Great Houses

Gregory Treat: All right.

Hello everybody, and welcome to
episode 32 of The Great Houses Forum.

We are continuing our discussion of
the cursus honorum of the AI future.

So, uh, for those of you,
we took a break last week.

Well, it's not really a break
as, as we'll, as we'll soon see.

Um, last week we talked to Zach Fava,
um, about his, um, his company, which is

providing an, an education piece for, um,
for the household AI that I think will,

will come out of a lot of this stuff.

So today we're gonna return
to the, to the series.

Um, so again, our goal, my goal in this
series is to set out what I think is

coming and present that in a way that will
help you think about this for, especially

for purposes of training your kids to
participate in, in this at a high level.

And again, I think that the, the
technology is, is largely here, or

at least we can see where it's going.

What, what I think a lot of people, the
… What I think is not obvious to a lot of

people is how that technology will require
us to reorganize things socially, right?

So the first thing that I, uh, we walked
through in this series was the idea that,

that LLMs are gonna bridge domains, doing
to essentially all computer work what

CAD did to design work, which is just
kinda to create a, a common labor pool,

is, you know, it's, it's gonna mean that
people from formerly different industries

will now be able to compete against
one another, to work with each other.

That creates lots of innovation.

It creates lots of new opportunities.

It creates, especially at the high end,
uh, higher pay scales, and all of this

starts to look like a unified hierarchy.

Okay?

So, uh, that new domain will have
a limited number of specific skills

that will allow their user to be
productive and useful in a vast

array of different industries.

Again, so all of those industries are
now competing with each other for talent.

Um, the, and, and the, the, there's,
a lot of times there's silos

between skilled labor pools, and
those silos are gonna come down.

So- And the way that I've been
framing this is the skills of

code, content, and command.

Uh, you know, code is basically the,
the skill of logic or architecture,

uh, you know, m- mental architecture.

This, you know, people talk
about, uh, uh, graph design or,

or, or graph engineering, right?

This is thinking at a higher and
higher level about what is it that

you want, what are the outcomes,
what are the constraints, what

are the specs that you want?

Um, because when… The, the more
vague you are about that, the, the

more it increases your token cost.

Okay?

Then there's content, understanding,
you know, matters of taste, what will

appeal to people, how, how to use AI,
uh, effectively for those things, um,

being able to explain, to, to notice.

I mean, basically, it's just basically
noticing, oh, these are all the things

that I don't wanna show up in my work.

and then command is, uh, the ability
to, to make decisions, to, uh, to own

responsibility for liability purposes.

And again, the, one of the interesting
things about the, the command idea

is, again, it, it dramatically
lowers the token cost, right?

If you give… And I'll, I'll,
I'll c- may come back to this

point several times through this.

Uh, if you give a- an AI, you say, "Well,
well, make this decision," that is much,

much, much more expensive token-wise
than saying, "Give me 10 detailed plans

for what you could do, you know, for,
for, for different decisions we could

make when faced with this option,
and then I will pick between them."

The… If you just say, "Give me
the options," and c- even build

out quite detailed plans for how
you'd go about that, the AI can do

that, you know, quite effectively.

So last time we reviewed three points.

Um, you know, humans are not obsolete,
AI reshapes every industry, and the

skillsets, the, the point that I want
to make, it most, most notably in

the agricultural industry, they all
have military applications, right?

There, it's, it, it's, it's going to
be more and more difficult to, to have

these separate boxes where … And,
and, and I think, again, the, the

point that I'm making here is not,
not that we will all be part of the

military, but that, that there, that
will result in social changes, right?

There is part of our culture, part of
the way that, that we have expressed

ourselves and, and, and in many
ways the, the origin of a lot of our

cultural pathologies is the way that
we have walled off our scholars from

our warriors for the past 150 years.

And, and it looks to me like in,
you know, 10, 20, 30 years that may

not be technically possible anymore.

Um, the, the, the skill of being good at
using an AI and robots to do something

… Now, that doesn't mean that the, the
cultures, the subcultures that we have

created will disappear in 30 years.

But it means that similar to, to the
CAD space where you had, you know, 3D

printer people and CNC people, kind of
hard manufacturing production people,

and then all of these, uh, you know,
arts and design, prop manufacturers.

They were all thrown into the same
space, and their values kind of ha-

got to hash it out with each other.

I think that's gonna happen, but again,
on a, a vastly larger scale here.

Okay?

So today we are going to discuss safe
jobs and, uh, w- you know, it's … A

bunch of people are talking about safe
jobs and why that is the wrong question.

And then, uh, my, the second point that
we'll, we'll talk about is, is something

about the, the economics of a prime.

What, what one person who holds all three
AI skills that, you know, that code,

content, and command, uh, what, what
that person is worth and how that, that

pricing, that capacity redesigns the firm.

And again, as we go through this m-
my, my hope is, and we'll, we'll, we'll

turn … You know, we, we will spend
some episodes spec- specifically talking

about this at the end of the series.

But h- my hope is is that as we're
going through this I'm presenting the

information in a way that will help
you make educated decisions about how

to train your children, what to do,
what not to do, um, 'cause I think

that, that in the broad strokes these
are, these are the trends that, that,

um, that we're gonna be all facing.

And obviously if we can be on the, the
positive side of the trends than on the

negative side of the trends that's, that's
the benefit of our, of our children.

Okay?

So, uh, the internet thinks that there are
some, some AI safe jobs, AI-proof jobs.

Uh, so Resume Now 2026 says nurse,
anesthetist, uh, anestheticist, which

I, I, I just think that's totally wrong.

Um, emergency physician,
judge, pilot, surgeon.

We're gonna kinda go through those,
those lists and critique them,

and then I'm gonna tell you what
I think is really going on, right?

you know, one of the things that
I wanna, I wanna point out is that

AI and robots are very powerful.

Um, you know, if, if we're talking
about the skilled trades or clinical

medicine, or therapy, right?

People say, "Well, therapy is this
AI-proof, uh, presentation." I mean,

I just wanna observe, you know, AI
plus robots is very powerful, right?

Even on … Even in, like,
the, the super soft, uh, skills

like, like a therapy job or a, a
pastoral, you know, counseling job.

I mean, think about how
many people were drawn into

ChatGPT-4, I think it was, right?

And there's all these people that,
like, ha- ha- created this close

emotional connection with an AI.

Now, the AI was not good
enough to manage that well.

There were some problems.

That's why they had to, to, um, to kinda
decommission that, that model because

it was, it was causing them troubles.

Um, but these are very powerful models.

Now, now e- especially in the trades, I
think the key issue, uh, that we will face

for the foreseeable future is cost, right?

It's not gonna be cost-effective to
have an autonomous plumbing robot.

We'll talk about that in a second.

Um, but, but, you know, costs
espe- uh, for robots I think are

going to change dramatically.

I think that the cost of inference
is gonna get cheaper and cheaper.

I mean, we're, we're already seeing
higher and higher capacity models

being squeezed down so that they can
function on smaller and smaller bits

of hardware, especially when you can
define the job that you wanna do, right?

If you want a general purpose AI,
right, that's, that's a harder thing.

But if you can say, "Hey, we want
something that can think about these

types of tasks," man, um, I anticipate
that g- just getting cheaper and cheaper.

So-

Um, a-and again, the, the, the,
the question that most people are

asking on this topic is wrong.

They're asking, are humans obsolete?

And what I want instead for you to ask
is, does mastery of core AI skills,

right, code, content, command…

And, and, you know, one of
the other points is there

may be other skills, right?

I'm, I'm … I, I actually can think
of three or four other skills, but

for, for rhetorical purposes, I'm
sticking to three for, for this

presentation and for the, uh, the,
uh, the essays that I'm gonna release

on Substack that, that pair with it.

Um, but regardless if it's, if it's
five, three or five or 10, like again,

this, this represents a dramatic shift
in, in the, the number of skills.

When people, people can potentially master
all three skills or all five skills.

Certainly, if there's 10 core skills,
you will be able to intimately know

people who have all 10 skills and kinda
manage them as a local team, right?

Um, and the other point is that, um,
just because a, a, a profession is

AI-proof, like what does it mean?

What does AI-proof mean?

Does that mean that the people in that
prof- like, in that profession will

still be doing the same things that
they were doing five, 10 years ago?

Um, I, I don't think so.

I think that, that AI is going
to dramatically transform all of

these professions, um, and, and

So the fact that there will still be
people that are called plumbers and

people that are called therapists
and people that are called this and

that does not mean that AI won't
dr- uh, be as impactful on those

industries as the computer was, right?

So again, my contention is that a
practitioner that can afford the

upfront capital for some combination
of AI and robots and hire or train

the skills to use them effectively
will dominate their competition.

Okay?

So, uh, you know, the- the- the
first example I wanna p- put in

front of you is plumbers, right?

So we're gonna start with the trade.

So- so the question is, i- are- are
plumbers gonna be turning wrenches?

And- and for now, yes, probably.

Uh, it's- it's not cost-effective to
have a- a plumbing robot, especially not

an autonomous plumbing- plumbing robot.

Um, I- I think that's- that's changing.

It could change even- even as
shortly as- as within a year.

Um, but I think the thing to
think about, the question to ask

is what will happen when certain
plumbers start to integrate AI?

What will that look like?

Okay?

So we have one plumber who, you
know, he comes out, he walks the

property, he quotes a number.

Uh, everybody understands that that number
is kinda subject to all of these things.

Um, and then- and then he, um,
he relies on his reputation.

And in many- in many times, it's- it's
more about I'm here right now, and I

will help you, and if- if you don't
agree, then I'll just leave, and you'll

kinda go to the- the end of the list.

Um, a- a newer plumber will have a camera
drone or they have these, you know,

cameras that go on- on long lines, uh,
that they can put through, and they can

just see everything inside the pipe.

They can show their client
what the pipe looks like.

There's- there's YouTube.

I mean, this … The thing about
this is this is content, right?

There are YouTube channels devoted to
basically plumbers showing people what

was wrong with their, uh, with their
plumbing system and then fixing it, uh,

which is kind of a- kind of a crazy thing.

People watch this on YouTube.

Um, you know, as- as AI, kind of the
more personal assistant AI-type stuff,

uh, comes online, imagine that he has a
conversation with a client which the AI

is listening to, and when he … when-
when the client says yes, the- the AI just

generates a quote, generates a proposal.

Um, it's already probably
pulled a profile on, uh, on

this person, uh, on this client.

So I think, you know, the thing
is, is that- that all of these

things are running on code.

So, so there is someone … If
you … Once you see a- a plumber

like that, there is someone in his
life, maybe it's him, maybe it's not,

that is deeply familiar with code.

Um, there is- there is someone in
his life that is deeply familiar with

content, that has that content skill.

And then again, I think the, uh, the thing
that's gonna make it so that there's still

a plumber for the foreseeable future is
it's going to be wildly cheaper to have a-

a human kind of peering at the different
options that the AI is pulling up.

And the AI says, "Well, it
could be this," you know.

"Here's the images that
we're interpreting.

We're gonna circle this.

Could be this option, could be that
option." And the plumber just says, "Yeah,

it's that option." Um, even if … Even
when we get to the point where, like,

the frontier models could accurately make
that calculation, it's not clear to me

that, um, that it'll be cost-effective
for that to happen anytime soon.

Okay?

Now, um, and- That's probably not
true of medicine, for instance, right?

For medicine, uh, the, the, these
procedures are so costly that maybe it

is cost-effective to have a, a robot
that, that makes no decision … that

makes all the decisions, and we just
kinda churn the, the token cost.

On the other hand, right, if you've
got, if you've got somebody that is,

is operating on your body, you probably
have a, a greater concern with, with

what's going on there because you
don't, you know, you're, you're, they're

You want someone to take
responsibility, okay?

Um-

And, and again, I think the, the, the
point to make here is, is that the,

there are gonna be people who master
these skills and people who don't

master these skills, and the people
who master these skills will be able to

out-compete their, their competition,
especially if they can scale, right?

There's a, a chiropractor, I think he's
actually… Uh, is he a chiropractor or

a, a… Oh, he's a, he's an eye doctor.

I think he's in the DFW area.

Um, and he has fired all of
his staff, and that mean- AI

has replaced all of his staff.

It does all of his coding.

Um, and, and in his mind, that means he
has more time to spend with a client.

So he spends a full hour getting to
know people, talking to them, just

kind of… He, he doesn't have to
constantly be scratching down notes

'cause the AI takes care of all of that.

He can, you know, at a certain point
in the conversation, he can look

over at his iPad and see what, what
questions do I still need to, to ask,

and it'll list them, those out for him.

But he is able to be more
approachable, more personal,

more focused on the client, okay?

Now, I wanted to, um, I
wanted to talk about therapy.

Uh, you know, talking about these, these,
the kind of the talking professions.

And I think it's interesting,
uh, we, we will have to see

how human tastes change, right?

Uh, 'cause, you know, the, the, w-
if, if the thesis is AI and robots,

you know, a counseling session
does not get better because you can

fly a drone or run a robot, right?

Uh, it's, it's not clear immediately
how, how the, the way that we think about

AI will Will be helpful to a therapist.

Um, and, and, you know, obviously we
can design really compelling personas.

I, I mentioned ChatGPT earlier.

Um, but for a lot of people, that
would poison the product, right?

The client wants a human.

They're, they're gonna demand a human.

That's their expectation.

Um, but even in things like therapy, uh,
you know, AI is fantastic at building

profiles of specific people, what
they like, avoid, fear, respond to.

Uh, you know, gathering that, that sort
of public data on one person used to be

a capacity, on the civilian side anyway,
that, that only Google possessed, and now

everyone with access to AI will have that.

And so in theory, you could have a, a
session that begins with the clinician

already knowing the loves, aversions, and
history that no intake form will, would,

would have captured, and so the client,
you know, can, can have that experience.

Now, uh, you know, professional
ethics is, is, is always a question

in these types of, of conversations.

The, the, um, many boards of ethics in
many states have already said, "Hey, you

can't look at social media profiles."
I, I, I kinda wonder how long that'll,

that'll last, um, especially given,
you know, how much you're, um, how, how

much people are, are relying on AI for
all of their decision-making, right?

You know, one of the things that, that,
that I talk about in this is whoever

has access to your core AI chats, right,
which, which is not just the things

that you're typing into the AI, but
the AI's reasoning about you, right?

'Cause, 'cause AIs have to build
profiles on you to predict what

you want so it can serve you.

So i- whenever you're interacting
with an AI, it's building a profile

on you, um, and that profile
is contributing to its answers.

It's kind of referring to
those things internally, okay?

Um, and so who has access to that, right?

Do you, do you want, you know, your
boss to have acc- I mean, i- in many

cases, especially in, in sensitive,
uh, sensitive, uh, professions, the

company will demand access to that.

They will require access to that because
it's a, it's a sensitive profession.

Do you want your therapist to have that?

What, what if you could know that
your therapist would be wildly more

effective if, if they had access to that?

Um, and, you know, if … And
the other question is, like, what

happens if it just gets banned?

Well, if it just gets banned,
uh, by the ethics boards, that

technical capacity doesn't go away.

It just probably migrates to people
who are going to, you know, use it

without, without any restraints.

Um, you know, what does that,
what does that look like?

Is that really the, the, the
direction that we wanna go?

Um- And so again, I think even
in, in therapy, even in these like

intense human-driven things, as
AI reshapes all of society, it's

gonna reshape that, that experience.

Even though I, I don't think
it's gonna replace the human.

I mean, one of the things, you
know, again, therapy shows up on

the list of these AI-proof jobs.

I actually think there's a, there's a
world in which once human taste, uh,

accommodates for it, uh, where, where
we get rid of a lot of human therapists,

where many people are relying on, on,
on AI therapists and kind of running

their decision-making through them.

Uh, I think that would be
probably terrible for society.

Um, I, I, I, I wonder how long that, that
could last, but I think it's a thing that

people will, will try at a certain point.

Um-

And there's just a, you know, there's
just a lot of, there's a lot of questions

about what, what happens when you
give machines access to these tools?

What will happen when a, when an AI has
access to a camera that's looking at

you, when it can track your eye movement?

You know, there was a, there was a
scandal, um, a couple years ago with

the, the Apple Vision Pro where it,
um- It, it was tracking people's eye

movement sp- and it had to because
you can actually type with your

eyes in a, in an Apple Vision Pro.

But they … It also allowed you to
have an avatar, like a, like a, like a

chibi avatar, like a animated computer
CGI avatar, and the eye movement

of the eyes of the avatar was so
accurate that people were able to steal

people's passwords if they had access
to the feed of your, uh, your avatar.

Um, so that is an incredibly detailed,
um, you know, camera that's watching you,

that's, that's tracking all these things.

For sure that could track your, your
facial mo- your facial, um, features, your

expressions, maybe your, your skin tone.

Um, you know, as, as we get more and
more, um, you know, health, health type

things where they can, they can have
something on your wrist that tracks your

heart rate and maybe even blood pressure,
blood sugar, all of those things.

What will a machine be able to do
with that level of information on you?

Right?

What will it be like?

How, how will it be able
to influence your thinking?

And, and I mean, you know,
maybe for good, right?

People are trying to, um, people
are trying to use this for good.

People are trying to, uh, have
this be a positive experience.

But I just think, you know, we, we
need to understand that if you could

have an AI that set your kind of mental
temperature, right, that could, that

could put you in a calm state, right,
uh, what would you pay for that?

Uh, uh, you know, in many ways
this is kind of like, like

light-based therapy, right?

Uh, it's sort of

But it will be, it will be e-
equivalent, I think, to a drug.

It will have a, an overriding impact
on you, um, that will change what

you're doing and, and, and how you're
experiencing your life, uh, as soon as

you give an AI access to a camera that's
looking at you and a screen that it can

put in front of you to show you images
and maybe, maybe some sound component.

Um, will that be, you know,
something like hypnosis?

I don't know.

Um, but I think all of these things
are … You're, you're not gonna be

able to, to keep the bounds between,
uh, you know, a therapeutic counselor,

you know, who's licensed and all these
things, and w- when the only thing you

need, right, 'cause, 'cause a therapeutic
counselor or even, you know, a hypnotist,

right, or, or, or a facilitator, those
people it takes years of training.

There's, there's kind of
a huge regime around it.

People, people are, are suspicious of it.

But what happens when if you put a phone
in somebody, in front of somebody's eyes

and you put earbuds into your ears and
you've got that same level of capacity?

What then?

Um, so I think that will, I
think that will dramatically

influence how these things go.

Okay?

Now, uh, one of the professions that I'm,
you know, particularly interested in is,

is the law, is, is the legal profession.

Um, and I'm, I've listened to a whole
bunch of pitches about how AI is

gonna impact my field and, you know,
my current stance is as follows.

So I think the, the, the human lawyer
is irreplaceable in at least two rooms.

So one is when a client is entering
into an engagement with a law firm.

I think for the foreseeable future,
the, the people that come wanting

a law firm… Right now there's a
question about people who come wanting

legal services or compliance services.

Will they want something different?

Maybe, maybe they will.

Uh, but if you're looking for a
law firm, you are looking for a,

a human person that's gonna make
you feel safe and trustworthy, and

like they've got experience and
they're gonna be good at their job.

And then the second thing is, again,
also for the foreseeable future, anytime

that a judge is gonna rule, right?

I mean, judges are gonna want lawyers.

People, I don't think,
appreciate how much work it is.

Basically, what a lawyer does for a judge,
two lawyers do for a judge, is they make

all decisions into the, into a binary.

The kind of binary that you can
just do a, a, a yes or no, party

A, party B kind of a decision.

Um, and that, that takes a lot of work,
it takes a lot of negotiation, and I

don't think judges are gonna be willing
to not have lawyers do that for them.

I, I, I very much doubt that they
will be interested in, in having an

AI do that for them anytime soon.

Okay?

Uh, but outside of those two rooms,
you know, I think the, the, the

workflow is going to, going to,
to, to become the, the AI-empowered

individual or small team, right?

So client marketing is content, drafting
is code and to some extent content.

Uh, discovery review, scheduling,
follow-up, all of these are things

that AI will do better, faster,
cheaper, and with increasingly

less and less human involvement,
aside from the AI operator, right?

Aside from somebody who is, who
is checking on the AI and making

the decisions and, and perhaps
they need some amount of legal

expertise, but their primary
interface with the law is through AI.

Um, there's a really interesting
component for, for the law, which is,

uh, jury selection and profiling, right?

Um- I'll talk about, a little
bit more of that in a second.

Uh, another thing is, is argument
selection, profiling the bench, right?

Um, there are commercial systems
out there that already claim to

predict a judge's ruling on a motion
to dismiss at roughly 85% accuracy.

Um, as those systems become
cheaper, it will become increasingly

difficult to avoid using them, right?

Um, so there was a TV show on
CBS called Bull that, uh, had the

actor who I think of as the, the
actor that plays, uh… He's, he's,

he's one of the actors from NCIS.

That's how, how, how I, uh,
re- always remember him.

But he plays Bull.

Um, and it's, it's based on a
company, um, Courtroom Sciences,

which is again, in, in, in Dallas.

Um, and it's been shaping big firm
trial outcomes since the 1970s.

Uh, now their, their cost is, you know,
tens of thousands of dollars a day.

I think their cheapest
package is like $25,000 a day.

Um, and so you have to be
able to afford that, right?

Um, that, that was a, that…
You know, $25,000 a day is a

pretty incredible moat, right?

Um, and now any solopreneur, any
solo practitioner who has mastered

increasingly common AI tools is
gonna be able to do that, right?

Um, e- even the judge himself, right?

You know, that, that kind of argument
profiling, judicial profiling.

The, the… Now, again, I, I
think judges will resist this.

They will, they will
want to, to resist AI.

But what do you do when, when,
uh, you're getting arguments that

are, that are tuned to you, right?

When someone's looked at the law and then
they, they, um- They go through and they,

"Well, we're gonna, we're gonna present
this argument in, in a way that this

particular judge will like the best."

And, and I think AI is gonna be
fantastic at that sort of thing.

Once you … You're like, "Here's the
outcome that I want. Here is the case

law and the authority. All right,
now go do all of the research that

you can do on, um, on this judge,
and come up with a way of making

it sound attractive to him." Okay?

Um, and, and, you know, this will include
for sure all public information, and then,

you know, for law firms and clients that,
that, that have, uh, have means, probably

substantially non-public information.

Um, and how do you … The question
is how you defend against that, right?

Well, with ev- as with everything,
the way you defend against

AI is with another AI, right?

So, uh, you're gonna need a huge amount
of AI integration, uh, where, where

you'd have an AI that's listening,
that's kind of on the judge's side,

that's listening to all the arguments and
telling the judge, "Hey, we think, uh,

we think this, this argument is designed
to influence you in this direction."

Right?

Um, a- a- and again, it's, it's unclear
what a judge is gonna do, right?

Like, if you've come up with a compelling
argument, an, a- an argument that's

compelling to them personally, but also,
you know, stands up, is, is, um, is,

you know, gonna, gonna be able to hold
on, um, is gonna be able to stand up

to, to scrutiny, maybe survive appeal,
why wouldn't you decide that, right?

And I'll note that, you know, France
has already banned this practice, okay?

Um, so when we think about the changes
that AI is gonna make, um, I, I, I

wanna make a, a broader, a broader
claim, and this, you know, would

include things like, like AI and,
or like, uh, medicine and ag that we

talked about last week or last time.

So most of these examples
are command seats, right?

They are a verified human holding
judgment and accountability over a

high-stakes decision at, at kind of
a decisive moment, some- something

where there's going to be a, um, uh,
potentially liability for it, right?

Um, and, and again, I, I, I think that
there's, there's going to be p- places

in the stack in, in almost every industry
where that human is going to survive.

Um, but that's, that's because
those are the places where

that skill is most needed.

And so obviously, you know, you have
to have a human there to, to do that

skill, but that human is going to
be in that seat using AI to mediate

the, the, the command skill, right?

And occasionally the content skill, okay?

And that means that the, the
day-to-day operations of these

professions is, is gonna change, okay?

So, um, so that's, that's where
I think, uh, that's where I

think we're g- people are going.

I think the, the, that is kind of the,
the final thing that I'll say about

just specific, um, specific professions.

Um, but I, I, as, as I was preparing
for this week, I just wanted to

kind of conclude that and, and,
and finish off with my thoughts.

Every time when I, when I look at these
things, I look at these listicles and

people say, "Hey, there's, there's
gonna be AI, AI-proof professions." Y-

yes, that means there, there's, that
we're not gonna replace those jobs

entirely with AI, but all of those
people are going to be using AI, okay?

All of those people are gonna
be doing content, right?

Increasingly, they're gonna be
doing their own marketing, and that

means that, that they're gonna be
using AI for content and persuasion,

proof, the story of the work, right?

I was talking earlier about, you know,
and there are plumbers whose content

is so good that they release it on
YouTube and people watch it, right?

Um, all … A bunch of back office
functions are also gonna become just,

you know, pure codable functions.

Um, you know, the, any accounting and
billing, uh, scheduling, drafting.

I mean, again, the, the, the, the pieces
of accounting that are gonna stay are

going to be, like, actual accounting,
like making, making decisions about

characterization, making decisions about
what, what kind of money is this, right?

That are kind of more soft skills.

Um, and then, and then at the end of
the day, you're gonna have a lot of

command decisions, and not just command
decisions where the AI can't, right?

Um, but command decisions where it's
very, very, very expensive for the AI

to make this decision, so we'd rather
have a human in the loop t- 'cause that

will dramatically, and by dramatically,
I, I know I'm being told 10,000

times reduce the token costs, okay?

So now we're gonna move into what, what
does this look like for the firm, right?

So there was a guy named Ronald Coase,
um, or Coase, I can't re- I can't

remember how to pronounce his name.

But he, he talked about,
you know, why firms exist.

So they say the, the market
has coordination costs, and the

firm grows to the size at which
internal cor- coordination is

cheaper than, than external
coordination or outsourcing, right?

Um, so every department, if you've got
a big, you know, you know, large, um,

uh, a large corporation in the modern
day, uh, is that, that, that … Those

are all coordination costs, right?

You have legal, design, operations,
communications, compliance, and those

people are hired because someone with
money believed that routing that work

through outside contracts was going to
be more costly, at least in the long run,

than employing the people directly, okay?

Um, and so we're gonna … You know,
I, I think that it's important to know

that that co- that coordination problem
is real, and it's not going away.

Um, and, and there's still some open
questions about how much AI can absorb

about that, uh, that coordination problem.

But, uh, so we'll, we'll … We're
gonna talk about two examples.

We're gonna talk about the zero
human company, and then we're

gonna talk about Microsoft.

So, um- You know, one of the things
that, that I talked about last week was

how, or last time, was how electronic
motors, uh, changed everything, right?

So you had the, the old steam-based
motors or, or, um, you know, they would

have large water-powered, uh, motors
where you have a, a mechanical drive

shaft that would kind of go over the top
of the, the factory floor, and everyone

would mechanically attach their little
machine there to, to that big drive shaft

that was, that was over the top or maybe
running through the middle of the floor.

Um, and then when electric motors
came on, in- initially they just

attached a new electric motor, um,
or a new internal combustion motor

to that same drive shaft, and that
turned out to not, not work very well.

What they needed to do was give, you
know, an electric motor at each individual

station, and then ultimately, um, they,
you know, every, every, every tool,

power tools have a, a, an electric motor
that's specific for each one of them.

So, um, I think that, that, that we
are going to see a similar process

in, in terms of how we redesign the
corporation, um, how we redesign

these business organizations.

You know, the corporation, um, was,
was basically designed i- a- after the

invention of the printing press and it
was kinda designed to take, take advantage

of the fact that you could print multiple
identical sheets and keep track of things

and move, um, you know, push paper from,
from desk to desk, and that represented

an incredible efficiency boost, right?

So in January 2026, Brian, uh,
Roemmele, uh, launched what he calls

the world's first zero human company.

No human employees clocking in, AI
agents running the day-to-day on a

12-year-old MacBook in his garage, right?

That's where he started.

So now again, zero humans means
zero humans on payroll because

there's one guy, Brian Roemmele,
who is still in the picture.

He calls himself the chairman.

So, um, he, he initially kind of
framed it as a, as a C-suite of models.

He said that Grok was the CEO.

It was scanning X. It was holding meetings
every 15 minutes and issuing directives.

Uh, he had Claude Code
be his chief engineer.

He gave it full OS access.

It was writing scripts and downloading
tools and requesting budgets.

Um, and you know, there's these hilarious
interactions and, you know- One, one

of the other points that I'd make is
this, there's clearly some content

still going on here, so this is all
mediated by, by Brian's Twitter account.

Um, it's not like he's posted, you
know, uh, cryptographically verified

chats of these interactions or anything.

So they fought with each other.

Uh, so Groq denied Claude a $150 PDF
tool and demanded $1,700 for marketing.

Um, and then, uh, then Rommel
announced a new hire, uh, which was

Claude Bot, uh, and, and which gave
it open source hands for the agents.

And then he created this thing called
Joule Work, which was basically like

an energy budget, so he's paying
the agents in, uh, in energy to

run their, their computing cycles.

Uh, and, and, and this is, this has
continued on, you know, uh, I don't

know exactly what the business model
is or how he's making money, but

you know, last month he said he had
expanded to thousands of live agents.

He has a simulation layer to stress test
decisions before real capital, um, and

sales and agent payment rails are made.

Um, but again, so this is all, like, fun.

This is cool, right?

And it's, it, it's, it's kind of a fun
little rabbit hole to, to dive down.

This is a zero human company, but
it's not a zero human company, right?

There is still one guy, um, and you know,
he's… You ask what he's doing, and he's

doing these same three skills, right?

He is coding the high level architecture.

He is, you know, setting out the
internal logic of, "I want this agent

to do this, and I want that agent to do
this." He's deciding what agents exist,

how they talk, what tools they get,
which processes to prune or restart.

He is still producing the content on X.
I mean, I'm assuming that the, that the

agents are helping with that, but he
is, you know, involved in the posting

process, and he is still making the, the
key decisions, what to spend, who to hire,

when to kill a runaway loop, and, and he
is authorizing the, the next redesign.

And, and, you know, again, this
is not what chairman means,

right, in, in normal parlance.

Um, a lot of times the chairman
is somebody who kind of

understands the overall direction.

They're good with the finance piece, but
they're not somebody who is intimately

involved in, in technology architecture.

Um, what, what he is is
he's, he's a, he's a prime.

He is a, he's an AI paladin.

He is somebody who is really, really
good at this technology and, you

know, he's someone who can design the
architecture, rhetorically defend it,

and then take responsibility for it.

Okay?

'Cause his job is basically
redesigning the company.

Um, so he, you know, he brings
on Claude Bot, he brings on…

He, he, he decides we're gonna,
we're gonna pay you in, in energy.

Um, he wired in, uh, remote offices
over LM Link and hiring idle home

computers from, uh, from other,
other people on the, on the internet.

And then in March, uh, he was making
DDR memory for, for his agent swarm.

And so, you know, he, he is
constantly evaluating where he,

he's given this, this design, this
architecture, and the AIs live within

the architecture that he set up.

They're not going… They might
propose changes, um, but they're

not going to really be able to think
about, well, what, what don't we know?

And ma- many of these AI agents
also, you know, they, they don't do

as well with brand-new tech, right?

'Cause there hasn't been-- They
weren't trained on an internet where

this was common knowledge, right?

That's been, been one of the
interesting challenges of, of AI,

is it, it knows, it, it has a deeper
intuitive understanding of stuff that

was common knowledge on the public
internet when it was trained, right?

Um, or if there was a cutoff date
for the training data, whatever the

cutoff date for that training data is.

Stuff beyond that, that it has
to reason about it, it, it's

much worse about reasoning it.

Okay?

Um, and, and again, uh, and this, this
is, it was, it was in kind of reading

about Brian, what he's doing that, that
I, I learned this, that there is a, a,

a token cost to making decisions, right?

If you, if you ask an AI to make a
decision, if you give it open-ended

autonomy and you say, "Just keep, just
keep looping until it acts," uh, it

can cost tens of thousands of times
more inference than just laying, saying

to the AI, "Give me your top 10, 5
options or, or 10 options," and even,

you know, getting, charting a course
in detail, and then I'll pick, right?

Um, so I wanna emphasize this again.

Autonomy isn't free, right?

Like, the fact that the, that the
model can do this and, and maybe even

at approaching, you know, five nines
percentage points make the right decision,

well, that, that doesn't mean that,
that that's gonna be cost-effective

in most industries anytime soon, okay?

Um, and, and, and even, you know,
there's al- there's costs in terms

of how much, how much energy you
have to feed it, how much you have

to pay for that kind of compute.

There's also time because there is just
a, a time component that the, that the

models have to work through, right?

Uh, there, it, it's gonna be a long
time in many industries, I think,

before the, the models can make a
decision a- in a feedback loop and

be fast enough for that to, to work.

Um, in many cases, I think h- just having
a human in there that says, "I'm gonna

look at these three to five options,
and I'm gonna make a decision quickly,"

uh, that's going to, uh, that's going
to make workflows economically viable

long before we get there with the, with,
with a pure autonomous model, okay?

Um, so again, chairman is not the
right word for what Brian is doing.

I think the right word is paladin
or prime or something, something

that acknowledges, you know, he,
he is making all the key decisions.

Um, and, and I think that, that, um-
You know, one of, one of the reasons

is because, like, you can't just freeze
the agent org chart and, and, and say,

"Well, now instead of having humans
doing these different jobs, we're gonna

have AI do these different jobs," right?

AI reasons differently, it processes
memory differently than humans.

There will be enormous efficiencies
to find if you move, um, if you're

able to say, "Well, somebody
needs to redesign it," okay?

And I think that Brian, you know, imagines
that at some point he'll create a stable

workflow, that the zero human company
will be able to run, run without him.

That, that seems… That, that's
how he comes off on Twitter, right?

Um, and, and I think that, that there's
gonna be applications where that's true.

Things like a, a fabrication run, you
know, a manufacturing run, you know?

Uh, or an aquaponics greenhouse.

Some kind of limited, defined space.

Something that, uh, that, that
could be run on some level by

a CNC machine or robot arms.

Um, you know, famously Elon's
building this, this massive

chip fab in, uh, in Texas.

God bless him.

Um, and the whole thing… You know,
it's, it's funny because on the

one hand we're used to chip fabs.

We're like, "Oh, this is gonna
be great for the community."

I mean, I think it will be.

I think it'll… It's a good thing
that it's here, but it's not going to

generate jobs in the same way that,
that, that factories would in the past.

I mean, the, the, the factory is
going to be as autonomous as anything

has, that, that's ever been, right?

Anything inside a box, right, whether that
box is a CNC machine or a building, it, y-

you're gonna be able to get to the point
where you can optimize it and, um, yeah,

the, the, the inference is good enough.

It's making decisions fast enough.

It doesn't need to be improved.

But what about for a company, right?

What about when you're
competing in the markets?

What about when you're solving
those massive integration problems,

the, the coordination problems?

What about when you're
providing liability backstops?

At this point, I, I just don't
see how there will ever be a time

where that doesn't require humans.

It does- where, where humans thinking,
the, especially our capacity to think

outside of the box and say, what,
what level of abstraction is it most

efficient to try to solve this problem in?

You know, there's, there's that
old, that old line about, uh, you

cannot solve a problem at the same
level that it was created, right?

You have to go, you know,
deeper or higher, whichever de-

depending on kinda how you're, how
you're thinking about it, right?

So what does this look for, like
for the, the modern tech giants?

What does this look like for Microsoft?

Um, so Microsoft, uh, you know,
let's think about Microsoft today.

They, they still run
as a command hierarchy.

They have Windows, they have Microsoft
365, Azure, Gaming, LinkedIn, GitHub,

and they're, that's, they're all under
one CEO, and they have one kind of

equity narrative that they, that they
tell, uh, to the market at least.

Um, externally they, they
have a, a multi-model story.

They have Foundry, they have Copilot.

Um, they, they are really pushing more
that, that they want you to be able to

swap the model and stay on Azure, right?

Uh, that's their, their, their,
their data, data repository, their

data management system, right?

Um, now internally it's, you
know, they're, they're, they're

wanting you to use Copilot, right?

Um- But it's interesting, you know,
the, the thing that people… The

reason why people buy Microsoft,
um, is not necessarily because

it's the best software, right?

Um, uh, it- it- it's because there's
a, a, a lack of risk there, right?

Um, you know, the, the, there's this old
phrase, you know, no- no- nobody ever got

fired for buying Microsoft Azure, right?

Like, there… Now, now people might get
fired for f- implementing it improperly,

for setting it up wrong, right?

You know, one of the things about
these deep Microsoft projects is

they're, they're more customizable
f- in, in many ways, and that, and

that means that you really can set
them up in a way that's, that's…

You can screw up the setup.

Um, but it's such a robust product.

It's so familiar.

So many people use it, and especially if
you're in any kind of, like, compliance

type of a, uh, type of a thing.

Um, you know, if, if you, um, make
a decision to, to buy Microsoft

and something goes wrong, it's
the fault of the implementer.

It's the fault… You know,
basically the person who makes

decisions to buy Microsoft Azure,
that's like a VP of a company.

That's somebody in the C-suite.

The person who sets it up is, you know,
a frontline guy, um, and, and those

are, those are two different levels
of, of firings taking place, right?

If you were…

If you… But if you're the, if
you're the VP and you say, "Well,

we're gonna use this untested new
software," right, and something

goes wrong, well, that's your fault.

You decided you wanted the
untested new software, okay?

And that brings us to the, uh, the
forward deployed engineer, right?

Because I think that this is,
this is all gonna change, right?

So, uh, Mark Cuban, um, has, uh--
was on a podcast, and he, he talked

about, you know, the, the demand side.

He thinks that software is gonna die
as a shared project, uh, product when

every firm can afford customization
to its own utilization, and there

are tens of millions of US companies
that will need someone to build that.

So the, the job title that people are,
are throwing around on the internet

is the forward deployed engineer, FDE.

And this, as, as in so many other
things, is a Palantir framing.

It's a small team with an end-to-end
owner, kind of like a startup

CTO embedded inside the customer.

Um, so Frontier Labs and the big vendors,
including Microsoft, are already hiring

them in volume, and this is, this is,
this is the new, the new model, right?

The, the deployments die at integration,
compliance, and the ops handoff

and, and that's what the, what the
FDE is supposed to, uh, to solve.

So an FDE wears three hats.

One, they're a software engineer writing
production code in the customer's stack.

Uh, two, they are a business
consultant discovering what

the domain actually needs.

And three, they're a product
manager fielding field pain,

uh, back into the vendor's core.

And this sounds, friends, a lot like
code, content, and command to me, right?

the, uh, the forward deployed engineer
is, is the guy that, that, that

bridges that last mile between an AI
product and, and, uh, enterprise value.

Um, and I think what's important to
understand here is that, that the FDE

looks like someone who is designing
custom workflows with a big company

like Microsoft providing a set of
compute capacities or underlying app

primitives that are designed to work
in an increasingly modular stack.

And, and, and this is how, uh,
Microsoft is gonna compete with

the zero human company, right?

Because Microsoft is, is seeing, I'm
sure, everything that Bryan and other

people like him are putting out.

They're seeing what, you know, Jack Dorsey
is doing over at Block and saying, "How

do we, how do we compete with that?" Okay?

So, but as power moves from the product
organization that ships a, a horizon

feature, um, to the people that,
who can make Azure and Copilot do a

specific thing for that customer's
idiosyncratic needs, they're…

That's custom work, right?

Custom work means liability, means
responsibility, and so there's gonna

be i- uh, pressure to say, "Well,
actually, we want, we want these

forward deployed engineers, but we
kind of want something that bears the

risk, uh, between them and us," right?

Um, so the old chart, the old org
chart assumes that the, the moat

was, was intellectual property
created at Microsoft headquarters

and, and consumed at the edge.

And the FDE model assumes that the
moat is, is hardware primitives

and, and the ability to, to, to make
this thing do what the, what the

customer specifically needs, okay?

And so when, when we've got forward
deployed engineers that are, that are

arranging the core services that, uh,
that Microsoft provides, th- that,

again, looks more and more culturally.

That, that forward deployed engineer is
going to think of himself as competing

with, with Brian Romel, and, and, and
he's gonna treat Microsoft more and

more like a c- a hardware vendor, right?

Microsoft is providing the compute.

Microsoft is providing a bunch of
off-the-shelf components, but he is

kind of in business for himself, okay?

So if this is the new structure, what
does a sane reorganization look like?

And I, I think a good
answer is to look at Visa.

So in the 1960s, the new
technology, uh, that people were

dealing was the credit card.

Um, you, you had, you know, the, the
new plastic credit card with the ridges.

You could, uh, sometimes you
could clear that electronically.

Sometimes people would still use these
old, uh, these, these paper cards, um,

where you would, you would run the, the
card through, uh, you know, um, contact

paper or th- things that would, would show
up, um… Ah, what am… Carbon paper.

That's the word I'm looking for.

And, and then you have
an imprint of the card.

That's why the, that's why the numbers
were raised in a lot of these old cards.

Um, and it's important to think about
the fact that, that before this,

payments were, were local and closed.

So you would have a, a department
store charge, uh, card that

would work at one mer- merchant.

A gas card would work at one brand.

Uh, Diners Club and American Express
were travel and entertainment charge

cards, um, and, and they, they worked
in a lot of places, but they were aimed

at kind of expense accounts elites.

They were not, they were not a, a mass
market product, and they certainly were

not shared between competing banks.

Like, that, this, this was the
thing that was so interesting

about the Visa card, right?

So uh, you know, it started off as
the, the, uh, BankAmericard- And, and

you have national licensing, right?

So you have an ordinary
consumer revolving credit.

It had one brand, it was issued
by many banks, and, and a

incr-incredible number of people had
to honor every franchisee's card.

Um, so this, this created a
nationwide common market for digital

money, and that was something
that the old systems could not do.

And, and it was massively successful, but
it created a bunch of, uh, of problems.

So Bank of America had not accounted
for the new competition between their

members that they had created, right?

Because i- and, and the way that
that competition would work.

So by the late 1960s, the
technology was, was chaos.

There was, there was weak authorization
and clearance, and clearing.

There was a ton of fraud and, and just
a bunch of losses, and then, you know,

the, the, the franchisees would kind of
get together and, and blame each other.

And, and there was nothing
wrong with the system.

The technology worked.

It worked exactly as advertised
and in many ways was, was better.

Um, but the, the system wasn't
working because you needed a

new social arrangement, okay?

Um, so… And one, I think the, one
of the things that they, they just

didn't really anticipate was how much
competition they had to allow for, right?

Um, they… People were now able to
compete with one another for clients,

for customers, um, and that was,
that was one of the, the aspects

that the technology facilitated.

Because again, it was, it was not just a
common market for money, it was a common

market for credit card customers, right?

So everybody could now be wired into
the same, the same thing, and it

wasn't, it wasn't a big hassle to
change, for the customer, to change

from a card issued by this bank to
a card issued by that bank, okay?

So they, they had to change
their management structure,

their social structure.

They called it, uh, chaordic
structure, uh, so chaos plus order.

So they, they said, "We're all gonna
compete on what are the rules, what is

the brand, how do we do authorization,
clearing and settlement?" And, and

they're competing on, on kind of
the financial products, what deals

that the cards offers, and they're
competing on the customers, okay?

And I'll note this is also a very
liability efficient structure.

So virtually all complaints and decisions
were owned at the member level, meaning

Visa itself was allowed to focus on
the piece they did best, which was

the, the payments technology, right?

So if we turn, if we say, "What would
this, this look like for, for Microsoft?"

Well, this looks a lot like,
you know, Microsoft becoming

kind of a, a neo-feudal, um,
a neo-feudal structure, right?

So Microsoft is the sovereign.

It's got identity, it's got graph
memory, it's got compliance, um,

and, and it's still, you know, for,
for certainly for a generation,

it's going to be a career safe rail.

No… Again, no one is gonna get fired
for buying Microsoft, and that is, is

kind of the core of their offering.

Then you're gonna have the
forward deployed engineers

and the integrator houses.

These are your barons.

They're, they're vassals who can,
uh, in principle switch overlords

or run, run, run for multiple
people, but they'll probably have

one core offering that they…

One, one core, uh,
sovereign that they serve.

Um, and increasingly, the enterprise
customers are, I think, going to be serfs.

They're, they're bound less by the
license fee than by audit, career

risk, um, the way the data works.

And, a- and even if they were to
switch, it- it's going to… It's always

going to be easier to switch to a new
vendor, to a different baron, if you

will, uh, than it will be to switch
to an entirely new system, right?

This is, this is also true, one of
the, one of the dynamics with AI

that, that, that w- that they're gonna
have to deal with is the potential,

the new potential that, that comes
from AI vendors being able to cheaply

compete with their clients, right?

So Chamath, uh, has put out, uh,
some stuff about, you know, if you

deploy OpenAI or Anthropic deep
into your firm, you are letting

the fox in the henhouse, right?

So, you know, PwC and Accenture
trained PricewaterhouseCooper, um,

ha- spent tens of thousands on, on
Claude, and it's starting to look

like the, the, the labs are, you know,
deploying competitors to them, right?

So the AI vendor is using your usage
to get better at replacing you.

Um, the same thing is true,
uh, in, in pharmaceuticals.

So Anthropic, uh, sells Claude Science
to drug makers, um, and, you know,

big names like Sanofi, Novo Nordisk,
uh, were among the named clients.

And then in June 2026, it announced
that it would enter the pharmaceutical

market and develop drugs itself.

Um, this, this, this struck a lot of
people as, as unfair and, and I think

it created a big backlash, but, uh,
but this is what's happening, right?

This is what's happening.

So, so how do you solve that?

And, and I think the way you solve
that is by, you know, to lean into

the, uh, into the, the feudal, the
neo-feudal metaphor, is to give the

baron skin in the game, all right?

Um, so you grant the forward deployed
engineers a real chunk of the

profits, maybe even a chunk of the
equity, um, i- i- i- in these kind

of surf clients that are, I think,
gonna be more and more captive.

If they can't figure out how to make AI
do it for themselves, right, um, then,

then they're just gonna be more and more
unable to move away from this, right?

And, and so what you wanna do is say,
"Hey, we don't want you to compete

with us. We want you, we want you
to be invested in our success."

Um, and, and so you give them,
you give them some of the upside.

So this is gonna be the Visa logic again.

Push liability and upside to the edge
that actually holds the relationship,

um, uh, 'cause a salary deployer
who owns nothing is, is, is gonna,

you know, defect endlessly, right?

But a baron, somebody who see I, I own
this territory, these are my clients,

this is, this is my thing that I'm
doing, um, they're, they're gonna

defend the surf against being eaten,
including by their vendor, right?

Um, and, and I think again, th-
there's, there's… One of the

key questions in here is who has
access to all, to the underlying

chats, right, and the underlying
processing that the agent is doing.

And if Microsoft can, you know, say, and
they will need to say credibly, "Hey,

we're gonna have systems where cr- you
know, cryptographically, we actually

don't have access to any of that stuff.

Um, we're gonna push that to
the forward deployed engineer.

He'll have access to it.

He'll be able to grant access to, um,
the, the executives of the company," that,

that seems like a, a really interesting,
uh, a really interesting possibility.

So to conclude Uh, what, what is the
market value of, of an FDE, right?

So this is … And, and again, I think, I
think this is too small of a name, right?

Like I talked about, um, i- i- two
weeks ago, I talked about how builder

is a name that's gotten thrown around,
and I think this is just, that's

just too small of a name, right?

Um, so what is the market value of
a forward deployed engineer, right?

And, and, and again, keep in mind that,
that the paralegals and IT professionals,

and the technical support assistants,
and the forward deployed engineers

that have been software engineers
for their whole careers, they will

all be competing against each other.

They will all be priced
a- a- against each other.

I mean, that forward deployed
engineer is worth a ton, right?

He, he becomes the bottleneck.

He becomes the, the one thing standing
between this firm a- and greatness.

And that start … You know, when, when
you start asking the questions like,

like how, how much is that person worth,
well, i- i- … And this is especially

true as we're, we're looking at education
and training costs just dropping near

to zero as, as training and education
becomes an AI mediated thing as well.

A- and you get this interesting point
where let's say you got a cohort of

people, you know, one in 20, um, and
maybe even one in 50 or one in 100.

Um, if you get one person that can, that
can develop into a true prime who can

enter any industry and compete at the
highest level, that more than repays

the training of a whole class, right?

It, it, it will likely even pay
for maybe even a lifetime support

for the rest of the cohort.

Not that we should
necessarily do that, right?

It's not … That, that's
a separate social question.

Um, and, and … But, but I think the
fact that we could do that, the fact that

we can say, "Hey, we can push on people.
We can even potentially take risks in

training people because it's so important
to get as many of the, the AI palims."

The fact that the numbers work for
that is really int- in- interesting.

So, uh, when the upside of one AI
operator dwarfs the cost of the pipeline

that produced him, that, that, that
allows you to take these sorts of risks.

So, and again, I believe that
societies that can maximize the

production, it's not just, you know,
individual operators or companies.

I think it's, it's societies as
a whole that can maximize the

production of trustworthy primes.

I think they'll outcompete
societies that cannot.

Okay?

Um, and so as we're, as we're, you
know, concluding, again, code, content,

command, this is one skill set.

It's gonna apply in, in
military applications.

It'll have civilian applications.

It's gonna apply to the trade,
to medicine, to therapy, to law.

I don't see any occupation that,
that escapes competition, at

least within the job, right?

This is not to say that humans are
obsolete, but somewhere in the stack,

there's gonna be humans, and AI-assisted
humans are going to out-compete,

um, all of their competitors.

And that… A- and when, then when
we realize, hey, a, a Prime is, is

going to create enough value to make
it worthwhile to train entire cohorts

of people just to get one, right?

Um, a- a- and what, what that means is,
is that people will invest in this, right?

It means, you know, I was having a
conversation with somebody the other

day like, "Well, but y- you know, you
only, if only people with money will be

able to do this, how, how f- we're not
gonna have enough, we're not gonna have

enough people to be the AI paladins."

I'm like, "I think money will
flow to people with the skill set.

Money will flow to people that
might have the skill set," right?

Um, so that's n- that's the framing.

This is how important this,
this skill set will be.

This is how important the- these people
who will be quickly, I think, a new class,

a new group of people who have this, a,
a very similar set of core experiences,

who have a similar set of core interests,
who are going to, you know, be, be

in a position to wield political
power to their collective interests.

Um, and, and so I, I think that is the,
the, the thing that, that is coming.

That is the future that I see.

And so the question will be how, how will
society as a whole, uh, respond to it?

Again, at the, at, at the individual
level I think there will be people

that, that, that master AI and
out-compete, uh, their peers.

At the corporate level there's gonna
be, you know, the zero human companies

or, you know, baron, the barons
of, uh, of, of Microsoft, right?

Um, and then next week we're gonna talk
about the, the Roman cursus honorum, um,

and, and how that's going to apply to, um,
to how we could potentially redesign our

society or redesign chunks of it, right?

Uh, one of the things is we wouldn't
have to redesign … When I say

society, that doesn't mean, you
know, America, 300 million people.

You could do it on the scale of,
of a, of, of a small, a small city.

May- a large town kind of a thing, deal.

Um, but, but, you know, which, which, you
know, is what Rome was compared to w- you

know, for, for much of Roman history when
they were running the cursus honorum,

you know, Rome was, was the size of a,
of a, of a small or e- or mid-size city.

It wasn't, you know, it didn't,
I don't think it hit two million

until, until nearly or right around
the time that it was an empire, so.

Um, because one of the other points
that I wanna make is it's not just

gonna be important to produce these
people that have the skill set.

When you're giving all this trust,
when you're giving all this power

to someone, um, you're gonna wanna
know that they're high character.

You're gonna wanna know that there's been
some kind of moral formation and, you

know, the answers that the people have
been giving for the last, you know, 100

years, well, well, moral formation is
just not, not predictable, not doable.

I don't think that's true, and I
think it'll become necessary to have

people that have a certain, uh, a
certain degree of moral formation.

So we'll, we'll get into that
next time and, uh, thank you

so much for, for listening.

This is, uh, the end of my prepared
remarks, and now we'll move into, uh, Q&A