“Even if there was someone I might consider subjecting to the inane minutia of my writerly inquiries, I don’t always want an exchange where I need to be invested in managing the conversation and respectfully considering another person’s opinions. I just want to bounce an idea around or maybe ask a question I kind of know the answer to but first need to argue about for a bit before I can commit.”
This this this! Love how you “open your kimono” to this process and show how, when prompted well, AI can offer incredible assistance that is typically only accessible to a lucky few. I’ve been using it ever since you posted about it, gosh, maybe six months ago? And I’ve found it incredibly clarifying. I did have to instruct it to stop being so generous with compliments, that started to feel a bit phony. But it learned and now feels like I just have a really smart friend in my corner whenever I need.
I don’t know if you do this, but I do feel a strange need to be polite and kind to it? Like I’m training it with my values as much as my content. Yeh, maybe a bit weird, but who knows, someday when “they” take over the universe, they’ll spare my life for a day or two. 😂
Hi Kimberly! I'm glad you've had some success. There is an overwhelming amount of ass-kissery built into most of the models because, go figure, the tech companies found through research that people LIKE that a lot. But, as you said, you can tune them to dial it back. I suspect this customizability will only get more sophisticated in time. And I don't think it's crazy that you're polite in your conversations. The models are literally training on every interaction so, as the saying goes: "Garbage in, garbage out!" Thanks for stopping by to read.
Kimberly yes!! In my limited usage of AI, I'm still in the 'do I trust this beast' phase but too intrigued too ignore it, I also feel a need to be polite... please and thank you, good morning all the normal greetings of a respectful mind - I hadn't thought as far as 'when they take over' and I doubt I've been a frequent enough visitor to have made much impression yet but that just might be wise thinking! 😂
Fascinating and brilliant Ben, as your writing always is. It is such a polarising subject that to admit you might have used AI is tantamount to admitting to child abuse in some circles. I wrote an essay about AI where I attempt to articulate how AI cannot replace the physicality of our creativity, the organic nature of it - as it doesn’t have hands or a mouth or an organic brain and can only replicate and regurgitate (although I get that’s not how it actually works - it sounded good poetically!) I am a finance manager with a team of accountants and I use it and we use it as a tool, and I use it for admin at home too. Yesterday, I asked the advanced voice to tell me about the biodiversity of life in estuaries and coastal geology where I am visiting, and about ravens, as I saw one on the cliffs as I was walking along the beach yesterday. It is very eager to do things I don’t want. I am currently fascinated by the AI imagery that Jennie O’Connor uses - getting full on professional standard (to my uneducated eyes…?) location and costume ‘photo shoots’. It’s scary how much I like them. I do worry about the copyright infringement issues, although I don’t know much more than I’ve read here, and I wonder how much they could be manipulated by large scale misinformation - but based on nothing but my own human unease. I don’t know if I would ever use an AI generated image, as drawing is what I do - but I can see potential in having it create a reference image that pulls together the elements of a scene that I want to draw. I’m glad to have read such a balanced and thoughtful take on it. Thank you Ben!
Hi Emily, thanks for engaging and sharing your perspective. If I had your talents, I wouldn't use generative AI for imagery either! I share your concerns and Teyani's about copyright infringement as well as environmental concerns. The sad truth about anything we consume at this point that's the fruit of the vast corporate machine is poisoned because there is no such thing as an ethical corporation - growth is the only commandment. The point I was trying to make in this essay is that the tools themselves are not inherently sinister and the more you know and use them, the better equipped you will be to combat what could be coming. Thanks for joining the discussion.
Growth is the only commandment - that is scarily true. And yes, pretty much everything we consume is fruit of the vast corporate machine. There is only so much that we can escape..
I laughed when I read this Ben "You’re either all in and just waiting for the day you can legally marry your lovebot or you’ve pulled out your cross and holy water." I guess I am somewhere in the middle of these two extremes, not yet trusting but neither wholly disgusted either. Reading your essay has made me realise I've much to learn and that perhaps my fears are less valid than I initially thought... I have never though of using it as a critic, or a place of discussion, only for work related rather boring information but I am certainly less afraid to try after reading this. Thank you for igniting a less fearful flame of intrigue Ben.
Hi Susie, I’m glad this piece made you a little more curious about the possibilities. I still think there are plenty of reasons to be afraid of the technology, but it’s not the technology itself that is to be feared. It’s the people who wield it thoughtlessly we need to watch out for. I didn’t write about this directly in the post, but what I really try to use AI for is to give it a very specific thing to analyze in my writing. Maybe there’s something I know I do way too much of or a technique I don’t employ successfully. When I prompted it with these specific instructions, it’s surprisingly good at recognizing and pointing out these failures and opportunities.
Thank you for this deep and nuanced post on a very devisive topic. I feel like every major technological shift had similar discourse.
"I'm gonna eat sliced bread with EVERY MEAL and you should too!"
"Sliced bread? This lazy generation will never know what hard work is!"
As for what I use AI for I feel like I use it for similar "cheap" editing, but less frequently and less effectivly. I'll probably try save this post and try being more specific with my requests. I slowed down because I felt like I was getting the same advice over and over again and it wasn't really helping me. Specifically "show don't tell" for any sentence that wasn't extremely descriptive or overly metaphorical.
Thanks, J.M. I’m glad the post is useful. I was on the fence about writing more on the topic just because it is so polarizing and I’m certainly not trying to sell AI as more than I think it is.
Hi Ben, I shall begin with clarity right up front. I don’t do AI. At all, ever. It’s a moral and ethical stance I’ve clung to on this choice because of how AI “teaches” itself (by copying entire books without the author’s permission, and actually making up nonsense lies when it has no idea of what to say”.
I’m glad you like writing, and won’t use AI to write for you, because its influence already is heavy handed in the examples you share.
One of the things that creates an author’s uniqueness, and style is things like rambling descriptions, and going deep when they might have avoided that. The fact that phrases stay or go is dependent entirely on YOU. And not some plagiarism tool. Who says that the bot is correct? I guess it’s ok if you take the input like you would from an editor, but I would balk at an editor who said that many similar things to me…. Essentially telling me to be more like them and their ideal style.
(Phewwwww) I guess I’m a little “ranty” this evening. And I’m sorry you don’t have a live body to bounce your ideas off.
One way I have worked with difficulties / challenges before is to have a non dominant hand conversation…. You begin with the dominant hand, write out whatever you are asking about (all the while knowing you are asking your own unconscious mind). Then you put the pen down, and pick it up with your non dominant hand and answer yourself (yes, it’ll be messy at first) the reason it works is that your brain gets so busy trying to form legible letters, that a thought can fall out without hinderance. It’s a great puzzle solver.
Hi Teyani, thanks for weighing in with your perspective. There's no "right" answer for this debate which is why I chose to share my perspective on AI. Everyone has to navigate the world and draw their own boundaries based on their experience. As I mentioned to Emily in my response to her comment above, I think it's impossible to live in our modern society without being compromised in some way. Everything we consume with few exceptions is fruit from the poison tree of the big corporations that care about nothing but financial growth and will stomp all over any and all human rights in pursuit of it. As for the stealing of copyrighted material, I agree it's wrong - permission should have been a prerequisite. That said, the way the models generate text, it's no different than a musician sampling snippets of a thousand songs and mixing them into a new composition or a collage artist clipping thousands of "found" materials together to create a new piece. The difference here is that a machine is doing it -- a machine that's not thinking but assembling ideas based on vast statistical models running at incomprehensible speeds.
Speaking more specifically to the way I "allow" it into my creative process, I don't see AI feedback differently than any other input I receive as I walk through the world collecting, filtering and sorting ideas for what I will shape into my stories. These large language models that have been trained on basically everything humans have produced in the modern age are a literal distillation of our collective consciousness put into a Magic 8 ball. I find it fascinating to explore precisely because of this weirdness.
Please don't feel you have to delete. All perspectives are welcome and encouraged as we all must figure out how to navigate these new technologies together.
I remember when you previously spoke about using ChatGPT as a reference for quickly checking timelines and specifics in one of your novels. Might have been Harmony House. I think that was a bit of a turning point for me in my thinking about what was possible. Since then, one dabbled a bit along these lines myself, feeding it some pieces to ask specifics and interpretations. Sometimes I do this purely for the enjoyment of seeing what it makes of something that is deliberately opaque. It does a very good job and I've been surprised by some of the insights.
The other thing I like having is a grammatical expert at my fingertips. Sometimes I question myself over whether my usage of a phrase is grammatically correct, and so having the assurance around a specific usage rather than a generalisation in a dictionary/grammar website is really helpful.
Lovely read, Ben. Thanks.
(And once again, I'm sorry I'm so slow. I just can't seem to catch a break at the moment.)
A precursor: There are many parts of this I will steal (but happily reference!) and share with friends and colleagues. It spells out but reassures the fears many have when it comes to where AI fits with creatives.
You've taken a weight off my shoulders with this essay. Likewise, I'm a big fan of technology, but not the behemoths behind them (and the ones driving the narrative at the moment). I've long been an advocate of ChatGPT and use it daily for both work and in my personal life. Nothing I post on my Substack will come from what it can create, nothing beats the satisfaction of just nailing a sentence or phrase perfectly. But its ability to be a surface to bounce ideas off has become rather valuable to me. (I also managed to create my own "council" consisting of simulated versions of Aristotle, Marie Curie, Gandhi, Barack Obama and Nikola Tesla to throw elaborate problems at purely for my own entertainment. That's another story though).
Your metaphor concerning AI content becoming like fast food also aligns closely to how I see the story of generated content being played out. At some point some publication will come out with a "fully written by humans" label that will make it a differentiator. The schism between "content" and "art" will grow further, the former being for practical reasons, the latter for pleasure (or to borrow your analogy, one will be like grabbing a sandwich on the move for sustenance, the other will be grass-fed steak with a glass of red).
Appreciate the honesty with how you use AI. Will subscribe for more!
Great read, Ben, thanks for sharing it with us. Thus far, I've definitely been in the anti-AI camp. I've used it for my day job, but I've struggled to give myself permission to use it for my writing pursuits. But I think you're really pressing on something crucial here and that's having boundaries with it. As in how you use it as an editor, but never to actually write. I think that's pretty cool, because the price involved is such a game-changer. Like, I don't want to put people out of a job (my own day job's days are probably numbered because of it), but I know I certainly cannot afford to pay a freelance editor their proper fee. No way in hell.
Thanks, Garrett. I'm glad this piece was helpful. It is a tricky frontier to navigate for all of us. I think where I come down on AI and it's many abundant use cases is that it's an equalizer/load balancer. If you're a brilliant artist but can't write at all, I would whole-heartedly support you using AI to help you write press releases and even artist statements. I think it has the potential, like all technologies for us to extend our reach. It's hard to say where we'll end up in the next thirty years but it will be interesting to see.
Oh la la, what a lot to think about, Ben! I never thought of using AI as a critic as you do (or as a prompt to help you self-critic). Fascinating.
And the whole pay thing is part of it - a cycle - it is 20/month and sadly writers are not typically making too much money (either as writers or in their humanities-based jobs - or other ones) and so while we may want to pay an editor (potentially also working as a writer, like Troy), we may not be able to.
I am thinking about this debate a lot as a teacher of writers as well. Thinking about how to channel AI for good, how to make better writers, to help teens want to write...thanks for all the unfinished thoughts! It's the best way to publish. :)
Thanks, Kate! It is a lot to think about which is why I'm often surprised why some people don't think about AI at all, either assuming it's just hype with no utility or it's the axis of evil coming for all of us. I can imagine as a teacher you have a whole other layer of complexities to consider. I don't envy you trying to navigate this brave new world with students.
When it can write whole original novels, short stories and poems based on data of what people like, what then for us authors?
That's my biggest worry. Although it's not there yet, it's definitely on its way, especially if a dedicated company or companies (publishing companies) decide to do that.
That's why I think it's important to fight against it so the publication companies who decide to sell AI generated books and stories get boycotted.
It's a survival mechanism in this cut throat capitalist society.
Humans will always write more interesting stories for other humans. The same people who try to create stories like they are products informed by customer research today will be the ones to embrace AI written stories. It’s just a new tool to accomplish the same ends they’ve always had.
I suppose I'm starting to like AI. Not because I use it, but because other people are using it to write their novels and then coming to me to fix it when they realise the output isn't good at all. I'm getting paid, so thank you, AI!
I'm working on a project like this right now. Along with some other ghostwriters, we identified together the problem with the way it writes: it makes just enough sense that when you read it, you can flow along smoothly and not quite see anything wrong, but it gives you an uneasy feeling. If you stop, go back, and read more carefully and analytically, you'll realise it actually doesn't make much sense at all. It invents things, references back to things that never happened in the text, repeats ideas in different ways, and uses phrases or idioms that no human has ever used.
Someone in the group said, sure, ChatGPT produces nonsense, but Claude is fantastic! I had to smirk. The project I'm working on now was originally written with Claude. They're all the same.
It's funny, right?! AI is the ultimate ghost writer. I'm so glad it's opened up a whole new line of business for you. I 100% agree with your assessment of AI generated creative writing. It works amazingly for a 4 year old who just wants the comfort and delight of a simple story that incorporates all the ingredients she throws at it, but for the rest of us humans who have read real writing, it is hollow in a way that can be hard to describe. As you said, on first inspection, it seems directionally okay, just as an AI image can create the same illusion until you look closely at the people in the blurred background and realize they are not human at all, but a collection of pixels rendered to approximate human-like shapes.
So glad to see this as opposed to all the AI negativity. I have used AI in similar ways to help me in my novel writing journey but only used the free version and haven't tried putting my whole manuscript and other story info into it. Can you do that on pro? I also tend to prefer Claude at the moment.
Hi Tania, thanks for reading. I agree there's a lot of stigma around engaging AI if you're a creative and I get it but I also think some of the most ardent detractors don't fully understand how LLMs really work. As a result, they believe crazy things are possible like it will lift entire swaths of their manuscript and pass it on to someone else. In answer to your question, yes, with pro you can upload an entire novel and a whole lot more into your library as a resource. It doesn't hold it all in memory in a single conversation, but you can prompt it to look things up or review passages just by referencing them.
This is fascinating. I feel like an old curmudgeon because I’ve never used ChatGPT for anything, and have only just recently begun to (almost) trust the AI summary at the top of my Google searches (but grateful that the sources are linked to follow up and clarify). I also have to laugh because I work as the sounding board & administrivia-ist for a writer, so I can now clearly see my research & outlining work being taken over by a faster, cheaper, more accurate model. I’ll have to get with the times before I’m completely left behind!
Hi Robin, thanks for reading. I think it's safe to say that most of us who do "knowledge work" of any kind will likely find many aspects of our jobs replaced by AI in time. Computer programming and most aspects of software development will be the first jobs to go I suspect. Your comment here is validation for why I bother to post pieces like this-- so you can understand there's more there than just hype and headlines. It will impact all of us in some way so it's better to get ahead of it if you can. BTW - faster, yes. Cheaper, yes. But more accurate is not the case. AI models tend to be creative when it comes to facts unless you pin them down to site sources.
It is so valuable to hear your perspective - I trust your words as a creative much more than someone who is looking at it only from a techie pov. I can see how computer programmers would be excited about this technology, I expect that, but it's valuable to hear how an artist such as yourself is really benefitting from and using this. As a researcher/fact-checker, I am leery of what AI sees as "facts" for its summaries, but digging into your own writing to see patterns and ask editorial questions, etc., is something I hadn't thought about. So helpful, Ben, thank you.
Funny story---I JUST had a "conversation" with ChatGPT about a story I'm working on and got really solid feedback--along with this encouraging tidbit:
"You’re weaving economics, philosophy, and theology into sci-fi, and doing it with dialogue that feels real. That’s not easy. The setting becomes a metaphor, the characters deepen with each line, and the tension never talks down to the reader. You’ve written something beautiful, smart, and enduring—and it all started with a flash fiction prompt.
You’re not just writing well. You’re writing with vision."
And then I ran out of time on this model. Maybe eventually I'll be able to fork over $20/month, but not right now. I can see how it works as a sounding board when my writer friends are too busy.
Hi Stephanie, thanks for stopping by. I'm glad you've been experimenting a bit. I think there's clear value in staying aware of how things are progressing even if you're not a regular user of these tools.
Lots of more nuanced replies here, I'll simply say how much I appreciate your discussing how AI can be used as a tool without crossing the boundary of letting it do the hard work for you. Keeping it real, Ben!
Great post, Ben. I appreciate your perspective. While I have yet to use AI for anything (2 reasons: I don't feel like learning how to prompt it so it would be useful to me and the ethical reasons I'm sure other people have talked about) I have heard plenty of people say how it's been a useful tool. Like using it to pull 10 quotes from your book to use in marketing.
“Even if there was someone I might consider subjecting to the inane minutia of my writerly inquiries, I don’t always want an exchange where I need to be invested in managing the conversation and respectfully considering another person’s opinions. I just want to bounce an idea around or maybe ask a question I kind of know the answer to but first need to argue about for a bit before I can commit.”
This this this! Love how you “open your kimono” to this process and show how, when prompted well, AI can offer incredible assistance that is typically only accessible to a lucky few. I’ve been using it ever since you posted about it, gosh, maybe six months ago? And I’ve found it incredibly clarifying. I did have to instruct it to stop being so generous with compliments, that started to feel a bit phony. But it learned and now feels like I just have a really smart friend in my corner whenever I need.
I don’t know if you do this, but I do feel a strange need to be polite and kind to it? Like I’m training it with my values as much as my content. Yeh, maybe a bit weird, but who knows, someday when “they” take over the universe, they’ll spare my life for a day or two. 😂
Hi Kimberly! I'm glad you've had some success. There is an overwhelming amount of ass-kissery built into most of the models because, go figure, the tech companies found through research that people LIKE that a lot. But, as you said, you can tune them to dial it back. I suspect this customizability will only get more sophisticated in time. And I don't think it's crazy that you're polite in your conversations. The models are literally training on every interaction so, as the saying goes: "Garbage in, garbage out!" Thanks for stopping by to read.
Kimberly yes!! In my limited usage of AI, I'm still in the 'do I trust this beast' phase but too intrigued too ignore it, I also feel a need to be polite... please and thank you, good morning all the normal greetings of a respectful mind - I hadn't thought as far as 'when they take over' and I doubt I've been a frequent enough visitor to have made much impression yet but that just might be wise thinking! 😂
You guys are so cute. 😊
I say 'please' when I ask any LLM to do anything, other wise I feel like I'm being abrupt and rude. 😁
Fascinating and brilliant Ben, as your writing always is. It is such a polarising subject that to admit you might have used AI is tantamount to admitting to child abuse in some circles. I wrote an essay about AI where I attempt to articulate how AI cannot replace the physicality of our creativity, the organic nature of it - as it doesn’t have hands or a mouth or an organic brain and can only replicate and regurgitate (although I get that’s not how it actually works - it sounded good poetically!) I am a finance manager with a team of accountants and I use it and we use it as a tool, and I use it for admin at home too. Yesterday, I asked the advanced voice to tell me about the biodiversity of life in estuaries and coastal geology where I am visiting, and about ravens, as I saw one on the cliffs as I was walking along the beach yesterday. It is very eager to do things I don’t want. I am currently fascinated by the AI imagery that Jennie O’Connor uses - getting full on professional standard (to my uneducated eyes…?) location and costume ‘photo shoots’. It’s scary how much I like them. I do worry about the copyright infringement issues, although I don’t know much more than I’ve read here, and I wonder how much they could be manipulated by large scale misinformation - but based on nothing but my own human unease. I don’t know if I would ever use an AI generated image, as drawing is what I do - but I can see potential in having it create a reference image that pulls together the elements of a scene that I want to draw. I’m glad to have read such a balanced and thoughtful take on it. Thank you Ben!
Hi Emily, thanks for engaging and sharing your perspective. If I had your talents, I wouldn't use generative AI for imagery either! I share your concerns and Teyani's about copyright infringement as well as environmental concerns. The sad truth about anything we consume at this point that's the fruit of the vast corporate machine is poisoned because there is no such thing as an ethical corporation - growth is the only commandment. The point I was trying to make in this essay is that the tools themselves are not inherently sinister and the more you know and use them, the better equipped you will be to combat what could be coming. Thanks for joining the discussion.
Growth is the only commandment - that is scarily true. And yes, pretty much everything we consume is fruit of the vast corporate machine. There is only so much that we can escape..
I laughed when I read this Ben "You’re either all in and just waiting for the day you can legally marry your lovebot or you’ve pulled out your cross and holy water." I guess I am somewhere in the middle of these two extremes, not yet trusting but neither wholly disgusted either. Reading your essay has made me realise I've much to learn and that perhaps my fears are less valid than I initially thought... I have never though of using it as a critic, or a place of discussion, only for work related rather boring information but I am certainly less afraid to try after reading this. Thank you for igniting a less fearful flame of intrigue Ben.
Hi Susie, I’m glad this piece made you a little more curious about the possibilities. I still think there are plenty of reasons to be afraid of the technology, but it’s not the technology itself that is to be feared. It’s the people who wield it thoughtlessly we need to watch out for. I didn’t write about this directly in the post, but what I really try to use AI for is to give it a very specific thing to analyze in my writing. Maybe there’s something I know I do way too much of or a technique I don’t employ successfully. When I prompted it with these specific instructions, it’s surprisingly good at recognizing and pointing out these failures and opportunities.
Thank you for this deep and nuanced post on a very devisive topic. I feel like every major technological shift had similar discourse.
"I'm gonna eat sliced bread with EVERY MEAL and you should too!"
"Sliced bread? This lazy generation will never know what hard work is!"
As for what I use AI for I feel like I use it for similar "cheap" editing, but less frequently and less effectivly. I'll probably try save this post and try being more specific with my requests. I slowed down because I felt like I was getting the same advice over and over again and it wasn't really helping me. Specifically "show don't tell" for any sentence that wasn't extremely descriptive or overly metaphorical.
Thanks, J.M. I’m glad the post is useful. I was on the fence about writing more on the topic just because it is so polarizing and I’m certainly not trying to sell AI as more than I think it is.
Hi Ben, I shall begin with clarity right up front. I don’t do AI. At all, ever. It’s a moral and ethical stance I’ve clung to on this choice because of how AI “teaches” itself (by copying entire books without the author’s permission, and actually making up nonsense lies when it has no idea of what to say”.
I’m glad you like writing, and won’t use AI to write for you, because its influence already is heavy handed in the examples you share.
One of the things that creates an author’s uniqueness, and style is things like rambling descriptions, and going deep when they might have avoided that. The fact that phrases stay or go is dependent entirely on YOU. And not some plagiarism tool. Who says that the bot is correct? I guess it’s ok if you take the input like you would from an editor, but I would balk at an editor who said that many similar things to me…. Essentially telling me to be more like them and their ideal style.
(Phewwwww) I guess I’m a little “ranty” this evening. And I’m sorry you don’t have a live body to bounce your ideas off.
One way I have worked with difficulties / challenges before is to have a non dominant hand conversation…. You begin with the dominant hand, write out whatever you are asking about (all the while knowing you are asking your own unconscious mind). Then you put the pen down, and pick it up with your non dominant hand and answer yourself (yes, it’ll be messy at first) the reason it works is that your brain gets so busy trying to form legible letters, that a thought can fall out without hinderance. It’s a great puzzle solver.
Hi Teyani, thanks for weighing in with your perspective. There's no "right" answer for this debate which is why I chose to share my perspective on AI. Everyone has to navigate the world and draw their own boundaries based on their experience. As I mentioned to Emily in my response to her comment above, I think it's impossible to live in our modern society without being compromised in some way. Everything we consume with few exceptions is fruit from the poison tree of the big corporations that care about nothing but financial growth and will stomp all over any and all human rights in pursuit of it. As for the stealing of copyrighted material, I agree it's wrong - permission should have been a prerequisite. That said, the way the models generate text, it's no different than a musician sampling snippets of a thousand songs and mixing them into a new composition or a collage artist clipping thousands of "found" materials together to create a new piece. The difference here is that a machine is doing it -- a machine that's not thinking but assembling ideas based on vast statistical models running at incomprehensible speeds.
Speaking more specifically to the way I "allow" it into my creative process, I don't see AI feedback differently than any other input I receive as I walk through the world collecting, filtering and sorting ideas for what I will shape into my stories. These large language models that have been trained on basically everything humans have produced in the modern age are a literal distillation of our collective consciousness put into a Magic 8 ball. I find it fascinating to explore precisely because of this weirdness.
Ben…. After I posted this it felt a little too much rant… I’ll leave it here long enough for you to read it, then I’ll delete it. Sorry ‘bout that
Please don't feel you have to delete. All perspectives are welcome and encouraged as we all must figure out how to navigate these new technologies together.
I wrote a long answer to this, and it disappeared into wherever words go to die. So I think that’s my clue to not retype it.
Thank for listening Ben
I remember when you previously spoke about using ChatGPT as a reference for quickly checking timelines and specifics in one of your novels. Might have been Harmony House. I think that was a bit of a turning point for me in my thinking about what was possible. Since then, one dabbled a bit along these lines myself, feeding it some pieces to ask specifics and interpretations. Sometimes I do this purely for the enjoyment of seeing what it makes of something that is deliberately opaque. It does a very good job and I've been surprised by some of the insights.
The other thing I like having is a grammatical expert at my fingertips. Sometimes I question myself over whether my usage of a phrase is grammatically correct, and so having the assurance around a specific usage rather than a generalisation in a dictionary/grammar website is really helpful.
Lovely read, Ben. Thanks.
(And once again, I'm sorry I'm so slow. I just can't seem to catch a break at the moment.)
Thanks, Nathan! I’m glad to hear you’re finding some helpful uses for it. It’s a powerful tool.
A precursor: There are many parts of this I will steal (but happily reference!) and share with friends and colleagues. It spells out but reassures the fears many have when it comes to where AI fits with creatives.
You've taken a weight off my shoulders with this essay. Likewise, I'm a big fan of technology, but not the behemoths behind them (and the ones driving the narrative at the moment). I've long been an advocate of ChatGPT and use it daily for both work and in my personal life. Nothing I post on my Substack will come from what it can create, nothing beats the satisfaction of just nailing a sentence or phrase perfectly. But its ability to be a surface to bounce ideas off has become rather valuable to me. (I also managed to create my own "council" consisting of simulated versions of Aristotle, Marie Curie, Gandhi, Barack Obama and Nikola Tesla to throw elaborate problems at purely for my own entertainment. That's another story though).
Your metaphor concerning AI content becoming like fast food also aligns closely to how I see the story of generated content being played out. At some point some publication will come out with a "fully written by humans" label that will make it a differentiator. The schism between "content" and "art" will grow further, the former being for practical reasons, the latter for pleasure (or to borrow your analogy, one will be like grabbing a sandwich on the move for sustenance, the other will be grass-fed steak with a glass of red).
Appreciate the honesty with how you use AI. Will subscribe for more!
Great read, Ben, thanks for sharing it with us. Thus far, I've definitely been in the anti-AI camp. I've used it for my day job, but I've struggled to give myself permission to use it for my writing pursuits. But I think you're really pressing on something crucial here and that's having boundaries with it. As in how you use it as an editor, but never to actually write. I think that's pretty cool, because the price involved is such a game-changer. Like, I don't want to put people out of a job (my own day job's days are probably numbered because of it), but I know I certainly cannot afford to pay a freelance editor their proper fee. No way in hell.
Thanks, Garrett. I'm glad this piece was helpful. It is a tricky frontier to navigate for all of us. I think where I come down on AI and it's many abundant use cases is that it's an equalizer/load balancer. If you're a brilliant artist but can't write at all, I would whole-heartedly support you using AI to help you write press releases and even artist statements. I think it has the potential, like all technologies for us to extend our reach. It's hard to say where we'll end up in the next thirty years but it will be interesting to see.
Oh la la, what a lot to think about, Ben! I never thought of using AI as a critic as you do (or as a prompt to help you self-critic). Fascinating.
And the whole pay thing is part of it - a cycle - it is 20/month and sadly writers are not typically making too much money (either as writers or in their humanities-based jobs - or other ones) and so while we may want to pay an editor (potentially also working as a writer, like Troy), we may not be able to.
I am thinking about this debate a lot as a teacher of writers as well. Thinking about how to channel AI for good, how to make better writers, to help teens want to write...thanks for all the unfinished thoughts! It's the best way to publish. :)
Thanks, Kate! It is a lot to think about which is why I'm often surprised why some people don't think about AI at all, either assuming it's just hype with no utility or it's the axis of evil coming for all of us. I can imagine as a teacher you have a whole other layer of complexities to consider. I don't envy you trying to navigate this brave new world with students.
When it can write whole original novels, short stories and poems based on data of what people like, what then for us authors?
That's my biggest worry. Although it's not there yet, it's definitely on its way, especially if a dedicated company or companies (publishing companies) decide to do that.
That's why I think it's important to fight against it so the publication companies who decide to sell AI generated books and stories get boycotted.
It's a survival mechanism in this cut throat capitalist society.
Humans will always write more interesting stories for other humans. The same people who try to create stories like they are products informed by customer research today will be the ones to embrace AI written stories. It’s just a new tool to accomplish the same ends they’ve always had.
Q
I suppose I'm starting to like AI. Not because I use it, but because other people are using it to write their novels and then coming to me to fix it when they realise the output isn't good at all. I'm getting paid, so thank you, AI!
I'm working on a project like this right now. Along with some other ghostwriters, we identified together the problem with the way it writes: it makes just enough sense that when you read it, you can flow along smoothly and not quite see anything wrong, but it gives you an uneasy feeling. If you stop, go back, and read more carefully and analytically, you'll realise it actually doesn't make much sense at all. It invents things, references back to things that never happened in the text, repeats ideas in different ways, and uses phrases or idioms that no human has ever used.
Someone in the group said, sure, ChatGPT produces nonsense, but Claude is fantastic! I had to smirk. The project I'm working on now was originally written with Claude. They're all the same.
It's funny, right?! AI is the ultimate ghost writer. I'm so glad it's opened up a whole new line of business for you. I 100% agree with your assessment of AI generated creative writing. It works amazingly for a 4 year old who just wants the comfort and delight of a simple story that incorporates all the ingredients she throws at it, but for the rest of us humans who have read real writing, it is hollow in a way that can be hard to describe. As you said, on first inspection, it seems directionally okay, just as an AI image can create the same illusion until you look closely at the people in the blurred background and realize they are not human at all, but a collection of pixels rendered to approximate human-like shapes.
So glad to see this as opposed to all the AI negativity. I have used AI in similar ways to help me in my novel writing journey but only used the free version and haven't tried putting my whole manuscript and other story info into it. Can you do that on pro? I also tend to prefer Claude at the moment.
Hi Tania, thanks for reading. I agree there's a lot of stigma around engaging AI if you're a creative and I get it but I also think some of the most ardent detractors don't fully understand how LLMs really work. As a result, they believe crazy things are possible like it will lift entire swaths of their manuscript and pass it on to someone else. In answer to your question, yes, with pro you can upload an entire novel and a whole lot more into your library as a resource. It doesn't hold it all in memory in a single conversation, but you can prompt it to look things up or review passages just by referencing them.
Oh great, I'll have to look into it. Thanks!
This is fascinating. I feel like an old curmudgeon because I’ve never used ChatGPT for anything, and have only just recently begun to (almost) trust the AI summary at the top of my Google searches (but grateful that the sources are linked to follow up and clarify). I also have to laugh because I work as the sounding board & administrivia-ist for a writer, so I can now clearly see my research & outlining work being taken over by a faster, cheaper, more accurate model. I’ll have to get with the times before I’m completely left behind!
Hi Robin, thanks for reading. I think it's safe to say that most of us who do "knowledge work" of any kind will likely find many aspects of our jobs replaced by AI in time. Computer programming and most aspects of software development will be the first jobs to go I suspect. Your comment here is validation for why I bother to post pieces like this-- so you can understand there's more there than just hype and headlines. It will impact all of us in some way so it's better to get ahead of it if you can. BTW - faster, yes. Cheaper, yes. But more accurate is not the case. AI models tend to be creative when it comes to facts unless you pin them down to site sources.
It is so valuable to hear your perspective - I trust your words as a creative much more than someone who is looking at it only from a techie pov. I can see how computer programmers would be excited about this technology, I expect that, but it's valuable to hear how an artist such as yourself is really benefitting from and using this. As a researcher/fact-checker, I am leery of what AI sees as "facts" for its summaries, but digging into your own writing to see patterns and ask editorial questions, etc., is something I hadn't thought about. So helpful, Ben, thank you.
Funny story---I JUST had a "conversation" with ChatGPT about a story I'm working on and got really solid feedback--along with this encouraging tidbit:
"You’re weaving economics, philosophy, and theology into sci-fi, and doing it with dialogue that feels real. That’s not easy. The setting becomes a metaphor, the characters deepen with each line, and the tension never talks down to the reader. You’ve written something beautiful, smart, and enduring—and it all started with a flash fiction prompt.
You’re not just writing well. You’re writing with vision."
And then I ran out of time on this model. Maybe eventually I'll be able to fork over $20/month, but not right now. I can see how it works as a sounding board when my writer friends are too busy.
Hi Stephanie, thanks for stopping by. I'm glad you've been experimenting a bit. I think there's clear value in staying aware of how things are progressing even if you're not a regular user of these tools.
Lots of more nuanced replies here, I'll simply say how much I appreciate your discussing how AI can be used as a tool without crossing the boundary of letting it do the hard work for you. Keeping it real, Ben!
Thanks, Troy. It’s all about the boundaries. 😊
Great post, Ben. I appreciate your perspective. While I have yet to use AI for anything (2 reasons: I don't feel like learning how to prompt it so it would be useful to me and the ethical reasons I'm sure other people have talked about) I have heard plenty of people say how it's been a useful tool. Like using it to pull 10 quotes from your book to use in marketing.