Dude, we are living in the idiocracy

version of the AI timeline. Trump

ordered agencies to stop using Anthropic

with a six-month rampdown period, and

Hegsth has moved to label it a supply

chain risk. This is not normal vendor

drama. This isn't this isn't normal at

all. This is precedent setting AI power

politics.

Here's the sequence of events. Then

we'll talk about what the heck they're

even arguing about. Apparently,

Department of War and Anthropic, as well

as other major AI vendors, have been in

discussions about using their LLMs for

national defense. It goes south when

they hit a stumbling block. So, Dario

Amod posts a press release through

anthropic site stating what they will

and won't do. And this escalates quite

fast. Trump through official White House

comms as well as Truth Social says,

"Effective immediately, federal agencies

stop use. we will be observing a

six-month rampdown of use of anthropic.

Then the Pentagon moves to declare

anthropic a supply chain risk to

national security. They also make some

bold claims about who can and can't use

anthropic even outside of government

specific business. And then Anthropic

issues a statement to that saying

actually your claim is toothless and

legally you can't do it. This is the

inflection point because this is going

to determine who the keys of AI power

are handed to for the future. And don't

get me wrong, I am an anthropics court

on this and I'll explain why, but

ultimately I believe AI should be in the

hands of the people. I'm a big proponent

of running local models. I think there

needs to be more focus on that. That's

the reason that Open AI was founded in

the first place. But let's park that

because it's right now anthropic versus

the Pentagon. This all kicked off when

the Pentagon asked for models for quote

all legal purposes in the Department of

Defense. I haven't heard as of this

point any dispute coming from OpenAI,

from XAI, from any of the other major

model providers. We don't know if they

are complicit with this, if they're in

agreement, or if they will be contesting

this as well, but we haven't heard

anything from them. So, now would be the

time to speak up. Still haven't heard

anything, though. It's 8:00 p.m. on

Friday night in Phoenix and uh and still

nothing. Anthropic in their press

release says all legal, all lawful

purposes. Sure, fine. All good with

that. Uh, hold up. Two exceptions,

though. We don't want you using it for

mass domestic surveillance of US

citizens, and we also don't want you to

fully automate killchain decisions. And

again, I'm a proponent of local models

of roll your own. I'm very libertarian,

but at least in this circumstance,

anthropics response seems non-emotional.

It seems very wellreasoned. And I think

those are two noble things to abstain

from. Like I I would have great moral

qualms about participating in the mass

surveillance of US citizens

domestically. That seems like something

that we should do everything in our

power to prevent against. That actually

seems like something that the government

should prevent against that they haven't

done a very good job preventing against.

So it's actually good for wants to see a

corporation saying we're we're not going

to surveil US citizens. We're not going

to do that. That's not within our moral

framework. Fully autonomous weaponry is

a bit of a thornier one. And that could

be a whole YouTube video on its own. I

won't get into it, but I see the

reasoning from Anthropic here that they

don't want to be involved in these

missionritical killchain decisions. And

I can see a very solid argument for it

because essentially what you're doing is

you are trusting an algorithm to make a

decision about a human life. Ethically,

I don't know as a society where we

should stand on that. I know where I

stand on that, but I don't know where

we'll end up. So, I am very much in

Anthropic's corner on this debate. If

you fervently disagree with me, I want

to hear about why in the comments. Let's

have a rational discussion about it.

This is something that we should all be

talking about to figure out where the

line is for our society. This degrades

pretty quickly into a legal argument.

And usually I wouldn't bore you with too

much of it and I'm not going to bore you

with the details, but it's a very

interesting legal case shaping up

because you have on one hand the

Pentagon which has like an army of of

lawyers. At least like oneif of the

Pentagon is probably lawyers. That's

like one side, one whole side of the

Pentagon. They probably got the lawyers

in there. It's probably like the bottom

right side of the Pentagon. That's where

all the lawyers sit in a in a big long

row with each other. They have a lot of

lawyers and those lawyers have worked in

department of defense type law for a

very long time. They have a lot of

experience there. Right? On the other

hand, you have probably some of the most

cracked lawyers in the world in

anthropic because they have an insane

amount of money to throw at people. An

insane amount of money. They're they're

attracting better lawyers than the

Department of Defense with their

salaries. And they also have AI,

probably better AI that's available to

peeons like you and I behind the scenes

that they can just set loose on the law

books and find technicalities and

loopholes and all sorts of craziness

like that. And yeah, like yeah, the

Department of War does not have that.

I'm not a lawyer. I really cannot tell

who's in the right or wrong here, but I

can tell you what they're specifically

debating about because that's the

interesting point here. Pete Haggsath

and the Department of War contest that

if uh software is labeled a national

security supply chain risk like they are

moving to declare anthropic. What that

means is that if your company is working

with the Department of War on anything

even business totally unrelated to the

US government outside of that you're no

longer allowed to use anthropics. So

it's not just like a ban while you're

working on stuff that you make for the

Department of War. It's a total ban for

your entire company. That would be like

if you built refrigerators. You have a

big refrigerator business. You sell

them, you know, through Home Depot,

Lowe's. You're well known in the space,

but Department of War calls up and

they're like, "We need this particular

heat sink that you manufacture for your

refrigerators. We need it." You're like,

"Okay, fine. Great. Great. I want to

support the American war fighter. I'm

going to quint quintuple the price

because I know you can pay for it with

that defense budget, but yeah, I'll sell

you as many as you need." The catch

there is you use Anthropic for your

business everywhere. Like most

businesses are using that or open AAI

now. You're using it all over the place.

You're using it to aggregate customer

feedback. You're using it as a support

chatbot and you're using it to design

these heat sinks. So Anthropic's

argument is that you can no longer use

it to design the heat sinks for the US

government. That's that's clear. It's

out of the question. Anthropics argument

goes on to say, "You are still well

within your rights to use it to design

your refrigerators and for any other

business uses outside of what you're

doing for the Department of War." Hegath

and the Department of War contest that

no, you cannot use it for your entire

business, refrigerators, what, whatever.

You can no longer use Anthropic at all.

It'll be interesting to see how that one

shapes up, but both of them are assuming

full authority on this question already.

The latest press release from Anthropic

says, you know, hey, if you're

concerned, if you're doing some business

with Department of War, you can still

use it outside of that. And if you're

concerned, you can call our sales or our

legal team. Uh they're here to help you.

Uh so both of them are doubling down on

this, and we'll see who blinks first.

The capitalist craziness on top of all

of this, if this was not already a

nuttiest situation, this is maybe the

weirdest part of it. Anthropic is a

company. It's got to make money, right?

OpenAI is a company. It's got to make

money. You got to be wondering if Daario

and Sam are on the phone saying, you

know, blood packed, like we're going to

not allow our LLMs to be used for this

thing and that thing. Or if they're just

like, hey, business comes first, baby.

And you know, I'm maybe Sam Alman's

saying, I'm going to totally compromise

the morals of my LLM. We'll let them use

it for whatever. We do not care as long

as they have the money to pay for it.

Not saying that Sam Alman said that, but

I'm saying as an example. And uh all of

a sudden, I mean, you know how big the

defense budget is. It's massive. If you

don't know, go look it up. I mean, it is

it's bigger than dwarfs any other

spending in the US. It's preposterous.

And so, you hope you hope that people

will take a stand about what they're

designing and what they're building and

say, you know what, like this is this is

my boundary and you have to respect my

boundary. And that's exactly what

Anthropic is doing. I don't know if

OpenAI is doing it, but you have to

think if you're Daario, if you're the

board of directors, if you're an

influencer in that space, you got to be

thinking like if OpenAI keeps doing

business with them, that could mean

we're ruined in terms of profit. Like

we're going to bring in a lot less money

if the other AI companies cave and start

working with the government. It's a

gross angle to think about, but

undoubtedly this is going on when

there's that much money at stake. The

administration is largely framing up

this argument on the basis of

patriotism, morality, ideology, and

military readiness. All of which think

things I agree with. Anthropic is

framing this uh quite intelligently, I

might add, on model reliability and

constitutional rights. The model

reliability we want to put a little

asterisk to and as a clarification

there. They're saying they are by their

own admission saying we do not trust our

own models enough to say that yeah sure

you can use it to determine whether to

blow that guy up or not. We don't we

don't think our models are good enough

which I think is probably a pretty

responsible take if you've used the

models. It's concerning to me that the

Department of War wants full access to

these. This is not a tech company. Uh

Pete Hgsth I don't think he's ever been

a programmer. I haven't really looked

into his history, but he doesn't strike

me as the type that spends a lot of time

with computers or books for that matter.

And Anthropic is caught a stray bullet

on this one. Uh they're going to be made

an example of whichever way this goes

down, but this is going to set the

template for all AI defense projects

moving forward. So this is far far far

bigger than Anthropic. And the precedent

that is going to be set is can private

model providers like Anthropic, like

Open AI, can they define what you can do

with the model, what the government can

do with the model? You, you and I don't

get models without safeguards cuz we're

we're too dumb. You know, they got to

protect us from ourselves. This is

absolutely insane timing for this to

really come to a head before the weekend

because we're going to have to wait

until Monday to see does anyone file

anything in court on Monday, later in

the week. But that's what's going to be

next is there's going to be some court

filings, some legal proceedings that

kick off. But what I'd be really

interested to know is if you have a

company, right, that's working like that

refrigerator company. They're working a

little bit with the Department of War.

You know, I wonder, do they say like,

"Okay, we're we're terrified of of legal

proceedings, so we're just we got to

drop anthropic on Monday. We're not

using it uh anymore at the company on

Monday." Or do they say, "F it. We ball.

This is this model is tops. It is Opus

4.6 six is amazing. We are not going to

stop using it for a refrigerator

business. I don't know what's going to

happen there. My hunch is the former

because there's a lot of money coming

out of the Department of War and people

will tend to go with that unfortunately.

If you got a source there, let me know.

I would I would love to know how that's

actually uh hitting the market on Monday

and hitting corporations. But I know I'm

glad I'm not a lawyer at one of these

companies because uh I am not going to

work this weekend and they are going to

work a lot this weekend. I'll be

tracking all of that. Rest assured,

nobody is going to change their terms of

services without me catching it and

reporting it on the channel. We're going

to have all of the coverage of this

moving forward. If you haven't already,

subscribe to the channel, click the bell

to be notified when new videos drop,

when new news drops on this, and also

sign up for the newsletter. You want the

facts and figures behind this, sign up

for the newsletter, get that in your

inbox. Thank you for watching.