“in interaction with others[ ]we bring meaning into the world”
September 3, 2024 6:42 PM Subscribe
I’m not saying that art has to involve tedium. What I’m saying is that art requires making choices at every scale; the countless small-scale choices made during implementation are just as important to the final product as the few large-scale choices made during the conception. It is a mistake to equate “large-scale” with “important” when it comes to the choices made when creating art; the interrelationship between the large scale and the small scale is where the artistry lies. [Ted Chiang, New Yorker]
mentions Emily Bender [medium] & François Chollet; recently
mentions Emily Bender [medium] & François Chollet; recently
craniac, thank you! also, there is more Ted Chiang previously
posted by HearHere at 6:58 PM on September 3
posted by HearHere at 6:58 PM on September 3
(The NaNoWriMo thread from yesterday ended up including some discussion of this essay after someone linked to it in a comment. Only one person really pushed back against its premises, but in doing so -- as someone else later pointed out -- they kind of fell right into one of the pitfalls the essay warns against.)
posted by nobody at 10:00 PM on September 3 [3 favorites]
posted by nobody at 10:00 PM on September 3 [3 favorites]
... the interrelationship between the large scale and the small scale is where the artistry lies.
I'm reminded of a passage from the book Chaos: Making A New Science by James Gleick:
A Beaux-Arts paragon like the Paris Opera has no scale because it has every scale. An observer seeing the building from any distance finds some detail that draws the eye. The composition changes as one approaches and new elements of the structure come into play.
Great article. The author makes some arguments that aren't made often enough when talking about AI, and especially AI and art.
posted by Johnny Lawn and Garden at 11:14 PM on September 3 [3 favorites]
I'm reminded of a passage from the book Chaos: Making A New Science by James Gleick:
A Beaux-Arts paragon like the Paris Opera has no scale because it has every scale. An observer seeing the building from any distance finds some detail that draws the eye. The composition changes as one approaches and new elements of the structure come into play.
Great article. The author makes some arguments that aren't made often enough when talking about AI, and especially AI and art.
posted by Johnny Lawn and Garden at 11:14 PM on September 3 [3 favorites]
I think he’s right, and the reason the small choices matter in writing is that each occurrence of a word does not have exactly the same meaning and implications (or implicatures). In context, the way words are used tells us things beyond their dictionary definition. Notably this includes stuff about the attitudes and understanding of the characters (not least the narrator) and one of the core projects of the novel, for example, is to explore and illustrate conscious human experience.
So the choice of the next word in the text string that constitutes a novel may look like a fairly free choice between the grammatically acceptable options, but as Chiang acknowledges, it is really much more complex than that.
posted by Phanx at 11:19 PM on September 3 [7 favorites]
So the choice of the next word in the text string that constitutes a novel may look like a fairly free choice between the grammatically acceptable options, but as Chiang acknowledges, it is really much more complex than that.
posted by Phanx at 11:19 PM on September 3 [7 favorites]
I really liked this commentary (critique?) of Ted's essay:
https://open.substack.com/pub/fivegoodhours/p/escaping-a-hostage-situation?utm_campaign=post&utm_medium=web
Though, for the life of me, I can't figure out the name of the actual author ¯\_(ツ)_/¯
posted by newdaddy at 1:20 AM on September 4 [4 favorites]
https://open.substack.com/pub/fivegoodhours/p/escaping-a-hostage-situation?utm_campaign=post&utm_medium=web
Though, for the life of me, I can't figure out the name of the actual author ¯\_(ツ)_/¯
posted by newdaddy at 1:20 AM on September 4 [4 favorites]
> "what’s wrong with all of this? Nothing -- unless you actually take it seriously" Ben Davis' critique of Deleuze's cliché is worth contrasting with fivegoodhours. as Five's explicit goal is to be unoriginal, one might say at least there they're successful? yet, that is also... a choice
posted by HearHere at 3:31 AM on September 4 [1 favorite]
posted by HearHere at 3:31 AM on September 4 [1 favorite]
I had been trying to write a comment on the NaNo thread about the ultimate effect of the program and how it ties in with a certain segment of the writing community's embrace of AI, but it felt mean-spirited and useless so I didn't pursue it. But in this thread with its echoes of that earlier one--and maybe especially with the useful five good hours commentary, and yes, we'll throw on Bacon and Deleuze as well--I can put words to this feeling.
"Make a painting in the style of Francis Bacon" is a prompt you can use. "Grow up gay in the 20s, experience the pleasures and dangers of the underworld, absorb the culture around you, consider how the world turns humans into flesh machines, enjoy the fun of shock, think about if Picasso were working with cut-up chickens rather than paints," is not. Much as Chiang talks about choices large and small, one's art comes from too many places to fit into a prompt.
I liked NaNo for a while, although I hardly ever actually managed to participate--it was more common that I'd be in the middle of a book when November rolled around, or outlining something I wouldn't get started until mid-month. I think I got the little certificate once? But with the idea that everyone can write a novel--that everyone can sit down on November 1 and by averaging 1667 words per day, come out of it on November 30 with a 50,000-word book--wasn't simply a democratic approach to fiction. It was also part, I would say, of a cheapening experience. Something that cut out the heart of fiction and replaced it with something else.
You can see why I wouldn't want to say this on the original thread; too many people were sharing positive experiences, and it would be wrong to say, "This was a horrible program that did bad things to how we think about art." Wrong ethically, wrong factually, but correct gesturally, because you want to take in not only NaNo but the burst of the self-publishing business that happened alongside it--the idea that not only might you have a novel in you, but you could make a million dollars from it.
For its inspiration, this movement took up pulp writing rather than art. It flattened everything. Are you a plotter or a pantser? We know that's not the only way to think of how we approach a book but that distinction accreted to the question of how to write, sclerosed the discussion. It was more interested in advice in tweet or blog form, than digging into history and asking what became of all those pulp writers you're trying to emulate without reading them or studying them. The answer would have been too depressing: Most of them died in the heated stench of a downtown alley, fingers permanently crooked up from the pressure of typing 1666 words a day. A few made it out but when they made it out they weren't writing pulp anymore, they were writing memorable stories of depth that exposed their love and research.
Consider a specific limited question of technique: How do you make a reader feel something with a scene? What are ways to elicit particular emotions? Go out onto your favorite search engine, see what people say, and you will be shocked. More disturbing than any Bacon painting will be the simplicity and certainty of the advice. "Show rather than tell"--no, you don't say. "Build sympathy." "Kill a beloved character." Over and over again, repeatedly, the same answer in blog form. The responses are so general that you sense a troubling truth: They don't actually know how to do it. They haven't really thought about it. They haven't studied emotionally moving works, haven't tried a thousand things to get it right. They have read a blog and now are passing down the certainty of the lesson they have learned--not from the writing, not from the horror of trying to breathe life into a cold dead scene--it's gossip, it's a game of telephone, except the message doesn't change, doesn't evolve. Call it a game of fax instead.
The five good hours response to Chiang is just doing Bloom's anxiety of influence; artists fear the influence and put their fingers into it, trying to decide what to pull closer and what to push away; college students facing an essay feel anxiety without the influence, they don't have anything to get their hands into; AI embodies the influence without the anxiety: What if you could have everything and never have to worry about getting it right and making it yours?
"Make a painting in the style of Bacon." "Grow up watching horror movies where the physical constraints of latex and fiberglass create a visual catalog of what may be done to the human body against its will, then in college discover there was this painter who was already there long before you, and feel some strange kinship with the suffering figures melting onto his canvas. Think about suffering, about how you feel when you're older and the excitement of those horror movies has faded, replaced with a sad dullness that wants to feel something deeper. Consider how many people also want to feel something deeper. How do you give that to them, when your only tools are suffering and curiosity?"
I do think choice is a way to look at it, I do think rejection or pushing against cliche is a way to look at it. The cheapening of the fiction-writing process (and let's be historical for half a second: there has always been this cheapening, you have always been able to find a cookie-cutter approach to writing, what we have now is bigger and more expensive and more culturally penetrating but it's not new) happened side-by-side with the democratization of fiction-writing but they did not have to be the same thing. We emerged from everyone wanting to be a novelist into a new age where we're promised nobody has to be a novelist, it'll all be done for you. Our interest in this--in flattening ourselves, in removing our own capacity to choose or to reject--in turning ourselves into objects rather than subjects--is worthy of a shuddering thought.
posted by mittens at 4:18 AM on September 4 [25 favorites]
"Make a painting in the style of Francis Bacon" is a prompt you can use. "Grow up gay in the 20s, experience the pleasures and dangers of the underworld, absorb the culture around you, consider how the world turns humans into flesh machines, enjoy the fun of shock, think about if Picasso were working with cut-up chickens rather than paints," is not. Much as Chiang talks about choices large and small, one's art comes from too many places to fit into a prompt.
I liked NaNo for a while, although I hardly ever actually managed to participate--it was more common that I'd be in the middle of a book when November rolled around, or outlining something I wouldn't get started until mid-month. I think I got the little certificate once? But with the idea that everyone can write a novel--that everyone can sit down on November 1 and by averaging 1667 words per day, come out of it on November 30 with a 50,000-word book--wasn't simply a democratic approach to fiction. It was also part, I would say, of a cheapening experience. Something that cut out the heart of fiction and replaced it with something else.
You can see why I wouldn't want to say this on the original thread; too many people were sharing positive experiences, and it would be wrong to say, "This was a horrible program that did bad things to how we think about art." Wrong ethically, wrong factually, but correct gesturally, because you want to take in not only NaNo but the burst of the self-publishing business that happened alongside it--the idea that not only might you have a novel in you, but you could make a million dollars from it.
For its inspiration, this movement took up pulp writing rather than art. It flattened everything. Are you a plotter or a pantser? We know that's not the only way to think of how we approach a book but that distinction accreted to the question of how to write, sclerosed the discussion. It was more interested in advice in tweet or blog form, than digging into history and asking what became of all those pulp writers you're trying to emulate without reading them or studying them. The answer would have been too depressing: Most of them died in the heated stench of a downtown alley, fingers permanently crooked up from the pressure of typing 1666 words a day. A few made it out but when they made it out they weren't writing pulp anymore, they were writing memorable stories of depth that exposed their love and research.
Consider a specific limited question of technique: How do you make a reader feel something with a scene? What are ways to elicit particular emotions? Go out onto your favorite search engine, see what people say, and you will be shocked. More disturbing than any Bacon painting will be the simplicity and certainty of the advice. "Show rather than tell"--no, you don't say. "Build sympathy." "Kill a beloved character." Over and over again, repeatedly, the same answer in blog form. The responses are so general that you sense a troubling truth: They don't actually know how to do it. They haven't really thought about it. They haven't studied emotionally moving works, haven't tried a thousand things to get it right. They have read a blog and now are passing down the certainty of the lesson they have learned--not from the writing, not from the horror of trying to breathe life into a cold dead scene--it's gossip, it's a game of telephone, except the message doesn't change, doesn't evolve. Call it a game of fax instead.
The five good hours response to Chiang is just doing Bloom's anxiety of influence; artists fear the influence and put their fingers into it, trying to decide what to pull closer and what to push away; college students facing an essay feel anxiety without the influence, they don't have anything to get their hands into; AI embodies the influence without the anxiety: What if you could have everything and never have to worry about getting it right and making it yours?
"Make a painting in the style of Bacon." "Grow up watching horror movies where the physical constraints of latex and fiberglass create a visual catalog of what may be done to the human body against its will, then in college discover there was this painter who was already there long before you, and feel some strange kinship with the suffering figures melting onto his canvas. Think about suffering, about how you feel when you're older and the excitement of those horror movies has faded, replaced with a sad dullness that wants to feel something deeper. Consider how many people also want to feel something deeper. How do you give that to them, when your only tools are suffering and curiosity?"
I do think choice is a way to look at it, I do think rejection or pushing against cliche is a way to look at it. The cheapening of the fiction-writing process (and let's be historical for half a second: there has always been this cheapening, you have always been able to find a cookie-cutter approach to writing, what we have now is bigger and more expensive and more culturally penetrating but it's not new) happened side-by-side with the democratization of fiction-writing but they did not have to be the same thing. We emerged from everyone wanting to be a novelist into a new age where we're promised nobody has to be a novelist, it'll all be done for you. Our interest in this--in flattening ourselves, in removing our own capacity to choose or to reject--in turning ourselves into objects rather than subjects--is worthy of a shuddering thought.
posted by mittens at 4:18 AM on September 4 [25 favorites]
Ted Chiang is a so good, he's the pinnacle on the The Science Fiction Writer’s Hierarchy of Doubt.
posted by signal at 5:41 AM on September 4 [5 favorites]
posted by signal at 5:41 AM on September 4 [5 favorites]
I can see the argument, for the most part. But I still feel like defining what is and what is not art is an intractable problem.
Duchamp's "The Fountain" didn't require more creative choices than writing a LLM prompt. A lot of art intentionally replaces chance with choice, too. And is a novel "more art" than a short poem which requires a lot fewer choices but where every choice is critical? Hmm.
Certainly the statement that "AI isn't going to make art" needs the caveat that it's talking about present-day technology rather than a hypothetical general artificial intelligence that makes conscious choices (whatever consciousness is).
posted by Foosnark at 5:48 AM on September 4 [1 favorite]
Duchamp's "The Fountain" didn't require more creative choices than writing a LLM prompt. A lot of art intentionally replaces chance with choice, too. And is a novel "more art" than a short poem which requires a lot fewer choices but where every choice is critical? Hmm.
Certainly the statement that "AI isn't going to make art" needs the caveat that it's talking about present-day technology rather than a hypothetical general artificial intelligence that makes conscious choices (whatever consciousness is).
posted by Foosnark at 5:48 AM on September 4 [1 favorite]
Duchamp's "The Fountain" didn't require more creative choices than writing a LLM prompt.
Lincoln Michel this morning with a similar point: "Tangential to recent AI discourse, but it seems incorrect to make 1 to 1 comparisons between visual art and novels. 'Rothko didn't make many choices in a painting! Duchamp didn't!' Maybe not, but we understand their works as part of a larger artistic projects full of intention."
posted by mittens at 5:55 AM on September 4 [5 favorites]
Lincoln Michel this morning with a similar point: "Tangential to recent AI discourse, but it seems incorrect to make 1 to 1 comparisons between visual art and novels. 'Rothko didn't make many choices in a painting! Duchamp didn't!' Maybe not, but we understand their works as part of a larger artistic projects full of intention."
posted by mittens at 5:55 AM on September 4 [5 favorites]
“ with the idea that everyone can write a novel--that everyone can sit down on November 1 and by averaging 1667 words per day, come out of it on November 30 with a 50,000-word book--wasn't simply a democratic approach to fiction. It was also part, I would say, of a cheapening experience.”
I feel like a broken record here, but wrt the more than “democratic” experience — the “cheapening” experience, as you said — I think it comes about because the democratization of art here isn’t that everybody with something to say can write a novel. It’s that everybody who has time to write 1,667 words per day, every day of November, can write a novel. And those are two different people, in many, many cases.
posted by toodleydoodley at 5:58 AM on September 4 [3 favorites]
I feel like a broken record here, but wrt the more than “democratic” experience — the “cheapening” experience, as you said — I think it comes about because the democratization of art here isn’t that everybody with something to say can write a novel. It’s that everybody who has time to write 1,667 words per day, every day of November, can write a novel. And those are two different people, in many, many cases.
posted by toodleydoodley at 5:58 AM on September 4 [3 favorites]
> see when will the art world recognise the real artist behind Duchamp's Fountain? [guardian]
posted by HearHere at 6:28 AM on September 4
posted by HearHere at 6:28 AM on September 4
Here's the linked version of Five Good Hours's (FGH) Escaping a Hostage Situation from newdaddy's comment upthread.
FGH takes Chiang and argues that art "composed entirely of free “choices” made by the artist, is inconceivable." And while I don't think Chiang is fully asserting that, FGH is making me think about effort/practice/choice as methods of learning vs modes of intelligence.
posted by zenon at 8:20 AM on September 4 [1 favorite]
FGH takes Chiang and argues that art "composed entirely of free “choices” made by the artist, is inconceivable." And while I don't think Chiang is fully asserting that, FGH is making me think about effort/practice/choice as methods of learning vs modes of intelligence.
posted by zenon at 8:20 AM on September 4 [1 favorite]
Complementary critique referenced earlier in the discussion
posted by craniac at 8:48 AM on September 4 [2 favorites]
posted by craniac at 8:48 AM on September 4 [2 favorites]
I like Chiang's perspective on this, because for me, the "countless small-scale choices" are the only way I ever get anything written. I am completely unable to outline fiction, despite having tried many, many times. It's hard for me to imagine anything interesting that might happen in the middle of a novel, or the end of a novel - it's like trying to plan a dinner party for strangers. But once I start writing I start making small and almost arbitrary choices about background details, and when I look back I find that those small and arbitrary choices seed what comes next, like the grit in an oyster that eventually develops into a pearl. And suddenly it's less like trying to plan a dinner party for strangers, and more like trying to plan a dinner party for friends, when you already can at least make educated guesses about what food one person will like or what music another person will like.
posted by Jeanne at 8:56 AM on September 4 [8 favorites]
posted by Jeanne at 8:56 AM on September 4 [8 favorites]
Duchamp's fountain is a collection of lies stacked on theft stacked on lies on appropriation stacked on borrowing and influence. Which, I suppose makes it relevant. For starters. If you've ever seen the thing sitting in a museum, what you have actually seen is a third rate copy of the "original", which was lost. The copies were commissioned and signed (but not made) by Duchamp in the 50s or 60s. Those look like garbage, but they are indisputably his work (except that he was prompted to do so by others). The original was maybe Duchamp maybe Freytag-Loringhoven, or maybe Stieglitz, who took the main photo of the thing. His photos definitely added something, and the other images we have of it hanging in the background of Duchamp's studio don't really give the full sculptural effect.
Except, no. Fountain didn't spring, fully formed out of the ether. The urinal had a designer, who presumably worked at the J. L. Mott iron works and designed the thing with beauty, cost, and functionality in mind. But wait! That isn't true either. Because the J. L. Mott urinal was a copy of a urinal designed in, IIRC, Cornwall. That design was likely a refinement of other, earlier designs. A *lot* of unattributed work had to go on to get to those 1950s fountains.
People will say that the thing that matters is that Duchamp gave those late copies his capital A Artist imprimatur, and all the work that went on previously is mere design and craft. I can see the point, I guess, but I do wonder if the real lesson we should draw here is that the modernists were a bunch of thieving charletans, and what we should value is the design and craft.
posted by surlyben at 8:58 AM on September 4 [9 favorites]
Except, no. Fountain didn't spring, fully formed out of the ether. The urinal had a designer, who presumably worked at the J. L. Mott iron works and designed the thing with beauty, cost, and functionality in mind. But wait! That isn't true either. Because the J. L. Mott urinal was a copy of a urinal designed in, IIRC, Cornwall. That design was likely a refinement of other, earlier designs. A *lot* of unattributed work had to go on to get to those 1950s fountains.
People will say that the thing that matters is that Duchamp gave those late copies his capital A Artist imprimatur, and all the work that went on previously is mere design and craft. I can see the point, I guess, but I do wonder if the real lesson we should draw here is that the modernists were a bunch of thieving charletans, and what we should value is the design and craft.
posted by surlyben at 8:58 AM on September 4 [9 favorites]
I find some delight in the fact that the copies that exist in museums are purpose built to be sculptures of urinals, which subverts the idea of a "readymade" that the original expressed.
posted by surlyben at 9:12 AM on September 4 [3 favorites]
posted by surlyben at 9:12 AM on September 4 [3 favorites]
If there is anything that can get through to a young student inclined to use ChatGPT, he’s nailed it—using AI to write “is like bringing a forklift into the weight room.” That, at least, any kid can understand. The teacher doesn’t want to read about, say, the use of the eyes of Dr. T.J. Eckleberg in The Great Gatsby. The teacher wants to see that you have the skills to process and analyze a text! You have to develop those by sucking at them and gradually getting better!
posted by Countess Elena at 9:34 AM on September 4 [4 favorites]
posted by Countess Elena at 9:34 AM on September 4 [4 favorites]
...doesn’t want to read about, say, the use of the eyes of Dr. T.J. Eckleberg...
T. J. Eckleburg Follows Up with His Branding Agency [mcsweeney's]
surlyben, i appreciate your analysis & would add that extractive processes are also significant when considering the tech. Kate Crawford's Atlas of AI [g] summarizes this (& explores other avenues of thought) well:
the work of art in the age of mechanical reproduction [wiki]
posted by HearHere at 1:31 PM on September 4 [3 favorites]
T. J. Eckleburg Follows Up with His Branding Agency [mcsweeney's]
surlyben, i appreciate your analysis & would add that extractive processes are also significant when considering the tech. Kate Crawford's Atlas of AI [g] summarizes this (& explores other avenues of thought) well:
The mining that makes AI is both literal & metaphorical. The new extractivism of data mining also encompasses and propels the old extractivism of traditional mining. The stack required to power artificial intelligence systems goes well beyond the multilayered technical stack of data modeling, hardware, servers, and networks. The full-stack supply chain of AI reaches into capital, labor, and Earth’s resources — [ ] From the perspective of deep time, we are extracting Earth’s geological history to serve a split second of contemporary technological time [31]Complementary critique :
the work of art in the age of mechanical reproduction [wiki]
posted by HearHere at 1:31 PM on September 4 [3 favorites]
"Such a person inadvertently reveals that they think formulating sentences is a nuisance rather than a fundamental part of storytelling in prose. Generative A.I. appeals to people who think they can express themselves in a medium without actually working in that medium."
Chiang isn't saying much that's new, but he's saying it very well, and he comes right to the heart of the matter: writing is work, and writing is communication from a human being.
Writing is a skill— a remarkably democratic one, in that anyone can learn it; but it does require time and practice. If you have to write a business memo and don't have that skill, I can see the appeal of LLMs. But I'd say the question to ask is not "Can I make the LLM write 'my' novel?", as "Why does business require so much drivel?" Rather than have the LLM turn my bullet points into a boilerplate letter, why can't I just send the bullet points?
posted by zompist at 2:46 PM on September 4 [4 favorites]
Chiang isn't saying much that's new, but he's saying it very well, and he comes right to the heart of the matter: writing is work, and writing is communication from a human being.
Writing is a skill— a remarkably democratic one, in that anyone can learn it; but it does require time and practice. If you have to write a business memo and don't have that skill, I can see the appeal of LLMs. But I'd say the question to ask is not "Can I make the LLM write 'my' novel?", as "Why does business require so much drivel?" Rather than have the LLM turn my bullet points into a boilerplate letter, why can't I just send the bullet points?
posted by zompist at 2:46 PM on September 4 [4 favorites]
Many novelists have had the experience of being approached by someone convinced that they have a great idea for a novel, which they are willing to share in exchange for a fifty-fifty split of the proceeds. Such a person inadvertently reveals that they think formulating sentences is a nuisance rather than a fundamental part of storytelling in prose. Generative A.I. appeals to people who think they can express themselves in a medium without actually working in that medium.
Well put
posted by St. Peepsburg at 3:19 PM on September 4 [2 favorites]
Well put
posted by St. Peepsburg at 3:19 PM on September 4 [2 favorites]
Is there a similar opportunity to make a vast number of choices using a text-to-image generator? I think the answer is no.
This is a reason I don’t find image generators (or music generators, even more so) very interesting or enjoyable to use right now, but it seems unlikely that it will remain true. There are already some interesting advanced techniques available to shape image generation, and of course there’s good ol’ Photoshop to reshape and collage the outputs.
I suspect this might be because Miller was using dall-e for something it’s not intended to do
… a huge part of the history of art in every medium.
posted by atoxyl at 5:34 PM on September 4 [1 favorite]
This is a reason I don’t find image generators (or music generators, even more so) very interesting or enjoyable to use right now, but it seems unlikely that it will remain true. There are already some interesting advanced techniques available to shape image generation, and of course there’s good ol’ Photoshop to reshape and collage the outputs.
I suspect this might be because Miller was using dall-e for something it’s not intended to do
… a huge part of the history of art in every medium.
posted by atoxyl at 5:34 PM on September 4 [1 favorite]
I think it almost goes without saying that “generative AI” won’t replace Art as a socially situated exchange between humans but that seems orthogonal to the interesting questions about what it will do. I think this essay is dead on about what sucks about the current generation of AI tools and a certain strain of ideology behind them. But I think it would be a massive underestimation of human creativity to conclude that people won’t figure out how to use them to make Art nonetheless.
posted by atoxyl at 5:38 PM on September 4 [1 favorite]
posted by atoxyl at 5:38 PM on September 4 [1 favorite]
I’m wondering what would spur a kid to use a ChatGPTalike instead of writing to express their own thoughts. When does the risk begin? My not that informed hunch is that a kindergartener will not be tempted but a middle schooler might, so, sometime between those points, I imagine that something changes. I ask because I wonder what feature of school it is that makes writing to express oneself not seem worthwhile.
posted by eirias at 5:47 PM on September 4 [1 favorite]
posted by eirias at 5:47 PM on September 4 [1 favorite]
One of the first use cases I ever saw for GPT was AI Dungeon, made-up-on-the-fly (and rather surreal because this was in GPT-2 days and the model had a very small context window) interactive fiction. That kind of thing is where generative text opens up interesting artistic possibilities.
posted by atoxyl at 5:48 PM on September 4
posted by atoxyl at 5:48 PM on September 4
When does the risk begin?
"Haraway brings the cyborg corpus of women of colour to the forefront due to language politics pervading the struggles of women of colour. She highlights the importance of writing for all colonised groups, to be specific cyborg writing. She writes that cyborg writing is about the power to survive and to ‘seize the world of the tools that marked them as Other’. Likewise, she cites the example of Cherrie Moraga [bomb: “How have you managed to balance the risk-taking with what I would assume is a certain backlash or misunderstanding about your feminism or the indigenous perspective in your work?”] and how her chimeric identity subverts colonial narratives" [medium]
posted by HearHere at 7:37 PM on September 4
"Haraway brings the cyborg corpus of women of colour to the forefront due to language politics pervading the struggles of women of colour. She highlights the importance of writing for all colonised groups, to be specific cyborg writing. She writes that cyborg writing is about the power to survive and to ‘seize the world of the tools that marked them as Other’. Likewise, she cites the example of Cherrie Moraga [bomb: “How have you managed to balance the risk-taking with what I would assume is a certain backlash or misunderstanding about your feminism or the indigenous perspective in your work?”] and how her chimeric identity subverts colonial narratives" [medium]
posted by HearHere at 7:37 PM on September 4
I ask because I wonder what feature of school it is that makes writing to express oneself not seem worthwhile.
I would guess two things snuff out the spark of wanting to explore, experiment and express yourself: Assessment and boredom.
Assessment, when the fear of punishment / judgement outweighs the joy of learning.
Boredom, when the thing you're learning about doesn't interest you.
Would kids turn to LLMs if they were allowed to write about something they're passionate about, with the knowledge that their writing would not be assessed? Maybe, but it seems less likely?
posted by Zumbador at 9:00 PM on September 4 [3 favorites]
I would guess two things snuff out the spark of wanting to explore, experiment and express yourself: Assessment and boredom.
Assessment, when the fear of punishment / judgement outweighs the joy of learning.
Boredom, when the thing you're learning about doesn't interest you.
Would kids turn to LLMs if they were allowed to write about something they're passionate about, with the knowledge that their writing would not be assessed? Maybe, but it seems less likely?
posted by Zumbador at 9:00 PM on September 4 [3 favorites]
I thought we were done with Haraway's cyborgs and had moved on to monsters. Even so! It's interesting to ask in this age of extended consciousness whether an LLM can even count as a cyborg appendage, or if it is exactly the opposite, a sort of techno-psychic amputation.
posted by mittens at 4:25 AM on September 5 [1 favorite]
posted by mittens at 4:25 AM on September 5 [1 favorite]
> [larb:] We also need to figure out again and again who “we” are
posted by HearHere at 6:33 AM on September 5
posted by HearHere at 6:33 AM on September 5
I would guess two things snuff out the spark of wanting to explore, experiment and express yourself: Assessment and boredom.
I mean, sure, but also: people are inherently pretty lazy. I'm sure some kids are worried about the assessment part of it, but I think it's more likely that taking the easiest route to being done quickly with a baseline level of "quality" is incredibly attractive. Unfortunately, it's not very helpful to the actual task of learning. Especially given the fact that many tasks like writing can be even more boring, tedious, and hard when you haven't done them much and aren't very good at doing them yet. So I'm sure you could try to come up with fun, freer, more exciting prompts for writing, but no, I very much doubt that is a silver bullet solution to kids using LLMs to avoid doing the work.
To the point of what is art: I think the fact that we're still all talking about Duchamp to this day, arguing about what his work is, what it means (or doesn't mean), proves that he was an incredibly successful artist. But to me part of the definition of art is that it drives people to want to discuss and analyze it with one another, that it is so beautiful you want to understand how, or so weird you need to ask someone else what they think, or even sometimes so unexpectedly banal that it makes you write many paragraphs about how the artist is a fraud, more than a century later. When someone has a really weird, basic idea that they execute with an LLM that makes everyone talk, and keep talking, and keep talking for years and years, they will absolutely have created art (and I am sure this will happen). But it will be the genius of the weirdness of the idea that makes it art, which yes, requires (I think) a human understanding of the context of human conversations at the current point in time.
posted by ch1x0r at 9:59 AM on September 5
I mean, sure, but also: people are inherently pretty lazy. I'm sure some kids are worried about the assessment part of it, but I think it's more likely that taking the easiest route to being done quickly with a baseline level of "quality" is incredibly attractive. Unfortunately, it's not very helpful to the actual task of learning. Especially given the fact that many tasks like writing can be even more boring, tedious, and hard when you haven't done them much and aren't very good at doing them yet. So I'm sure you could try to come up with fun, freer, more exciting prompts for writing, but no, I very much doubt that is a silver bullet solution to kids using LLMs to avoid doing the work.
To the point of what is art: I think the fact that we're still all talking about Duchamp to this day, arguing about what his work is, what it means (or doesn't mean), proves that he was an incredibly successful artist. But to me part of the definition of art is that it drives people to want to discuss and analyze it with one another, that it is so beautiful you want to understand how, or so weird you need to ask someone else what they think, or even sometimes so unexpectedly banal that it makes you write many paragraphs about how the artist is a fraud, more than a century later. When someone has a really weird, basic idea that they execute with an LLM that makes everyone talk, and keep talking, and keep talking for years and years, they will absolutely have created art (and I am sure this will happen). But it will be the genius of the weirdness of the idea that makes it art, which yes, requires (I think) a human understanding of the context of human conversations at the current point in time.
posted by ch1x0r at 9:59 AM on September 5
« Older To "accelerate the process" | Marvellous surprise as rare endangered coastal emu... Newer »
posted by craniac at 6:45 PM on September 3 [2 favorites]