How to survive the age of AI, fake images and psyops

A photograph used to mean something. Wherever you encountered it – on the front of a newspaper, in a gallery or on the internet – there was an assumption that it represented something real that had happened in the world, in a real place and time. Of course this wasn’t necessarily true. From hoax photos of fairies at the bottom of the garden, to Joseph Stalin erasing purged officials from group portraits, image manipulation has been used to mislead viewers throughout history. But, even with tools like Photoshop at our fingertips, this process took enough time and effort that we could safely assume most photos were authentic. Then generative AI came along, and everything changed.

The artist Trevor Paglen remembers the precise moment he noticed this shift, which he refers to as the “indexical flip” (the term “index” commonly referring to the idea that an image serves as a record of the reality it depicts). “The day after the US started launching air strikes against Iran, I’m looking at the Twitter feed, and I’m trying to figure out what’s actually happening,” he tells Dazed. “And I just noticed that, instead of trying to identify the crap and throw it away, I was trying to identify what’s real. I’d gone from assuming things are ‘real’ to assuming things are ‘fake’.” 

This was a problem that Paglen had already predicted in his newly-published book How To See Like A Machine. “Everything has become a potential AI hallucination,” he writes. “In a post-indexical world, all media become superpositional.”

Bringing together several strands from across the past decade of Paglen’s career, the book weaves together wild-but-true stories about UFOs and government psy-ops; thoughts on magic, algorithmic feedback loops, racist AI used in policing, and firsthand insights into the early days of machine vision. And what do stories about the US Air Force leading alien investigators on a wild goose chase in New Mexico have to do with the image world we’re living in today? Quite a lot, actually. As Paglen tells it, the rise of technologies like generative AI and machine vision has created a world where “photographs are more akin to magic tricks or acts of conjuring”. They no longer function as “representations” of reality, but “activations” that can change reality itself. Think: a photo taken by a speed camera that sets in motion an automated legal response, but also the AI-generated slop that’s primed to activate specific parts of our brain and steal hours of our attention.

As many artists try to push back against these emerging technologies, or distance themselves from them – “which I totally understand,” Paglen notes – he puts them under the microscope, giving them the time and attention they deserve as one of the main forces shaping our lives today. “I naturally gravitate towards trying to understand how the world is changing,” he explains. “And so it’s more coincidentally I end up working with a lot of experimental tools.”

Below, Paglen digs deeper into How to See Like A Machine, touching on the old magicians’ tricks used by today’s AI chatbots, how to spot the covert techniques that shape your sense of reality, and why UFOs make for such a powerful meme.

In a recent performance-lecture at Tate Modern, you described computer vision and generative AI as comparable to the invention of photography and linear perspective. What does this comparison mean to you?

Trevor Paglen: When we look at the invention of photography, or the invention of perspective, or any number of other moments, the fundamental question of what an image is was changed. With photography the image became a thing that had a connection to the world outside of itself, and the invention of perspective did something similar. For me, the invention of computer vision and generative AI are on a par.

With computer vision, you are removing a viewer. Historically, an image required a human to look at it in order to exist, and that’s not the case anymore. When we talk about the world of computers making images for other computers, we’re talking about a fundamentally new kind of image that is natively invisible to humans. That’s very bizarre, right?

“When we talk about the world of computers making images for other computers, we’re talking about a fundamentally new kind of image that is natively invisible to humans. That’s very bizarre” – Trevor Paglen

Then, with the advent of generative AI, you have on the one hand an untethering of images from representation. This really complicates photography in particular, in the sense that you can have photographs where there was no camera. You can have artworks where there was no artist. This is another one of these monumental shifts. 

What’s more, when we’re talking about generative images – and text too – the normal context that you experience those in will be some kind of feed or algorithmic engine. And the engine behind those images is watching you, monitoring how long you are looking at an image for, what kinds of images seem to be more exciting to you. So the engine that’s serving you those images is, in a very real sense, trying to both create a world for you and harvest something from you in the form of attention.

You’ve mentioned that the idea of images being connected to a ‘real’ world, outside themselves, might be newer than we think. Historically speaking, ‘representation’ might be the aberration. And, in that case, why does it matter today that images are no longer tethered to reality?

Trevor Paglen: That’s a big question that I try to start to unpack towards the end of the book, thinking through the relationship between empiricism, modernity and democracy, and how big a role images play in that. Specifically photography. From the highest satellite viewpoint – and this doesn’t exist in real life – the theory of democracy, as I understand it, is that we have a shared set of facts that we discover through empirical means, and then we deliberate about what to do about those facts. It requires a collective belief in empiricism. 

If you think about the photographs from Abu Ghraib, those photographs came out and collective common sense dictated: ‘That actually happened.’ You can argue about what the politics of that are, what to do about it or whatever, but there was a creation of a shared reality. I’m not sure that that would happen today, in this post-indexical flip moment. So is that dissolution of indexicality a bigger deal politically than we might think?

“Why do we spend time looking at 15-second clips of people eating spaghetti on Tik Tok? This makes no sense. There seems to be something neurological going on, where the image is speaking to your subconscious” – Trevor Paglen

Can you tell us more about ‘operational images’ and where they fit into this landscape?

Trevor Paglen: Totally. ‘Operational images’ is a phrase that the filmmaker Harun Farocki came up with. What he was talking about were images that were used in some kind of a control system, so an image that you would use to calculate the trajectory of a ballistic missile, or an image that a robot might make for itself in order to navigate. In other words, images that were part of some kind of software, or industrial or policing circuit. 

The point of those images wasn’t to tell you a story so much as to have something do something. I build on that and talk about this general shift from representations to activations. And that can include things like the operational image as understood by Farocki... but I extend that to neurological effects. We don’t have to think about operational images as just existing within a software or industrial environment. You can think about it as inducing cognitive activations as well.

Why do we spend time looking at 15-second clips of people eating spaghetti on TikTok? This makes no sense. This is not a reasonable thing to do. There seems to be something neurological going on, where the image is speaking to your subconscious and activating your cognitive apparatus in such a way that you feel compelled to pay attention to these things, even though the rational part of your brain cannot understand why. And there’s a deep bench of media forms that have tried to exploit that.

You referenced the lizard brain in your Tate lecture, and made a tongue-in-cheek connection to the lizard people conspiracy theory.

Trevor Paglen: Insofar as there’s an argument in the talk, it’s that we live in a world of images that are predatory on these preconscious parts of our brain, our lizard brains. And the more and more that we live in an environment where we are guided by those impulses, the more and more we become lizard people. And what I mean by that is creatures that are motivated by different desires than the rational. In the background, I’m thinking a lot about what you need to do to succeed in the Trump administration, or what you need to do to succeed in a lot of pretty ugly corporations. In a way, you have to become inhuman to do that.

You compare the way images are wielded to access those parts of our brain to stage magic, as well as psy-ops by intelligence services. How conscious do you think these connections are?

Trevor Paglen: I don’t think it’s conscious at all. Marketing, psy-ops, magic... these are all fields that have studied perception and how you manipulate somebody’s sense of reality. I’m not sure that whoever is building the TikTok algorithm is looking at the army field manual on psy-ops for ideas. But I think there’s techniques that you find over and over again when you’re in the business of trying to manipulate somebody or craft their reality.

One technique you refer to is the poison coin, referring to an invention by John Mulholland, the CIA’s ‘staff magician’ from the 1950s. Can you tell us more about that?

Trevor Paglen: The idea is that you use the quotidian to hide something sinister. They used to sell those little coke cans that you would unscrew and put your weed in it. You just kind of throw away the image, but it’s actually hiding something else. This is a technique that magicians have used for a long time. In the book, I talk about how Mulholland worked for the CIA to develop weapons and covert operations techniques using principles of magic. And when we come to the present day, the big one is where you teach or when you have computers speak and emote as if they were humans.

This is a kind of poison coin technique because you’re creating this wrapper around a machine that is emotionally engaging and tells you how smart you are. We’ve seen this chatbot psychosis stuff, which I think is very real and much more prevalent than people make it out to be. But you see this kind of thing all over the place. If you have a robot vacuum in your house there’s a good chance that vacuum is making measurements and cataloging everything in your house, and sending that to some kind of data merchant.

Do you think there’s a way for artists to intervene in these kinds of systems, to make meaningful change?

Trevor Paglen: No. I don’t think you can influence that infrastructure in a meaningful way. What I think you can do is learn how to spot the techniques that are being used against you. Doing this research [has] made me see how the media environment that we live in now is [different] in very significant ways to the one that we lived in 30 years ago, and certainly before that. It’s brought to my attention the need to have better analytic tools to think through the present. The analytic tools that we have to describe media and images from 50 years ago are describing a different world than exists now. And that was part of the idea of writing the book, to help develop a vocabulary to understand what media is now.

Let’s talk about aliens. Were aliens and UFOs part of the picture since the very beginning?

Trevor Paglen: It started when I was in graduate school. I wrote my geography PhD dissertation on secret military bases, about places that quote-unquote ‘don’t exist’. And I wrote a book called Blank Spots On the Map, which was about how you build those secret military bases. Of course when you start looking at Area 51 there’s all this alien mythology around it, but I was looking at Area 51 as a real military base, and what I noticed is that there’s a whole bunch of conspiracy theories and disinformation around that. I was constantly talking to people who, in retrospect, I think were trying to do disinformation operations on me.

“UFOs are essentially a meme that psy-ops folks have used for a very long time” – Trevor Paglen

There is a part of the psy-ops world that tries to make it like seem you’re a crazy person if you’re interested in secret military stuff, and I think I was the target of that. I don’t think it was successful because intellectually I knew, ‘There’s all this disinformation I’m probably being fed. I need to bracket that out, because it’s too complicated, so anything that I can’t find a primary source for we’re not going to use.’ But then when I started doing this latest body of work, thinking about psy-ops as a lens to understand contemporary media environment, I was going back to that work, thinking ‘okay, what is that world of disinformation? What is that world of psy-ops that the younger me had to ignore in order to do good work, and for my own sanity?’

The second you dive into that, it’s UFOs and all kinds of weird stuff. UFOs are essentially a meme that psy-ops folks have used for a very long time, going back to the 1950s. That’s certainly one of the big ones, particularly in relation to the Air Force and some CIA stuff.

Why do you think UFOs are such a powerful meme?

Trevor Paglen: For a lot of reasons. There’s a huge religious component to it. It’s not so different than people seeing angels or seeing the face of the Virgin Mary in the sky. One of the things I’m interested in is how UFO photography in particular is predatory on that indexical common sense, predatory on the presumption that a photograph of something depicts something that happened in the world. There’s also spirit photography. There’s all these edge cases of photography that prove that indexicality is very, very complicated. UFOs are the most famous example, and as such the genre of UFO photography prefigures a lot of the contemporary image space that we’re in.

That’s the super basic take on UFOs, but you can spin it out as far as you want to go and really get into some of the theological and metaphysical implications of these types of images. To what extent do we conjure UFOs into existence through our collective belief in them? To what extent does that collective belief create a situation where the meme itself is generating its own story?

I want to believe...

Trevor Paglen: Exactly. Jesus, for example, is very real, whether he existed as a person or not.

What do you make of the Trump administration’s loud, public disclosures of supposed UFO images, and the counter conspiracies that they’re trying to distract from negative headlines in the ‘real’ world?

Trevor Paglen: To me, UFOs are perfectly consistent with the image world of the Trump administration, in terms of creating memes of him as Jesus or the Pope or Superman or whatever, or making these AI-generated videos of him flying bombers and dropping shit on protesters. To me, it’s perfectly on brand for the image space that this administration is creating. Also I don’t think it’s a coincidence at all that, in the age of generative AI, everybody’s seeing UFOs.

How to See Like A Machine by Trevor Paglen is out now. The artist is also set to co-curate the new Zero 10 sector at Art Basel 2026.