A museum explores how deepfakes alter our information landscape – Grid News

Get the best of Grid in your inbox, every day.

A museum explores how deepfakes alter our information landscape

Deepfake technology is an artificial intelligence method that manipulates audio or video to make a person appear to do or say something that they did not.

Previous examples that have gone viral include a video in which the actor Jordan Peele imitated Barack Obama, and a recent TikTok where an uncannily real-looking Tom Cruise plays golf and performs sleight-of-hand magic. In the documentary “Roadrunner,” filmmakers created audio of the late celebrity chef Anthony Bourdain speaking lines he never said aloud.


Hear more from Anya van Wagtendonk about this story:


Experts have issued warnings about how this technology might be misused to discredit public figures, perpetuate fraud and spread realistic misinformation. And now, a new exhibit at the Museum of the Moving Image, in Queens, N.Y., explores how easily our eyes can be tricked.

One art installation enacts a midcentury living room, with a television displaying President Richard Nixon as he gravely announces that the Apollo 11 moon mission has failed. Of course, that did not happen. Neil Armstrong, Buzz Aldrin and Michael Collins made it to the moon and returned safely. But the artists employed deepfake methods to investigate the ways in which the presentation of truth in our information landscape — from museums to the media — is increasingly blurry.

ADVERTISEMENT

The exhibit raises broader questions about the specific role of museums in shaping trust and narratives, said Ian Elsner, an independent museum exhibit creator and software developer who has worked on interactive exhibits across North America. He highlighted the deepfakes exhibit in a recent episode of his podcast, “Museum Archipelago.”

Elsner spoke with Grid about the implications of such an exhibit, and what it tells us about how museums share information — and can spread misinformation, too. That’s because museums present objects and narratives in ways that are often unchallenged by the viewing public, Elsner argued. And at a time when trust in institutions is on a steep decline, trust in museums has remained relatively stable.

“Enjoying such a position of power over the visitor is something that I think everybody in the museum world needs to take seriously,” he said in an interview with Grid. “Because of the example of deepfakes and many other types of technological change, [the museum world] needs to be prepared for a future when that level of trust is not automatic, and is not given freely when you walk into the door and buy your admission ticket.”

This interview has been edited for length and clarity.

Grid: How does the use of deepfake technology evolve the use of technology in museum-based learning?

ADVERTISEMENT

Ian Elsner: If we take a step back, one of the things that I think a lot about in my work is this idea of, “The medium is the message.” And the medium of museums is a very powerful medium. It says truth, where many other media don’t. Just the simple act of putting something [on] a pedestal, just the idea of a well-lit display case, just the idea of a piece of art hanging on a white wall that’s the focus of the room, indicates to a visitor, in many cultures, that this is important, this is worth saving or this is important to know about.

So starting from that idea, I’m interested in ways to think more carefully about that trust that museums have.

G: Tell me about the deepfakes exhibit. Why did it pique your interest?

IE: Yeah, so this is a this is a very cool project that’s currently at the Museum of the Moving Image in New York City. It’s part of an exhibit called “Deepfake: Unstable Evidence on Screen.” The centerpiece of this exhibit is a piece by artists Francesca Panetta and Halsey Burgund that uses deepfake techniques to create a video of Richard Nixon delivering a speech that he never delivered.

In 1969, during the Apollo 11 moon landing, members of the Nixon administration realized that they probably should have a contingency address prepared in the event that something went wrong. Specifically, if the Apollo 11 lunar lander crashed on the moon, and didn’t have a way to get back, the way that the world would learn about this was probably through President Nixon. And so they wrote this real speech that Nixon was to deliver in the event of this contingency. The title of the memo is actually “In Event of Moon Disaster.” And that’s where the name of the artists’ piece comes from.

ADVERTISEMENT

So it presents a really interesting way to talk about this deepfake technology, because here we have a speech that was written and was supposed to be delivered by this historical figure, but never was because the mission was a success. What the artists did was they took the speech and deepfake technology — including a voice actor, who did many, many takes to try and imitate Nixon’s voice, and synthetically enhanced it to make it sound more like Nixon — then take actual footage of Nixon giving an emotional speech, and altering the way that his lips and mouth and teeth look to make it look like he’s delivering the words that were written for him. And the resulting video is a very convincing facsimile of the truth. Here we have Nixon delivering a speech that he never gave.

And of course it is presented in a museum context. So it’s a chance for the visitor to interrogate what a “deepfake” is, how to look out for it. Did you recognize it? Could you recognize it if this wasn’t with the surrounding context?

G: What do you think are the pitfalls or drawbacks to using such technology?

IE: I interviewed someone named Julie Garcia, who wrote her dissertation on the role of the Creation Museum in the modern American creationist movement. And she writes about how the idea of a Creation Museum latches on to the general public’s trust of museums as an institution and uses it to present a version of science and a version of history that is far outside the scientific consensus.

This is a trend that is fairly recent. It’s only within the past few decades that “young-Earth creationists” employed the museum in their quest for scientific legitimacy. What they’re doing is borrowing the style of natural history museums and science centers that the public might be familiar with.


I think it’s so important, whenever we discuss museums, to not separate museums into “museums that tell the truth” and “museums that don’t.” Because, obviously, the truth is much more complicated than that. And even museums that are seen as legitimate have a lot of problematic ways that they present truth, [such as] the presentation of empire in the British Museum, widely considered to be the first public museum which acted, very conveniently as a way to show the average British subject living in London the wealth, power and vastness of the empire through objects collected from a whole bunch of other places. And then that becomes the dominant narrative.

I can’t help but think about a conversation I had at the Tiagarra Cultural Center and Museum, which is in Devonport, Tasmania, Australia. I spoke to the director of the museum, David Gough, who is a Tasmanian Aboriginal. He grew up going to that same museum before he became director of it, when it was controlled not by Tasmanian Aboriginals but instead by white scientists. And those white scientists had concluded that, well, there’s not really any Tasmanian Aboriginals left in this area, and so this is the best we can do, is we can create a museum about them.

He’s the chairperson of the Six Rivers Aboriginal Corporation there, which is one of the governing bodies. Once he was able to take ownership of the museum, it was very important to him to start to put masking tape over the inaccurate museum information displayed on the panels, and instead write with a Sharpie the accurate information as he knew it.

If you think that these sorts of things are an outlier — like, “Oh, that’s an interesting story in a very particular context” — I’d invite you to think more about a lot of the assumptions of the dominant culture wherever you are. Think about how what is accepted as fact, in a scientific context or a historical context, is just the dominant cultural narrative, presented ad nauseum, and buttressed by institutions like museums.

G: So, I started this conversation asking you about deepfakes, but it seems that you’re arguing that even without that explicitly manipulated/manipulative technology, that the line between information and misinformation can be blurred within a museum setting, through more subtle, less explicit ways.

ADVERTISEMENT

IE: Yes, exactly. Once you get that idea in your head, that sort of a full motion, audiovisual experience might not be real, then I think it calls into question a lot of other forms of media. I certainly know as a podcaster how easy it is to manipulate audio and to cut it up and to make people say something that they never said, just with audio.

The interesting point that I want to underscore is that — maybe because of the acknowledgement that images can be manipulated now; the idea that videos can be manipulated pretty easily, and increasingly so — trust in many, many institutions has dropped. [But] the American Association of Museums publishes stats every year indicating that, for whatever reason, museums still score incredibly highly on most trustworthy institutions.

It’s curious why museums continue to enjoy that place of cultural prominence. I think that it might be a delayed reaction: that members of the American public spend a lot more time with friends and family and newspapers and TV and the internet than they do with museums. That would delay the downward slope that that I would expect.

The way that deepfakes fit into all of this is that it gives us yet another point to start to question what we see. And I think for the creators of “In Event of Moon Disaster,” they freely acknowledge [that] there’s many, many other low-tech ways to create misinformation than deepfakes. Deepfakes are just another tool in the arsenal.

G: As a museum expert, are you at all concerned about the possibility that this high level of trust will erode? Or does that feel more exciting and promising?

ADVERTISEMENT

IE: I think that it’s important that the high level of trust erode. I think that the level of trust is undeserved. So often, classic museums undergird the institutions in cultural life that have the most power and the narratives that have the most power. You see this in a lot of Southern museums with how they present the American Civil War, where the narrative that’s presented is the narrative that’s the easiest to present, or the one that doesn’t require changing the way the dominant culture feels.

If you are an adult who doesn’t take classes, where else do you go where you might expect to learn something about the world than a museum? In some ways, it is a very open process. And I think that, unintentionally, and intentionally, museums have historically taken advantage of that openness.

  • Anya van Wagtendonk
    Anya van Wagtendonk

    Misinformation Reporter

    Anya van Wagtendonk is the misinformation reporter at Grid, focusing on the impact of false information on policy, elections and social behavior.