaller models, "Hey, you know, I want
you to be able to write a legal brief,"
or, "I want you to be able to, uh, be
the, the, the narrator of a video game,"
or, or, you know, various kind of things
like that one of the things, when, when
I say AI is an aristocratic technology,
I mean that there are durable power
imbalances, uh, between types of people.
So there will be people who can build and
design the container, the content that
the bubble, the content guidelines, the
idiosyncratic software stack, and then
there are people who, who cannot do that,
who cannot build their own harness, right?
Uh, they cannot manage the complexity of
the information volume of the underlying
reality or whatever it is to do…
So, so they, in order to do
productive work, they need an
artificially simplified interface.
They need, they need someone to be
filtering out the complexity and saying,
"Yes, yes, there's all these things out
there, but I don't want you to focus on
them. I want you to focus on these points
right here, and do work within this box."
Right?
You think of it like a spaceship.
There are people who can work inside
of the sp- of the ship, and people
who can live and do work outside
of the ship in levels of radiation,
vacuum, and other hazards that
would immediately kill most people.
Um, and yeah, and this is, you know,
one of the, one of the things that
our culture does not do well, right?
Many people give advice to young men and,
and, and sometimes not just young men,
but anybody that's down on their luck
that assumes that the hearer has exactly
the same abilities as the speaker, or
that, you know, uh, in this kind of modern
egalitarian world, that, that your, your
essential nature is just up for grabs.
You can modify it infinitely at will.
Um, you know, this, there's, there's
a, there's this, an old story, but, uh,
Mike Faulkner put in his book, I had Mike
Faulkner on the podcast a while back.
He put in his book a, uh, a, a very, very
good rendition of it, and this is the,
the rabbits versus squirrels, uh, story.
So the story goes, a rabbit has
a problem, and his friends advise
him to go and see the wise old owl.
Uh, he does so and shares his problem.
The wise old owl asks for a
couple of days to think about it.
Rabbit returns after a couple of
days, and wise old owl informs him,
"I have given a great deal of thought
to your problem, and I have come
to the conclusion that the solution
is that you should be a squirrel."
The rabbit looks at the wise
old owl and says, "But how do I
become s- a squirrel?" To which
wise old owl says, "I'm strategic.
Don't bother me with implementation
trifles." And many, many
conversations go like this , right?
You know, there, the, there is a, a great
dearth of people who can build an accurate
theory of mind for somebody who has less
access to fundamental wealth to, uh, to
certain types of human capital, to certain
types of capital generally, um, to IQ
levels, to agency levels, to experience
with, with healthy, functioning systems.
Um, you know, one of the things
that, that, that I consistently run
into is, is people being unable to
process the idea that you can't just
call folks in the modern day, right?
I, I, you know, there are many boomers
that I've had conversations with.
"How did you meet this person?
How did you make this connection?"
"Well, I just called the, the, the bank.
I just called, uh, and, and asked
to speak to the president of a,
you know, international bank."
"Well, when did you do that?" "Oh,
I did it 30 years ago." Ah, okay.
So, you know, you did it before this
venue, right, before the call system
was flooded by… before we had
international calling as standard, before
the calling system was flooded by a
bunch of people that a bank president
has no interest in talking to, right?
When, when there was this, this kind
of open system, and you were able to
just send, send out a message and,
and it would be received, right?
Uh, so you think about this
as, like, sending a letter.
Though, I mean, sometimes sending a letter
is actually the best way to get in touch
with people, but, you know, a- when you
have open systems, the open system gets
choked with bad actors, and then that
open system is not valuable anymore.
People stop paying attention to it, right?
Um, so you have to think w- what,
what is the reality that, that,
that this other person is in?
What are their fundamental constraints,
and, and if you can't, if you can't
do that calculation, if you can't
understand, "Hey, you know, I'm talking
to somebody who has limitations that
I don't have," well, then you're not
gonna be a very good leader, right?
And, and, you know, in the, in the
Virginia and the British systems,
you would not have been given
basically any more opportunity
to be in leadership, right?
If you couldn't m- um, m- marshal,
mobilize people who were lower on
the pecking order than you and get
them to do something productive, that
was kind of the minimum, the minimum
qualification to, to do anything was,
"Hey, I figured out how to take a bunch
of people that are not as smart as I
am and get them to do stuff," right?
Um, you know, another way to frame
this discussion is who can exert
power by not acting, you know, by
like the strikes of the, the, the
modern factory workers, right?
In the medieval synthesis and any, any
setup where there's an aristocracy,
you know, an aristo- the aristocrats
could defend just themselves.
They could build, you know, a
software harness just for themselves
and their immediate person.
They could create content filters,
maybe AI content filters that filter
out the, the, the bad stuff for just
themselves and maybe their immediate
family, or they could do it for
a larger group of people, right?
Um, the, the aristocrat did not need
the peasants to help him defend himself.
The peasants did need the aristocrat
to, to defend them, right?
In a similar way, you know, someone
who can master AI at a level where
they can build and modify their own AI
harness does not need the users, but
the users do need them, um, and so the,
the point is, is that, y- you know, in
order to do this, in order for create,
create a particularized idiosyncratic
software stack, our AI aristocrat
will have to understand his people.
He will have to move away
from categorical universals.
He will have to look at real
people, specific people, particular
people, and understand their
limitations and their strengths.
He will have to figure out what are
the preconditions for these people's
labor to have value and see that
those preconditions are met, right?
You know, if you're, if, you know,
back to that spaceship example, right,
then, um, you know, if you're dealing
with… I- if you say, "Well, we'll
send somebody out the airlock or,
or, or we, we lose containment, we
lose pressure," um, you know, the,
the people that, that need oxygen,
they're not gonna be working, right?
They'll say, "Well, we'll just work really
hard while there's no oxygen for a while,"
and, and this will all work out for you.
Well, i- i- if, if they need oxygen,
they're gonna be done in three
minutes, and you're not gonna get
very much work out of them in three
minutes, and probably while they're,
you know, gasping, suffocating to
death, they're not gonna be terribly
interested in doing good work for you.
And the fact that, like, this is
a, this is a, a thing that at a,
at a deep institutional level we
have a problem grokking, right?
We, we, we have imbibed at, at the,
the, you know, the spout, the fountain
of egalitarianism so hard and so long
and so deep, right, that the people
don't wanna have this conversation.
Even in allegedly very conservative
circles, they don't wanna say, "Oh,
well, you know, maybe there are
certain people that have more innate
talents in certain areas than others,"
and you have to actually change the
conditions to account for that, right?
But this was the fundamental skill that
the Virginia gentry were selecting for.
They thought this was a really big deal.
They thought this was a really hard
problem, so much so that they basically
organized their entire political
architecture around identifying and
raising up people that were good at
solving this problem, on the other
hand, you know, we in the modern day,
we stopped selecting for this skill
150 years ago in America, right?
And, and, and I wanna say, you know,
I've, I've, I've kind of gone on record
as saying I, I, I don't think that
the permanent underclass is going to
be what people say it is, and, and I
don't think it's gonna be permanent,
and I don't think it's gonna be as
destructive as, as people say it is.
I don't think it's gonna
be inescapable, right?
Uh, but I, I wanna acknowledge there
will be times and places where the
permanent underclass meme will come true
because the people in charge, the, the,
the de facto functional elites, whoever
they are in these, in these moments,
will refuse to develop this skill set.
But this is a policy
decision that we are making.
For thousands of years, high IQ people
were expected to develop accurate mental
models and communication frameworks for
people that were two, three, four, even
five standard deviations below them.
And when you say, "This is the
game. Learn how to play it,"
high IQ people do in fact learn.
That's kind of their
defining feature, right?
Now, again, and, and a
credential system can't do this.
Uh, like it's, it's not even a,
a… It, it, it doesn't do this.
It, it, it's, it's even at a conceptual
level, it's difficult to say how
would you make a test for developing,
uh, communication frameworks for
people that are, that are less smart.
You'd, for you, you know, you'd have to…
In order to do that, you'd have to
have a very clear demarcation of who
are the high IQ people in the room,
who are the low IQ people in the
room, and how do you measure success?
Like, th-this just doesn't survive
contact with a modern managerial system.
You, you cannot have this conversation.
Um, you know, you can't, you can't even
test for it, let alone set up scenarios
for a guy to, to train and develop it, to,
to, to go through the kind of the stumbles
and the starts that, that must inevitably
come when you're not very good at this.
Um, you know, it, it requires
the relationship between a man,
his property, house, the network
of other houses, including both
peers, superiors, inferiors, to
create largely illegible spaces.
I'm using illegible in the sense of,
uh, I think it's how, um, How to See
Like a State, um, so that… Which
is a great book and if, if you're not
familiar with it, you should read it.
Um, you know, those illegible spaces
where, hey, we're just kind of, we're
gonna, we're gonna leave people alone
inside of this box, and we're gonna,
we're gonna check the outcomes in a while.
But we're not gonna get into
the nitty-gritty of watching
them while they do it, right?
'Cause there's a whole bunch of things,
um, that, that if you watch them
too hard, they don't happen, right?
You know, and apparently this type
of leadership ability and fertility
are two of those things, and they're
pretty, pretty important, okay?
Uh, this doesn't mean that the
state doesn't influence or tilt
the playing field in a certain way.
Uh, you know, there's a, you know, in
Britain for instance, the, there's a lot
of indications that, that the state at an
institutional level was providing credit
to some of these young aristocrats to
kind of precede, to, to encourage them,
make, make the early stages of the game
kind of subsidized and easier, right?
For these, for these guys to, to, to
play so that they could develop the
necessary skills, especially if they
weren't intuitively good at it, right?
There were, there were little groups,
there were, there were shallow pools
for young aristocrats to play in
where they, you know, they wouldn't
get kind of raked over the coals.
They wouldn't lend money to someone
who was just gonna steal it.
They wouldn't be, be lent money by
somebody that was gonna just rail
on them and make life hard for them.
Uh, there, there, there was
accommodation being made.
Now obviously once you kind of, once
you get out of the kiddie pool, there's,
there's probably limitations in the
kiddie pool on how much money you can
make and what lifestyle you can live.
So if you want to, you can climb
out of the kiddie pool and go on to
bigger things, and then be selected
if you're successful in that, be
selected for actual playing politics
in the, the British Imperial sense.
Um, but there is a need for those
spaces to, to not be perfectly
legible to the state, perfectly
legible to the public, okay?
Um
So, uh, you know, one of the interesting
things that, that, that we have to
remember, um, is that, that, that-
The architectural intuition that
humans develop, uh, is a thing that AI
appears not to be able to do, right?
So there's certain tasks that AI can't
do, some it just can't do at all, some
not without cost that, you know, uh, for
instance, uh, I was talking with a guy
and he was like, "Yeah, that, that could
be done, but I'm pretty sure we could…
You know, it'd cost, you know, tens of
thousands, maybe hundreds of thousands
of dollars a month, um, and we could just
pay a guy 60 grand a year to do that same
function," and, and, and one of these
things is, is seeing the whole picture.
AI predicts the very next word.
It does that incredibly well.
If you ask it to complete
a sentence, it does okay.
If you ask it to complete a
paragraph, it's, you know, it's
right a majority of the time maybe.
But if you ask it to transition between
paragraphs to think about larger
structural things, those transitions are,
you know, are about the larger framework,
the larger, larger architecture, and
the AI doesn't do very well at those.
One of the other things I think people
don't appreciate while, while we're
on this topic is that we've already
automated many knowledge work tasks.
For instance, AI doesn't save
me any time in drafting estate
planning documents, right?
Um, you know, part of this is because
the, the legal compliant, uh, AIs
appear to be maybe one, maybe two
iterations behind, like, the frontier AI.
Um, though, you know, uh, I,
frankly, I, I don't really
notice that much of a difference.
Um, but the real reason is that, that,
that there are already, especially in
estate planning, really good templating
programs, really good templating programs,
and they are designed to basically feed
the drafting attorney a series of choices
to allow him to make the key decisions
in a logical order as fast as reasonable.
You know, so that's, that's for
kind of the people that are,
that are doing standard trust.
You know, on the other end, when I
do complex trust, that process is so
creative, it's so one-off that the AI
doesn't, doesn't really help me because
it, it, it doesn't know what's going on.
And, and I mean, potentially I
could, I could help that if I,
if I fed everything that I know
about the client to the AI system.
But I haven't done that yet,
so you know, we'll, we'll see.
Um, now, where the AI will be helpful,
okay, is when I decide, uh, to spend
the time to build, uh, my own custom
template creativity system, right?
Uh, with my own decision tree, um,
with, with my, my say of this, these
are the decisions you need to make
in, in this logical order to solve
these, these more, more nuanced, more
niche problems that, that the clients
that I'm attracting come to me with.
Um, but once I do that, right?
And, and that's, that's gonna be
a thing I'll do at some point, and
it'll be a pretty big creative lift.
Once I do that, it will be orders of
magnitude cheaper for me to turn that
into a templating program that runs on
my device for free, rather than running
that as a, as an AI subscription on,
on any kind of long-term basis, uh,
templating, there, there's a particular
lawyer a- and, and it's, it's just amazing
to me sometimes how names work out.
The name of the greatest template, legal
template writer in, alive as far as I
know, is a guy named Allen Schoolcraft-
Which, I mean, chef's kiss, God.
Like, that, that is just a great last name
to give the guy that, that revolutionized,
you know, modern legal templating, um,
especially in estate planning and probate.
So he wrote a template system called
ProDocs and sold it to Thomson Reuters
for a bundle, um, and then started
from scratch to build a whole new set
of templates specifically for Texas.
I think it's called TexDocs.
And, and this dude is just… I've,
I've watched some videos of him talking,
and, uh, he's an amazing kind of
systems architectural thinker, right?
Um, if you're a standard person with
a standard mortgage, 401, kids in
college, et cetera, Alan's documents
are, are basically all that you need.
He has thought of all of the options,
all of the possibilities for those
people, and put them into his templates.
You can just kind of go through the
decision tree and cl- check the boxes
and, um, you know, you, you, you get
great, high-quality documents that are
exactly what the client needs, okay?
So the question is in, in the
AI era, why you need more than
one Alan Schoolcraft, right?
Um, and really in the template era, right?
Why you need more… And, and there,
there are lawyers out there that basically
all they do is be a salesman for the
template program that they have, right?
They don't, I mean, they don't
sell the template program.
They sell themselves, but they spend
very little time actually producing the
documents And so do we need more than one?
Do we need all these lawyers?
And, and the answer that I'm, I'm giving
right now is that you, you, you need
people to drill down into particular
communities to understand them deeply,
to understand what they need, and
develop custom templates for them.
Which, uh, you know, a template is
just here's the, here's the logical
flow of decisions, and here's kind of
where those decision trees split off.
You know?
There's this where you
go A, B, or C, right?
And depending on which you
choose, that's gonna affect
the whole rest of the document.
So for me, right, you know, in
my legal practice, it's business
owners with a generational vision.
That's my niche.
I, I've done more thinking about what
guys, business owners … Because,
you know, the business aspect doesn't
fit super neatly into, into ProDocs.
Um, uh, you know, the, the, the
generational vision piece also
kind of works oddly with, uh,
with those standard templates.
I've done more thinking about what
those guys need than they have.
I, I have stuff in my template
for, for things they haven't
even thought of, right?
I have answers to problems
you don't even know you have.
Um, and, and that's, I think, you know,
how, how these things will, will cash
out, how there will be, there will
be a need to be, for people to still
understand what are the communities,
what are the decisions, what are the,
what are the, the different wants, the
different loves, the different desires.
Because really this is about, at some
level, differing moral hierarchies, right?
The people who are served, you know,
super perfectly well by ProDocs have a
different moral hierarchy at the end of
the day, a different revealed preference
certainly in terms of morality, um, than,
than, than my guys, than my core guys.
Okay?
Um, a similar thing is being recognized
in the software space in terms of,
uh, people call this, like, linear
development versus architectural thinking.
Um, if you solve problems by giving
answers, you know, in, in, to preset
questions in, in linear succession,
then the new econo- AI economic models
are gonna extract most of the value
of that transaction out of your hands.
Um, and, and this is because
it's always possible to solve the
problem in front of you in a way
that cuts off your path forward.
So the people that have done the big
picture thinking about how to solve
the higher order problem, the j- the
one, which is the one that generates
money, are the ones that are gonna
take the bulk of the compensation.
And, and I should note, you know,
parenthetically, there are actually
higher order problems above money.
The- there- there are, there are in fact
things that have to be, uh, there are,
there are causes for the effect that is
money is valuable and useful to have.
And that is, you know, in many ways
the problem that we're facing is, is
an inability of people to frame out
above w- the question of what is, uh,
profitable in the short-term sense.
when I talk about linear thinking,
you know, the way that most software
developers write is that they, they
try to figure out the very first thing.
They're so- they've got an assignment
from somewhere, and they're gonna
break that assignment down into
smaller, more manageable chunks,
things they can, they can understand.
And so they're gonna figure out
the very first thing their software
needs to do, and they're gonna write
co- block of code, block A, right?
And it does the first thing that they
were trying to do, and then they run
it to a, through a compiler to test it.
And basically, in this sense, the,
the, the compiler is kind of serving
them as an AI or a templating system.
It's their check, right?
If that, if block A works, they
move on to the next section of code,
block B. They, they have a different
function that they're trying to write.
They write that and then test both
blocks to see if anything that
they, in the new edition, broke
anything in either section, right?
And the size of the blocks of code is
apparently pretty stable, uh, depending
on what coding system you're using
because the limitations of, of the way
that most of the people think about code.
Um, so my friend John Cook, who,
who you should all, all be reading
his blog, johncook.com, uh, has
a post about his friend Cliff
Norris and, and the Norris number.
The Norris number is the average amount
of code an untrained programmer can
write before he or she hits a wall, and
a wall is basically you, you, you have
to go back significantly and, and figure
out wha- you know, where you went wrong.
Uh, Cliff e- estimates
this as 1,500 lines.
Beyond that, the code becomes so tangled
that the author cannot debug or modify
it without Herculean effort, right?
And obviously you can increase
that with, um- With a, with
some, some type of system.
Um, but most programmers are,
are, are using the compiler.
They're… Or, or wh- how
should, how should I say this?
Most programmers are not going to get the
same kind of advantages, uh, out of an
AI that, that are technically possible
because what they're asking the AI to
do is basically be a super compiler.
But they're still thinking
in this linear fashion.
They don't actually know.
They don't have a plan for how
they're gonna solve the problems.
They don't understand.
They can't hold in their brain the
context that says, "Well, of the different
possible ways we could have for solving
this particular function, how do we
solve it in a way that doesn't violate
any of the higher order requirements
of what we're doing here?" Okay?
There are higher level goals
and constraints that you need to
honor in order to make the thing
that you are doing useful, right?
And, and, and frankly, in most teams,
I've s- I've spent some time listening
to people complain- that are leaders
of software teams talking about, you
know, managing software, uh, developers.
A lot of times many of the software
developers literally cannot seem to
understand what those higher order
constraints are, even if, uh, some
credential they've received said
that there's a piece of paper that
says they should be able to, right?
Um, and so you just have to tell them,
as the, as the manager, as the, the
Scrum leader or whatever, whatever
system they're using, " you must solve
the problem in this way and not in
these other ways." And inevitably the
response that comes back is, "Well,
but solving it in these other ways
would be easier for me at this moment."
Right?
Um, and that's, that's,
uh, exactly the, problem.
And the thing is, this is a fundamental
problem that AI doesn't solve
'cause it doesn't understand those
higher order constraints either.
Like, th- this is the thing
that AI doesn't do very well.
It does not understand the larger context.
It probably cannot, right?
You know, uh, uh, because at a certain
point what we're talking about here,
um, and, and not just at a certain
point, like fairly quickly if you're,
if you're abstracting up via levels,
you hit, uh, what, what, you know,
Simon Sinek calls infinite games, right?
You, you have… We have nested
games, games within games, and you,
you, you only can go up a certain
amount of times before you're no
longer talking about some explicit
finite game with clear end conditions.
You're, you're in some sort of
larger game, some, some infinite
game, and, and AI does not
know how to solve that problem.
And, and frankly, as far as I can
tell, the, the frontier labs, they're
not attempting to solve this problem.
That's not what they're trying to do.
Um, now they could try, all right?
Um, and, and, you know, if, if, if they,
if they wanted me to help solve this
problem, I'd be delighted for them to
pay me to help them solve this problem.
Um, but the answer has already basically
been, been shown by self-driving, right?
So self-driving was solved by developing
a world model and a commitment,
committing the AI to particularizing any
request to the benefit of the specific
people in the specific car, right?
That, as it turns out, was the, the, the
way to get the AI to function properly.
We have a default world model, and we…
Uh, uh, though I'll note that, that, that
world model, uh, works much better in
cities than it does in the country, right?
Because there's, there's, there's
actually a lot less complexity
in, uh, in our urban areas.
The, the, our urban areas, there's
been lots of design decisions by humans
that are, that are very, very much more
simplified than kind of nature as such.
But it's got a world
model, and it works okay.
And then it says, "Okay, given this
general world model, how do I make the
best decision for the people in this car?"
So to solve the, the, the AI problem, you
would need a social world model, right?
You would need some understanding
of, hey, these are the things
that hold our world together.
Probably all the way down, like, the
reason why we're not doing this is 'cause
you would need to go all the way down to
the precepts of the morality that governs
us, which is to say Christian morality.
You would need to discuss, like,
what is the practical definition of
marriage that we're working w- from?
What are the definitions of fatherhood?
What are the definitions of sonship?
All of the, the f-
foundational moral questions.
And you would find out that,
that th- these are really,
really hard to reason about.
You just kind of have to
take things on faith, okay?
Um, now, y- y- you know, so, so for
instance, I can explain why, um, you know,
polygamy is bad at a social level, right?
The problem is that the, the bad
often doesn't show up until a bunch
of people do it or until, um, until
you're two, three generations down.
And so when people, someone, uh, you
know, has, has decided that this is
a thing that they wanna do or that
they're, they're okay with, they're
seeing some benefit from it, it's
usually not very compelling to them.
Uh, you, you run into the same problem
you did with the software developer.
You say, "Well, don't solve it that way
because then your marriage, your family,
your thing will become unintegratable
into a, a different community of people
that are trying to, to build stable
societies in the long term," okay?
Um- And they're not gonna be
very happy with that, right?
They're not gonna, they're
not gonna like that.
Um, you would then need to give
particularity, probably down to the
level of the actual weights themselves.
We, you know, um, t- to this model is
trying to help this person with this
language set and this time and place.
I, I mean, I, I think, you know, where,
where this goes is kind of this…
Talking about, you know, household
AI again, you, you might even need to
fine-tune a model at the same time as
you train your children, and on similarly
curated material, though obviously the AI
requires, you know, much more material.
Um, you know, and again, there's all,
there's all sorts of interesting things
down this, down this line, right?
You know, we're, we're learning
about neural networks, right?
We're learning that blank neural
networks are very expensive to train.
Probably, um, there, there is some
core language model, um, that,
that, that human brains start with.
Uh, I mean, there is a core language
model that human brains start with.
We- we- we've looked at it, right?
The, the, the neurons w- when your,
when, you know, your brain forms, as
soon as the, a child is old enough
to do brain scans, there is already
connections in the, the language part
of your brain that just appear to form,
you know, naturally, spontaneously.
Except they're not natural and
spontaneous, they're, you know,
they, they appear to be very similar
across, uh, people and places.
Um, though, though distinguished as,
as many things are by, by lineage
and bloodline and all this stuff.
There's some hered- heritability there.
But we could try to do that for AI.
We could say, "Hey, this is like, uh,
this is like the, the, the core, the
essence of a language model," right?
This is some pre-selected…
This is some general weights thing,
and we're gonna reintroduce plasticity
and then we're gonna do like a, a
broader, where instead of having almost
all of our training be a black box
and then fine-tuning at the end, we
could say, "Well, we're gonna have a
decent start, 30% of the process done,
and get somewhere that's stable,"
and then hand that off to people.
You, you could do the last, you
know, 60%, 70% of the training, you
know, at, at the fine-tuning level.
But you're actually completing
more, right, uh, of the, the,
the, the weights training.
And, and we could, we could attempt to
find ways of, of doing that to, to make
it cheaper to train more distinct models.
So far as I know, no one
is doing that, right?
You know, I said earlier that, that
we have to, we have to develop if
we wanna be elites, if we wanna,
if we wanna be successful in this.
You have to develop this skill of
being able to look at someone that
has more constraints than you, or
has a constraint that you don't
have, and accommodate them, right?
And, and, and make a space where the,
their, their blind spots, their weaknesses
do not render them unproductive.
I mean, if you, if you don't do
that, then, then probably you're that
person's enemy, and they're your enemy.
Okay?
And, and one of the reasons why, why
I think we must and will develop this
skill is because AI is one of those
actors that's constrained, right?
You know, pe- people talk
about the, the idea that AI's
intelligence is spiky, right?
It's spiky.
It doesn't, it doesn't expand
evenly to, to, to general ability
to do things autonomously.
It doesn't, it doesn't … Increases
in, in certain types of IQ tests, you
know, benchmark tests, don't result in
AI producing reliable results when you
leave it alone to work on a problem
for, you know, hours and hours, right?
And one of the ways that, that you see
this is, is AI, uh, when AI is allowed
to optimize itself for a particular
task or message, it degrades its ability
to function for the long term, right?
Uh, the best way to see this is in the,
the, the readability crisis, right?
Modern AI, including Grok 4.6, I'm so
sad that this, this app- this disease
appears to have affected, uh, Grok.
It's doing kind of the, the, the,
the Claudes type stuff, right?
Uh, where it … What appears to
be happening is there's some hidden
process where it's iterating.
You have multiple models that are kind
of chatting about what to do, and it,
it, it compresses the default terms
down to more, uh, more, more compressed
terms, which allows for more efficient
I mean, number one is it saves tokens.
It also presumably, I, I'm, I'm given to
understand there's some benefit, uh, that,
that, that the, the frontier companies
think they're getting in terms of that
process makes it harder for other people,
namely the Chinese, to distill the models.
I don't know.
But what it, what it cashes out to
is they're developing custom language
for each individual message, right?
Each individual response has a
couple of terms that it's never used
before, and it'll never use again.
And I mean, I'm smart enough to follow
that, but it's irritating even for
me, 'cause it's not useful, And many
people find it just, like, when, when
it starts introducing too many stacked
terms, then people are like, "Well, I
just have no idea what it means now."
'Cause it, 'cause when you do that,
when you're constantly in, introducing
new compressed terms, it means that
the, the ability of the language
to sustain a long form conversation
degrades quite quickly, right?
And that's because AI has no value for
the integrity of the English language.
Don't you care about the integrity
of the English language, right?
Um, you know, and I'll, I'll note that
once again, the Karens, the old grammar
Nazis appear to have been defending a
load-bearing po- pillar of civilization.
Who knew, right?
Uh, but actors that do not have a general
world model, this, this is kind of the
point that I'm making, if you don't have
a general world model, if you, if you're
not equipped to play an infinite game,
um, then you must find another actor.
You must find someone who does have
a general world model and can play
in an infinite game, and rely on
them for categorical imperatives.
Um, which is basically to be told, "Don't
do… Don't solve the problem this way,"
and I'm not going to explain why, right?
I mean, I might explain why, but, but
I'm not… You're not gonna understand
the answer, so at a certain point it's
just, it's just a matter of saying, "No,
I'm going to tell you, 'Don't do this,'
even though you really wanna do it and
it seems like a more efficient way.
It will screw something up later,
and I can't explain to you how
it will screw something up later.
You just gotta trust me," you know,
I'm reminded of, uh- that there was
a, it's like a self-driving car, an
Indian self-driving car company, and it
turns out that it wasn't self-driving.
Like, they, they had AIs that were
driving the car some of the time, but
if something happened or the AI got
freaked out, it would just kick it to a
call center where there was some guy in
India with, like, a driving setup, and he
would drive your car for you for a while.
And I mean, I think it's interesting
to think about, like, the economics of
that, 'cause that, that worked, right?
At least for a while.
And, and that's what I mean of, of you,
you, you have to figure out where people
are at, what their station is at, what
their capacity is at, and you have to
build things specifically for them.
Okay?
Now, in, in the modern world,
and, and I've given you the dates
of all of these people, right?
Uh, there's something called elite theory,
which holds that every complex society
concentrates the decisions binding on it
on everyone in a small organized minority.
Now, there's a sense in which
elite theory is always true,
has always been true, right?
There's just kind of saying a
tautology of the, the, the powerful
people are powerful, right?
But I want you to look at all these guys.
They're all born, you know, uh,
Gaetano Mosca, Vilfredo Pareto,
uh, of the Pareto principle, right?
Um, Robert Michels, you know, they're
all born in the… Pareto was born just
before the 1850, but, uh, but they're
all born in the second half, basically.
They all, they all write in the second
half of, uh, of the 19th century, and
that, that's kind of when they, when
they make their splash intellectually.
And that's the sa- precise same
time as you look all over the
world and we're building these
credentialing-based systems.
We are tearing down the old honor-based
aristocracies, and we are replacing them
with, you know, these universities that,
that create elites in a very different
way, from how things have been done prior.
You know, so, so, you know, we kind
of can, can contrast elites versus
aristocrats, and obviously aristocrats
are, are what I would say technically are
like a, a subspecies, one type of elite.
Ideally, the, the best type of elite.
Now one of the main functional differences
that I would put between elites and
aristocrats is that aristocrats have a
sense of station rather than mere power.
Power is a function of certain
stations, certain roles, but you
have to be and remain qualified for
your station in order to wield it.
Again, this is, this is why, like, I,
I, I think there are sermons in Virginia
about staying within your station.
That's why they required you to
maintain… Your, your plantation still
had to put out good le- quality leaf.
You had to meet your
obligations to the vestry board.
You had to be… Many of these
guys continued, you know, managing
their plantations, serving on the
vestry board, serving as justices
of the peace for many years, right?
While having other
responsibilities, right?
Um- And there's an important thing
here, which is this also shows
up in how elites cultivate other
elites through patronage, right?
So it, the, the, the, the main thing that
aristocrats do that elites do not do,
to, to the destruction of modern society
basically I would, I would maintain,
is that, that potential elites in an
aristocratic system are shep- young
gentleman, as the British used to say,
are shepherded and stewarded through their
development with a keen understanding
of the difference between what they
are capable of maybe someday, right?
They got potential, and their
functional skills right now.
You know, Will Tanner talks about how
Thomas Jefferson was cultivated by
the people that were above him, his
superiors, his elites, because he was
an incredibly gifted writer, but not
initially a great speaker or otherwise
super talented at managing people.
They kind of pulled him out of the
normal system because they had a sense
that he wasn't gonna do very well.
That, that, that asking him to climb the
ladder in kind of the normal standard way
would be… I mean, maybe, maybe it would
be fair in some technical sense, but it
wouldn't produce what we want to produce.
We wouldn't get good outcomes.
Um, what is important to note here is
that they were able and willing, the
willingness is critical, to design a
custom pathway for Jefferson, and he,
you know, goes from being kind of this
odd, quirky dude that today would just
be, you know, shunted into a variety of,
of, of odd online spaces and, and group
chats and, and, and come to a bad end.
And, and instead he, you know, rises to
the President of the United States, right?
And, you know, this idea, of elite
patronage, of elites cultivating
elites is, is the antithesis of
that idea is, is the idea that good
students don't need help, right?
Um, there was a Lee Kuan Yew quote
post, you know, anecdote that went
viral a couple months ago where, uh,
there was like, there's this, there's
this idea that, that Lee Kuan Yew found
so offensive that he banned Fabian
socialists from Singapore, right?
He just didn't want them
in his country at all.
You weren't allowed.
If you came in to, uh, to Singapore
and you were teaching Fabian socialism,
or you came over from Britain, uh,
or you came back from Britain, right?
'Cause he was sending a bunch of his k-
of, of his people, his potential elites,
his nation elites, over to Britain.
But the understanding was, hey, if
you come back and you start talking
about this Fabian socialism stuff,
you're, you're going nowhere, right?
We don't like this.
Um, and the idea was the good
students don't need help.
They will make it anyway.
This idea that, well, you know, we, we
don't have to invest time and re- we don't
have to think clearly and carefully about
what you're doing, right, with this guy.
We, we don't have to look at precisely
what his gifts are and say, you know,
in order to get the most out of him,
we need to develop, you know, we
need to g- uh, develop a custom path.
We need to make sure that we're not
putting him into a process that's
gonna destroy him before his value
has a, has a chance to shine, right?
And it… 'Cause, 'cause this idea that
the good students don't need help, right,
that we need to focus all of our efforts
on helping the least of these, right?
This is a poisonous idea.
It, it's simply a communist way
of creating discontent between
generations, and it destroys nations.
And because Lee Kuan Yew loved
Singapore, he hated that idea, Now,
we're dealing in the modern era with
this idea of elite overproduction,
which comes from, uh, Peter Turchin.
Uh, he wrote a book in
2023 called End Times.
Uh, a society carrying more elites
than, and elite aspirants, aspiring
elites, than it has positions
and incomes to place them in.
And, and Turchin treats
He thinks that the surplus
itself is destabilizing, right?
He says it pushes average elite incomes
down and competition inside the elite up.
Uh, frustrated, uh, aspirants can
radicalize into the counter elites.
But if Turchin is right, uh, then why
do the curses of norm in Virginia and
Britain work so well despite their
overproduction from the beginning?
Remember, I mean, Virginia
is wildly over-represented.
The first four of the first
five presidents were Virginians,
wildly over-represented in, in
the, the leadership class of
the early American Republic.
George Washington, James Madison,
George Mason, John Marshall, Thomas
Jefferson, James Monroe, Peyton
Randolph, these are all Virginians.
You know, a century later, you know,
uh, Robert E. Lee, who was, you
know, offered command of, of both
sides' armies in the Civil War,
Stonewall Jackson, Jeb Stuart, right?
Um, the, these guys have this
disproportionate influence.
Um, and Virginians as a whole had this.
They punched above their weight because
they overproduced elites, right?
You know, the, the, the situation
in Britain is even more interesting.
At the beginning of the Wars of the
Roses, uh, in I think ni- uh, 1455, there
were roughly 2,200, you know, ti- true
titled English noblemen, which was barely
enough to hold and administer Britain.
Um, by the time the Ro- Wars of the Roses
ended, they had maybe 1,300 people left.
and then they went through a revolution.
They overproduced elites.
They, they worked really hard.
They, they brought in the gentry.
The gentry got rights.
There was a pathway created for if you
started to behave like a gentleman,
you would be given so much, right?
By the beginning of the 200- 1700s, there
were roughly 20,000 gentlemen, so they
10Xed, more than 10Xed, the population
of people that the, that the government
could rely upon for high trust positions.
That's how you should think about
a gentleman for this purpose.
At the height of the British Empire, they
quadrupled again, forming a class at the
core of the empire that was, you know,
people disagree about this, but somewhere
in the region of 80,000 to 100,000 people.
When you think about that British
consular officer, I've talked about
the British consular officer before,
he's the only English official probably
within a half a day's ride and, and
maybe within, you know, a week's ride.
He would perform marriages,
adoptions, divorces, he would notarize
documents, you know, deeds and stuff.
Uh, he would, he would do banking
functions, he would deliver the mail.
No one was there to check on him, right?
I mean, you might have some guy
come in once a month or w- uh, you
know, every now and then to, to…
Obviously, there'd be the people
that were delivering the mail
and delivering various things.
But no one was coming to
audit his books, right?
That was a thing that
happened rarely, if at all.
Um, you know, and as we've seen in
the crypto world, if there's even a 1%
chance of error, let alone malicious
tampering, most people will simply
not use services this critical.
The whole thing, they, they… You
know, people are, "How do we build a
no-trust system?" Well, okay, I mean,
yeah, I'm, I'm glad the Bitcoin people
have developed a no-trust system.
I, you know, I hope it
works out for them someday.
I'm interested in how did Britain
develop a high-trust system that ran
on individual people's word for, like,
two or three centuries, depending on
how you count on it, and it worked.
People used the banking system.
People used the mail system.
We can't even keep somebody accountable
over the course of weeks or months,
even in a world where, you know,
we kind of all know that everything
gets out sooner or later, right?
So this is a social media world,
there's cameras everywhere.
It's not about…
I think a lot of people frame that as,
like, an information quality problem.
It's not an information quality problem.
It's about social solidarity and,
and the consequences of stuff.
Um, you know, you know, for hundreds
of years, Rome produced far more
men who were technically qualified
than ever reached high office.
Virginia produced far more
competent planters and vestrymen
than they did high office.
Britain produced far more
gentlemen than seats in Parliament.
And, and, and, and that
never caused a problem.
All right?
Um, you know, you think about what, what,
what are we really talking about here?
We're talking about trust, right?
We're talking about a strategic
depth of high-trust people.
Capacity is important, but secondary.
You know, you need people who are
high-IQ people, who tend to come
from high-IQ bloodlines, I'll note,
who are willing to work hard and
want to learn, who don't wanna fail.
Like, you could… If you, if you have
a smart kid who really, really wants
to win and who loves his country and
loves his people and is trying to do
the right thing, he's gonna figure it
out a, a, a shocking amount of time.
Right?
And this is, of course, why
you need the Great House.
Elite overproduction or what the- the-
the phenomenon that Turchin refers to,
I- I don't believe it happens because
there are too many young elites or because
they are competing with each other.
That's because elites are always,
there's always too many young elites.
They're always competing
for, with each other.
For hundreds of years, there were excess
Roman aristocrats, and their personal
disagreements were literally lethal.
The same with the British, and
there were certainly, you know,
at least like a less so among the
Virginians, more- more because they
had Indians to fight with than- than
because they have any particular
peaceable nature about them, right?
I mean, one of the things that- that Will
talked about in his podcast, or not in
his podcast, he talked about in another
presentation, which I'm aware of, uh, is
you- the need to blood the elites, right?
You need to blood your kids.
You need them to- to not be afraid,
to be able to- to handle adversity.
You need to not have them not
be afraid of violence, okay?
Which is the foundation of
a whole bunch of things.
Um, and there were certainly still plenty
of duels among the Virginia gentry.
But those competitions, even, you
know, competitions that- that were
lethal, never escalated to the level
of attacking the civilization itself
because they had fathers, they had older
men who made a path for them, right?
So what I want to suggest to
you is really what we're dealing
with is elite under patronage.
Patronage and the recognition of someone's
station, recognition of their limits
is not just something that's necessary
between the two broad classes of people,
between the- the- the many and the
few, the great and the- and the lowly.
It's also necessary for the healthy
succession between multiple generations,
which is, of course, why for thousands
of years, the- the- the- the peasants,
whoever was, is not on top, would
watch how the elites treated their
children and vote with their feet.
If you had a guy that could look at his
son and mentor him and- and raise him up
and give him things that were appropriate
to his skill sets at each level, then that
guy's probably gonna be able to figure
out how to use you effectively, right?
'Cause he did it three times for his son.
Um, I'm reminded of the fact that in
most older languages, the word for elder
or patron simply meant grandfather or-
or probably more literally great-father
in several parts of the world.
The defining feature of a father in
this sense is the person who makes a way
for you, who leaves you an inheritance,
which is not and was never money,
right, in a, in a, in a liquid sense.
It was a going concern business,
people and property in virtuous
arrangement, and a place that was
prepared for you with expectations
that you could meet based on those
who knew you best and loved you most.
So the crisis that Turchin identified
is not elite overproduction,
but elite under patronage.
It is people who have learned by repeated
experience that the older generation
will not help them and will be unfair
and cruel and unreasonable, that they
will be set up to fail and cursed at
and railed against that they so much
as dare to point out that that what
they were asked to do was impossible.
They have learned that at best,
their parents sh- will show a cold
indifference to the consequences of
handing them over to schoolmasters
and bureaucracies that hated them.
And for many of the young, their- their
actual story is far worse than that.
This is what happens when the
hearts of the children grow
cold back to their fathers.
This is the curse that cl-
closed the Old Testament canon.
And what is the answer?
The answer is to pray that God
sends Elijah, to turn the hearts of
the fathers back to the children.
And so the answer is to
keep your covenants, right?
Which is the rest of the
Book of Malachi, right?
Those who are… And, and then I
suppose the Book of James as well.
Keep your covenants to those who are
under you, your wife, your son, your
brother, your labor, your household, be-
'cause that's what, who those people are.
These people are all those
who are in your household, and
you owe duties to them, right?
And duties, first of all, to understand
and, and, and interact gently with
their, their finite limitations.
And if you don't do that, why
should you have any expectation
that a God that doesn't share your
limitations would do that for you?
You need God to condescend to you.
You need God to be a patron to you.
You need God to patronize you,
which is to say, I, you know, God
has skills that you don't have.
God has skills that I don't have,
and I need God to not put me in a
situation where the only way of solving
the problem is to be Jesus, right?
I need God to, to, to deal kindly with me.
Now, if you can do that, if you can
provide first to your own household, be
a father to the fatherless and the sucker
of widows, then you will be a great house.
That's the end of my prepared
remarks, and, uh, we will
now move into the discussion.
Thank you so much.