00:02:35.680Yes. So as a journalist, I had been writing about technology for a long time.
00:02:41.380And in 2017, I wrote a profile of Sam Altman, who was would go on to become the CEO of OpenAI. And when I was reporting on that, he was talking about OpenAI and I did a little reporting on it. But then a couple of years later, I started hearing about this thing called GPT and then it was called GPT-2 and then it was called GPT-3. And it was this technology that could produce language.
00:03:07.260And it wasn't a chat bot then. The way it worked then was like you would type some words and press a button and then it would kind of continue that as if it were you.
00:03:16.460And so I still had his phone number from when I had reported on that profile.
00:03:20.820So I texted him and said, hey, can I I heard you're letting researchers and journalists try out this GPT thing.
00:03:27.180Can I try it out? And he connected me, you know, with OpenAI's PR person or whatever.
00:03:32.380And they said, sure, you can have access.
00:03:34.920And so I was thinking as a journalist, I was thinking a lot at the time, as I always do, about like what tech companies promise us, like what their marketing message is, basically.
00:03:47.140And I was like, well, what is the promise of a technology that can like complete your sentence?
00:09:17.800I'm, like, just thinking back to, like, my grad school days of, like, reader response theory and, like, oh, the meaning is made in this moment.
00:09:24.800But what if the text itself is like there is no author, right?
00:09:29.000The author is dead and all of the meaning is coming from how we receive it in that moment.
00:09:41.140So we both grew up screwing around on the Internet.
00:09:45.560And I think sometimes, like, I just, I was watching my friend's kids the other day try to find something that, like, a video that they were looking for on YouTube.
00:09:56.180And they just, like, they're just not playing as much.
00:09:58.600They're not just, they're just not being as weird as we were.0.90
00:10:00.980And I think that we had this opportunity to really kind of, for lack of a better word, like, fuck around with the technology.
00:10:06.860Like, try to make it work in weird ways.
00:10:11.100I think, I mean, I think that continues to happen.
00:10:14.940I would say I am like you're doing well no but I'm thinking about TikTok which I I'm like not
00:10:21.420on some of these platforms right but I'm thinking about the way in which like people do like invent
00:10:27.100these new ways of talking or behaving or interacting with each other like on platforms
00:10:33.380like TikTok where you know you have the freedom to do that and yet at the same time it's all
00:10:38.580like mediated by algorithms in a way that was maybe I was going to say less true for us but
00:10:46.100maybe not even at all true for us because the internet didn't work that way so whereas back
00:10:50.960then you and I could put all kinds of weird experiments onto the internet or like be versions
00:10:57.460of ourselves in interesting ways um and it was just us interacting with other people through a
00:11:03.780screen now there's always this algorithm in between right so now you're always performing
00:11:09.660for an algorithm in addition to for other people and I would say that that like corrupts the ability
00:11:16.700of like 22 year olds today to feel as fully free as we might have felt as 22 year olds or 13 year
00:11:24.260olds right right and just feeling like you are passive to it instead of an active manipulator
00:11:31.480of it. Yeah. The one thing I do want to do right from the top is a lot of people online get cranky
00:11:38.980about using the bucket term AI to talk about some of the things that we're going to talk about
00:11:45.040today. So when you think about the very like the various containers within this larger designation
00:11:52.120of AI, what do you think of? Because like I'm thinking of LLMs and chatbots and that's different
00:12:00.240than, I don't know, like predictive text when you're texting? Is it also different from like
00:12:10.000using Grammarly? Yeah. How do you think about it? Yeah. So, I mean, I think there are some of
00:12:15.880these things that are similar. Like I would argue that using Grammarly actually maybe is in a
00:12:20.640similar category as using ChatGPT. However, people will sometimes refer to like AI that can do the
00:12:28.340work of a radiologist, like can read a scan, right? And that's something different. So maybe
00:12:34.740we say for the purposes of this conversation, when we use the shorthand AI, what we're referring to
00:12:40.200to use just simple language is AI that can generate text and AI that can generate images
00:12:48.260and videos. Yeah. Okay. I like that. I like us to have our terms at the beginning. And then the
00:12:54.620other thing we often in this show like try to kind of put our cards on the table in terms of
00:13:00.140positionality and usually that's like how we relate to a text like how did you did you read
00:13:06.600this growing up like do you have feelings about this like we did an episode about little house
00:13:11.240on the prairie and i'm like this is my positionality with little house on the prairie
00:13:14.640but i i think it's important here too so i'll just say that i am very resistant to
00:13:20.160using any sort of LLM predictive text tools in any of my work to the extent that I
00:13:28.100have all of these complex settings turned off in my Chrome, my Google searches, so that I don't
00:13:35.420get even the summary when I do a Google search. I don't like that. To me, it short-circuits
00:13:40.340something. But I'm frustrated with it because now this Google search doesn't work the same
00:13:46.140way that it used to. So it doesn't give me the good stuff that I want. I also feel very frustrated
00:13:51.800with the way that people have conversations about AI online because it's very black or white. Like1.00
00:13:57.260it's either like pure boosterism or if you use LLMs in any capacity in your work, like you should
00:14:06.540be canceled kind of situation, right? So how do you, how are you feeling right now about it?
00:14:13.300Yeah, I know. I love that opportunity for us to just get to lay it out that way. I myself have used AI in my work specifically as a way to show how these products function and try to shed light on the power dynamics and the market dynamics inherent in how these products work.
00:14:43.300So for me, like I'm always trying to I think of myself as using these products as like a way to critique them.
00:14:49.780That said, you know, some people might say like one thing I could potentially say is like that's how I use them.
00:14:56.320And that's the only way I think anybody should use them.
00:15:00.000I actually don't have a position on like how people should or shouldn't use AI.
00:15:06.740I think for us as artists or writers or journalists, it does feel really important. And I think I would kind of draw a line in the sand around transparency and disclosure. It feels really important that we tell people whether and how we are using these products. Maybe it's important that we tell people what the goal is in using these products if we do.
00:15:31.140So I wouldn't say, like, you know, you can't be an artist, you can't be a journalist, you can't be, you know, a professional in the world if you use these in any capacity.
00:15:39.800But I would encourage people to sort of, like, think critically about how and why they are using these products, what the goal is, what the benefit is, what the cost is, not just to them, but more broadly of using these kinds of products.
00:15:53.320Yeah, just, like, think much more contextually and sociologically.
00:15:57.960Well, then I, you know, we can add one more thing actually that I also just wanted to say sometimes when we talk, just again, as we like are defining our terms, when we talk about AI, usually we're talking about AI products from big technology companies.
00:16:12.100So, like, Google and OpenAI and Anthropic. And I do know a lot of artists who are using AI in their practice, but they're, like, building their own small models. They're collaborating with other artists in a way that doesn't use environmental resources the way that the really big corporate models do that aren't bound up in labor exploitation or exploitation of other people's creative work in the way that the big models do.
00:16:41.220And that's an interesting distinction to make, too.
00:16:43.880Yeah. And the other thing, like, I always think we all are like we're part of all like as authors.
00:16:50.080We have a stake in this, a literal financial stake because of the anthropic settlement.
00:16:55.020Like was your book in there? Were any of your books in there?
00:16:57.760Mine weren't because they were all pre the cutoff date.
00:17:02.400Yeah. My first three books, all of them, all in there.
00:24:32.600Today's episode is brought to you by Ollie.0.68
00:24:35.560August is National Dog Poop Month, which I think is a made up thing, but I get it.
00:24:41.440Sometimes I'm just like, all I do is count the number of dog poops that I've picked up in a day.
00:24:48.220And sometimes Charlie and I like report back about the consistency of the poops.
00:24:52.940It's a way that you talk about your dog's health.
00:24:56.320And the one thing about when your dog is eating Ollie, first of all, they're smaller, so you just have to deal with less mess.
00:25:03.880Sometimes I feel like I have a bouquet of poop bags when I'm coming home from a walk.
00:25:09.200Because let's say both of the dogs have two poops for some reason, and you're just like, oh my gosh.
00:25:15.040It's much better when your dog is eating high-quality food that creates fewer poops that are also not messy and stinky.
00:25:23.520And if anyone gets being this invested in the quality of your dog's poop, it's Ollie.
00:25:27.540They deliver fresh, human-grade food that is gently cooked to maximize your dog's nutrition and also make their poop smaller.
00:25:35.240Ollie's fresh recipes are developed by real chefs with gut-friendly ingredients that support better digestion, energy, skin health, and better poops.
00:25:43.440They have flavors like fresh beef and sweet potatoes or fresh turkey and blueberries,
00:25:49.180And you'll notice the difference fast.
00:25:51.340Smaller, firmer, less stinky poops, and a much happier belly.
00:25:55.580From the moment you start your subscription, everything is tailored to your pup.
00:25:58.800So you get the perfectly portioned meals.
00:26:00.880You also get this pup tainer with a scoop for easy storing and serving so you don't
00:26:05.100have the gross tin can in your fridge, and also a smooth food transition guide so that
00:26:09.720their digestion stays on track as you transition to Ollie.
00:26:12.860This summer, treat your pup to better food and perfect poops. Head to ollie.com slash culture, tell them all about your dog, and use code culture to get 70% off your welcome kit when you subscribe today. It's super easy and risk-free with their obsession guarantee. That's O-L-L-I-E dot com slash culture and enter code culture to get 70% off your first box. Ollie, feed the obsession.
00:26:38.680Today's episode is brought to you by Blissey.
00:26:41.520Melody, how would you rate the frizz factor going on with your hair right now?
00:26:45.960I'm experiencing a lot of breakage right now because pool season, as we've discussed before,
00:26:52.160and also my CPAP mask causes a lot of breakage.
00:34:14.060You've got kids or 20 year olds in your classroom and you need to imagine like what the world
00:34:18.660is going to be and what they're going to need.
00:34:20.260And so if you can plant this message that you know what the future is going to be and it's going to be a future in which AI is dominant, then you can convince administrators who can then try to convince teachers like there's this future that we know is happening.
00:34:35.420In order to be successful in that future, your students need to know X, in this case, need to know how to use AI products.
00:34:42.960The problem with that, of course, is that nobody knows what the future is going to hold.
00:34:46.920There's some evidence that we're in an AI bubble maybe that's going to burst, that the AI companies actually haven't found a way to be financially sustainable. And so maybe there is no business model here. Maybe this technology will persist in some way but won't have the dominance that the technology companies predict it will.
00:35:06.820And if that's true, then it means that if you're teaching students to use these products and rely on them, you're actually doing them a disservice in preparing them for a world of the future. So that's like one thought exercise I like to keep in mind.
00:35:20.860And my answer to that first question about, like, is there any benefit to having AI be part of the creative process for students working on personal essays, for example, it's a somewhat of a simpler answer, but I would say that it is useful to be cognizant of, like, what communication, what the written word is, right?
00:35:43.720it's never been a fixed thing. It always evolves. So to the extent that like chat bots and their
00:35:50.280language production is a part of how we experience language now, I do find it interesting and useful
00:35:57.940to ask whether there are ways to engage creatively with these products that sort of like speak to the
00:36:07.040role of these products in our lives. It's a really specific case, though. So what I'm describing is
00:36:15.200not, you know, saying to a chatbot, hey, I'm trying to write an essay about my grandmother's
00:36:20.940death. Like, here are 10 details about what happened. Can you draft it for me? That's not
00:36:25.820at all what I'm saying, because that's not like a conscious investigation of how these products
00:36:31.140function. It's just like using these products to spit out some text that might be serviceable for
00:36:36.140you. So in that like much broader context, which I think the person who asked the question is
00:36:41.560asking about, I struggle to see any benefit. Yeah. When I think of that prompt, right,
00:36:47.440like I have to write an essay about my grandmother and here are 10 facts about her life and her
00:36:52.620death. What I would find interesting is like exploring that space between, and you do this
00:36:58.880in searches, like exploring that space between what it proffered and what my actual memories were.
00:37:05.500And how were my prompts maybe imprecise? Maybe I was writing them in a way that, like, did not actually reflect or could not reflect my actual feelings about my grandmother, right? Like, so delving into those corners, that's a cool essay. But that's not what these students are using.
00:37:27.920That's right. And I actually it makes me think about Auburn's question and comment earlier, actually, Anne, because when you describe that, something that comes to mind is like, and what would happen if like the person who's writing that essay about their grandmother went to their best friend who also knew the grandmother or their sister and said, like, here are the 10 things like what comes to mind?
00:37:52.580What am I missing? Right. Like that's that's definitely going to be like a more fruitful, interesting conversation that's going to lead to something generative for the writer than whatever ChatGPT has to offer.
00:38:04.380One thing that came up in other questions from educators is that like some of the people who are shouting the praises of ChatGPT are people who already developed that creative muscle to even think of these two prompts that we thought of, right?
00:38:18.100Like these two different approaches. But if you have never had the space to develop that creativity, it's not as useful. Do you subscribe to that?
00:38:28.120I can see the logic in thinking about it that way. I think about how, for example, when I wrote that essay Ghosts, right, where I was like offering, I was writing something and then it would fill it in, fill in the rest for me.
00:38:41.680And then I'd say to myself, oh, that's not right. And I delete it and I do it again myself. Basically, that process was me having like the critical faculty to see that what this product was producing was did not reflect my own self-expression.
00:38:56.440And I think I'm only able to do that with 25 or whatever years of experience as a writer. I think that would have been difficult, maybe impossible for me to see when I was 18 or 25 and still developing as a writer.
00:39:13.220That said, I think the value of having these products express things for us or like, quote unquote, help us see things that we might not have seen about our own self-expression is like there's potentially a value there.
00:39:28.660But there's a huge potential cost having to do with the way in which the use of these products can like corrupt our own way of thinking and can maybe like sidestep investigative routes and creative routes that we might have gone down if we hadn't been in conversation with a chatbot.
00:39:48.020Not to mention, as Auburn again mentioned, these overarching costs having to do with environmental resource use, labor exploitation, and so on.
00:39:59.920That sort of like just naturally comes with any use of these products.
00:40:04.760I want to pick up on the part, the democratizing creativity part, because we got a question specifically about that.
00:40:11.900apparently they were talking about democratizing creativity through ai at a recent design
00:40:18.120conference i find this framing interesting because so far capital d design has been protected
00:40:25.200but creativity in general is something that anyone can try why would it need to be democratized
00:40:31.900maybe this points to a general shift in understanding creativity as being more about
00:40:38.080the thing created and not the process of creating? Or maybe it's an additional element of bread and
00:40:45.140circus trying to throw the average Joe a bone while they install their surveillance apparatus
00:40:50.760at great expense to our communities. What do we think they're attempting here?
00:40:55.540I love this look. Like, it's really a rhetorical analysis that it's quite good. Also, just like
00:41:02.240the use of bread and circus to describe democratizing creativity. What do you think
00:41:07.500of that phrase? What is it trying to do? I mean, it's funny. I will say with every single one of
00:41:12.540these questions, like the analysis embedded in the question is so valuable. I'm like, let's just do
00:41:17.800whatever we need to do to get that analysis on air so people can hear it. That's oftentimes the
00:41:23.980case. Yes. Yeah. Yeah. I mean, I think there is a real desire on people's parts to be part of like
00:41:33.620an artistic or creative discourse. I think that's real. I think that is being exploited by tech
00:41:40.780companies who see that desire. I don't think it's just marketing messaging because I wrote this
00:41:47.740article about AI in the romance writing world. And I talked to some people who never considered
00:41:59.680themselves writers but like really wanted to be involved in romance writing and found that by
00:42:06.020using AI products they were able to produce these texts that they felt were partly their own that
00:42:11.680they wouldn't have been able to produce without AI. At least one person who I included in this
00:42:18.420article I wrote described anti-AI sentiment when it comes to creative production as being ableist
00:42:25.640because the idea being that one might have a disability that makes it difficult to express
00:42:32.580oneself in conventional ways and the use of AI could allow a kind of expression that might have
00:42:39.560felt inaccessible, which I find really fascinating, you know, and I think the fact that people feel
00:42:46.360that way requires that we kind of sit with it rather than dismiss it. I don't think it's just
00:42:51.920propaganda agreed and yet um it raises questions i think i think the thing to do is maybe try to
00:42:58.880dismantle like what we are prioritizing when we when we talk about creative production like i
00:43:04.820think it probably is a very i think it is a very real thing that our creative industries are ableist
00:43:12.440and racist and sexist and put up barriers that make it difficult for people um to enter who are
00:43:19.340not from like the the mainstream majority culture or background but i think maybe the next question
00:43:26.560is like well how can we dismantle that so that there is more varied self-expression including
00:43:33.140from all kinds of people rather than um saying let's keep things as is and just use chatbots
00:43:39.540so that people who don't come from that background are able to quote unquote able to produce text or
00:43:46.220images that conform with what we've decided is okay. Yeah. And I think of like, you use the
00:43:52.460phrase creative production. And I think much of the goal with using AI in these forms is to up
00:44:00.520your creative production, right? Produce more, have more creativity. And I often go back to like
00:44:08.040these scenes in Mad Men where Don Draper is just like going to the movies, getting drunk at lunch
00:44:14.700and sleeping, but like taking naps, walking around, not doing much, and then comes up with
00:44:20.260a smart idea or a clever idea, whatever it is. Right. And the way that we just like completely
00:44:28.240like our work, so much of our work is trying to compress creativity, trying to make it into
00:44:36.240something that can be like optimized in some capacity. And so I hear democratizing creativity
00:44:43.000And I hear optimizing creativity and it can't be like what what I'm hearing, too, is just like the eagerness to adopt these tools is a frustration that there isn't enough time in our lives with our hungry jobs and our intensive parenting practices and our precarity and our constant messaging from all sides.
00:45:06.120there isn't enough time for creativity, but we want it because it is human. And so we are turning
00:45:13.820to these tools. Absolutely. Yeah. Yeah. I mean, we don't live in a culture that values the novel
00:45:23.020that took 13 years to write like mine did, or makes it easy to do that. And yeah, I agree fully
00:54:16.600Like all of those smells combine to form the master smell, which is just a particular smell that like it's it's unspeakable.
00:54:30.420And I just hate it so much. So I've spent a lot of time in my laundry life figuring out the best way to combat it.
00:54:38.000And I just have to say that, like, I've tried lots of things that come in plastic jugs.
00:54:42.420So I was so excited when Blueland came up with this new laundry detergent tablet that is specifically designed to tackle really stinky stuff and also stains.
00:54:53.360So it's a dual-layer tablet that combines advanced stain-fighting enzymes with odor-fighting technology, gives a much deeper clean, but also, like all Blueland products, is plastic-free and microplastic-free.
00:55:06.600And also, these are pre-measured, mess-free tablets, so you just drop them directly into the drum and wash as usual.
00:55:13.260Blueland is a certified B Corp and Leaping Bunny Cruelty-Free Certified.
00:55:17.300Their formulas are EPA Safer Choice Certified, and many products have also earned Cradle to Cradle's Gold Material Health Certificate.
00:55:24.340So if you're looking to make a small change in your laundry routine, you can get 15% off your first order at blueland.com.
00:55:32.020slash culture. That's 15% off your first order at blueland.com slash culture.
00:55:45.560Okay, our next question is about something that we talked about at the beginning of the episode,
00:55:49.280which is about the difficulty of having these discussions online, or just having these
00:55:54.740discussions broadly. So this comes from Whitney. I'm an artist who works at an arts organization,
00:56:00.220and I'm also politically left. I interact with both of those communities online and in person
00:56:06.440and I find that the discourse around AI is really different depending on where I have the
00:56:12.200conversations. Online there's absolutely no great area about AI. People won't even consider
00:56:20.280discussing potential uses for organizing projects or tasks and no one wants to admit that they have
00:56:25.960dabbled with it for fear of cancellation. In person, however, I'm talking to all sorts of
00:56:31.700creative and progressive people who are using AI to help them streamline the tasks that they're
00:56:37.160not really good at and organize projects and otherwise make it possible for them to do the
00:56:42.820part of life that they are good at. So what do you make of this divide? Once again, it feels to me
00:56:48.980like online community is more interested in purity tests than giving people the benefit of the doubt
00:57:04.140I love that question because I've noticed that too.
00:57:06.640And I always like find it interesting and surprising,
00:57:08.680like the disconnect between what people disclose
00:57:11.300in online spaces or in their own writing
00:57:14.260versus what they say privately in conversation.
00:57:17.620It's making me think of something I posted
00:57:22.100about something problematic about ai and then this um this like bookseller in seattle reposted it
00:57:31.480and said like i think i i mean but i i can't i think it was in response to a post of mine like
00:57:37.160oh it's really rich that this writer who uses ai in her work is like complaining about ai you know
00:57:44.940Um, which I think is like, like on the one hand fair, but on the other hand, like doesn't account for like the way in which I'm hoping to actually critique AI with the work itself, you know?
00:58:33.120I think it's a really interesting critique, but it wasn't the critique that this person was making.
00:58:37.660And so I think the Internet, because of algorithms which which amplify anything binary or polarizing, are always like the Internet is always social media is always going to be a place where like very black and white opinions or polarized opinions.
00:58:55.480are going to be amplified. Um, and I, I think like, I don't know, this is going to be a very
00:59:01.560blanket statement to make. I wonder if you agree. And, but I think like binary views of things are
00:59:08.300like almost never the most interesting ways to look at things as artists or journalists or
00:59:14.120creative people. Um, and so I'm always like very skeptical or suspicious of anything that's like
00:59:22.440black and white x statement about ai in either direction you know i always like want to know
00:59:29.300where the nuance is yeah same and i think like i the algorithms have only become worse um in terms
00:59:36.780of surfacing like once uh whether it's on threads or it's on like tiktok or whatever like you find
00:59:44.440yourself in the hole and the hole keeps getting deeper um yeah and i do i found the most interesting
00:59:51.320conversations that i've had about ai have certainly not been online they have been with quote-unquote
00:59:56.980normies who just talk to me about what they're seeing in their workplace and some of the and0.97
01:00:03.920they talk very frankly about like this thing that's being pushed on us is stupid and like0.98
01:00:09.620it's like that's just they're trying people that's just our organization trying to like0.94
01:00:13.800seem modern and like throwing around ai stuff and then i talked to other people who are like
01:00:18.820this is a technology that is like you know this has been a sticking point in the way that we have
01:00:24.120worked for such a long time and this is simplifying it and we don't we have to figure out like
01:00:29.400especially my friends who are lawyers they're like what's the future of paralegals like what
01:00:34.120are we going to do like this is a foundational component of the way that we have done work for
01:00:40.140so long and they are people who we love and that have been parts of our lives and parts of our
01:00:46.000workplace. Like, what is that future? And they're thinking about it. And I think it's not it's not
01:00:51.160as simple as like, oh, we're just going to cut costs and get rid of all the paralegals. And0.99
01:00:55.920they're also incredibly wary of hallucinations, in part because there are consequences for
01:01:01.440citing cases that don't exist. Yeah, I'm thinking about I have a really close friend who runs a
01:01:09.300regional office of a nonprofit that is doing really, really important and super under-resourced
01:01:17.800and politically challenged work in the world right now. And this friend sometimes uses AI for things
01:01:25.760like grant applications and reports on how resources are being used. And going back to
01:01:33.820Auburn's framing about cost and benefit, I think it's really interesting to think about like
01:01:39.740a use case like that, that's like a little more complex and confusing to talk about than like
01:01:48.020some, you know, big car manufacturer getting rid of all these jobs because they're using AI. Like,
01:01:53.660what does it mean for my friend who's working, you know, 60 to 80 hours a week to do really,
01:01:59.620really important under-resourced, politically challenged work to use AI to generate a report
01:02:06.800that might have taken 10 hours and is now taking one hour of her time when now she can use the0.59
01:02:14.280other nine hours to do really important direct service work that she otherwise wouldn't have
01:02:19.880had the time for, right? How does that affect that cost-benefit analysis? And I don't have
01:02:24.320an answer to that question. But I think, like, those case studies are trickier and more interesting
01:02:29.780to grapple with than the more obvious ones. Yeah, there's a question that we don't have
01:02:34.080enough time for it, but that I thought was really interesting is from a sculptor who said, like,
01:02:39.640in order to do my creative work, in order to have those expanses of time to do the physical,
01:02:45.580like, the thing that, like, absolutely makes me feel human, the thing that I love,
01:02:49.620and the thing that is my job i use ai for like some of my grant writing you know basically like
01:02:56.060the stuff the stuff that historically either uh people had rich people who funded them to be able
01:03:04.300to do or they came from money or they had wives that did it right like they had people who did
01:03:09.480this work for them so that they could focus on the creativity and that to me is also another one that
01:03:14.720makes me like that I want to rattle around in my brain and be like yeah how are we thinking about
01:03:19.640this interestingly I've been on the other side of that recently in a way because I'm reading
01:03:25.420applications to Hedgebrook the residency up there near where you live right right and most um there
01:03:32.800there have been very few applications that have sort of like triggered my AI radar but there have
01:03:38.600been a few where like the language of the residency application like reads AI-ish to me and I think to
01:03:45.620myself okay like this person was saving time like applying to residencies isn't a core part of their
01:03:50.920work they're writers however I do read it differently and I don't you know it does not
01:03:57.860have carry like the power that something that reads as written by a real person has you know
01:04:05.240Totally. And I think of that like, you know, the the tautology of like students writing papers that are made with AI and then teachers have too many papers to grade.
01:04:19.400So they start grading using AI. So it's basically like AI writing papers for AI, like how you get into this weird space.
01:04:26.980And like, is this the future of grant writing? Like, is it computers writing grants that then computers will sort through the grants to decide who gets the grants? Like, and how will they learn to talk to one another in terms of like surfacing that? And what are the weights and balances that we are maybe not perceiving in terms of how the computer will perceive that?
01:04:47.620Totally. And what are like the biases that are ingrained in that kind of approach, right, that is maybe trained to prioritize like certain historically mainstream ways of expressing oneself?
01:05:03.860Totally. Totally. Or just like what if you chose not to use AI to write your grant application and it just gets thrown out with the bathwater?
01:05:10.740Right. I use that phrase baby with the bathwater because our last question is about throwing the baby out with the bathwater, but also about baby names. I love this question. It's from Kellyanne.
01:05:24.300I am the most anti-AI person you could imagine. Am I throwing the baby out with the bathwater?
01:05:30.860Yesterday, my brother said, this is a job for AI, as we were jokingly predicting my other brother's
01:05:36.320next baby name, based on his current kid's names. I feel instant revulsion when somebody suggests
01:05:42.440AI to me, especially for a task that is imagination, family, and love-centered. But am I wrong? He's
01:05:50.220probably right that AI would be better able to synthesize the names that are similar in popularity
01:05:54.520and style to what my brother has already chosen, I still feel that revulsion. Is there actually a
01:06:00.320place for it in our lives? How do we continue to exclude it in the ways that matter and are
01:06:04.880meaningful without becoming an insane Luddite? This comes back to Auburn's question, don't you
01:06:10.800think? Well, something I love that's like inherent to this question, actually, and this like connects
01:06:16.680to your work also is, like, what the question is actually telling us about, like, the values and
01:06:23.800interests of the person asking the question, right? So this is, like, two siblings talking
01:06:29.720about a third sibling who, like, has some kids. Like, gently razzing them a little bit, right?
01:06:34.260Yeah. Like, it's a little bit of intimacy. Yeah. Yeah, exactly. Like, there's going to be another
01:06:38.300kid. They're trying to figure out if they can guess the kid. So, like, there's so many potential
01:06:43.020dynamics here maybe it's like these two siblings who are thinking about their other sibling who is
01:06:47.260either like very conventional or very unconventional or like names every kid after like another
01:06:54.100relative that they know or something like that right there's like something very specific to
01:06:58.580that family and the family dynamics inherent to that conversation maybe it's also that like
01:07:03.720each of the siblings are having this like subconscious sibling rivalry dynamic where
01:07:09.580they're like, ooh, which of us knows this other sibling of ours better, right? Like, which of us
01:07:13.860can use the names of the existing kids to figure out what the next one is? If we each come up with
01:07:18.860one, who's going to win, right? What does that say about our relationship with our other sibling?
01:07:23.320What does it say about our relationship with each other? And so, like, everything that makes that
01:07:27.600question about predicting the next name interesting or valuable have to do entirely with the requirement
01:07:35.580that it's the people in the conversation
01:07:37.640who come up with the guesses of the names, right?
01:07:40.580Like, it ends up actually being meaningless
01:09:20.160Like, what if they showed the list to their brother, right?
01:09:22.740Like, after the baby was born or something like that.
01:09:24.920Like, that's another funny thing to do.
01:09:27.060But how do you feel in scenarios like that when someone's like, well, is using AI for
01:09:32.540a joke like in this kind of kidding scenario is that worth using the resources you know like yeah
01:09:41.060is that worth it yeah well i mean the reason the reason i suggested an answer like that is because
01:09:47.940it then allows the person asking this question with the revulsion um to not use ai right like
01:09:56.180this person is like not participating in a decision to use AI. I also just for the record,
01:10:02.200I also would not consider that a use of AI that would feel worth the cost to me. However, like
01:10:08.980people are allowed to make their own decisions about, you know, how they use technology or
01:10:14.400whatever. So we could imagine a world in which rather than saying to the brother, like, I'm
01:10:18.980going to disown you as my brother if you use AI, right, to say like, listen, you're going to do
01:10:23.860what you're going to do, like, go ahead and do what you're going to do, and then potentially,
01:10:28.320like, use that as, like, an interesting, like, rhetorical moment, right? Or just let the brother
01:10:34.280do what he's going to do, explain their own reasons for not wanting to do that, and let it go,
01:10:42.060right? But it allows, like, a kind of integrity on the part of the questioner where they don't
01:10:45.960have to be part of the, like, okay, now we're using AI. Well, and I think it brings us to
01:10:50.400something we touched on a little bit earlier, and that I think a lot of us have learned of
01:10:53.820the course of the last 10 plus years which is that like shame is not a an effective rhetorical
01:11:00.200device yeah like if you want someone in your life to think a little bit more about the way that they0.99
01:11:05.240are using ai telling them that they're killing the planet or that like it's making you stupider
01:11:12.200like those things are not going to change their minds if anything it's going to make them more0.81
01:11:17.000resistant to thinking more deeply about it because it has been a place where they have encountered
01:11:22.360to other people shaming them, and they have to harden their position in order to justify their
01:11:26.540use. So I think ways that you can open up the conversation about it to talk a little bit about
01:11:32.700why you feel the way that you do, or not in like a hectoring way, right? Just in a,
01:11:39.520when it's imagination, family, and love centered, like these are things that I want to keep
01:11:43.700imagination, family, and love centered and not deflect that.
01:11:48.640Yeah. And by the way, to be clear, and you know this, Anne, because you read my book,
01:11:52.160but some listeners might not like I write about using chatbots. I write about how I sometimes use
01:11:58.280them and and and in a way that isn't about critiquing these products. Right. And part of the
01:12:04.780reason I do that and I do it with respect to other things, too. So I critique social media and then I
01:12:10.020say I use social media. I critique Google and I talk about using Google. I critique Amazon and
01:12:15.240talk about using Amazon. In some ways, that's the subject of the book. It's like for me, the subject
01:12:20.260of the book is like this back and forth between reading a chapter where I'm critiquing something
01:12:25.200and then seeing a chapter where I'm actively using it. And my goal there in some ways was
01:12:31.480a rhetorical goal. My goal was to like show that hypocrisy in myself that I hoped readers would
01:12:39.800also see reflected in them. And so, you know, I wanted people to like read the book and feel like,
01:12:46.960OK, it's not just me. Like, I feel less alone because this person is enacting this thing on the page that I'm also familiar with. But I'm also hoping that, like, in putting myself in, like, this hypocritical, maybe even villainous role, I'm inviting readers to see that in themselves. Right.
01:13:06.760And so that way, it's not me sort of like shaking my finger at a reader and saying like, all these things are horrible. If you're using them, you're horrible because I'm using them, too. Right. It's more this rhetorical move of saying like, hey, I all these things are horrible. I'm still using them. Maybe you are, too. Isn't this horrible? Right.
01:13:27.520And also like an invitation for introspection, which I think like great art asks of us. And hopefully that's what this podcast has done for others. Not that this podcast is like great elevated art, but it is an invitation for introspection. So, yeah.
01:13:44.340Absolutely. No, I agree. And I really appreciate the conversation. And I feel like we have, hopefully, we'll see what Auburn says in the comments.
01:14:41.020Head to patreon.com slash culture study.
01:14:43.720It's just five bucks a month or $50 a year, and you'll get ad-free episodes, an exclusive
01:14:48.580advice time segment, and weekly discussion threads for each episode.
01:14:52.100It's also, if you already have a subscription to the Culture Study newsletter, it's just a little add-on.
01:14:56.760And be sure to follow the show wherever you get your podcasts.
01:14:59.360We have so many great episodes in the works, and I promise you don't want to miss any of them.
01:15:03.800If you want to suggest a topic, ask a question about the culture that surrounds you, or submit a question for our subscriber-only advice time segment, go to our Google form at tinyurl.com slash culturestudypod or check the show notes for a link.
01:15:17.380The Culture Study Podcast is produced by me, Anne Helen Peterson, and Melody Rowell.