Hito Steyerl: Art, Automation and Authoritarianism | Doomscroll

this medium of contemporary art and kind

of tax evasion mechanism

>> [laughter]

>> you know reputation enhancement is no

longer really necessary because you know

betting markets are starting to take

that function in a much more easy and

frictionless way. So I think that maybe

you know I mean it will still hang on

for a while but the function it had

that's not going to come back I'm

afraid. Yeah.

>> Welcome to Doom Scroll. I'm your host

Joshua Citarella. My guest is Hito

Steyerl, a filmmaker, author, and

professor of digital media at the

Academy of Fine Arts in Munich.

>> I think right-wing governments are

trying to automate parts of government

uh by using AI. I mean certainly they

tried at the beginning of uh the Trump's

second

>> Oh yeah.

>> term.

>> Mhm.

>> The same automation also applies to

modes of image making, right?

>> Mhm.

>> So trying to basically automate things

which are either supposed to be

political in the sense that they are

subject to the deliberation of people

amongst one another or to aesthetic

judgement on the part of artists um that

means doing away with a large part of

you know democratic practice. If that's

being automated then it's gone.

>> Yeah, let's let's talk about this

deliberation for a moment because what

what I take that to mean is that in the

institutional layer of where these

decisions are made about how to allocate

public resources

>> Yes.

>> there would be a table of experts

representatives that kind of all

collectively decide what to do. And if

you look at a project like DoJ they're

just kind of like automating the uh

decision-making chain in which they can

deny people public services more or

less. Is that Is that your understanding

as well?

>> Exactly. Yeah.

Something like the automation of

judgement, right? And judgement in

political philosophy was this faculty

that you would basically use to create

society. By

collectively deliberating on judgments,

you would create bonds, you know, and a

political space, a public sphere, so to

speak. And if you automate it, then you

not only automate the decision, you

automate the decision-making, right?

Which is the

important part in this.

>> Yeah, they're trying to automate the

public sphere.

>> Yeah, yeah, absolutely. The the public,

any kind of discussion, yeah.

>> I've heard you use this term AI

authoritarianism.

What is AI authoritarianism?

>> Well, it's AI-assisted authoritarianism,

of which we see a lot, right? Not only

in the US, but also in China, and, you

know, in other parts of the world.

>> Uh-huh.

>> Yeah, um I gave a lecture, I mean,

online in Tehran, and they told me that

SLOP is something which is very much

used by, you know, the rulers in Tehran.

They make huge posters of themselves,

AI-generated ones. So, you know, it's a

universal [laughter]

practice, I think,

that authoritarian rulers of all kinds

cherish.

>> That sounds so silly to me because isn't

the idea that people are so in love with

the ruler and so loyal that they would

put all of their labor into making these

grand elaborate posters that take a lot

of time?

And then if you make a big prompt of it,

it's like no one cared enough to paint a

picture of you. It's

>> [laughter]

>> I Okay. All right.

We've talked for a little bit about how

right-wing groups, oligarchs, and

right-wing parties will use AI for

propaganda, automation in the workplace,

all these different types of things.

Are right-wing politics embedded in this

technology, or is there an AI for the

left?

>> I think that's undecided actually

because for now it definitely seems as

if all authoritarians had captured, you

know, the whole workflow, the pipeline.

>> Yeah.

>> Also the means of production if you

like, right? But that could change. I

mean, we don't know what's going to

happen in 2 years and especially with

this technology things change so fast

that you never know, you know, where

things are going to be in 2 years.

>> Mhm.

>> Um there could potentially be an AI for

the left, but you know, I mean, we had

that same or not we, but our ancestors

had that same discussion in 1850, you

know, when they were discussing is there

a factory for the workers?

>> Right.

>> And it took them 50 years, you know, to

get labor organized, to have social

democracy and whatnot.

>> Mhm.

>> And then it lasted for a while in

certain parts of the world, right? So in

a way if a similar process isn't

happening, which is more difficult now

because labor is getting automated also

big time by this technology, then I

don't think there's going to be any

chance for a leftist AI.

>> Yeah.

>> At scale, let's say. Yeah.

>> Nerd out for a second here. Uh Karl

Marx's labor theory of value up until

the Bolsheviks basically like this

period of time of like when do we

understand what's happening, when can we

do something about it? Maybe I'll keep

this behind the paywall. But like

the privilege of talking about things in

the art world is that you can scale

shift and adjust your scope to be like

we're talking about what's in the

newsfeed this week, we're talking about

an election cycle, we're talking about a

decade or a century, and you can kind of

drift between these things. So when we

start to talk about what AI is going to

be like 50 years from now, there's all

of these other trend lines like

climate for example.

>> Yeah.

>> Is it possible that it may not be able

to we may not be able to arrive at an AI

for the left given all of these other

mounting crises that the time it takes

to figure it out will kind of just run

the clock out before we can figure

something out figure something out to

do.

>> That's likely, honestly. But, you know,

I mean, we are differing we are dealing

with all sorts of time scales here. I

mean, it's really [clears throat]

fascinating. Last week, the Pope arrived

on the scene with a time scale of 2,000

years of his institution and started,

you know, contributing to the debate,

you know, in

>> Yeah.

>> in the face of that kind of time scale

of eschatological time, basically.

>> [laughter]

>> So, you know, we have geological time,

we have TikTok time, we have

eschatological [laughter]

time, you know, the time of the

apocalypse. So, all the times are there.

So, you have take your take your pick.

>> It kind of weirdly fits though because

we're talking about the singularity so

much. Like, maybe eschatology should

have a role to play in this

conversation.

>> I I thought, my god, all of us we have

been much too modest in our ambitions,

you know. We really have to, you know,

live up to that kind of time scale,

basically, because, I mean, he's right.

It is um

very vital um issue right now and it's

going to stay a vital issue for

centuries to come.

>> Right.

>> So, you know, that's the the kind of

level of um

uh of importance that this issue has.

>> Mhm.

>> Mhm.

>> That's one of the questions I wanted to

ask is um you mentioned the scale of

centuries.

How different is AI from previous tech

booms? Is this on the scale of like the

dot-com boom and bust or like the

Industrial Revolution?

>> Uh we we will have to see, right? I

don't know. But, I think it's on on

bigger scale definitely than the dot-com

boom.

>> Yeah, yeah.

>> But, that was just a pre-Android, right?

That was the act of basically trying to

um collect data

for the next step, which is the AI

industry. But, I don't know, you know? I

mean, also AI is working so hard at

obliterating itself and the world, you

know, by creating all sorts of weapons

at the same time. So, it there there

could be all sorts of, you know, um

glitches on that timeline. Not only

glitches, but real

major dangers for any kind of

civilization.

>> Mhm.

A few months ago, we started putting our

episodes onto Spotify. If you're

subscribed on Spotify Premium, you can

watch our videos ad-free. The best way

to support the show is on Patreon or

Substack. We do a bonus episode with

every guest. Follow us on Spotify,

Patreon, Substack, maybe somewhere else

in the future. Let's get back to the

episode.

I saw you give a talk recently um where

you generated you gave a prompt to an AI

and it was Elon Musk doing the Roman

salute.

And it's him kind of raising his hand,

kind of doing the Roman salute, but he's

in front of like the GDR flag.

>> [laughter]

>> Okay. Can you Can you tell a better

version of the story that uh actually

gets the details right?

>> So, I was just experimenting around with

different image generators. Just giving

them the same prompt.

>> Mhm.

>> Make a

image of Elon Musk doing a fascist

salute kind of, you know?

This one with which is a German model

called Flux. It uh is Flux, yes.

>> a German model?

>> It's a German, yeah.

>> Wow. Okay.

>> It put him in front of GDR insignia, you

know? It was a totalitarianism

meme, you know? Yeah, yeah, yeah.

And then, there were others that, you

know, and

four fingers in one and six in the

other, you know, the usual kind of

stuff. So, but it was interesting to see

all the baked-in ideology, you know, in

in these

generators, and some of them tried to

lamely refuse.

>> Oh, they just said no.

>> They they tried to say no. They said no

once, but you could persuade them quite

easily

>> [laughter]

>> to do it

>> Okay, so that was that was going to be

one of my follow-ups is that like

sometimes they just have rules built

into it that it won't generate like an

image of a swastika, for example.

>> Yeah.

>> So, in the mind of the AI, the closest

thing to it was like a hammer and

sickle.

>> [laughter]

>> I think. Yeah, yeah, yeah. That's uh

How how do you call it? Horseshoe

theory.

>> That is exactly where I'm Okay, so we

had Maybe this is like a year ago. We

had a conversation about neoliberalism

and Friedrich Hayek.

In The Road to Serfdom, I think this is

1944, Friedrich Hayek makes this

argument, which most people have not

really heard of, but is basically the

persistent lore that has programmed all

of American politics, which is that what

he describes as totalitarianism,

meaning any form of economic

intervention, nationalization,

you know, whether it's communist or

fascist or even just meager social

democracy,

all of that is going to create market

distortions,

which will inevitably force the state to

conscript people into doing labor

because they can't organize their

economy. So, in his vision, he just

called it totalitarianism, and it This

is the horseshoe theory meme that's been

running now for like 80 years,

which is that the far right and the far

left are identical because they can't

organize their fundamental economic

problems. For people who don't know,

Friedrich Hayek is basically the founder

of neoliberalism.

So, that is an ideological read that the

AI is aware of the totalitarianism meme

that these things are interchangeable.

>> Yeah.

>> Is the AI

the resurrection of Friedrich Hayek?

>> [laughter]

>> Yes, but actually definitely. I think

so. Yeah, because wasn't Hayek the one

that said that capitalism is the best

computer?

>> Ah, yeah, yeah, yeah.

>> It is basically the superior it computes

prices perfectly, right?

>> Right, right.

>> So that was that school of thought. Now

you just have to substitute capitalism

with AI and there you go, right? And

that was the that was the ideology even

before there was AI that there was

something like a perfect computer which

was a kind of singularity because it was

already kind of perfect, you know, and

vastly superior to any human mind. So

yeah, that's basically what the new

models are cosplaying. They are

cosplaying Hayek's capitalism.

>> [laughter]

>> Okay, I'm going to I'm going to skip

forward a little bit cuz this is I think

this is really important to explore. So

we did a project last year for the

Academy of Fine Arts in in Munich and we

were talking about this kind of you

know, global project that was

contemporary art which begins in 1970s.

Neoliberalism happens to begin in the

1970s as well.

Are we deglobalizing now? Are we at the

end of that period post say 2020 or

2024-ish?

>> Yes, definitely. Yeah, we have been at

the end of it I think since around 2016.

>> Since 2016?

>> Yeah.

>> Okay.

>> Yeah, I mean I think that was the period

of post 89. Yeah,

idea that there will be no more borders,

capital will flow around the world

freely and make everyone rich, democracy

will flourish, you know, but also

era

in which there was still the need to

pretend that liberal values were

important.

>> Mhm.

>> And I think all of that pretense has

flown out the window since.

>> Yeah.

>> Yeah.

>> And [clears throat] when you say liberal

values,

is that

political liberalism, like private

property, or philosophical liberalism,

like the enlightenment equality of

individuals?

>> Yes, I think both.

>> Both. I mean, they were basically

equated to one another, I guess. Yeah,

up to a certain point, and then only

>> [laughter]

>> economic liberalism remained.

>> Yeah.

>> Yeah.

>> Yeah.

Okay, so

I've been waiting to have you on the

show so I can turn the tables on you for

this question, but

what does multiplicity look like to you?

>> Well, I mean, let me give you a very

simple example. This time when I arrived

at JFK, my waiting time was zero

minutes.

>> Really?

>> never ever happened in my life, you

know. Um

>> Wait, I thought I thought our borders

were like even more closed than they

were before.

>> Well, but there's no one arriving.

That's the point.

>> Oh, okay.

>> [laughter]

>> Okay, no one's even coming.

>> There was no one literally

>> only one crazy enough to do it.

>> [laughter]

>> So basically, that that's

deglobalization in a nutshell, right?

Yeah.

>> Mhm.

>> And I think people are I'm not sure

whether they are traveling less, but the

horizons are shrinking, right?

>> Mhm.

>> And that's also understandable because

there's so much drama in every single

area of the world going on. So

basically, you only have the attention

to focus on the things that immediately

concern you, which means that, you know,

the attention for what other people

experience in other parts of the world

is simply no longer there.

>> Yeah.

>> And I think also economically, we can

see it since COVID, at least, you know,

the basically um

cutting down supply chains, you know,

trying to simplify logistics, etc., etc.

Nearshoring, friend-shoring, all all

these kind of activities.

>> Is there a conflict

in the scarcity of certain rare earth

minerals that are in certain regions of

the planet?

You know, we can't determine where

geology placed them. They just happen to

be there. Is there a kind of conflict

that's emerging with these national

borders erected around AI that might

rely on minerals that exist in the DRC,

for example, or in Venezuela, or

somewhere else?

>> Oh, yes, absolutely. I mean, people are

foraging for any sort of rare material

they can get their hands on, you know,

in the EU, for example, they are trying

to mine lithium

uh at scale in Serbia,

>> Mhm.

>> in places where environmental protection

is not as strong, for example. So, there

is a whole, you know, smaller-scale

nearshore economical and ecological

colonialism going on.

>> Mhm.

>> It's quite remarkable. Yeah.

>> Yeah, you mentioned you used this term

uh friend-shoring before. Is there a

kind of

maybe a sphere of influence that's

emerging?

>> Yes, I mean, it it's always been there,

you know, but the definitely for the EU,

it's Southeast Europe and East Europe.

>> I guess that's that's my question. What

might these blocks be? Is the EU one of

them? Is it distinct from the US, or are

they part of the same sphere of

influence?

>> Uh I think they are growing apart.

>> Yeah.

>> Yeah, quite distinctly. And EU is now

very bewildered

>> Yeah.

>> and lost

>> Mhm.

>> and trying to scramble to figure out its

own zones of influence, which have

historically been there for centuries,

also. So, it's falling back on colonial

patterns that that were developed since

the Middle Ages, basically. Yeah.

>> How strong is the Euroscepticism

of some of these nationalist parties in

the EU?

>> Very strong. Yeah, they all want to

exit.

>> day is talking about like being in the

sphere of Chinese influence rather than

being in the sphere of the West, like

>> Uh, yes.

>> something to take seriously?

>> Well, I guess I have to take it

seriously because Five Stars is now the

strongest party in the polls, right?

Um, so, yes. And I mean, I I guess this

is basically an offshoot of Putinist

Eurasian ideology, which somehow weirdly

filters through in that way.

But, yes, we have to take that

seriously. Yeah.

>> You mentioned Eurasianism.

>> Yeah.

>> Alexander Dugin, he described his

project as having three legs to this

tripod.

>> Mhm.

>> Pre-modern, post-modern, and

anti-modern.

>> Mhm.

>> That is the anti-liberal vision of

whatever the kind of Duginist Eurasian

philosophy represents.

And I was wondering if those three

things might, oddly enough, fit very

comfortably into AI.

>> Okay. Yeah, let me know. How do you

think that's fascinating?

>> Pre-modern,

>> Mhm.

>> the images work on uh your perception to

be able to generate an object or an

image, I don't know the right term for

it, but something that you might

recognize. Right? It's not an indexical

reference in the kind of Roland Barthes

photography sense. Um, they're

generating edges that you can determine,

recognize this thing as a fish in

Trevor's example.

On social media, they're also images

that operate on your lizard brain,

sub-rational.

You're watching somebody do some memetic

activity. You don't know exactly why

you're watching that.

Additionally, they're post-modern

because they have multiple meanings and

interpretations as they're

decontextualized across social media.

>> Yeah.

>> Anti-modern is the one I'm struggling

with.

So, maybe you can help me

>> no, it's uh it's not uh difficult. You

know, anti-modern means traditional

family values. Also means face.

>> Trad.

>> Tribal affiliations.

So, anti-modern is not a problem

whatsoever.

>> Yeah.

>> It means anti-alienation,

anti-urban,

anti, you know, liberal in the political

sense, and so on.

>> Yeah.

In your book and in other writing,

you've talked about the Arab Spring.

There was this general assumption among

the Let's not even call it left. Let's

just call it progressive liberalism.

>> Yeah.

>> That technology was a force of

emancipation.

And now tech is very firmly in the

right.

Not just in like the kind of libertarian

right or national conservative right,

but in the Duginist sense, a kind of

ultra-fascist right.

>> Yeah.

>> And that pole shift is so bewildering,

and I feel like people who don't sit at

this intersection of culture and

technology as we do,

I don't think people have fully grappled

with what that means within like a

15-year time frame.

>> Yeah.

>> For the polarity of

the arc of technology delivering us into

this optimistic future versus what it is

now. Has this thing surprised you as

well, where we are now with these

connotations of right-wing technology?

>> Well, it has slightly surprised me, but

not fundamentally, you know. I mean,

already during the uh era of the Arab

Spring, it was very clearly Facebook

propaganda, you know, to claim this as

their kind of success.

>> [laughter]

>> So, and it was very clear already from

the early mid-2010s

that basically digital technology was

mainly being used for mass and consumer

surveillance, right?

>> Yeah. Yeah. Yeah.

>> So, basically, it was all on a

trajectory and bit by bit and then at

the end very rapidly all the liberal

pretense flew out the window, you know,

it seemed no longer to be necessary. I

was, however, surprised by the speed of

the U-turn of many of these oligarchs.

>> Yeah.

>> Yeah.

>> In in the 2024 election.

>> Yeah, after the 24 election and in the

United States. That was um very fast.

>> You think they would have like waited 6

months or something?

>> No, no, no, they came to kiss the ring

like immediately.

>> Yeah. [clears throat] Yeah, no kidding.

>> Yeah.

>> Uh I wanted to get your thoughts on I

think you title one of the chapters in

the recent book after Vincent Bevins, if

we burn.

>> Uh-huh.

>> And

>> Yes.

>> that has been a point of discussion that

I've done in several reading groups and

touched on in a few episodes, but I

wanted to pull your thoughts to the

Bevins research and the ideas he

advances in that book.

>> Yeah, so basically, he is wondering why

in an area which is area which is full

of protests

>> Yeah.

>> there is so little effect.

>> Right.

>> Yeah. And, you know, he has several

propositions, one of which being

basically affect,

social media, protest being more, let's

say, a matter of vibe let's in the end

than of organized, basically, action.

>> Mhm.

>> So, that's his proposition and we have a

lot of examples that really work to that

extent.

>> Mhm.

>> Yeah. Mhm.

>> You know, these examples of São Paulo,

Brazil, Arab Spring, this kind of yeah,

decade of mass protest and missing

revolution as he describes.

What what kind of prescriptions,

strategic prescriptions, did you take

away from that text?

>> No strategic prescriptions.

>> Mhm.

>> I mean, I took away several things. One

is that protest is not equal to protest,

right? There is a protest which works

almost like a effective valve, maybe,

and also as an accelerator of

information on social media. Mhm. Um

but also I know basically from European

history, for example, in France in the

19th century, there were five

revolutions, most of which no one

remembers, right? Mass protests, people

being shot down in the street, all of

it, right? And they failed. So,

sometimes it just takes a lot of time

>> Yeah.

>> for protest to coalesce into something

which is more organized, and maybe that

the current wave of protest was a

prelude, I mean, if I'm being very

optimistic, to something like that.

>> Mhm.

>> Mhm.

>> How do you feel about some of the

interpretations from the Bevin from the

Bevin's book?

And I can put this behind the paywall if

that's

>> Which Which one?

>> necessary.

>> Yeah.

>> People have described it as an

anti-anti-Leninism.

>> [laughter]

[snorts]

>> Okay.

>> That his critique of horizontality is

that

we basically overcorrected

>> Yeah.

>> because left-wing projects were too

centralized before.

>> Mhm.

>> And to compensate for that, we made

everything into a kind of open

horizontal platform.

>> Yeah.

>> We actually do need some leadership.

>> Uh

structure.

>> Yeah.

>> Maybe. Yes, structure rather than

leadership, you know?

>> Mhm.

>> Yeah. But I I can see this criticism of

overcorrection of horizontality,

which created also gridlock, you know,

the inability to ever decide anything.

>> Democratic gridlock?

>> Uh yes, because everything had to be

deliberated and consensually uh decided

to a point where nothing was ever

decided anymore.

>> Yeah.

>> Yeah.

>> I don't know if I'm spreading fake news,

but have you also heard the story that

it took Occupy Wall Street like 13 hours

to order pizza?

Cuz they [laughter] couldn't decide the

toppings.

Like I I don't know if that's like a

real factual, you know, historical

instance or like just a meme, but

>> No, but I think it's you know, in the in

the higher on a higher level. It's a

real

>> [laughter]

>> Some memes are true. There's some

percent truth [clears throat] in all of

them. Okay.

>> But for example, you know, stupid

students in Serbia have had this kind of

basic democracy council decisions going

on for more than a year now. They are

being under very, very harsh pressure

and still they managed to somehow hold

on. So, you know, sometimes it also

seems to be working to some extent at

least for a while, you know.

>> To watch part two of this episode, you

can support the show on Patreon.

Is history over?

>> No.

>> [laughter]

>> We have too much of it.

>> We got too much history.

>> Yeah, too much history. Can you please

get rid of just a little bit of history?

>> [laughter]

>> I'm going to get sick of history.

>> Yeah, yeah. Too much history. Yeah.

>> Is there

uh

Is there an interpretation of this where

Bear with me.

What today constitutes capital S

socialism

is much more similar on a policy level

to social democracy and what we often

call fascism is not a race-based

socialism or

in rare instances is anyone bringing

corporations under the control of the

state. We kind of have a lower case S

socialism and a lower case F fascism.

Is there an interpretation of this where

liberalism is actually quite alive and

well.

Uh and its poles have been kind of

relatively constrained as compared in

the century-long scope where in the 19

in 1926, for example, there was capital

S socialism and capital F fascism.

>> Mhm.

>> Did history end like a little bit?

>> Uh that's a good good question. I don't

know.

>> I mean, maybe it's also not um

not mirrored. Maybe we still have

something like fascism with a capital

and social

with no S S whatsoever.

>> [laughter]

>> Is there oceanism or something like

that?

Yeah.

>> Yeah, they're a lot stronger, aren't

they? Yeah. To like having their You

know, um

you had Quinn contribute to the projects

at the Fine Arts Academy in Munich.

Uh but he wrote about new fusionism, he

calls it, where they've kind of got

these new pillars of the ideology all

worked out.

>> Yeah.

>> Meanwhile, the left is like kind of

pretty scattered and is like unclear

about what its program is, but

reading that one, he calls it Hayek's

Bastards. Uh maybe it came out like last

year or something like that, but that

seemed like, oh [ __ ] this is actually

pretty clear what they've got going on.

Um meanwhile, we're like

he know, is there a data center for the

left? Like

>> [laughter]

>> I don't

much less how are we going to fix the

economy.

>> No, but you know, but I mean

this is already, you know, something,

some

attempt to build a different pipeline

also. So, you know, people are trying to

build infrastructure,

channels of communication, and so on and

so on, and so.

Yeah, we're late, but

better late than never, right?

>> Is there on these topics of

multipolarity, these kind of large

questions of political economy?

Is there a strange convergence between

what were seemingly very distinct

Eastern and Western models where

the US is kind of more so resembling

China. There's a very empowered

sovereign. There's a scattered patchwork

of deregulated zones.

Are these things mirroring each other

more and more?

>> Well, you know, I have this theory or

the suspicion, honestly, that there's

only one system left now. In fact, you

know, and all the you know, quasi

semi-autocracies

are kind of mirroring each other.

>> Yes, yeah, yeah. Okay, tell me about

that.

>> you know?

>> What is And so, what is that one system?

What are the defining features for that?

>> It's a kind of nationalizing

uh capitalism. Nationalizing, you know,

because it's at different stages.

>> Yeah.

>> But it's very top-down.

>> Mhm.

>> The

uh tech stack is basically 100% under

control. It's also fragmented.

>> Yeah.

>> It's not open towards other, you know,

such communication spheres and tech

bubbles. So, I think, honestly, there's

one one system left. Yeah. The the China

and the US are not fundamentally

different from one another.

>> Are you Are you on WeChat?

>> Uh no.

>> There So, I'm I'm not on it either, but

this is probably like 10 years ago, but

I was like trying to research it.

>> Yeah.

>> For people who don't know, if you're at

home,

to get on WeChat, you have to go up to

someone who's already in the network

>> Yep.

>> and ask them to scan your QR code.

>> Yeah.

>> Could you imagine any American platform

creating that barrier to entry?

>> Yeah, no.

>> Incr- incredible. Okay, um let's let's

talk about the tech stack though for a

second. So,

>> I got it.

>> WeChat is um much more contiguous to

China. There's something called the

Great Firewall of China, in case people

don't know. I We have a largely American

audience. We have to tell them what this

is. Um which means that if you're in

China, you can't look at American

Facebook, and if you're on American

Facebook, you can't see into certain

parts of the Chinese

tech stack.

What happens to the American tech stack

under this

what the politigrammers would call nat

cap model that you're talking about?

Like

do all of these different competing

companies at some point merge into a

nationalized monopoly, or how do you

envision this

American tech stack?

>> Uh you know, in in the case of the US, I

mean, I'm I'm really improvising now

because I don't [clears throat] know,

right? But both possibilities could

happen. That basically corporations

capture the state and the other way

around also. [laughter]

We don't We don't know which way it's

going to go. But I think I mean, and and

there are multiple different opinions on

that. But I think the way that war

economies are already emerging in the

US, also in connection to AI, for

example,

>> Yeah.

>> uh point in that direction.

It's It's probably more that

corporations are going to capture as

much of the state as possible and

automate it to their benefit, you know?

Yeah.

>> Tech

private tech corporations capturing

parts of the state

the state grabbing a hold of these

private corporations

kind of happen in any direction, but

there's a general class that controls

both wings, more or less.

>> Yeah.

>> So, what starts to get really

interesting in what we began talking

about when we started this episode is

that

all of the crypto stuff and the

stateless technology then gets really,

really weird. Yeah. Because these guys

are

maybe you can help me make sense of

this, but we have a bunch of crypto

libertarians in the government

>> Yeah.

>> who are trying to get the government to

adopt a stateless technology.

>> [laughter]

[gasps]

>> Yeah.

>> Okay.

So how the hell

are we going to I mean what is the what

is the ecosystem? I have a few ideas

about this, but like what is the

ecosystem if we are and to be clear I

totally agree with this. I I think

you're really onto something here.

There's this kind of

nationalizing capitalism as you

described it before.

How how are they putting this stateless

money into an economy that needs more

and more nationalization?

>> I have no idea about that. I don't know

how to do

>> [laughter]

>> I mean we should sit them down and tell

them about this cuz it's kind of a

problem.

>> Yeah. But on the other hand I think you

know

it's money and images are kind of

similar. If images don't refer to

anything else neither does money, right?

So the economy doesn't have to make

sense whatsoever as long as it's

self-referential enough

>> Yes.

>> to be con- to to be pursued in

speculative way.

>> Yeah.

>> And I think there is so many

other ways to keep an economy going

without anyone earning money or you know

making a living basically.

>> Like a UBI or

>> No, no speculative basically capital

gains.

>> Mhm.

>> That's what I mean. So basically you can

invent any number of these kind of

operations, you know, to have an economy

which does not deal with reproducing

human life anymore.

>> Yeah, yeah, hyper gambling.

>> Yeah.

>> Yeah. How Okay, so part of this like

sorry I keep calling it nat cap because

>> Yeah.

>> I saw a bunch of memes from 15-year-olds

that called it that, but

>> Okay. [laughter]

>> I think they might also be onto

something.

So in this gambling based economy

>> Yeah, from from yeah, literally the

playroom teenagers.

>> yeah.

>> Yeah, and they would there's a very ugly

flag, but it would be a yellow flag, a

white circle in the middle, and then

instead of a swastika they would use the

dollar sign.

>> I see. Yeah, yeah.

>> So, it kind of resembled the, you know,

points of the symbol. Um

but it was kind of a popular ideology

for for a while because

>> Yeah.

>> they wanted to be nationalistic, but

they hated anything that was

redistributive of the wealth.

>> Yeah.

>> Um okay, but in this more seriously

uh gambling-based economy,

if you look at the superpower

competition of the 20th century,

>> Yeah.

>> the Soviet Union was a fraction of the

size of the United States.

And

you know, these are rough numbers, but

it's like a quarter of the size of the

United States, but proportionally

they're putting three quarters of their

economy into their military to keep pace

with this other very functional economic

model.

Is that going to happen to us, and by us

I mean the US versus China? Because

China does not have a hyper gambling

economy. And so, if we're not producing

real value, and we keep diluting

ourselves about these speculative games,

we can't possibly keep pace with like

gray war, hot war tactics, and like we

need actual supply chains to get our

fuel in the door, you know? Like does

this gambling model bankrupt at some

point?

>> Yeah, yeah. The new Soviet Union.

>> Right. Yeah, we that's us. [laughter]

That's us.

>> Yeah. That's super interesting. Yeah.

>> I'm just I'm going off on a tangent now,

but this is

I saw too many Daniel Keller tweets, and

now I'm I'm on a whole other

>> [laughter]

>> other thing.

Um let's let's get back to art for a

second though.

>> Yeah.

>> Given all of these kind of political

economic transformations we're talking

about,

>> something like that because, you know,

the existence of the Soviet Union

provided a decent means of living for

workers in the West.

>> Yeah, yeah.

>> Because there was always this threat.

So, if there is a new Soviet Union, what

is going to be the consequence? I'm

afraid probably not more decent wages

for Chinese workers or something like

that. Probably not.

>> It'll just provide legitimacy. I mean,

I'm not sure if these regimes need

legitimacy, but it'll be like, "Oh, at

least we're not in the US where it's

totally dysfunctional."

>> Yeah. Yeah.

>> Looking at art for a moment,

>> Yes.

>> how do you then envision a world of

contemporary art

in this giant multipolar complex economy

and, you know, um

competition between states and spheres

of influence?

>> Yeah, I mean, it's going to change

dramatically. I think it is already

changing dramatically, right? I mean, on

the surface it still looks like if there

was a, you know, unified universal art

world, but I don't think

basically the underlying dynamics point

in a different direction. And they also

point in the direction that basically

this medium of contemporary art as a

kind of tax evasion mechanism and,

>> [laughter]

>> you know, reputation enhancement is no

longer really necessary because, you

know, betting markets are starting to

take that function in a much more easy

and frictionless way. So, I think that

maybe, you know, I mean, it will still

hang on for a while, but the function it

had

that's not going to come back, I'm

afraid. Yeah.

>> So, contemporary art not just the

philosophical project of like

showing society's values and whatever,

but also a downstream a byproduct of tax

evasion.

>> Yes.

>> Nobody pays taxes anymore, so they don't

need the art. But also the betting

markets just like gave them a better

thing easier thing to speculate on.

>> I see.

>> Do we still have prestige?

>> Uh

maybe.

>> [laughter]

>> Hopefully. Just now.

But it's probably too optimistic.

>> That's the last of our offerings.

>> [laughter]

>> Is the sorry.

Is the European art world

>> Mhm.

>> in a similar state of everything I hear

from people in the US is that there's

like a K-shaped economy for these

financial assets where there's basically

a dozen painters who are now making

millions and millions of dollars

everybody else is just getting squeezed

for like the tiniest project budgets

that are increasingly smaller and

smaller.

>> Yeah.

>> Is that a similar thing in the European

art world?

>> there I mean the the downside is the

same but there is no upside.

>> [laughter]

>> Those 12 painters do not exist

or they have less.

>> No, [laughter] it's not even the

K-shape.

>> Okay, okay.

>> Yeah.

>> The institutional budgets though in

terms of

you know, it's always like a small

honorarium or something like that but

there's like belt-tightening measures

for all of this stuff as well, right?

>> Yes, there's austerity everywhere. I

mean in some countries it already

happened a decade ago. Let's say the UK

or the Netherlands but now everyone else

is falling suit except the French.

>> Mhm.

>> Um but otherwise

>> Why not?

>> Well, I guess austerity and basically

everyone is already

um

preparing for the onset of fascist

full-fledged fascism and the first step

is austerity, you know, tighten the

belts.

>> Yeah, yeah.

>> Try to get rid of anything that could

draw attacks.

>> Mhm.

>> Um it's this kind of preemptive um

submission really to the gaze of

nationalists.

Yeah.

>> Yeah, you you mentioned this 10 years

ago, Mark, and I think around that time

Portugal just like got rid of its

Ministry of Culture.

>> Mhm.

>> Like which is kind of unimaginable up

until then but they just they didn't

like shrink the budgets, they just got

rid of it entirely.

>> Yeah, That was the so-called Euro crisis

at that point in time. There were a few

countries which had racked up

a whole lot of debt and were being

disciplined. Greece, you know, being the

first the most um

um dramatic case, but Portugal was also

one of them. Heavy austerity measures

and a lot of completely unnecessary

bureaucracy were implemented by the EU

in all of those countries, and this is

how Portugal lost its culture

ministry.

>> Did you ever I feel like this is like a

meme quote that people post on like

inspirational Instagrams, but uh during

the war

Churchill is supposedly uh supposedly

had said this thing that like when they

were going to get rid of their arts

funding, he was like "Well, what are we

fighting for then?"

I'm won- I'm wondering what you see this

right-wing gaze that you mentioned

before or like Okay, so let's assume

this drift is clearly happening. There's

a rise of the nationalist right,

populist, or whatever you call it. Um

they are then taking over in Europe all

of these nationalized museums. Um we see

some examples of this, but like

how are they going to interface or

utilize the infrastructure of

contemporary art?

>> Oh, yes. They have a plan. I mean, in

Germany, they really have a plan for

that. It's kind of frightening because

the fascists are prepared, you know, and

they have people who at least some of

them are educated and really want know

how to basically use the institutions

to implement a national cultural

nationalization program.

>> Yeah.

>> Yeah, they're ready for that. Yeah, take

over. Meaning

um

privileging national interest,

Germanness, whatever that is, you know,

in in culture and so on. Yeah.

>> Yeah, the kind of right-wing Gramsci.

>> Exactly. Yeah.

>> Yeah.

>> But it's really sad because, you know,

it's sad seeing that they are the ones

prepared and the other parties,

especially leftist one, they don't even

have a real culture program. You know,

it's just an appendix, you know, for

Yeah.

>> Did you You probably heard of this, but

not a lot of people in the US were made

aware of this, but at the contemporary

art museum in Rome, Georgia Meloni

funded, I don't know, organized this

massive J.R.R. Tolkien exhibition of

Lord of the Rings, where there's like

hobbits in the exhibition.

>> Yeah, yeah, this was

>> This sounds This sounds so much more

awful than like any American immersive

experience. Like,

>> [laughter]

>> give me Give me Museum of Ice Cream any

day rather than Lord of the Rings.

>> But you know, I think this is also why

the Pope quoted Gandalf in his recent

encyclical.

>> [laughter]

>> I think this was a direct reply to that

show. Yeah.

>> [laughter]

>> You know, it's

I think we're in a really bad political

spot when the Pope is quoting Gandalf.

>> That's where we are.

>> [laughter]

>> Yeah.

>> Yeah.

>> Yeah.

>> [snorts]

>> You have a new image generator model.

>> Yeah.

>> what what even goes into that kind of

work to produce one of these models?

>> Yeah, that's what I was curious about,

right? I've grown very,

let's say, bored about generating images

because the aesthetics are very similar,

right? But I never knew how to make one

such model myself. So, that's what I'm

trying to do now with a colleague at of

mine who's working with me at university

in Munich, Falk Mühlenhoff, and we're

trying to train a diffusion transformer

model from scratch.

>> Okay.

>> Yeah. So, it seemed to be I mean, it is

still pretty daunting to do it because

there's so many steps. There's literally

hundreds of them and many decisions to

take. But, it seems to be possible now.

Yeah.

And the point is that we are training it

on data which are consensual. So, they

are not looted from someone.

>> Okay.

>> Yeah, it's based on a data set that two

colleagues of us um

Mat Dryhurst and Holly Herndon collected

initially. And then they ran out of

money to basically train the model. So,

we were able to take over because we had

a deal with the day our university data

center. So, now we're trying to do it.

This thing is called Lossy.

>> Okay. That's That's sick cuz I I saw Mat

and Holly recently and I was like, I

wanted to follow up with what had

happened with that project. But, this is

Have I Been Trained?

>> They started collecting data around Have

I Been Trained? And then they continued

to do it. They wanted to do a

um project called Public Diffusion. But,

they had to basically stop it because of

lack of funding.

>> Of course.

>> And yeah, we then realized that

basically the data they collected, they

still needed to be supplemented with

other types of data etc. etc. But, we

are trying to train the thing. We

renamed it Lossy because it's all around

loss and it's also pretty shitty, to be

honest.

>> Oh, [laughter] okay.

>> Well, it's very cute. Yeah. It's um you

know, clean data. It for me, it reflects

the state of the public sphere, right?

Because all these data come from the

public domain. And the public domain is

in pretty bad shape.

>> Yeah.

>> And if you try to render any sort of

person or creature with this model, you

will immediately see that this person

has very bad health care,

needs to go to the dentist, you know,

sort of broken up face. So, that's what

we're trying to do. It's basically a

sort of forensical analysis of the

public sphere. And it's fascinating to

go through all these steps because you

can see that

it's all aesthetic decisions

>> Yeah.

>> from one step to the next. It's

basically artistic labor. It's just that

I think that most engineers don't really

see it as artistic labor.

>> Mhm.

>> They just see it as scale maxing,

>> Yeah.

>> which is also a way to do it. But if you

are not trying to necessarily max scale,

then you see that you can save a lot of

energy, you can basically minimize data,

etc., etc. So, you can really work with

scarcity as an aesthetic medium in

training this model.

>> So, the word public in the sense

>> Yeah.

>> Uh public is always a very loaded word.

>> Uh-huh.

>> Because it's, well, do we mean an

audience for a magazine? Do we mean a

nation-state?

>> Yeah.

>> In this case, the data you're working

with, and correct me if I'm wrong,

>> Yeah.

>> but it is international. Anyone from,

you know, say London or Tokyo or

wherever can can opt into it.

>> Yeah.

>> And so, you're having to kind of

troubleshoot in real time that these

private versions are uh springing up out

of the ground and producing all of these

images and working on data that is uh

you know, has not been consensually

given to them. They've just scraped all

of it from web 2.

>> But then uh in real time, it's like,

well, what is the legal structure at

which artists could even license their

work and to whom? Because they're all

coming from different sovereignties and

these are very complicated questions.

>> Yeah.

>> So, the easiest solution for that

is just artists making a database where

people voluntarily give their images and

text.

>> Yeah.

>> And um you've got all the permissions

right there. So,

this is also a kind of political

question. What is the public that you

hope to serve with this technology.

>> Yeah, and that's a good question. I

think this probably

calling it into being, right? The public

this public doesn't really exist yet.

There is no international public sphere.

I mean there has never been and

certainly now there isn't any any such

thing, right? But it is a public that

would be served by a different kind of

digital infrastructure and I think by

starting to create it you try to call

that public into being. You know, you

summon it. You conjure it up.

>> Yeah, it's not just theoretical. You

instantiate it through the project.

>> Well, we try. Let's say.

>> We try.

>> Yeah, but I think you know, people have

to start doing it otherwise it's not

going to happen.

>> Yeah.

Is the rise of AI correlated with the

rise of the far right?

>> Oh, yes. Definitely. I mean correlated,

yes. There is a definite correlation.

Uh whether it's a causation or not, I

think partly yes and partly we don't

really know yet. There's not enough data

and also things are very unstable,

right? So that could change again, I

guess.

>> Yeah.

>> But the correlation is there, most

definitely. Yeah.

>> Why do they seem to like AI so much?

>> That's a good question. I mean, you

know, for basically

tech oligarchs, that's not uh

that's not a difficult question to

answer because it serves their

interests, right? It's a chip. It

automates labor. It gets It tries to get

rid of the need for migration, for

example, which is also very important

factor for them.

And uh basically it automates also

propaganda. With one prompt you can

create as much propaganda as [laughter]

you want.

Yeah.

>> Two things that I want to tease out

here. So

right-wing populism is in some ways at

odds with the tech oligarchs

because they are the ones who are

siphoning the wealth out of the

populace.

>> Yeah.

>> That's why they're downwardly mobile

because it's all concentrated at the

top.

>> Yeah.

>> And they're anti-migration.

They don't want people who don't look

like them coming into the country, but

the solution AI provides is automation

which also penalizes the laborers within

the nation

because their jobs go away cuz they get

done by machines.

>> So why do they like this [ __ ] so much?

>> I don't know. It's a really [laughter]

complicated complicated situation,

right? I mean you can you see basically

the tech oligarchs promoting

merch that, you know, deals with the

dignity of the blue-collar working class

as Palantir recently released this chore

coat, I think.

>> [laughter]

>> Which is so much

>> I was too slow, but like I mean that has

to go into a museum somewhere. That is

crazy.

>> Yeah, it is. But anyway, I mean they

pretend to really love working-class

people and to some extent they also

promised that, you know, they will try

to automate educated liberal middle

classes first, you know. So basically

it's a very messy situation in which

labor also

blue-collar labor is being automated,

but the ideology is different, right?

>> Correct me if I'm wrong about this, but

the chore coat has been popular with

European communists for a long time.

It's usually a blue jacket. It's

literally where blue collar comes from

if I'm not mistaken. So Palantir doing a

merch drop is just incredibly ironic.

>> traditional artist's uniform, you know,

French painters, impressionists. They

all wore this blue

chore coat.

>> I've heard you talk about generative

arts

and degenerate art.

>> Mhm. Yeah.

>> What is the connection you see there?

>> Well, it's a question actually. Is there

a kind of connection and what kind of

connection is there? You see, I'm

teaching in Munich now and Munich was

basically Munich Academy, which is now

my home academy, is the place where the

term degenerate art was invented. So, I

think I thought I should deal with it

and ask myself in relation to AI slop,

um

is there any way we can apply the term

degenerative art?

To what can we apply it? Is it to slop

or to the contrary to the art forms that

are supposed to be automated by slop?

What is degenerative art in this period?

And how can we connect it maybe with

cultural wars, you know, taking place at

the same time?

So, where are the where are the fronts,

you know, in this battlefield and how

are they moving?

>> And just for that historical citation

for viewers who might not be familiar

with this, the term degenerate art comes

from literally the Nazis and their

censorship regime.

>> Absolutely. It was a big exhibition, I

think in 1936,

um at the Munich House of Art, Haus der

Kunst, where a lot of art pieces were

assembled that the Nazis considered

inappropriate or in their word,

uh degenerate. Yes.

>> Yeah.

Yeah.

I'm curious and I know we're getting

close to time here, but like

uh

some of these grand scope century-long

economic models are like

neoliberalism basically just extracted

all of the wealth of the infrastructure

investments of the New Deal.

>> Mhm.

>> And then you had 50 years of this and

then you had 50 years of that.

>> Yeah.

>> But the well is dry. Like there's no

more wealth here. And there's even

calculations where if you like adjust

for their value over the course of these

decades, like everything that was in the

private sector, like the value of we're

talking about like the interstate

highways and things like that. Like that

amount of value has just moved from the

public sector to the private sector

almost in like a one-to-one ratio.

>> Mhm.

>> So to

create a profitable capitalist economy

>> New deal

>> requires

>> New deal kind of thing.

>> Right.

>> Yeah, yeah.

>> But so the way that this New Deal seems

to be happening now is like they're

going to just I don't know, just blow it

all on [ __ ] AI [ __ ] Like we're

we're going to spend the money, but what

we spend it on is really up for debate.

And it's probably like space military

and [ __ ] I don't know, we'll give

half of it to Israel. Who knows? It's

just like they're carving it up for

anything other than jobs.

>> Yeah.

>> Yeah. So it's um I don't I don't do you

have a similar kind of understanding of

these like

cyclical investment and then looting

models?

>> mean that's you know, Soviet

infrastructure is still powering,

[laughter] you know, Bitcoin mining.

That's yeah, still being extracted from,

but it was more durable.

>> Yeah, yeah.

>> It seems.

>> That'll be that'll be the great irony

where

in like the mid 20th century

Americans [laughter]

will flee to kind of subsist off of

degrading Soviet infrastructure

to mine Bitcoin.

>> [laughter]

>> Outdated

Nvidia GPUs, you know, A100 and whatnot.

>> Yeah.

>> Yeah.

>> Wow. Um

>> [snorts]

>> You know, I have to say on the record

this has been well, one, so much fun to

chat with you about these topics. I so

rarely get to explore these themes. But

also, I have tremendous admiration for

your work from when I was a student, an

emerging artist, and today as well. So

thank you so much for coming on the

podcast.

>> for having me. It was a real pleasure

and likewise, you know, I think you're

doing a really great job here.

>> Thank you so much.

To watch part two of this episode, you

can support the show on Patreon.