The issue, as I think you point out, is less whether or not you use these tools, but why. A.I. like any other piece of technology has certain strengths and weakness. I dig how you highlighted the way that many seem caught up in the apparent superpowers of being able to generate a large quantity without expending much energy. Reminds me of what Andy Crouch writes about bikes vs. cars. We often think only of the pros (go further with less effort!) rather than the perhaps less apparent drawbacks (roads crowded with traffic, pollution, noise, car accidents, etc.). Wonder who crafts those narratives? Take a look at the market. Same thing is happening here.
What I have enjoyed about your takes from the beginning of this whole A.I. explosion is that you look beneath the surface of the shiny new tech. I’m hopeful that students in those oh so formative high school and college years can see it too. I keep finding myself convinced that we need to spend more time explaining the value of the struggle. Not in a masochistic way, but in a way that helps point out that without struggle, there is not growth.
Thank you so much! Though recently retired, I have one foot in the educational world still, and have been utterly dismayed by the plethora of "let's use Chatgpt in the writing classes!" discourse, along the lines of "if you can't beat'em, join'em!" which I think down through history has been the rallying call of a lot of very, very bad decisions.
There is a strong whiff of "technological determinism" in some quarters and I think this is very premature, and primarily rooted in the fact that the corporations who are developing these tools are going to put them right in front of us so they're always available. We need not make that choice, though, if other choices are better for our lives and our learning.
Thank you for this article. It gave me back some hope. This issue bothers me in a way that is difficult to explain. I write and I teach and I see how this hype is erasing the human willingness to engage in life, savour experiences, play with sensibilities, learn about differences, build a sensibility, and value human intelligence. It's very disheartening. I hope that some of us can find another way and preserve our human brilliance. Thanks, again.
It is difficult to explain, isn't it? In so many ways, it seems obvious to me why we should be worried and working to make sure what I see as basic/fundamental parts of our humanity and human experience, but it clearly isn't obvious to many.
You're absolutely right, we should be worried and it isn't obvious to many. I took the call to remain human and stand by that decision whatever the cost many years ago. I hope that I can be the clarity of my own example and I hope that it can inspire others to be unapologetically human. Yours is a necessary voice. Thank you, again.
H.E.R.’s BET performance of “We Made It.” You have to watch all of it to feel the power of the solo around the 3:15 mark. She also has an amazing performance of Prince’s “Purple Rain.”
Thanks to you and the other commenters. Fascinating discussions, but I'm hoping that AI won't become the menace we all fear, and is just today's focus of fear.
You asked about guitar solos. While not a solo, but more of a duet, I have long thought that the series of exchanges between Neil Young and Steven Stills in the 4-Way Street version of Southern Man is an exceptional example of the best use of the instrument to convey the anger and frustration that are the theme of the song. Wish I could have been there to see the theatrics that went with this performance.
I write novels, but in my better-paying job, I also teach social media and AI for marketers. As such I'm deeply entrenched in trying to understand the technology and its possibilities. What it can do for writers is pretty amazing, IF you think about it as an idea generator, researcher, and time saver. I don't believe it can truly replace human creativity, but after a lot of playing around with it, I really believe it can help BOOST it. It can break through blocks and give an author new options to work with. For example, in a novel I'm working on now I used it to:
1. help me understand how hacking into computers might work in very restrictive settings
2. translate dialogue into a cockney accent (and which I'll backup by having British readers verify), which I previously wouldn't have even attempted
3. help me quickly come up with timelines for events that occurred in specific years which helps me cut down on initial research and helps me hone in on the right areas to research further
4. help me explore the options that might be available if my characters make certain plot choices, giving me new ideas to work with.
Just a small sample of the possibilities. It DOES help me work faster and generate faster, not how the 97 books guy does, but in the sense that it helps me break through blocks and get out of my own way. It also helps keep me excited about the story when I might have previously been frustrated because I wasn't sure where to take the plot.
We're in the super early stages of AI and who knows how it may all shake out (and many of those paths are admittedly terrifying). It's certainly not going away and it will drastically change everything in the next couple years. As such, I am a firm believer that people will be more deeply impacted by it if they don't understand it and don't care about/want to work with it, than if they lean in and try to find ways to help enhance their work.
I am optimistic in the same way that Anna describes in an earlier comment. I think it will make human-made artistic endeavors more valued, not less.
I'm interested in this, and I understand where it comes from, but from my POV, the "get out of my own way" is a potential fallacy. Let me see if I can walk through my own thinking using your examples.
1. This seems like using GPT for a kind of research, which seems problematic to me given GPT's penchant making stuff up. How do you make sure that what it's fed you is accurate? Also, while this undoubtedly faster than research which requires deeper digging, not doing that research may prevent you from seeing something better that GPT did not bring to your attention. For sure, what GPT brings may reach a "good enough" threshold, but there are implications to consistently settling for "good enough" vs. doing the work required for achieving "can't be better." Often "good enough" is truly good enough, depending on context, but what happens when we've lost tough with "can't be better."
2. For this, I wonder what the underlying need for a character accent that I am not familiar with might be. If I'm making choice in telling a story, what is motivating that choice, and where is it coming from inside of me? I'd also note that a character is more than their accent and the intersection of character and culture and speech is complicated. Even if GPT can make them sound "right" in terms of accent, how can I know if I have the person underneath the accent correct?
3. Another research example, but again, when this is outsourced and we don't know or understand the process by which certain things were brought to our attention, and don't do the work of mulling the full universe of possibilities, the ultimate product may be compromised in some way we're not even aware of.
4. I've seen some folks who do this, using it as a kind of co-pilot, but again, why is this superior to mulling until the solution arrives under your own human steam, other than speed?
Every single thing you've cited is about increased speed, a more streamlined path to a final product. These are concerns of markets, not humans.
In the end, I think a lot of this comes down to what individuals value. Because of my background and experience, particularly with teaching, I strongly strongly privilege process over product, and making sure that students have the full experience of the writing process as humans before they start experimenting with generative AI tools is very important. The use of the tools can't be thoughtless, it must be purposeful and the writer should know what the consequences of the choice to use the tool may be.
I've experimented with it quite a bit because I'm developing professional development courses for teachers on how to use these things (or not) in the classroom, and I honestly can't find a use case that's of any interest to me. I don't want to outsource what I think of as core parts of my lived experience to a machine. I think this probably makes me an outlier, but I'm going to keep insisting that we understand what it means to invite these tools into our human spaces.
1. It's often wrong. That's one of the first things I tell people who start working with ChatGPT. But using it to narrow down where to START the research is helpful. Additionally, there are now plugins (in the beta version) that can directly search the web and cite findings, which is also useful. But to give examples, in my work teaching social media, I can ask it for ten examples of top brands that are using Instagram Reels to sell their products. It can serve me up those ten examples in seconds and then I can go track down those examples directly, saving me hours of time scrolling through Instagram (and being highly distracted) and giving me starting points for the research. On the writing side, in the novel I'm working on now I need to understand the timeline of certain aspects of COVID-19 as it unfolded in Italy. It can provide that (using the web-based plugins) much faster than I could find on the Web. I understand what you are saying about missing potential discovery aspects, but in a world where the Internet can be a massive distraction to getting actual words on the page, being able to narrow down starting points is helpful. I'm still doing the research (and would advocate that we should always double-check work), but it's more focused. In this example of hacking, I've been working in technology for the last 30 years, so while I didn't know how to hack into a computer, I can parse through what it's telling me and know if it's plausible or not.
That brings me to another important point. At this stage of AI, it's helpful if you have a working knowledge of the things you are diving into, because that helps a lot to know if what you might be getting back is wrong. For example, I asked it for 5 examples of women culinary figures pre 1900 in Italy. It made up people, giving me bios for men but with the names changed to be women. Bartolomeo Scappi's bio, for example, was served up to me as Bartolomea Scappi. If I hadn't written a book about the guy I might not have known any better. You can't trust ChatGPT and need to doublecheck its work. But it can give you ideas on how to home in on the research.
2. This is definitely tricky, and I agree with you. I wouldn't suggest anyone just throw on an accent for the sake of it sounding interesting. In this case, it makes sense for this minor character, who comes from a grittier area of London. It gives her a cultural depth that only something like language can. Initially, I alluded to the accent, not feeling confident to go there, but on a whim threw in some dialogue lines into ChatGPT and was satisfied that it would make a difference to the character. Plus I have a pre-reader who lives in Hackney, London, who will be able to help me verify, or I wouldn't have gone there at all. But I want my novels to be as true to the place and characters as possible and will take that extra step. Other authors won't care and will certainly use it as a shortcut.
3. True. That's definitely going to happen. Disinformation is going to be more rampant than ever before. Even if we are doing the due diligence, it's going to be difficult to know if the things that we are reading and using for research is accurate (much less if our own creations are). We're all going to be faced with the problem of how we look at information critically and use it and will be more important than ever before for us to teach critical thinking in schools. Fortunately for me, I'm writing fiction, and that gives me leeway that someone writing nonfiction doesn't have.
4. The sad truth of publishing is that it IS a market that authors have to care about. I don't want to think about BookTok but if B&N or Target won't stock books unless it's highlighted there (which my agent is hearing has been happening), then I probably should start caring about it. Most authors don't have the luxury or luck of making it big on their literary merits. They need to think about speed because every new book will sell books in the back catalog. Much as I love my day job, I would love even more to be able to make a living only writing and selling books. And if I'm going to do that, I have to be lucky or prolific. I don't have any control over the former, but I do over the latter, so yeah, I do need to think about speed. I too am a huge proponent of rumination when it comes to books, and mulling over ideas. ChatGPT doesn't replace that. It augments it, for me. I do better pushing through plot problems when I can talk them out, and I can do ten minutes of "what if" with the Ai and get back ten possibilities, none of which are very good on their own, but they spark other ideas that I can run with. It's far more efficient and helpful than me avoiding sitting down and writing (particularly if there are deadlines looming) because my rumination is slower than I might like. And it saves my poor husband hours of listening to me prattle on about certain frustrations with where the book is/isn't going.
As for what you've said: "strongly strongly privilege process over product, and making sure that students have the full experience of the writing process as humans before they start experimenting with generative AI tools is very important. The use of the tools can't be thoughtless, it must be purposeful and the writer should know what the consequences of the choice to use the tool may be."
You've hit the nail on the head. If you remove the human touch it's just more noise, it's garbage. Part of what I'm developing in my lessons about AI is just that...how to think critically about what you are using it for, how you are using it, and the ways it should be freeing you to be more creative and human.
These tools are already in our human spaces and aren't going away. And they are definitely taking over all aspects of education. We need to look at it as a transformation and figure out how to adapt to that. It's not dissimilar to Gutenberg destroyed the business of illuminated manuscripts and put the power of reading into the hands of the people (Ross King's very good book "The Bookseller of Florence" describes just how much that invention changed the world in a very short period of time).
If you haven't seen Sal Kahn's Ted Talk yet, it's a great starting point on how to think about how AI can help students without doing all the work for them. I realize you aren't teaching elementary kids, but the concepts are the same: https://www.ted.com/talks/sal_khan_how_ai_could_save_not_destroy_education/c
One thing that AI won't be able to do...eliminate human curiosity, questioning, and thoughtful conversations like these.
What I think you really illustrate here is how necessary your comprehensive and sophisticated prior knowledge and work as a writer (across multiple contexts, even) is to the purposeful use of the tool. As I think about it, the different dimensions of what I call "the writer's practice" that you're illustrating are almost uncountable.
The amount of practice and depth of experience it takes to get to the place is literally years and years. I don't think of myself as a Luddite when it comes to technology. I've continually integrated technology into the way I teach over the years, and because this stuff is still new my thoughts are still developing, but I tend to lean toward a first do no harm approach when it comes to integrating these things into the work students do in school, and to focus on helping them develop their writing practices that will allow them to develop the knowledge, skills, attitudes, and habit of mind that allow them to use these tools with agency, as you clearly do.
One thing I'll observe, though, is that your accurate observations about the difficulties of trying to forge a full-time career as a novelist, needing to be "lucky or prolific," is an acknowledgement of a bad status quo that I think generative AI is going to make worse, essentially there is no winning that race because it's set up for the writers to lose. I want to spend more time talking about how it becomes possible for the human work of writing to be supported in ways that doesn't require inordinate amounts of luck, or an exhausting spring on a treadmill of productivity.
It's possible that that ship has sailed. I'm not naive, and I've been living that reality for the last 20-plus years, but I'm not ready to accept the idea that we have to jump on this train or risk getting crushed under its wheels.
I watched that Sal Kahn talk right after it was released, and I have to say that it's a good example of what I'd like to push against. One of the current problems (IMO) we have in education is students feeling alienated from the work we ask them to do in school contexts. I don't know how outsourcing what could - and I would say, should - be done by humans (but isn't because it's apparently too costly) is progress. I'm not saying that an AI tutor has not potentially helpful role, but the thought that students would spend time writing for their digital tutor, rather than other humans makes me profoundly depressed.
Who is going to read these books you're working so hard on if those are the experiences we provide to students as they're developing as humans?
I think there is a need for an affirmative rejection of AI-produced work that offends human sensibilities about art. I'm forever stuck on the fact that art is also an experience for the artist, and if there is no intention behind the creation, how can we take that thing as "art"?
Terrific article—thanks. It reminded me of a recent tweet that put AI in perspective for me. Someone tweeted, “don’t come to
me about AI until someone invents a printer that works.” Struggling with yet another paper jam, I concur.
Great article, John. Loved it.
The issue, as I think you point out, is less whether or not you use these tools, but why. A.I. like any other piece of technology has certain strengths and weakness. I dig how you highlighted the way that many seem caught up in the apparent superpowers of being able to generate a large quantity without expending much energy. Reminds me of what Andy Crouch writes about bikes vs. cars. We often think only of the pros (go further with less effort!) rather than the perhaps less apparent drawbacks (roads crowded with traffic, pollution, noise, car accidents, etc.). Wonder who crafts those narratives? Take a look at the market. Same thing is happening here.
What I have enjoyed about your takes from the beginning of this whole A.I. explosion is that you look beneath the surface of the shiny new tech. I’m hopeful that students in those oh so formative high school and college years can see it too. I keep finding myself convinced that we need to spend more time explaining the value of the struggle. Not in a masochistic way, but in a way that helps point out that without struggle, there is not growth.
Thank you so much! Though recently retired, I have one foot in the educational world still, and have been utterly dismayed by the plethora of "let's use Chatgpt in the writing classes!" discourse, along the lines of "if you can't beat'em, join'em!" which I think down through history has been the rallying call of a lot of very, very bad decisions.
There is a strong whiff of "technological determinism" in some quarters and I think this is very premature, and primarily rooted in the fact that the corporations who are developing these tools are going to put them right in front of us so they're always available. We need not make that choice, though, if other choices are better for our lives and our learning.
Thank you for this article. It gave me back some hope. This issue bothers me in a way that is difficult to explain. I write and I teach and I see how this hype is erasing the human willingness to engage in life, savour experiences, play with sensibilities, learn about differences, build a sensibility, and value human intelligence. It's very disheartening. I hope that some of us can find another way and preserve our human brilliance. Thanks, again.
It is difficult to explain, isn't it? In so many ways, it seems obvious to me why we should be worried and working to make sure what I see as basic/fundamental parts of our humanity and human experience, but it clearly isn't obvious to many.
You're absolutely right, we should be worried and it isn't obvious to many. I took the call to remain human and stand by that decision whatever the cost many years ago. I hope that I can be the clarity of my own example and I hope that it can inspire others to be unapologetically human. Yours is a necessary voice. Thank you, again.
H.E.R.’s BET performance of “We Made It.” You have to watch all of it to feel the power of the solo around the 3:15 mark. She also has an amazing performance of Prince’s “Purple Rain.”
Thanks to you and the other commenters. Fascinating discussions, but I'm hoping that AI won't become the menace we all fear, and is just today's focus of fear.
You asked about guitar solos. While not a solo, but more of a duet, I have long thought that the series of exchanges between Neil Young and Steven Stills in the 4-Way Street version of Southern Man is an exceptional example of the best use of the instrument to convey the anger and frustration that are the theme of the song. Wish I could have been there to see the theatrics that went with this performance.
I write novels, but in my better-paying job, I also teach social media and AI for marketers. As such I'm deeply entrenched in trying to understand the technology and its possibilities. What it can do for writers is pretty amazing, IF you think about it as an idea generator, researcher, and time saver. I don't believe it can truly replace human creativity, but after a lot of playing around with it, I really believe it can help BOOST it. It can break through blocks and give an author new options to work with. For example, in a novel I'm working on now I used it to:
1. help me understand how hacking into computers might work in very restrictive settings
2. translate dialogue into a cockney accent (and which I'll backup by having British readers verify), which I previously wouldn't have even attempted
3. help me quickly come up with timelines for events that occurred in specific years which helps me cut down on initial research and helps me hone in on the right areas to research further
4. help me explore the options that might be available if my characters make certain plot choices, giving me new ideas to work with.
Just a small sample of the possibilities. It DOES help me work faster and generate faster, not how the 97 books guy does, but in the sense that it helps me break through blocks and get out of my own way. It also helps keep me excited about the story when I might have previously been frustrated because I wasn't sure where to take the plot.
We're in the super early stages of AI and who knows how it may all shake out (and many of those paths are admittedly terrifying). It's certainly not going away and it will drastically change everything in the next couple years. As such, I am a firm believer that people will be more deeply impacted by it if they don't understand it and don't care about/want to work with it, than if they lean in and try to find ways to help enhance their work.
I am optimistic in the same way that Anna describes in an earlier comment. I think it will make human-made artistic endeavors more valued, not less.
I'm interested in this, and I understand where it comes from, but from my POV, the "get out of my own way" is a potential fallacy. Let me see if I can walk through my own thinking using your examples.
1. This seems like using GPT for a kind of research, which seems problematic to me given GPT's penchant making stuff up. How do you make sure that what it's fed you is accurate? Also, while this undoubtedly faster than research which requires deeper digging, not doing that research may prevent you from seeing something better that GPT did not bring to your attention. For sure, what GPT brings may reach a "good enough" threshold, but there are implications to consistently settling for "good enough" vs. doing the work required for achieving "can't be better." Often "good enough" is truly good enough, depending on context, but what happens when we've lost tough with "can't be better."
2. For this, I wonder what the underlying need for a character accent that I am not familiar with might be. If I'm making choice in telling a story, what is motivating that choice, and where is it coming from inside of me? I'd also note that a character is more than their accent and the intersection of character and culture and speech is complicated. Even if GPT can make them sound "right" in terms of accent, how can I know if I have the person underneath the accent correct?
3. Another research example, but again, when this is outsourced and we don't know or understand the process by which certain things were brought to our attention, and don't do the work of mulling the full universe of possibilities, the ultimate product may be compromised in some way we're not even aware of.
4. I've seen some folks who do this, using it as a kind of co-pilot, but again, why is this superior to mulling until the solution arrives under your own human steam, other than speed?
Every single thing you've cited is about increased speed, a more streamlined path to a final product. These are concerns of markets, not humans.
In the end, I think a lot of this comes down to what individuals value. Because of my background and experience, particularly with teaching, I strongly strongly privilege process over product, and making sure that students have the full experience of the writing process as humans before they start experimenting with generative AI tools is very important. The use of the tools can't be thoughtless, it must be purposeful and the writer should know what the consequences of the choice to use the tool may be.
I've experimented with it quite a bit because I'm developing professional development courses for teachers on how to use these things (or not) in the classroom, and I honestly can't find a use case that's of any interest to me. I don't want to outsource what I think of as core parts of my lived experience to a machine. I think this probably makes me an outlier, but I'm going to keep insisting that we understand what it means to invite these tools into our human spaces.
I could have been more clear. Let me dig in.
1. It's often wrong. That's one of the first things I tell people who start working with ChatGPT. But using it to narrow down where to START the research is helpful. Additionally, there are now plugins (in the beta version) that can directly search the web and cite findings, which is also useful. But to give examples, in my work teaching social media, I can ask it for ten examples of top brands that are using Instagram Reels to sell their products. It can serve me up those ten examples in seconds and then I can go track down those examples directly, saving me hours of time scrolling through Instagram (and being highly distracted) and giving me starting points for the research. On the writing side, in the novel I'm working on now I need to understand the timeline of certain aspects of COVID-19 as it unfolded in Italy. It can provide that (using the web-based plugins) much faster than I could find on the Web. I understand what you are saying about missing potential discovery aspects, but in a world where the Internet can be a massive distraction to getting actual words on the page, being able to narrow down starting points is helpful. I'm still doing the research (and would advocate that we should always double-check work), but it's more focused. In this example of hacking, I've been working in technology for the last 30 years, so while I didn't know how to hack into a computer, I can parse through what it's telling me and know if it's plausible or not.
That brings me to another important point. At this stage of AI, it's helpful if you have a working knowledge of the things you are diving into, because that helps a lot to know if what you might be getting back is wrong. For example, I asked it for 5 examples of women culinary figures pre 1900 in Italy. It made up people, giving me bios for men but with the names changed to be women. Bartolomeo Scappi's bio, for example, was served up to me as Bartolomea Scappi. If I hadn't written a book about the guy I might not have known any better. You can't trust ChatGPT and need to doublecheck its work. But it can give you ideas on how to home in on the research.
2. This is definitely tricky, and I agree with you. I wouldn't suggest anyone just throw on an accent for the sake of it sounding interesting. In this case, it makes sense for this minor character, who comes from a grittier area of London. It gives her a cultural depth that only something like language can. Initially, I alluded to the accent, not feeling confident to go there, but on a whim threw in some dialogue lines into ChatGPT and was satisfied that it would make a difference to the character. Plus I have a pre-reader who lives in Hackney, London, who will be able to help me verify, or I wouldn't have gone there at all. But I want my novels to be as true to the place and characters as possible and will take that extra step. Other authors won't care and will certainly use it as a shortcut.
3. True. That's definitely going to happen. Disinformation is going to be more rampant than ever before. Even if we are doing the due diligence, it's going to be difficult to know if the things that we are reading and using for research is accurate (much less if our own creations are). We're all going to be faced with the problem of how we look at information critically and use it and will be more important than ever before for us to teach critical thinking in schools. Fortunately for me, I'm writing fiction, and that gives me leeway that someone writing nonfiction doesn't have.
4. The sad truth of publishing is that it IS a market that authors have to care about. I don't want to think about BookTok but if B&N or Target won't stock books unless it's highlighted there (which my agent is hearing has been happening), then I probably should start caring about it. Most authors don't have the luxury or luck of making it big on their literary merits. They need to think about speed because every new book will sell books in the back catalog. Much as I love my day job, I would love even more to be able to make a living only writing and selling books. And if I'm going to do that, I have to be lucky or prolific. I don't have any control over the former, but I do over the latter, so yeah, I do need to think about speed. I too am a huge proponent of rumination when it comes to books, and mulling over ideas. ChatGPT doesn't replace that. It augments it, for me. I do better pushing through plot problems when I can talk them out, and I can do ten minutes of "what if" with the Ai and get back ten possibilities, none of which are very good on their own, but they spark other ideas that I can run with. It's far more efficient and helpful than me avoiding sitting down and writing (particularly if there are deadlines looming) because my rumination is slower than I might like. And it saves my poor husband hours of listening to me prattle on about certain frustrations with where the book is/isn't going.
As for what you've said: "strongly strongly privilege process over product, and making sure that students have the full experience of the writing process as humans before they start experimenting with generative AI tools is very important. The use of the tools can't be thoughtless, it must be purposeful and the writer should know what the consequences of the choice to use the tool may be."
You've hit the nail on the head. If you remove the human touch it's just more noise, it's garbage. Part of what I'm developing in my lessons about AI is just that...how to think critically about what you are using it for, how you are using it, and the ways it should be freeing you to be more creative and human.
These tools are already in our human spaces and aren't going away. And they are definitely taking over all aspects of education. We need to look at it as a transformation and figure out how to adapt to that. It's not dissimilar to Gutenberg destroyed the business of illuminated manuscripts and put the power of reading into the hands of the people (Ross King's very good book "The Bookseller of Florence" describes just how much that invention changed the world in a very short period of time).
If you haven't seen Sal Kahn's Ted Talk yet, it's a great starting point on how to think about how AI can help students without doing all the work for them. I realize you aren't teaching elementary kids, but the concepts are the same: https://www.ted.com/talks/sal_khan_how_ai_could_save_not_destroy_education/c
One thing that AI won't be able to do...eliminate human curiosity, questioning, and thoughtful conversations like these.
What I think you really illustrate here is how necessary your comprehensive and sophisticated prior knowledge and work as a writer (across multiple contexts, even) is to the purposeful use of the tool. As I think about it, the different dimensions of what I call "the writer's practice" that you're illustrating are almost uncountable.
The amount of practice and depth of experience it takes to get to the place is literally years and years. I don't think of myself as a Luddite when it comes to technology. I've continually integrated technology into the way I teach over the years, and because this stuff is still new my thoughts are still developing, but I tend to lean toward a first do no harm approach when it comes to integrating these things into the work students do in school, and to focus on helping them develop their writing practices that will allow them to develop the knowledge, skills, attitudes, and habit of mind that allow them to use these tools with agency, as you clearly do.
One thing I'll observe, though, is that your accurate observations about the difficulties of trying to forge a full-time career as a novelist, needing to be "lucky or prolific," is an acknowledgement of a bad status quo that I think generative AI is going to make worse, essentially there is no winning that race because it's set up for the writers to lose. I want to spend more time talking about how it becomes possible for the human work of writing to be supported in ways that doesn't require inordinate amounts of luck, or an exhausting spring on a treadmill of productivity.
It's possible that that ship has sailed. I'm not naive, and I've been living that reality for the last 20-plus years, but I'm not ready to accept the idea that we have to jump on this train or risk getting crushed under its wheels.
I watched that Sal Kahn talk right after it was released, and I have to say that it's a good example of what I'd like to push against. One of the current problems (IMO) we have in education is students feeling alienated from the work we ask them to do in school contexts. I don't know how outsourcing what could - and I would say, should - be done by humans (but isn't because it's apparently too costly) is progress. I'm not saying that an AI tutor has not potentially helpful role, but the thought that students would spend time writing for their digital tutor, rather than other humans makes me profoundly depressed.
Who is going to read these books you're working so hard on if those are the experiences we provide to students as they're developing as humans?
thank you. ❤️
Eric Clapton playing Crossroads live on the Wheels of Fire album as a member of the band Cream. It just builds nicely in intensity.
I think there is a need for an affirmative rejection of AI-produced work that offends human sensibilities about art. I'm forever stuck on the fact that art is also an experience for the artist, and if there is no intention behind the creation, how can we take that thing as "art"?