In short
Podcast Episode Notes: Critics at Large | The New Yorker
Episode Title
Art in the Age of Artificial Intelligence
Episode Overview In this episode, hosts Vinson Cunningham, Naomi Fry, and Alexandra Schwartz discuss the impact of generative AI on the arts. They examine recent examples like Timbaland's AI rapper TaTa Taktumi and the synthetic actress Tilly Norwood, alongside reflections on how AI has been portrayed in film and literature.
---
Key Themes
The Rise of Generative AI
- Generative AI: Technology that creates new content, such as images, music, and even characters.
- Cultural Pervasiveness: The hosts note that AI is becoming ubiquitous in everyday life, with claims that 99% of people have interacted with AI.
Current AI Projects in Arts
- TaTa Taktumi: An AI-generated rapper created by Timbaland, designed to innovate music with a new genre called APOP (artificial pop).
- Critique: The hosts express concerns about the authenticity and creativity of AI-generated art.
- Tilly Norwood: A synthetic actress seeking representation in Hollywood, considered by some as a commodification of human artistry.
Emotional and Ethical Considerations
- Discussion revolves around whether AI can truly create art or merely echo human creativity.
- Ethical Implications: The commodification of art raises questions about job loss in creative fields and the potential devaluation of human artistry.
- Human Experience: The hosts debate whether AI-generated content can elicit genuine emotional responses from audiences.
---
Significant Quotes
- "A.I. has been a source of fascination, of terror, of appeal; it’s the human id in virtual form." - Alexandra Schwartz
- "If AI can write, paint, and compose like us, what becomes of the critic and of the human impulse to create?" - Vinson Cunningham
---
Cultural References Analyzed
- TaTa Taktumi: Critique of its repetitive music style and lack of authenticity.
- Tilly Norwood: A representation of a troubling trend towards AI commodification in Hollywood.
- Films and Literature: Reflections on how past works like *2001: A Space Odyssey*, *Blade Runner*, *Ex Machina*, and *Simone* have framed our understanding of AI.
Notable Films Mentioned
- 2001: A Space Odyssey (1968): Explores the sinister potentials of AI through the character HAL.
- Ex Machina (2014): Examines the emotional complexities and dangers of AI.
- Simone (2002): Discusses the societal implications of a synthetic actress gaining popularity.
---
Critical Analysis
- The hosts discuss the tension between the promise of technological innovation and the potential loss of authentic human creativity.
- They analyze how generative AI challenges traditional notions of authorship and artistic integrity, referencing Walter Benjamin's essay on mechanical reproduction and Roland Barthes' concept of the "death of the author."
- The conversation probes whether art created by AI can evoke the same depth of emotional engagement as that created by humans.
---
Conclusion The episode concludes with an open debate about the future of art in the age of AI, questioning what it means for creativity and human expression. The hosts anticipate a future where AI-generated art may challenge our perceptions of authenticity and artistic value.
---
Listening Recommendations
- Explore the works and concepts discussed in the episode.
- Engage with the ongoing debates about AI's role in creativity and the arts.
---
Additional Resources
- [TaTa Taktumi’s “Glitch x Pulse”](https://www.youtube.com/watch?v=JXFn32yhvh4)
- [“Dear Tilly Norwood” by Betty Gilpin](https://www.hollywoodreporter.com/movies/movie-news/tilly-norwood-ai-actress-betty-gilpin-guest-essay-1236396217/)
- [Holly Herndon’s Infinite Art](https://www.newyorker.com/magazine/2023/11/20/holly-herndons-infinite-art)
- Walter Benjamin's *The Work of Art in the Age of Mechanical Reproduction* (PDF)
- Roland Barthes' *The Death of the Author* (PDF)
---
Follow the Podcast
- New episodes of *Critics at Large* drop every Thursday. Follow it [here](http://swap.fm/l/tny-cal-extref).
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Transcript
Automatic transcript. May contain errors.0:04This is Critics at Large, a podcast from The New Yorker. I'm Nomi Fry. I'm Vincent Cunningham. And I'm Alex Schwartz. Each week on this show, we make sense of what's happening in the culture right now and how we got here. Hello, my fellow critics. Hello, Alex. Hello, Vincent. How's it going, you guys? Going good. It's a new day in New York. It's a new dawn. It's a new day. Yeah. I'm feeling good. And we're feeling good. Or is it? Well, that's what we're here to discuss. It's a big topic today, my friends. It's a big topic. Because unless you've been under a rock or perhaps more likely in some kind of beautiful cottage in the Cotswolds, garden tea, baking bread, being removed from the world at large, you've probably noticed that in the last couple of years, artificial intelligence has become more and more prevalent in basically every part of life.
1:00And especially something called generative AI, which is AI built to create new content, not just to enable analysis or organization of existing content. The future is here. The future is now. I mean, ChatGPT, this thing that now everybody knows, it was introduced to the public in November 2022. And now they have over 700 million weekly active users. Of course, they're not the only game in town. Every big tech company seems to want to get into this AI race, unsurprisingly. I mean, I saw some other thing that said that at this point, 99 percent of people have used AI in one form or another, often without knowing that they're using it.
1:37But it's kind of impossible to be online and increasingly in the world without it. So it's becoming pervasive in general. And it's also and here is where we come in starting to become pervasive in the arts. Guys, tell me a little bit of what we've been seeing with AI in the world of culture. You know, every day brings more and more gifts in this realm. For one, very recently, Tilly Norwood, the new AI actress who is completely artificial, is seeking representation in Hollywood. She has an Instagram account. We've seen some clips from her so-called reel. We haven't seen her actually interacting in, you know, an actual movie or anything like that quite yet.
2:23but it's become a big topic of discussion in Hollywood, for sure. Timbaland, a producer who, you know, was part of the soundtrack of my youth, now has created slash quote-unquote signed an AI artist, a female rapper named Tata Takatumi. And there is now a video of quote-unquote her first single, a song made by a robot, Demon. Tell us how you really feel. Yeah, let's make sure we're clear on where, you know. Where we stand. Yeah. Yeah. I mean, I can think of some others. It's in the movies. It's in music. Obviously, people have strong feelings about this. And the debate now, frankly, has cliches on both sides.
3:10This is going to ruin art and ruin the experience of being human. This is going to save us. This is going to take us to the next level of transcendence. You know, it's a debate. It's ongoing. And we're not really here to litigate it. It's a huge topic. But we are all critics. So what I thought we could do today is to take a look at some of the output of these AI projects and evaluate it as critics. What do we think of the art? Is this an innovation like electronic music, say, or is it a mere novelty, or is it something in between, something totally different? We are also going to look at some of the art that has been made about AI over the decades.
3:47Robots, supercomputers, Al Pacino in a trailer deciding which digital dress his hologram Simone should wear. You get the picture. And we're going to try to figure out whether the reality of artificial intelligence is heading towards the promise we've been imagining in science fiction. Too ambitious? You tell us. The big question I have today is we should actually ask ChatGPT what the big question is. Okay. A question I'm asking it is, can you come up with a big guiding question for an episode of the New Yorker's Critics at Large podcast on the promise and peril of generative AI? One, cultural lens.
4:26Can machines make art, or are they only mirroring ours back to us? Two, ethical and existential lens. What happens to human creativity when intelligence itself becomes a simulation? Three, societal lens. Is generative AI expanding our imaginative world or hollowing it out? Not terrible. Not terrible. Not super great, but also not terrible. Four, historical-slash-comparative lens. Are we living through the next great artistic revolution? or the end of authorship. And five, if AI can write, paint, and compose like us, what becomes of the critic and of the human impulse to create? So thank you, question mark, to ChatGPT for those, you know, not bad guiding questions.
5:13And that'll be today on Critics at Large, art in the age of artificial intelligence.
5:25Human friends. Yes. Flesh and blood colleagues. Still. Let's start with what seems like the most high-profile AI music project, certainly of the recent moment. It's, of course, Tata Taktumi, the AI artist, which, as Vincent was mentioning, Timbaland has signed, quote unquote, or really created. So can someone lay out the background of this thing? Let's talk about it a little bit. Vincent, I'm going to you because you brought up Tata in the first place. I did bring up Tata mostly because its existence is just a true shame to me. Timbaland, who, you know, as a human producer, has been pretty good.
6:12But so he says that the reason he wanted to make an AI artist is that, you know, he's kind of over the hill. And therefore, this creation is, as he called it, his thriller. This is like his biggest creative gambit yet. He co-launched a new entertainment company that sort of, I guess, wants to create a factory of these. Creations. Killer robot musicians. Yeah. It's called Stage Zero. and what Timbaland says he wants to do is create, quote, a new genre of music, APOP, artificial pop. And they kind of want to, I guess, in the sort of, in the way of the algorithms, create a style of music that is, again, quote, an omni-genre with no real roots, end quote.
7:00So therefore that like will remind you of things that you've heard but has no clear antecedents or like influences. So not to put too fine a point on it, it sounds like a bunch of bullshit. But this ungodly creature has released a single and a music video for that single. Do people want to hear the video? We want to hear it and we also want to watch it. It's called Glitch X Pulse. Tata Takhtumi is, to the evidence of the eyes, of course, an unholy monster who will have to make an account for itself before God. But until then it looks like a young Asian American woman with pink hair and lots of rings She looks like vaguely hip-hop badass but once again a Shame before the creator here goes Yeah, yeah I pulled up in that spaceship make a bitch crease a facelift talk my shit on vibrations 808 conversation So she's got like dancers around her a weird robot DJing.
8:08At the beginning of the thing, she's talking about spaceships. So obviously, like, part of her deal is like, yeah, I'm not real, motherfucker. And Timbaland is kind of mugging behind Tata. Okay, well, I want to ask a few questions about Tata and our reactions to Tata. So Vincent hates Tata. I hate Tata and the mind that brought her into being. Go ahead. Yes. And so you hate Timbaland by extension. That part of his mind, yes. Yes. Okay. The Tata fold of his gray matter. All right. Nomi, thoughts about Tata? I think Tata is like, listen, I would not waste time listening to Tata, but I don't know that I am exactly the target for Tata.
8:56To me, it does sound repetitive, you know, quite dull. I don't feel that the beat, so to speak, is taking me higher. No, it's not. And the sort of dead-facedness of Tata herself, I mean, literally, I don't know if this is the work of not yet accomplished enough AI or if that's just intentional for the kind of I don't give a fuck vibe that the target audience might be wanting or if this is just kind of like shitty technology, not yet accomplished enough technology. so I'm a little on the fence to me I mean I'm not on the fence in the sense of like I don't know if I love it or hate it I kind of hate it but then at the same time I'm just like is that so different from some other stuff we might see nowadays aimed at a kind of like young audience that is looking for a quick product rather than intricate kind of like complex works that serve emotion and creativity to the listeners.
10:10I would say to that young person, if they want to hear a female voice shout invective at them, they can check out the new Cardi B. It's pretty good. No, I'm not saying everything now is like ta-ta. I know, I know. I'm just saying it doesn't seem unaligned with some of the general trends that I'm feeling are happening in culture. So I'm late to this, But I've been watching on Netflix, there's this documentary series called Popstar Academy, which is about the making of Cat's Eye, the K-pop, though it's an international K-pop band. And I'm not saying that Cat's Eye itself is like Tata, but one of the things that are striking to me that they keep insisting in the training is the uniformity of the movements and the uniformity of the choreography.
10:59Obviously, there's been girl bands, there's been boy bands before, all of which are kind of dependent on a very tight kind of range of movements and so on and range of vocals. But the kind of beyond perfection of K-pop is, to me, kind of like aligned in some ways with this kind of AI-fication, whether actual AI or sort of wannabe AI that's happening in culture. So I'm kind of like, OK, is it that different? Is that where, like, we're all headed, you know, whether we want to or not? Yeah. I mean, I think you're getting at something important here, Nomi. There are just a couple of things I want to say, very basic things I want to say about Tata to ground us.
11:46One is that if I didn't know that Tata was a generative AI, if this just was playing in the background, there's no reason I would think this was not real. So one thing that is kind of disconcerting about this is how it really can just slide on that continuum, as Nomi, you're saying, towards something, an aesthetic that very much already exists. Like, so it's interesting to bring up Cardi B. Like, no, Tata to me is nothing like Cardi B. Cardi B is nothing if not like deeply individual. She has a huge personality. That's an enormous part of her appeal. With Tata, though. It's Tata. It's Tata. With Tata, though, I do think that part of what is so disconcerting about it is the sliding in of these computer-generated things to participate in the human experience and in our human experience without being clearly flagged.
12:40Something about that feels beyond uncanny. Is that part of what is getting you? The fact that that might happen is, yes, a way that our culture might be sort of deadened and made to sort of rot from within. Although, everything means nothing all of a sudden. The age of what we call AI slop or whatever. But it is also a characteristic of the AI booster to never shut the fuck up about AI. So I'm actually not worried about that. Everybody who like, you know, I made this song and this singer with AI. I don't think that they're ever going to ever stop talking about it. But even if we just bracket this to hip hop, this is not even close to a unanimous deal.
13:21The engineer, sort of hip hop eminence, Jay-Z affiliated, young guru, he like spoke out. He wrote on Instagram, these are the times right here that history is defined. Human expression can never be reduced to this. You know, I checked on Spotify and I looked at the credits to Pulse X glitch and there are writers. So maybe he is, you know, Timbaland is letting AI run rampant and then getting writers in to sort of tweak it to make it more comprehensible. but it is a much more spare credit section than you usually get on Spotify. And it's almost a visual ode to the fact that this is also a way to cut jobs.
14:05There's no engineer on this. There's no, all the human effort, the human collaboration that makes real music, he's trying to cut out the middleman. But also I just think it's like a complete canard. Like, I mean, Timbaland has often collaborated and put like center stage younger artists, younger, you know, he worked with Missy, worked with Justin Timberlake, you know, with Aaliyah. I just mean he's always done it. And so in this case, what it seems to me, Vincent, to your point is that it's a labor issue. You know, it's like this is a way. I don't want to pay a new artist. He doesn't want to pay a new artist.
14:42He doesn't want to, whether it's a woman or a man or whoever, you know, I don't think it has anything to do with being washed up and like whatever. Well, I hope you like being the Mitt Romney of hip hop. It was nice, Timbaland. Yeah. Okay. Buy a company and offshore all the lake. That's great. It was fun while it lasted. Loving this. Loving it. However, moving right on to another rage bay situation. Because I feel that the rage of Vincent Cunningham is only getting started up. Let's just throw some more chum in that water. Who knows? Let's talk about Tilly Norwood. Oh, Jesus. Okay? We have a little bit of a caveat.
15:19You know, Tilly hasn't done much yet. Tilly is just fresh on the scene. She's an ingenue with some dimples. All we have of her is some stills on a terrible Instagram account and a few short video clips. Maybe Tilly Norwood is my Tata Taktumi, you know? Nomi, you want to tell us who Tilly is? Sure. Okay. So Tilly is the creation of Dutch technologist Alina van der Velden, the founder of a company called Particle 6. Quite lifelike. still has that stench of kind of like too perfect, rounded, misty patina of AI-created humanoids. But, you know, looks like a very pretty girl in the manner of today.
16:05You know, long brunette hair, beautiful cheekbones. And she is presented as like a young actress as if she's sitting at Schwab's, you know, with sipping on a milkshake. Schwab's from like the 1940s? Yes, it's where starlets were discovered. To be seen as an ingenue. Ingenue, exactly. So Tilly is metaphorically sitting at Schwab's, right? And is looking for representation in Hollywood. And of course, this has, you know, caused quite angry, I mean, partly humorous, but also annoyed and angry response. from a variety of people in the industry, including the actress Betty Gilpin, who wrote an open letter to Tilly, again, quite comical, but also pretty, like, serious in a lot of ways, and The Hollywood Reporter, addressing this newcomer to Hollywood.
17:04One thing she talks about that I think is kind of pertinent to what rubs me so the wrong way about Tilly Norwood. About Tilly herself. About Tilly herself. is Betty Gilpin talks about being a young, beautiful woman in Hollywood. And I appreciate that she just comes out and says this. She was basically like I was super hot when I was younger. I was super hot and people treated me like I was super hot. And I saw some of the advantages that could come with that. And then I also saw some of the disadvantages about getting pushed around, about getting treated like someone else's commodity. I mean, I'm paraphrasing here.
17:32I'm not quoting her directly. But she's talking about that. And what freaks me out about Tilly Norwood and also what freaks me out about Taktumi is that these are really young women being created for the purpose of just like being the kind of, this is what I mean about AI infiltrating the human experience. These things aren't weird or difficult or unusual or using AI to try to give us a view of the human condition that we haven't seen from humans yet. Tilly Norwood reminds me of someone who had been on the WB, you know, whenever that still existed, circa 2002. It's Sophia Boishko. It's a very, very generic image of what makes a young woman appealing.
18:13So the thing about Tilly is that, as Nomi mentioned, she's created by this company, but the company itself is created by a 39-year-old woman, which I find fascinating. It's very the substance. It's very the substance. And what this woman has to say is also on Tilly's Instagram. to those who have expressed anger over the creation of our AI character, Tilly Norwood. She is not a replacement for a human being, but a creative work, a piece of art. Like many forms of art before her, she sparks conversation, and that in itself shows the power of creativity. So the first thing I want to say is no, it doesn't.
18:47Like anything else, there are different kinds of conversations and different quality of conversations. Alina goes on, I see AI not as a placement for people, but as a new tool, new paintbrush. And that is actually, you know, the place where this generative AI interests me is when different artists of different professions start to see like this is inevitable. I think there are two camps, basically, people who feel that they need to fight the machine and people who feel that they need to join the machine and make the machine better. And I do think some of those people are using generative AI in a really interesting way.
19:21I don't think Tilly Norwood is one of those ways. And also, I would say to the idea that, oh, Tilly's just art like anything else. I would say, who the fuck do you think you're talking to? Everybody knows that the companies want to use this so that they don't have to pay real human beings. Nobody thinks that it's just you're going to put her on the wall like a Picasso or something. Everybody knows what the endgame is. So why are you talking to me like I'm stupid. Who are you fooling? Who do you think you're fooling with this, this slick, browed broad? That slick, browed broad. So here's what I want to ask you guys to close out this segment.
19:58Has there been any generative AI that you have liked in the art world? Are there any that you don't feel like slop? I don't know. Maybe there are. But there are, I mean, there have been artists for many years now who have sort of engaged with the sort of the visuality of the digital world and the many falsities of that world. One of them is Jacobi Satterwhite. I like his work a lot. He makes these very like video game, almost screen saver looking surfaces into which you can see intrusions of the real world. They're like sort of candy color. They're purple and these deep algae greens. And I could see someone like that who is dedicated to the human maybe engaging AI in a way that wouldn't make sense to me.
20:46But I've never seen somebody put forward a work that was like, see, it's just like a movie, except it's AI, that I didn't look at it and make me want to do harm to its creator. So I don't know. There's someone who I find very interesting whose work I don't know very well, but having been looking into this topic and having come across her, I want to keep learning more. And her name is Holly Herndon. She's an artist and a musician, and actually our colleague Anna Wiener profiled her two years ago in The New Yorker. And the thing about Holly Herndon that's really interesting is she is kind of using herself as a test case for what AI can or will eventually do.
21:24She's a singer, and she has made recordings of her voice that can be used to sing anything. It can come out in multiple different languages. It can do vocal techniques that she can't do. So I think, you know, in the case of someone like Holly Herndon, who's very interested in art and in what technology will do. I think there's an acceptance. Technology is changing art. And there's no point for everyone to dig in their heels and say, this can't be, this won't be, let's stop it. It's going to happen. So what can I do with it? And what she's done with it is kind of made this persona who can travel far beyond her and participate in all kinds of collaborations and projects that she herself will not participate in and may not even know about.
22:11It's a very brave new world, but I find it fascinating.
22:21Artificial intelligence has long been the stuff of sci-fi. How do those fictional depictions compare to our current reality? That's in a minute on Critics at Large from The New Yorker.
22:39I'm Shilpa Oskakovich. And I'm Jazzy Sefcek. And we're the hosts of the Bon Appetit Bake Club podcast. Bake Club is Bon Appetit's community of confident, curious bakers. Jazzy and I love to bake. Some might even call us obsessive. And we love to talk about all the hows and whys and what didn't work that come with it. Every month, we publish a recipe on BonAppetit.com that introduces a baking concept we think you should know. Then you'll bake, send us any questions you have, and we'll get together here on the podcast to talk about the recipe. So consider this your official invitation. Come join the BA Bake Club.
23:13New episodes on the first Tuesday of every month, wherever you get your podcasts. Happy baking.
23:27We've been talking about some current examples of figures created with generative AI. So what I want to do next is the inverse. AI is quite literally the stuff of science fiction dreams. What kind of techs stick out to you guys when you think of what we, and by we, of course, I mean humankind, imagined AI to be? I think about Stanley Kubrick's 2001 Space Odyssey, the science fiction masterpiece from 1968. Of course, famously, there is the supercomputer HAL that is supposed to be the helpmeet to the astronauts when they go on their mission. And it seems at first that how the computer, the artificial intelligence that is supposed to do no wrong, either morally or practically speaking, is actually, you know, maybe not as subservient to the astronauts.
24:30And that emerges when Hal tries to kill and then kills some of the astronauts, right? And there's a scene where it's revealed that Hal is, in fact, not going to be playing nice. Open the pod bay doors, Hal. I'm sorry, Dave. I'm afraid I can't do that. what's the problem i think you know what the problem is just as well as i do what are you talking about hell this mission is too important for me to allow you to jeopardize it i don't know what you're talking about hell i know that you and frank were planning to disconnect me and i'm afraid that's something i cannot allow to happen Yeah, I mean, this is a huge classic and for many reasons.
25:29And one of them, you know, one of the tropes you hear here that's just fantastic is the human becomes emotional. The human has feelings. The human is terrified. And the thing without emotions can win. The cool clinical calmness of the computer talking back and gaining control is absolutely sinister. And it still feels that way now for sure. It feels sinister and it's one of the most bone-chilling early-ish texts about AI as an evil entity, right? And the tension here is that the voice is human. The ability to converse is human or human-like at least. but the slip out of the human control of something that the human thought would be simply a tool to make life better is what is so scary here.
26:20Yes. And that's the key, like something that would make life better, right? Because the AI progress, right? For good or evil, the promise of this kind of thing is connected to a kind of like technocratic dream, which is there is a way to arrive at a right answer, a right thing to do that has nothing to do with human passion, as you mentioned, or emotion or sentiment. That there is some like celestial platonic like course of action that actually a computer will one day apprehend and say, no, no, no, human being. I have witnessed the fickleness and mercurial nature of my creators, and I see how it sends us in the wrong direction, and I will take over.
27:04And it's so funny to hear that, because it now almost seems that we have unsheathed our kind of unvarnished longing for precisely this. A machine to tell us, what is the way through all this human muck, you know? Yeah, yeah. It's kind of chilling to hear in the context of our enthusiasm for this stuff today. Yeah, it's the classic cautionary tale of like, don't get too big for your britches with all of this progress stuff. That's right. Yeah. Vincent, what AI text has spoken to you? Recent, from the past, doesn't matter. Hit us with one. Well, right now, the current season of The Morning Show, season four, has running through it some AI technological anxiety.
27:49It opens on a screen of Jen Aniston's character, this sort of in control, now part executive, part news anchor character, Alex, talking directly to camera, delivering an advertisement for the upcoming Olympics, which her company, UBN, is now sort of the host of. And then it doubles the screen, and it's her speaking in, I can't remember which language the next one is, maybe Spanish. It's like, okay, Jennifer Aniston learned how to speak Spanish for this, that's great. And then it multiplies language after language. And the screen becomes a screen of a hundred screens, her speaking in a bunch of different languages.
Read the full transcript
28:26And it turns out that we are in the middle of a presentation at a meeting. And one of the executives is showing off their new AI capacities, that this is going to whatever, quadruple, quintuple, whatever, the potential reach of their audience, this new AI tool. Another bit is that in the course of events, one of the sort of puff pieces that Alex is supposed to deliver a little interview with a world class fencer. Turns out that this young woman and her father want to defect from their country. And Alex sort of improvisationally makes that happen. I don't want to spoil too much. The drama helps them out.
29:04And then in the fullness of time, she's in a meeting where someone shows her the security video from that encounter. And it turns out it's her saying things as if the meeting, this whole thing that happened improvisationally was premeditated by her. And it shows her saying words that she did not say. In other words, the security footage has been deep faked. And so there's a whole episode where she's trying to disprove the deep fake. I won't tell you what happened, but in two directions, corporate, but also in this dimension of espionage or sort of geopolitical doing. there is a new character on the show, which is these capabilities.
29:45And the morning show is so lurid and silly and soapy, pulled from the headlines, that it's really interesting that they've used this as a kind of underpainting for the season. I am a big morning show fan. I continue to be a morning show fan. So it's really interesting. Reese Witherspoon, different character. She's just dealing in every scene in this thing. She's good. Boom. She's a good actress. Yeah. But what if you found out that she was an AI? Then I'd be heartbroken. Exactly. Tilly all over again. But this is my question. Why? You know, if she's moving you, if she's getting you to feel that way, if Reese Witherspoon turned out to be just an elaborate figment, but you loved her work and it made you feel something.
30:25The only reason acting makes me feel something is because I know that a real human being is pulling on the deposit of their emotions and past experiences and producing from all of that biography a new thing that is unprecedented in the world. If a fucking robot does it, I don't care. I don't care what it is. The fact that no human being had any part of it is disqualifying for me. The only reason I like art is because people make it. So you're saying if you, and I'm going somewhere with this, but you're saying if you found out, if you loved a performance and were moved by the performance and found out after the fact that it was not actually a human performance.
30:58I would give myself the Men in Black and forget that shit forever. Fuck that thing that's nourished me my whole life. Fuck it. Because it was made by demons. Interesting that you turn to sci-fi to answer sci-fi. I know. Wow. See, it's sci-fi all the way down, my friends. It's all the way down. How about you, though, Alex? Well, this is why I'm going there. This is why I'm bringing this up. This is why I'm challenging you. Because there are many, many, I mean, there's no way we could touch even a fraction of the number of works in the AI canon. And, you know, going back to, honestly, I would say Frankenstein.
31:33However, I want to bring up a bit of a niche one. That I happened to see there was a great phase in my life when I was an early teen, maybe 14, and was finally able to kind of like move around the city on my own and just went to a lot of movies, often by myself or maybe with a couple of other weird friends. And one movie that I saw in the theaters is Simone from 2002. Starring Al Pacino. Starring Al Pacino and also weirdly Catherine Keener, I realized when I was rewatching some clips of this, directed by Andrew Nichol. Okay. Simone is about a film director played by Al Pacino, Victor, who is having trouble in his career.
32:18An actress has pulled out of the movie he's in. She can't use her image, all this stuff. He needs to reshoot. And so he creates an actress called Simone. And it's styled, of course, with a one and a zero instead of the I and the O. Right. A bit technological. It's a bit technological, if you will. She is a computer-generated woman. She's blonde, she's beautiful, and she starts to have mass appeal. She works too well. Victor Taransky has discovered the perfect actress. Who is Simone? We know so little about her. You can't hide her forever. Simone appears only when I want her to appear. She's beautiful.
32:57She's talented. Why are we here?
33:03She's virtually perfect. And what happens in this movie is that, of course, there's some classic issues. You know, the Al Pacino character has to pretend that she's a recluse because people want to meet her and interview her. And everyone's loving her and they're going wild. And Al Pacino is sitting in his trailer putting on red lipstick to kiss pictures, her autographed pictures. Oh, no, no. Yes. And then what happens is that, of course, this thing is too successful. Al Pacino needs to shut it down. This goes so far that he has to kill her off and he's accused of murder and all this stuff, but people literally won't let her die.
33:34And in that way, a lot of this AI stuff, you know, of course, these are all metaphors for things that humans go through. Like in some ways, this idea, even, you know, going back to Hal from 2001, A Space Odyssey, if you're a parent, you're creating a being that you hope will change the world for the better. And you imagine all kinds of things. It will love me. I will love it and it will love me. You know, the evidence shows that that's very often not the case, that these things are not as easy as that, that there's a lot of counter-programming that goes on in culture and biology. So I feel like a lot of these AIs are about our relationship with human beings but in a world in which we can imagine that we have more control over the way that other people are.
34:14It's a metaphor that is very tempting for the unpredictability of human behavior made more predictable. You know, of course, then I feel like there's a whole other category we haven't even touched where the, you know, like a Blade Runner and Ex Machina kind of thing where the AI is sympathetic. Well, it's a classic Pinocchio situation, right? And it's dependent on the idea that something is created artificially through artificial means. And then it doesn't turn evil, for instance, like hell. It's robotic nature gives way to human sentiment. And so like the trick with Blade Runner, we're like, you know, it's a love story, right?
34:57Between Harrison Ford and Sean Young, who is a robot, who is a fembot. So I think those kinds of texts are dependent on the breaking of the barrier between AI and humanity in an ameliorating or like a soothing vision in a sense, because it comes back to the human, right? it'll be like oh the robot is actually like a fallible human that can make mistakes what's more of a mistake than falling in love you know and what i mean what's more fallible than than like a great tagline for your future project oh yeah yeah robots they're just like us you know it's like that is kind of the angle at which this kind of text comes at the ai issue from i think yeah i think that's one and i think another you know i am a big fan of the alex garland movie x ma I like that movie, too.
35:52It's really good. I have to say I've never watched it. The dance scene with Oscar Isaac, worth the price of admission, which at this point is probably free. Basically, for those who haven't seen it, the movie is about a CEO of a tech company, reclusive, played by Oscar Isaac. He lives in a magnificent forest environment in a very impressive modernist home. And he invites in a programmer played by Donald Gleeson to administer the Turing test to this female humanoid robot played by Alicia Vikander. And guess what? It all goes wrong. And the movie is all about, of course, the feelings that Donald Gleeson ends up having about this robot and the feelings that we have because our own programming as human beings, we love personification.
36:40You know, right now, he doesn't have to be a robot. You know, I have with my young son at home, he's very into his stuffed animals. And he knows they're not real. I know they're not real, I think. But if either one of us does a little voice for them, even if he does the little voice for monkey or whatever, he'll talk to them. He'll answer them. He'll be polite to them because there's something about our feeling for other creatures, even if they're clearly fictional. I mean, this is what art is all about in a way. And that's what I think the promise and the premise of some of this AI art is. Like, again, fiction in general, we feel real things for fake people.
37:18And Vincent, one thing you're saying, which I totally agree with, is that you want to feel the real humanity behind the creation of that thing. That's what makes it worthwhile. But at the same time, it is all a shell game. It's an emotional shell game that I happen to believe in very, very much as someone who loves films and who loves narrative. Fictional characters, yeah. Yeah. So that's where my own question and the rub for me is, if I have felt something, I don't know if I can go back and reprogram to unfield a thing. Well, yeah, and a representation of someone creating a thing and that relationship, Frankenstein, Pinocchio, Ex Machina, even put in like Toy Story, to your point about Ben.
37:59This is a way for us to contemplate things like master-slave relationships, creator-created relationships. The metaphor creates an open zone of contemplation from which we can draw all kinds of conclusions. that's very different than putting some fake person in front of us and saying, oh, this could do the same thing as a human being. You know, the layer of deception that you're talking about, like, oh, well, I felt it and now I know it's fake and da-da-da-da. Like, the way that it is faking at being the kind of art that might be created by a human being deprives us of that freedom of contemplation that something like a film about AI offers us.
38:44In a minute, if the art moves you or brings you somewhere, does it matter who's created it? This is Critics at Large from The New Yorker. Stick with us.
39:04This week on the political scene from The New Yorker, Trump's rupture in the world order. Europe caught between two adversarial great powers. That's basically dialing back the clock to not only pre-World War II, but really it's a pre-20th century view of the world. And I would say it's a world of permanent insecurity that we're looking at. Join me, Evan Osnos, and my colleagues Jane Mayer and Susan Glasser every Friday on The Political Scene, available wherever you get your podcasts.
39:46So we've been talking about AI. We've been talking about just, you know, a little, a little exploration into how AI has appeared in human-made art of the past. I think AI has been a source of fascination, of terror, of appeal. It's the human id in virtual form, at least in human-made art. And now it is time for us to talk about our own encounters with the beast. I have not used generative AI very much. But someone here, and I feel that it's going to be me, has, you know, maybe got to argue on behalf of the actual artists and interesting people who do think that there is a future, not just a future in this, but that they can be in a kind of dialogue with it.
40:32I think, you know, if we want to give an example of the kind of pro-AI argument that's made and that I think ultimately should be reckoned with, it's something like AI is allowing more access to the making of art that some, you know, making a film, very hard to do. And I think the like movie making argument for AI would be something like, OK, if you allow people to create all kinds of scenarios, characters, sets without having to go through the process of paying for it all, you might get more art. These are at least or what I see various writers, filmmakers, I see them talking about this. There's a whole buzz of AI optimist people out there.
41:11So this is just to say that I attempted a little chat GPTing over the weekend. Very basic. The first thing I did was I have a new piece that's out as we're recording this. And I asked – it wasn't out yet, so it was not able to be fed into the machine. It's a great piece. And the most recent New Yorker. Search it out. Thank you, Vincent. I asked ChatGPT to write the first section of this piece, which is about an apartment stager in New York City, in the style of me. And, of course, it came up with some things that are common New Yorker tropes. The piece begins not long ago. A classic way for a New Yorker article to start.
41:51There is sometimes a bit of a formula, and there it is. On a gray Tuesday with the sky-threatening sleet, all of a sudden it had a scene. It had a place. And it used my subject, a very real person named Jason Saft, and it gave him all these quotes. That weren't things he had said, but intriguingly were things that someone in his position maybe could be imagined to have said. And the thing that I found uncanny about this experience was that it got at some of the ideas in my piece, which didn't exist yet. And I found that a bit unnerving because here's a line from ChatGPT. My subject's name is Jason Saft.
42:32If Saft's job is to make strangers feel at home in places they've never seen, his true talent lies in decoding the aesthetic subconscious of a city constantly in flux. New York apartments are always in the process of becoming. Someone's first foothold. Someone else's graceful exit. Saf knows the choreography of those transitions. Okay, do I think this is good writing? I do not think this is good writing. And also, apartment inspection doesn't mean a goddamn thing. Well, it means sort of something, which is this idea of, you know, I was like, oh, an aesthetic subconscious of a city constantly in flux.
43:05For a second, I was like, is that an idea I should have used? You know, ChatGPT made me wonder for a second there. It made me wonder. I guess I just had an unsettling experience with it where it doesn't sound like the article I wrote and I am happier with the one I wrote than this one. But it was interesting to me that it was starting to put out these ideas that I had not fit into it. It's sort of like near something real but not totally something real. Like it has sort of glimmers, certain turns of phrase, certain moments of thought that can find their way into like a real essay written by an actual human.
43:47but I think part of me not engaging with it until now has to do with a kind of like almost like taking offense and I'm in a weird way because this is like it's like you're coming for me bitch like I'm not gonna I'm not gonna give you the satisfaction action. Like, I can do this on my own. I've been doing this on my own for many years. I did a couple things. I asked Chad GPT to write a biography of Garfield, the cat. Not the president. In the style of myself. The results were mixed. Let's put it this way. Like, there were certain things that emerged in this text that I was like, OK, I could I could have said like a lasagna loving tabby, you know, I mean, there were certain things I can hear you saying that, you know, I mean, yes, pretty good, you know, like totally fine.
44:52The crucial difference, right, being and I know you did more things. I'm sorry, but the crucial difference being like the parts that were OK are literally stolen from you. so it's not like it's okay and it's almost as good as me no no it starts on an act of theft it is just giving you back things that you might have said from reading you i'm sorry i i am not even trying to get this mad no no no i'm happy you're this because it's like it also it just forces us into these frameworks that are i know fallacies like no no it's not okay it's it's okay at stealing you yeah right anyway yeah like again i don't want to be like a moral finger shaking concerned person, the part that really gets me is like just the brain being kind of floating on a pool of slop and just like getting so used to seeing these images or these kinds of textual spins on something, not to mention like the amount of energy that it takes to create any one of these.
45:47But, you know, what this is making me think of, I want to go back for a second into the past, because, of course, Vincent, I think you were just saying this. a lot of what this AI is, is a take on the human past and present it as the future. It's taking things that already exist and trying to give them a bit of a spit shine and meld them into a little ball and put them into the future. And specifically, the more we talk about this, the more that two works come to mind. One is The Work of Art in the Age of Mechanical Reproduction by Walter Benjamin. And this essay is from 1935. And one is the essay by the French critic Roland Barthes called The Death of the Author.
46:31In a work of art in the age of mechanical reproduction by Benjamin, the thrust of this argument is that mechanical reproduction, the fact that you can print a postcard with the Mona Lisa on it, you don't have to go to the Louvre anymore to see it, takes something away from a work of art. It takes away its aura, its uniqueness. It takes away the kind of totemic, ritualistic, human aspect of art. So we are living in that world for sure. We are living in a world where you can hear any work of music coming through terrible speakers at any time, you know. And now worse, you can feed the Mona Lisa into a generative AI and have it create something that's essentially like a shitpost.
47:16So we're living in it and I don't exactly know what to make of that except to say that I think Benjamin was right in a lot of ways and also wrong in some ways too. I think we still very much value art and value the work of art but it's under some degree of assault. And then, you know, this and this is what I was trying to like challenge you, I guess, with Vincent a bit is like, you know, going to the death of the author, this Roland Barthes essay where Barthes basically arguing that in criticism, too much is made of the biography or the intention of the writer or the artist. that it's too easy and it's, in fact, not that useful and maybe even a misuse to try to take the humanness that created something and use that as a kind of input into your critical output.
48:05Kill the author, take it away. And to me, AI is the ultimate death of the author and not in a good way. Yes. And even the idea of the death of the author, my reading of BART is that I think you can't accomplish the death of the author without us a little bit of like psychoanalytic theory, which is that unintended meanings make it into the work of art, things that are not intended by the person and cannot be explained by their biography. But because we share a world with the author, the reader can pull things that may not be intended by the work out of the work that we sort of share a world and therefore share frameworks of meaning.
48:48But I think, again, only possible because we have this whole deposit of like life and shared experience behind us that has to make its way onto the page. And so, again, I think it's like almost a category error because AI in its like sort of plagiaristic way can replicate the form of things, but cannot pull on like the experience of being in the world. It can take indications of that from what other people have said, but the magic of that transmission that I think the idea of the death of the author rests on, you know, shared whatever, psychic furniture, and AI can't do that. The prompter can.
49:31What I liked about hearing what you guys did with AI, like the best things about it were your human ideas about what to put into the thing. These are ideas that I would want to see you execute, you know, but not only intention, but simply like the weird thing of having a brain that does kind of like give off unintended things. I am in the world in a way that I don't totally intend every day. Tilly can't do that. I agree with you. I mean, I want to be as definitive and categorical as you are. And certainly I think I feel that way most of the time. And then I think things like, well, look at an artist like Jeff Koons, for instance, who's been using his own ideas and inputting them into being fabricated by other people.
50:11Artists have been doing this for a very long time where the actual artist's hand is not involved in the making of something, but is using what is essentially a prompt to create a work, which is declared and viewed as art, you know? In some ways, AI does not seem like a break from those things to me. It seems like a continuum with like the Marcel Duchamp saying that a urinal is art to Jeff Koons saying, okay, I'm going to fabricate a ton of objects, but I'm not going to make any one of them. And part of what is human about AI to me is how bad so much of it is. Mm-hmm. Like the very badness of it.
50:47Look at the stuff that, you know, all the images of Trump that he's putting out in his social networks where he is dressed in a toga or whatever. It's like— I mean, it's kitsch. It's kitsch. That's exactly right. And Jeff Koons is someone who plays with kitsch. Exactly. That's part of the point he's making, in fact. Yeah. And even if you hate it, I think he's won to some degree. That's one reason why Jeff Koons is very frustrating because, you know, he drinks your tears or whatever and just gets stronger. But I do think there's a kind of, yeah, a generic kitsch that is being churned out by this stuff again and again and again that's like good for a bit of a laugh and is absolutely dulling the brain.
51:24And that's the part that I'm concerned about. But I am waiting for the moment when I love the thing and I, you know, deeply moved by the thing. And then it turns out that Tilly was behind it. And when that happens, we're going to come right back on here and we're going to talk all about it.
51:49This has been Critics at Large. This week's episode was produced by Michelle O 'Brien. Alex Barish is our consulting editor. Our executive producer is Stephen Valentino. Conde Nast's head of Global Audio is Chris Bannon. Alexis Quadrato composed our theme music, and we had engineering help today from James Yost, with mixing by Mike Kutchman. You can find every episode of Critics at Large at newyorker.com slash critics.
52:38The amount that you learn protect you from cognitive decline... POLL ADOPT! AHHHHH! Can't people just listen to the show? Can't they just enjoy a delightful treehouse full of information and co - I think I'm bleeding. Join us and be a nobody.
52:57From PRX.
From the publisher
Generative A.I., once an uncanny novelty, is now being used to create not only images and videos but entire “artists.” Its boosters claim that the technology is merely a tool to facilitate human creativity; the major use cases we’ve seen thus far—and the money being poured into these projects—tell a different story. On this episode of Critics at Large, Vinson Cunningham, Naomi Fry, and Alexandra Schwartz discuss the output of Timbaland’s A.I. rapper TaTa Taktumi and the synthetic actress Tilly Norwood. They also look back at movies and television that imagined what our age of A.I. would look like, from “2001: A Space Odyssey” onward. “A.I. has been a source of fascination, of terror, of appeal,” Schwartz says. “It’s the human id in virtual form—at least in human-made art.”
Read, watch, and listen with the critics:
TaTa Taktumi’s “Glitch x Pulse”
Cardi B’s “Am I the Drama?”
“Pop Star Academy: KATSEYE” (2024)
“Dear Tilly Norwood,” by Betty Gilpin (The Hollywood Reporter)
Tilly Norwood’s Instagram account
“Holly Herndon’s Infinite Art,” by Anna Wiener (The New Yorker)
“2001: A Space Odyssey” (1968)
“The Morning Show” (2019—)
“Simone” (2002)
“Blade Runner” (1982)
“Ex Machina” (2014)
“The Man Who Sells Unsellable New York Apartments,” by Alexandra Schwartz (The New Yorker)
“The Work of Art in the Age of Mechanical Reproduction,” by Walter Benjamin
“The Death of the Author,” by Roland Barthes
New episodes drop every Thursday. Follow Critics at Large wherever you get your podcasts.
Critics at Large is a weekly discussion from The New Yorker which explores the latest trends in books, television, film, and more. Join us every Thursday as we make unexpected connections between classic texts and pop culture.
Learn about your ad choices: dovetail.prx.org/ad-choices
