Welcome to the Society of the Psyop

0:01What are we even doing now? Like, what does it mean to be making images or making art in a culture that is dominated by feeds, that is dominated by forms of media that are designed to manipulate you?

0:21Trevor Paglen:I'm Ben Davis, and this is The Art Angle, a podcast from Artnet News.

0:32Trevor Paglen:It is a time of disorienting, sometimes surreal shifts in the worlds of politics and technology. There aren't too many artists you might trust as a guide, but Trevor Paglin is one of them. Paglin has long functioned as an artist-researcher, with a fascination for pushing to the outer limits of what is seen or known. His work has often not just responded to, but anticipated major conversations around government surveillance, conspiracy theory, and the politics of Silicon Valley. And he has an important new book just out from Verso, How to See Like a Machine, Images After AI. It collects his writings on how the world of culture is mutating, exploring the secret networks of power that we can't see, but that govern how we see.

1:17Trevor Paglen:In addition, Paglin also has a new art piece, Voyager, currently on view in a show called Strange Rules in Venice. In essence, it is an artwork whose medium is hypnosis, Paglin has invented a device that guides you into a trance state to permit you to have otherwise impossible artistic experiences. I myself, jacked into Voyager and for a spell, was unleashed from gravity and levitated in the air, in my mind at least. Paglin joins me today to talk about how he's thinking about the forces reshaping culture now and also how this new type of art that he has dreamed up works and how it fits in with that theory.

1:59Trevor Paglen:Well, Trevor Paglin, thanks for talking to me today. So happy to be here with you today, Ben. Thank you. And where are you? I mean, in my head, you're always in the desert on some kind of mission. Where are you right now? I am currently in Northern California. I'm doing a few things out here. So I want to talk to you today about this book, How to See Like a Machine, which is a collection of, what is it, like 10 years of thought on your part? More than that? Yeah, about 10 years of essays. I mean, updated, but it collects all your thinking. And I guess before I talked about what's in the book, for a book of fairly esoteric thoughts about art and technology and politics, it has generated a lot of conversation.

2:39Trevor Paglen:Are you surprised by that at all? Yeah, I was totally surprised at how well it's been received and how many people seem interested in it. You know, I thought it was a pretty esoteric book for, you know, maybe some niche grad students. But yeah, it's been surprising to me how much interest there's been in it. Well, it's not, I guess, esoteric in form so much. It's very readable. But I mean, the subject matter is far out. I guess these are far out times. So people are looking for a guide. The text in the book, they are about everything from machine learning to brain imaging to chaos magic and UFO photos.

3:16Trevor Paglen:Is there a master theory you're working on? What ties them all together? Yeah. The thesis of the book is that we're sort of in the middle of two big revolutions that have happened in the history of images. If we think about the history of visual culture, we can point to moments like the invention of perspective or the invention of photography, kind of moments that really changed what an image was and by extension, how we see. And the argument of the book is that in the last 10, 15 years, we've undergone not one of those, but two. And the first of those would be the development of computer vision.

3:56And the second of those would be the development of generative AI, generative images.

4:02Trevor Paglen:That's something I really appreciate about having 10 years of your thought, because I feel like there's so much hype about generative AI right now, that kind of the first one almost gets skipped over. And it's the two together. Actually, the book, to me, makes the case they add up to a lot more than either on their own. So yeah, we're talking about these two revolutions in our relationship to images and literally in the metaphysics of images. First of those is computer vision. Second of those is generative AI. Now, technically, these are kind of like two sides of the same coin. You can build a computer vision system to be able to recognize a face, for example.

4:37And then what you can do is essentially run that thing backwards, and it will generate you an image of the face. However, I think culturally and in the kind of world of visual culture, these function sort of differently. When we think about the metaphysics of images, just like what is an image, the computer vision one is a really big moment in just thinking about what an image is. And the reason for that is that they're invisible.

5:09Trevor Paglen:You mean what? I mean, it's an image, right? Yeah. So historically, if you ask somebody on the street, what is an image? They might say, oh, it's a thing that you look at and you kind of have an experience of looking at that thing. And that's what an image is. It's essentially a relationship between a visual stimulus and your experience of that stimulus. And with computer vision, let's take a simple example, like a car that might be on a toll bridge that's taking pictures of license plates and reading those and then invoicing the people that drive through that. So that's a camera. It's making an image.

5:42It's a digital image. It's being read by a computer, by a machine. And then it is activating, you know, a billing system, for example. There's no human in that loop. And even though essentially what that camera is doing is a form of photography. There's literally nothing that you can look at, right? And so that's just weird, kind of like from a metaphysical standpoint or whatever you want to call it. And then on the other hand, it's really weird with the generative thing. We've entered a world where, you know, we essentially have things that look like photographs, but there's no photographer, there's no camera, things that look like artworks, but there's no artist, obviously, texts without authors.

6:23So there's something really weird going on there as well. And what I'm trying to do over the course of the book is kind of think through how to think about these revolutions. And what I kind of pretty early on recognized about this new visual culture that we're living in is that the tools I was taught to use in order to think about images, you know, in art school or what have you, we're like really not up to that task. It doesn't really make sense to try to use, you know, Susan Sontag or Roland Barthes to understand how a quality assurance computer vision system works or a policing computer vision system works, or for that matter, you know, how a image generator works or how a recommendation algorithm works.

7:13And so over the course of the book, I'm trying to find other discourses or other traditions of thinking about media that I think are helpful. And that brings us to this kind of really strange agglomeration of things like stage magic or military psyops or cognitive neuroscience or Old Testament theology or chaos magic or ufology. These are weirdly tools and traditions that have been helpful for me about thinking about what images are now.

7:47Trevor Paglen:Yeah, you say in the introduction, we have a lot to learn from the weird. And I want to get into the weird, but just to say a personal thing about what the book gave me is that I did get what you're just saying in this kind of paradigm shift where I think with a lot of the digital images stuff, people talk about a sense of reality, decay or collapse. and someone else will always come back and say like, well, images, it was always fake that they were related to reality, right? Like there's Photoshop and there's like, you know, all kinds of fakes, which you talk about in the book, you know, double exposure images or like spirit photos from the 19th century.

8:23Trevor Paglen:But it's really this idea that they're for machines, images for machines. That's the real break, not just that they're manipulable, but they're readable by something that's not human. So there's kind of autonomous world of images out there. That was like a very clarifying in terms of what the real epistemological break or break in the history of visual culture is. No, definitely. Like when I was in art school and studying this stuff, the digital thing had happened and people were kind of thinking about it in terms of postmodernism. There's no original and like everything's a copy and they're still using that kind of language.

8:56And what I started realizing was like, well, that is sort of a big deal. But what I think is a bigger deal is that the default state of a digital image is unreadable by humans. Like you can't look at it. Like you have to develop another piece of software in order to interpret that JPEG into something that you can see as a human. Again, if we go back to the metaphysics of the image, they're images for machines in the first instance that you can do an operation on them to make them readable by your eyes, but in their native state, they're not. And so I kind of run with that idea. And you're absolutely right that people are talking about the collapse of reality, these kinds of questions.

9:37When we think about, let's take photography, it's probably the best example, is that the common sense understanding of photography is that there's some light that bounced off something in the world and that created an impression on a photosensitive medium, a sensor or an emulsion or what have you. And so you have this relationship between an external reality and the image. We call this indexicality in photo world. And I think that the existence of photography kind of created a form of common sense where common sense dictated to you that when you saw a photograph, that something like that actually took place in the world, right?

10:23You saw a photograph of the napalm girl. You saw a photograph of a video of Rodney King. You saw images from Abu Ghraib. Common sense dictated that those things actually happened. Now, how you interpret that, different question, but common sense dictates that there's a brute fact that is recorded in that image. Now, we can intellectualize that and say, of course, that's not true and images have been manipulated, blah, blah, blah, blah, blah, blah. However, However, we're talking about the kind of collective common sense, right? And that seems to have eroded. That seems to be gone now in this age of generative AI when anybody can say like, oh, that's fake.

11:05Trevor Paglen:Like AI made that up. The default state is kind of like suspicion. Yeah, the default state is kind of suspicion. And I've been calling this the indexical flip. I mean, it's not in the book, but I sort of allude to it in the book. And I noticed it. I knew this moment was going to come, but I noticed like the day after the U.S. started bombing Iran, I was kind of scrolling through social media and I was trying to understand like what's happening here. And I realized that my default state on reading an image or a photographic type of image, my default state went from a common sense belief that that was a record of something that happened to, no, this is some AI crap that an algorithm is trying to feed me.

11:48And I was like, whoa, that's weird. So instead of trying to find the piece of junk and throw that out, I was trying to find something that had some veracity to it. So that's weird. Right. And that's a big deal.

12:00Trevor Paglen:Yeah, it's a very big deal for image makers, for artists. I think in some ways the default condition is that you look at everything like it's an artwork now rather than like it's a document that needs proof of its creative life in it. I think that's really interesting and going to change everything. And you kind of get that in the introduction following on from what you're just talking about when you talk about the switch from this idea of the language of images representing the world to the language of images activating the world. So what is that framework? Yeah, so this is another kind of concept that unifies the book.

12:36But I also think for me has been a helpful way of kind of having a concept that helps me think about this post-AI visual culture that we're in. So what I mean by activation, it's a little bit inherited from Farocchi's idea of operational images.

12:53Trevor Paglen:Heron Farocchi, who's a German filmmaker. So what I mean by activation is, you know, let's go back to the example we were using before about the camera that takes a picture of everybody's license plate and then sends you a bill or sends you a ticket or what have you. You have an imaging apparatus. It's making an image. That image is then activating some kind of commercial or policing system that is then triggering a result. Your credit card being billed, you being issued a ticket if you ran a red light, what have you. And so the image is not really something to be looked at. The image is part of this circuit, this policing circuit or this economic circuit, what have you.

13:34Now, when we get into the world of generative AI and recommendation algorithms, we can also think about those as activations in a different way. I spend a lot of time thinking about why the hell I'm scrolling the internet, why am I looking at a 20-second video of somebody eating spaghetti? This makes no sense, right? There's no story here. We're not going to analyze the picture of the person eating spaghetti as if it were a kind of mythological image that we needed to apply our skills that ideology critique to. There's something else going on. There's something neurological going on. So the idea is we can extend the concept of an image activating a circuit from industrial systems to a cognitive system.

14:22Here's a piece of media. It is going to induce this type of reaction in you. It's essentially activating your brain or your attention in a particular way. And to me, that's been a helpful way of trying to understand the feed or the kind of very online digital culture that we find ourselves within right now. I guess that's how I think about it, is thinking about activations as activating industrial or policing systems or activating a cognitive system.

14:52Trevor Paglen:Your interest in magic comes in or theories of the history of both magic spelled the way you think and magic with a K, which you'll have to make the distinction for me and why you think it's important. Yeah. So I was trying to think about forms of media that assume your perceptual system is a type of machine or a computer that can be hacked, right? And forms of media that take advantage of that. And I started kind of having this theory that stage magicians were the original cognitive neuroscientists, because what stage magic is about, the craft of it and like the media of it, we don't think about it as a media, but 100 % is.

15:37It's a form of media that is very aware of how you're going to perceive something and is essentially developing techniques to hack your perception, to make you see something that didn't happen or make you kind of believe something took place that was supernatural or that wasn't there. And it does that by managing your perceptual system. Now, stage magicians, when you talk to them, they will say, this is a lie. I am showing you things that are not real, right? And that's very much a part of the craft. That's what it is. Now, magic with a K is a little bit different. And if I can give a long answer, I think that this is also kind of grounded in neuroscience as well.

16:28To answer the magic with a K question, we would think about this as, you know, occult magic, right? Very different than stage magic. I actually want to ground that in neuroscience as well. And I want to ground it in a theory of visual perception that comes out of experiments on cats, eventually humans, what have you. So when you talk to neuroscientists, they think about vision and images really, really differently than I had thought about them. What we think happens when we look at something is that light bounces off something in the world, goes into our eyes, and our brain gives us a more or less accurate representation of what's out there.

17:18That's not how it's understood from this cognitive perspective. What happens instead is that your eyes ingest some kind of visual stimulus. Different patterns of neurons are activated in your brain. And then what you experience as an image is a, I think I use the word quasi-hallucinated, is a quasi-hallucinated experience that your brain generates for you based on those visual stimulus. And so this has a pretty long tradition in neuroscience of understanding vision this way, but I think it's been recently most popularized by Donald Hoffman, who uses the analogy that I can repeat where he says, you know, if you're looking at your iPhone and you click on the mail icon, that mail icon has nothing to do with what the computer is actually doing.

18:10It's just an interface that's been created for you in order to be able to manipulate what's going on the phone in ways that you want to do that. And the argument is that your visual cortex, your brain is doing something analogous to that. It's translating this, whatever pieces of raw data that it's collected about reality, and then creating essentially a GUI for you that you can use to manipulate that. The kind of philosophical implication of that is that we can only know reality through these hallucinations, essentially, that our brain makes for us. And whatever reality in itself is, we don't know.

18:49We don't have access to that. So, to bring this back to chaos magic, chaos magic kind of takes that as a given, right? It takes for granted that reality and our perceptions of reality are functionally equivalent, right? Whatever I perceive about reality, because I can't know reality, my perceptions are practical reality for all intents and purposes. And then it says, well, I can change my perceptions. And therefore, when I change my perceptions of reality, I'm effectively altering reality itself. That's the kind of basis of chaos magic and a lot of the magic with a K.

19:35Trevor Paglen:So, I mean, basically to sum up, magic is real. No. Or as you say in the book, like all perception is essentially a hallucination. Like your brain is on some base level creating reality as you see it. This is not like a metaphor. I think one of the examples I use is like, if you're a baseball player, you know, pitcher throws a ball at you, your mind has to take in that information, run it to the back of your brain, process it in your visual cortex, send it back to you, and then you can trigger an action swinging the bat. if you measure how long that takes for your brain to do, it's way too slow.

20:14It can be like hundreds of milliseconds. So what instead your brain does is predict what that ball is going to do, show that to you, and you're reacting to a hallucination that is your mind predicting what's happening in that situation in order for you to be able to hit that ball. So that's weird, But that's true. And that's kind of an example of this hallucination that your mind's doing.

20:39Trevor Paglen:That is all very cool and trippy. And on the other hand, sort of sinister. A serious part of the book is you going through the history of these collisions between CIA experiments with mind control, CIA actual interest in both stage magic and kind of crazy, more far out magic. because they're investigating some of the same things you are investigating, which is the nature of perception, how to manipulate people. And then that goes into developing these artificial intelligence, you know, machine learning stuff, right? So there's a dark side to this too. Yeah, I mean, I think that dark side of it, it's when you couple essentially techniques for manipulating people with capitalism, you're going to get kind of cultural engines that treat people as if they are machines that can be manipulated however you want.

21:35That's certainly what we're seeing. I call this the society of the psyop. Oh, that's my favorite part of the book by far.

21:42Trevor Paglen:Yeah. And I'm kind of riffing, obviously, I'm kind of updating Debord. Anybody who went to grad school learned about the society of the spectacle, which is how images modulate the way that we relate to each other. And then Foucault has the Society of Surveillance. With the Society of the PSYOP, the kind of argument is that the way we interact with most visual media now is probably on a feed or in relation to some kind of recommendation algorithm. So when you're interacting with a piece of media, that piece of media is also reading you in pretty much real time and is modulating itself in order to elicit some kind of reaction from you, right?

22:28Usually that reaction is, you know, sustaining your attention or trying to serve you some kind of clever advertisement, what have you.

22:35Trevor Paglen:Kind of like the most basic paranoid fantasies. If you talk to somebody who has paranoid delusions, you know, they're like, oh, they're watching my thoughts, you know, they're changing reality in order to screw with me. I mean, all that's basically real now. It's basically real. I mean, like, think about the number of times that you have a conversation with your friend and then you look at your phone and Instagram is showing you an ad for something you were just talking about. That's literally a kind of supernatural experience. We are now in the world of conspiracy theory and paranoia. Yeah. So I think the idea is going from a world in which we assume that a viewer essentially stood outside of images to being in a media environment where we ourselves are part of a circuit.

23:21Matt Dreierst and Holly Herndon were talking about for a while back, they had a book called All Media is Training Data. I kind of tried to add to that to say all media is training data and all media consumption is reinforcement learning with human feedback.