[CLIP: “A fair casting process: AI protection and fair compensation!”]
Sophie Bushwick: I’m sure you know what that is.
Tulika Bose: Yeah, for the first time since the 1960’s—both Hollywood actors and writers are on strike.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
[CLIP: “Now striking Hollywood actors hold their largest demonstration to date, and it happened in Times Square.”]
Bushwick: At the same time!
Bose: At the same time.
Bushwick: Then and now, the problem is about technology disrupting the world of Hollywood...
Bose: And the question of whether actors and writers are being paid fairly.
[CLIP: “We need to come to the table in good faith.”]
Bushwick: In the 1960s, the technology was television.
[CLIP: “Here comes Flipper!”]
Today it’s streaming platforms—but also the rise of powerful new artificial intelligence models.
Bose: Generative AI can make it way easier to use a performer’s likeness or voice without having that person there at all. And in that case, who gets to profit from the performance?
Bushwick: This is Tech, Quickly, the digital double of Scientific American’s Science, Quickly podcast. I’m Sophie Bushwick, tech editor at Scientific American.
Bose: And I’m Tulika Bose, senior multimedia editor.
[CLIP: Music cue]
Bose: So, Sophie, I actually kind of spoke to a few actor friends about what’s happening right now, and why people are striking.
Bushwick: I get the impression that some people are worried they’re going to lose their jobs to an AI.
Wolfgang Hunter:The worst case scenario would be there’s no more extra work, which is just like such a useful thing for a lot of like working actors, like even nonspeaking roles, extra work, like, that’s money, And if you can just AI generate a crowd and then doesn’t look like unrealistic—the thought of that is very scary.
Bose: That’s my friend Wolfgang Hunter, an actor, comic, and writer in New York City who has written for shows like Mr. Beast.
[CLIP: “I love the Midwest. Even the hecklers are nice people.”]
A lot of actors—even if you’ve seen them before—need to do everything they can to pay the bills.
Hunter: So it’s scary to think that if you sign a waiver, you could just like, be never getting work again, even as like an extra.
Bose: So Sophie didn’t you speak with someone about this too?
Bushwick: That’s right — I spoke with Hany Farid, a professor at the University of California Berkeley who studies this type of technology. He says this idea, that generative AI can take an actor’s image and insert it into any scene, is just putting a new name on existing tech.
Hany Farid: We used to call it deepfakes which was scary sounding and was largely associated with things like non-consensual sexual imagery and fraud and disinformation. Somebody in the industry did some very good rebranding and can now we call it generative Ai which sounds less scary and less awful. It is the same core underlying technology.
Bushwick: I tend to associate deepfakes with video. But generative AI can also be used to make still images, text, music, and other audio.
Bose: And this is also kind of worrying for actors who also rely on voice work.
Hunter: What's on the line right now is reoccurring work. So especially with like voice acting and stuff like that — AI can’t like generate a new voice or a new character, but it can mimic existing ones.
Bose: So this character that Wolfgang voices, for example...
[CLIP: “I don’t know TheLittle Mermaid. I don’t listen to hip-hop.”]
Bose: Could be easily sampled and replaced, right?
Bushwick: Exactly. And Hany talked about how easy this is with AI tools.
Farid: If I for example, Sophie record 2 minutes of your voice just 2 minutes nothing sophisticated just on this call right now: I can now go over to a commercial website and for $5 a month I can clone your voice and have you say anything I want you to say.
Bushwick: Of course, my voice clone might sound more realistic if the AI developers gave it more clips of my voice. In general, the more data you canprovide, the more realistic the digital double will look or sound.
Bose: Oh, digital double; kind of creepy.
Bushwick: Very creepy.
Farid: With a single image I can insert somebody who’s linking this into a video but because I've only got one view of them. It's going to be gonna be a little glitchy, and it will not hold up to Hollywood style. It's fine for Tiktok and Youtube—but it’s not gonna it’s not gonna pass muster for Hollywood.
Bose: The technology might get better eventually, but for now, actors have to go through something called the orb!
Bushwick: Okay, so it’s technically called a photogrammetry stage. But a video game actor who has performed on one of these stages, and who asked to remain anonymous, described it as “the orb” when he spoke to our tech reporting fellow, Lauren Leffer.
Bose: Very science fictional!
Bushwick: Oh, absolutely. I’m going to quote Lauren directly here: “Inside the orb, the world is reduced to a sphere of white light and flashes. Outside the orb’s skeletal frame is darkness. Imagine you are strapped into a chair inside this contraption. A voice from the darkness suggests expressions: ways to pose your mouth and eyebrows, scenarios to react to, phrases to say and emotions to embody. At irregular intervals, the voice also tells you not to worry, explains that more flashes are coming soon.”
Bose: Well, that sounds freakin’ terrifying.
Bushwick: Yeah, it does not seem like a pleasant experience! But if you’re a working actor, you have to put up with a lot of stuff like that for the job. And, on the one hand, AI is making it way easier to duplicate an actor’s image even without the orb—which just ramps up the pressure for performers to sign away their images.
Hunter: I am, like, such a guy that has such a bad habit of just signing something and not reading it. Absolute whatever. Yeah, cameras right there. Yeah, you're just like, lizard. But I think if I knew what I was looking at, I don't know. Because it's like, at the end of the day, that's all you have is yourself, you know, when you can't even get work or you're not even like in the industry, like fully. You only have this. You're basically selling, like, it's the closest thing I think we've actually had to selling your soul.
Bose: So one goal of the strike is for actors to protect themselves from having to sell their image—or to make sure that, if their digital double is going to show up in a bunch of future media, that they’ll be fairly compensated for that.
Bushwick: And it’s not just their own digital doubles that they need protection from. Remember how Wolfgang said AI can’t generate a new character?
Bose: Uh oh.
Bushwick: Yeah, I don’t know about audio-only characters. But AI can absolutely generate images of humans who have never existed.
Farid: I can now just whole cloth create the next generation of performers. That's not a replacement. That's a creation. I can decide, look, we're going to do market research and we're going to say, well you know a blend between George Clooney and Brad Pitt and fill in your favorite actors. There is you know? So maybe you create this sort of this chimera right? This hybrid.
Bose: Or, say, Timothee Chalamet?
[CLIP: “Janie’s hella tight. Maybe I’ll see you at the Deuce.”]
Bushwick: And if you have an AI-generated fake actor, it can be customized to different markets. Maybe there’s one version designed to appeal to Americans...
Bose: Another version for the Middle East and another one for Chinese markets.
Farid: So the studios, they may not even need performers. These things can be created whole cloth. Um, so it's not even a replacement anymore. It’s something different.
Bose: Okay, yeah, this is a lot worse than we thought.
Bushwick: I understand why actors are striking! But also, if the studios want to go all-in on AI, then sure, they might be able to cut out some of the humans in front of the camera. They might even be able to save some money. But it could also majorly backfire for them.
Farid: If this technology really gets good where I can go to ChatGPT and say, please write me a movie script with this storyline and these sorts of themes and then I can synthesize the voices of the characters and synthesize the videos and do all of that on my laptop, do we democratize access to filmmaking? It's an interesting question. Maybe it’s not going to be Hollywood studios right now. The performers are scared for their jobs, and I think they have the right to be. But what if the studio should be fearful? What if the very studios that are jamming up the performers—what if they’re next?
Bushwick: Science, Quickly is produced by Jeff DelViscio, Tulika Bose, Kelso Harper and Carin Leong. Our show is edited by Elah Feder and Alexa Lim. Our theme music was composed by Dominic Smith.
Bose: Don’t forget to subscribe to Science, Quickly wherever you get your podcasts. For more in-depth science news and features, go to ScientificAmerican.com. And if you like the show, give us a rating or review!
Bushwick: For Scientific American’s Science, Quickly, I’m Sophie Bushwick.
Bose: I’m Tulika Bose. See you next time!
