Never been more happy that I'm 'old school' and actually write my first drafts longhand on a lined pad. After I type it in, I print it out, and revise on the pages with my pen. Tactile and stimulates parts of my brain that come up with insights I don't get at a computer.
I've used AI for ideas for marketing copy (most of it was too superlative) but I got a few phrases and I'll use it for those kinds of tasks - never for writing. Thank goodness the Authors Guild now has the Human Authored emblem (using it on all my new books).
Well, I’ve never been more happy that I’m not only “old school” but actually old. A world where AI has infiltrated everything- even creativity is the ultimate dystopia for me.
Love the Brooke Barbie… and thank you for this very informative & yes, quite terrifying piece - and more importantly, for being so very Human - thoughtful, thorough, and no bullshit.
Thanks for this report, Brooke. Fascinating—and terrifying. Thanks for your vigilance and commitment to human authors! The one thing I’ll say in AI’s defense is about research. It can comb through millions of Google results in seconds to answer hyper-specific questions about historical details. Super helpful and a huge time-saver, especially for historical fiction writers trying to avoid anachronisms.
Yes, for sure. There are positives, as I noted. I'm note sure this is "defending" AI. But is it all bad? No, it's not. The issue is we have no agreed-upon usage for it. And yes, to Richard's point—it makes so many mistakes.
Yep. I was just reading the story about the summer reading list the Chicago Sun Times put out—confabulated fake books by real authors, courtesy of AI. Proceed with extreme caution, indeed.
I'm late to this comment thread, and after scrolling through all the comments, I'm quite surprised no one brought up the tremendous environmental cost of AI in terms of the massive, escalating energy and water usage. Even if the energy use gets more efficient, the overall usage will increase, and the cost to the warming planet is mind-boggling. Every time you play with it just for fun, you're part of the problem, in my view.
I'm so anti-AI that I never use it (except for occasionally making use of Zoom's AI-generated meeting summaries, which are comically bad), and I even habitually type "-ai" at the end of the google search bar, because that will eliminate the AI-generated summary and therefore, in a tiny way, reduce the energy usage.
But mostly, I too feel my heart sink at the cost to creativity and authenticity. Recently, I listened to an aspiring novelist who shall remain nameless, on a writing-oriented podcast, profess she doesn't use AI for her writing. But then in the next breath she said she turns to AI for help with query letters, used AI to help her write a high-tech part of her plot ("I still can't tell you what blockchain is about," she said, even though she put it in her book), and she uses AI to generate blurbs about their podcast (which sound generic, lacking any personality). To me, this is a slippery slope that I never want to get on, and if she doesn't have the voice, heart, and mastery of her story to write her own query letter, then she's got a problem that AI can't fix.
Such a valuable point, Sarah. Thanks for bringing it up. Of course this post is written through the lens of creativity and expression, but I'm also very very concerned about the environmental impact and distraught by our collective short-sightedness for future generations. But this is so widespread these days, it's hard not to live in doomland....
ChatGPT helped me with a client's messy references, produces fun images for Substack... plus some other useful stuff. But improve sentences? Nope! I have yet to see a better sentence. It can find synonyms, but so can a thesaurus. When it tries to help me rewrite, I tell it to shut up (I know, waste of time) because I want my writing to sound human not AI-ish. To its credit, it always acts contrite. But it's back at it at the next opportunity. Maybe authors who use AI to improve their sentences shouldn't be authors.
So curious, Jude, if you would say this about human editors. "Maybe authors who use human editors to improve their sentences shouldn't be authors." Or "Maybe authors who use a thesaurus to improve their words shouldn't be authors." That certainly would be the point an AI nonapologist would make.
Well, that's the purpose of a thesaurus, so I certainly wouldn't object to using one. Same for human editors. More to your point is the practice of ghostwriting. I have ghostwritten a number of books, and I think it can be a questionable practice, especially depending on how it's handled. Even so, it is still the human mind at work. Because, really, the ultimate problem here is yielding control of our culture and society to an artificial, nonhuman force.
Brooke, I began my adventures with AI a couple of years ago and am ready to share some of what I've learned. I'll be launching a 12-post series on Substack on the subject soon. Format: conversations with ChatGPT and me. They'll be on my AI Working Notes tab, along with a couple of other pieces I've written recently: https://susanwittigalbert.substack.com/s/ai-working-notes You're absolutely right: become an expert now, or forever wish you had.
Really interesting, Susan. I think this is a very innovative approach to understanding and self-discovery AND you are being transparent about what is yours and what is AI. So there's that—and thank you.
When I'm not writing books, I write e-learning and classroom courses for corporations and occasional nonprofit organizations. (Helps the work feel less soul sucking!) Last year, when one of my leads asked if I wanted to learn how to use AI to help gather content, I enthusiastically said, "YES!"
As you wrote, "If you’re not already interacting with AI and starting to become an expert in how to use it, you’re going to be left behind." Already, when being considered for new project work, decision makers are asking about my AI prompt engineering skills.
This is one of those tricky situations where as much as I hate that we have no recourse when it comes to our content being fed into language learning models, I can't just boycott AI. I have to learn how to leverage it or find a new job.
That said, it's really sad to hear you've received plagiarized manuscripts. That just sucks.
Brooke, Thank you for taking time from your many-hatted "Brooke Barbie" life to write this. I have been ignoring AI, partly because I do not love the values of the companies whose LLMs are the most popular ones (Chat GPT for instance, and Sam Altman's Open AI), but recently someone challenged me to use Claude for research, and I found it very helpful. So I'm proceeding cautiously into learning what it can do for me. I see AI as useful for the things that don't require creativity and a human touch, as in the examples you gave. But not for writing--that's my voice and my perspective, which is what makes my writing unique! Or photos--I use my own to illustrate my Substack posts--those images are my way of seeing the world, and that's what viewers respond to. The negatives you listed for AI are all very sobering, especially the idea that kids need to learn to create and learn and fail in order to grow up as capable humans. AI cannot be allowed to replace that. Your post reminds me that I need to remember to add a "Human Powered" statement to my posts on Substack. Thanks again for the education and inspiration. If I was collecting Barbies, I'd be lobbying for yours to be real, so I could have one! Blessings to you and yours.
Thanks for this comment, Susan. It's interesting to take stock of the ways that we are using and not using not just AI, but things like stock images. And to consider for ourselves what our own boundaries are. I like "human powered," and "human authored." I think these are good statements to claim to affirm our creativity apart from AI. xo
Thanks for this clear assessment of the AI situation. I would never let AI write for me although I have seen some of the ways it can help me. My biggest fear is that we are raising a new generation who won't know the difference and won't believe it matters.
Great assessment, Brooke. I, too, have found AI to be an excellent time-saving tool, especially when it comes to research. But I've given myself two rules: never ask AI to write anything for me. Never ask AI to edit a sentence. Actually, I'm not even tempted to break my rules. It's like asking someone to eat your chocolate ice cream. I did find myself "cheating" at one thing, though, and I have since realized how wrong it is. Since I grew up writing using the AP style guide, I am often unsure how the Chicago Manual of Style (used in book publishing) might treat a thorny grammar question. So for a while, I was asking ChatGPT, "how would Chicago Manual of Style edit this phrase: ..." I realized how this is cheating the good folks at Chicago Manual of Style, so I simply bought a subscription. And I'm teaching myself, which is something Chat can't do for me.
We should all have our own personal AI policies. It was suggested at the panel this weekend that all publishers have AI policies for their employees and for their authors. I'm at work on that!
Kathy, I too bought the full Chicago Manual of Style (very heavy) and am studying the relevant sections so my writing will pass a copy edit more quickly next time. Get flummoxed by the long dashes (which I love using but it wants me to use a colon instead).
Why Caroline, the online edition of CMS weighs very little, snicker snicker. (Fun to run into you here, my SWP sister.) There are times I really struggle to do things the CMS way. Just fights again 30 years of ingrained AP style.
"I will stop being a publisher before I allow AI to choose what manuscripts my publishing company publishes." <THIS. Thank you. I feel kind of weird about AI - in my first experience, it gave me 80% wrong info. BUT...when I was in school, I decided I wasn't going to memorize dates and kings - I could look them up. I feel that's a bit analogous to AI, although I have gone on to spend years studying various historical things. I enjoy deep dive. Another note: Sheila Heti published a New Yorker story based on "conversations" with AI. Of course, she was up front about it.
For nonfiction, you either need to be saying an old thing in a new way or saying a new thing. For either of those tasks, AI isn't really that helpful. If, on the other hand, you want to use your large platform to make money by saying an old thing in an old way, it's easy for me to imagine AI "flooding the [publishing] zone with shit."
I laughed when I got to the Brooke Barbie - it's adorable!
The first time we got in and played around with ChatGPT, my husband and I both had the same reaction. This is so incredible AND this is also scary. We both started imagining what people with bad intentions could do with it. He is in the research field and already has countless stories of bad actors who put their names on the hard work of others. There is a woman in my hometown who immediately started courses on how to give prompts to write your books. She and her clients started printing out books and selling them. Nowhere was it mentioned that they hadn't actually written them.
I'm hormonal today so I suppose I'm being more negative about things than normal. I'll keep the good parts you mention in mind and remember there are more ethical writers than not.
Brooke, Thank you for sharing those insightful & unsettling AI discussions from the conference. It’s hard to grasp just how extensively AI is expected to impact the publishing industry. It may generate more business in publishing, but less use for editors or designers, which is scary. Also, the effect on authenticity is deeply concerning. BTW~The Brooke Barbie is very cool!!
I have used AI to evaluate article titles and even article content, often finding it helpful to a degree. Using Claude, I was generally impressed with the level of detail of the analyses both on article titles and content. From that standpoint, it helped me consider ways to improve my work, but it also missed the mark in many ways, not understanding who my audience really is, and even miscalculating a word count. With article title evaluations, I found it was overly reliant on whatever algorithms or whatever it relies on, unable to focus on the human heart. I have found that when I asked it to edit for space, its limitations were clear: the edit was clunky.
Hi, Amy. You make a good sense here. I try to belive AI is not a monster but a good tool. If you master how to make it your wish-granting genie, then you may use it to serve your purpose. I just stopped short of becoming a Luddite.
Never been more happy that I'm 'old school' and actually write my first drafts longhand on a lined pad. After I type it in, I print it out, and revise on the pages with my pen. Tactile and stimulates parts of my brain that come up with insights I don't get at a computer.
I've used AI for ideas for marketing copy (most of it was too superlative) but I got a few phrases and I'll use it for those kinds of tasks - never for writing. Thank goodness the Authors Guild now has the Human Authored emblem (using it on all my new books).
Well, I’ve never been more happy that I’m not only “old school” but actually old. A world where AI has infiltrated everything- even creativity is the ultimate dystopia for me.
Love the Brooke Barbie… and thank you for this very informative & yes, quite terrifying piece - and more importantly, for being so very Human - thoughtful, thorough, and no bullshit.
Thanks for this report, Brooke. Fascinating—and terrifying. Thanks for your vigilance and commitment to human authors! The one thing I’ll say in AI’s defense is about research. It can comb through millions of Google results in seconds to answer hyper-specific questions about historical details. Super helpful and a huge time-saver, especially for historical fiction writers trying to avoid anachronisms.
Yes, for sure. There are positives, as I noted. I'm note sure this is "defending" AI. But is it all bad? No, it's not. The issue is we have no agreed-upon usage for it. And yes, to Richard's point—it makes so many mistakes.
Be warned Alisa, AI is an excellent "aggregator", and I use it too. But it makes mistakes
For sure! A good reminder for folks to proceed with caution. I generally verify with other sources once I find the nugget I'm looking for.
Sometimes it aggregates opinions : )
Humans do that too...
Got that right : )
Yes, except AI also lies & makes stuff up, so it's always important to fact check.
Yep. I was just reading the story about the summer reading list the Chicago Sun Times put out—confabulated fake books by real authors, courtesy of AI. Proceed with extreme caution, indeed.
I'm late to this comment thread, and after scrolling through all the comments, I'm quite surprised no one brought up the tremendous environmental cost of AI in terms of the massive, escalating energy and water usage. Even if the energy use gets more efficient, the overall usage will increase, and the cost to the warming planet is mind-boggling. Every time you play with it just for fun, you're part of the problem, in my view.
I'm so anti-AI that I never use it (except for occasionally making use of Zoom's AI-generated meeting summaries, which are comically bad), and I even habitually type "-ai" at the end of the google search bar, because that will eliminate the AI-generated summary and therefore, in a tiny way, reduce the energy usage.
But mostly, I too feel my heart sink at the cost to creativity and authenticity. Recently, I listened to an aspiring novelist who shall remain nameless, on a writing-oriented podcast, profess she doesn't use AI for her writing. But then in the next breath she said she turns to AI for help with query letters, used AI to help her write a high-tech part of her plot ("I still can't tell you what blockchain is about," she said, even though she put it in her book), and she uses AI to generate blurbs about their podcast (which sound generic, lacking any personality). To me, this is a slippery slope that I never want to get on, and if she doesn't have the voice, heart, and mastery of her story to write her own query letter, then she's got a problem that AI can't fix.
Such a valuable point, Sarah. Thanks for bringing it up. Of course this post is written through the lens of creativity and expression, but I'm also very very concerned about the environmental impact and distraught by our collective short-sightedness for future generations. But this is so widespread these days, it's hard not to live in doomland....
You just changed my life with this “minus ai” tip thank you! I totally spaced that this is an option 😅
ChatGPT helped me with a client's messy references, produces fun images for Substack... plus some other useful stuff. But improve sentences? Nope! I have yet to see a better sentence. It can find synonyms, but so can a thesaurus. When it tries to help me rewrite, I tell it to shut up (I know, waste of time) because I want my writing to sound human not AI-ish. To its credit, it always acts contrite. But it's back at it at the next opportunity. Maybe authors who use AI to improve their sentences shouldn't be authors.
So curious, Jude, if you would say this about human editors. "Maybe authors who use human editors to improve their sentences shouldn't be authors." Or "Maybe authors who use a thesaurus to improve their words shouldn't be authors." That certainly would be the point an AI nonapologist would make.
Well, that's the purpose of a thesaurus, so I certainly wouldn't object to using one. Same for human editors. More to your point is the practice of ghostwriting. I have ghostwritten a number of books, and I think it can be a questionable practice, especially depending on how it's handled. Even so, it is still the human mind at work. Because, really, the ultimate problem here is yielding control of our culture and society to an artificial, nonhuman force.
Brooke, I began my adventures with AI a couple of years ago and am ready to share some of what I've learned. I'll be launching a 12-post series on Substack on the subject soon. Format: conversations with ChatGPT and me. They'll be on my AI Working Notes tab, along with a couple of other pieces I've written recently: https://susanwittigalbert.substack.com/s/ai-working-notes You're absolutely right: become an expert now, or forever wish you had.
Really interesting, Susan. I think this is a very innovative approach to understanding and self-discovery AND you are being transparent about what is yours and what is AI. So there's that—and thank you.
When I'm not writing books, I write e-learning and classroom courses for corporations and occasional nonprofit organizations. (Helps the work feel less soul sucking!) Last year, when one of my leads asked if I wanted to learn how to use AI to help gather content, I enthusiastically said, "YES!"
As you wrote, "If you’re not already interacting with AI and starting to become an expert in how to use it, you’re going to be left behind." Already, when being considered for new project work, decision makers are asking about my AI prompt engineering skills.
This is one of those tricky situations where as much as I hate that we have no recourse when it comes to our content being fed into language learning models, I can't just boycott AI. I have to learn how to leverage it or find a new job.
That said, it's really sad to hear you've received plagiarized manuscripts. That just sucks.
Brooke, Thank you for taking time from your many-hatted "Brooke Barbie" life to write this. I have been ignoring AI, partly because I do not love the values of the companies whose LLMs are the most popular ones (Chat GPT for instance, and Sam Altman's Open AI), but recently someone challenged me to use Claude for research, and I found it very helpful. So I'm proceeding cautiously into learning what it can do for me. I see AI as useful for the things that don't require creativity and a human touch, as in the examples you gave. But not for writing--that's my voice and my perspective, which is what makes my writing unique! Or photos--I use my own to illustrate my Substack posts--those images are my way of seeing the world, and that's what viewers respond to. The negatives you listed for AI are all very sobering, especially the idea that kids need to learn to create and learn and fail in order to grow up as capable humans. AI cannot be allowed to replace that. Your post reminds me that I need to remember to add a "Human Powered" statement to my posts on Substack. Thanks again for the education and inspiration. If I was collecting Barbies, I'd be lobbying for yours to be real, so I could have one! Blessings to you and yours.
Thanks for this comment, Susan. It's interesting to take stock of the ways that we are using and not using not just AI, but things like stock images. And to consider for ourselves what our own boundaries are. I like "human powered," and "human authored." I think these are good statements to claim to affirm our creativity apart from AI. xo
Thanks for this clear assessment of the AI situation. I would never let AI write for me although I have seen some of the ways it can help me. My biggest fear is that we are raising a new generation who won't know the difference and won't believe it matters.
Me too, Sue. Very afraid for these kids and the impacts. The Ezra Klein episode is a deep dive on that.
Great episode.
Great assessment, Brooke. I, too, have found AI to be an excellent time-saving tool, especially when it comes to research. But I've given myself two rules: never ask AI to write anything for me. Never ask AI to edit a sentence. Actually, I'm not even tempted to break my rules. It's like asking someone to eat your chocolate ice cream. I did find myself "cheating" at one thing, though, and I have since realized how wrong it is. Since I grew up writing using the AP style guide, I am often unsure how the Chicago Manual of Style (used in book publishing) might treat a thorny grammar question. So for a while, I was asking ChatGPT, "how would Chicago Manual of Style edit this phrase: ..." I realized how this is cheating the good folks at Chicago Manual of Style, so I simply bought a subscription. And I'm teaching myself, which is something Chat can't do for me.
We should all have our own personal AI policies. It was suggested at the panel this weekend that all publishers have AI policies for their employees and for their authors. I'm at work on that!
Will be interested to know what you come up with, Brooke.
Here's mine, Brooke: https://susanwittigalbert.substack.com/about Love to see yours, when you're ready to share.
Kathy, I too bought the full Chicago Manual of Style (very heavy) and am studying the relevant sections so my writing will pass a copy edit more quickly next time. Get flummoxed by the long dashes (which I love using but it wants me to use a colon instead).
Why Caroline, the online edition of CMS weighs very little, snicker snicker. (Fun to run into you here, my SWP sister.) There are times I really struggle to do things the CMS way. Just fights again 30 years of ingrained AP style.
"I will stop being a publisher before I allow AI to choose what manuscripts my publishing company publishes." <THIS. Thank you. I feel kind of weird about AI - in my first experience, it gave me 80% wrong info. BUT...when I was in school, I decided I wasn't going to memorize dates and kings - I could look them up. I feel that's a bit analogous to AI, although I have gone on to spend years studying various historical things. I enjoy deep dive. Another note: Sheila Heti published a New Yorker story based on "conversations" with AI. Of course, she was up front about it.
For nonfiction, you either need to be saying an old thing in a new way or saying a new thing. For either of those tasks, AI isn't really that helpful. If, on the other hand, you want to use your large platform to make money by saying an old thing in an old way, it's easy for me to imagine AI "flooding the [publishing] zone with shit."
I think easy to imagine and it already is...
I laughed when I got to the Brooke Barbie - it's adorable!
The first time we got in and played around with ChatGPT, my husband and I both had the same reaction. This is so incredible AND this is also scary. We both started imagining what people with bad intentions could do with it. He is in the research field and already has countless stories of bad actors who put their names on the hard work of others. There is a woman in my hometown who immediately started courses on how to give prompts to write your books. She and her clients started printing out books and selling them. Nowhere was it mentioned that they hadn't actually written them.
I'm hormonal today so I suppose I'm being more negative about things than normal. I'll keep the good parts you mention in mind and remember there are more ethical writers than not.
and look what came up just now.....
https://substack.com/home/post/p-164023890
Yes, and read Kathleen Schmidt:
https://kathleenschmidt.substack.com/p/the-curious-case-of-the-ai-generated
Brooke, Thank you for sharing those insightful & unsettling AI discussions from the conference. It’s hard to grasp just how extensively AI is expected to impact the publishing industry. It may generate more business in publishing, but less use for editors or designers, which is scary. Also, the effect on authenticity is deeply concerning. BTW~The Brooke Barbie is very cool!!
I have used AI to evaluate article titles and even article content, often finding it helpful to a degree. Using Claude, I was generally impressed with the level of detail of the analyses both on article titles and content. From that standpoint, it helped me consider ways to improve my work, but it also missed the mark in many ways, not understanding who my audience really is, and even miscalculating a word count. With article title evaluations, I found it was overly reliant on whatever algorithms or whatever it relies on, unable to focus on the human heart. I have found that when I asked it to edit for space, its limitations were clear: the edit was clunky.
Hi, Amy. You make a good sense here. I try to belive AI is not a monster but a good tool. If you master how to make it your wish-granting genie, then you may use it to serve your purpose. I just stopped short of becoming a Luddite.