This is not the point of the article, but I feel it is very important to note that AI not having feelings is not something to celebrate. I’m not advocating for robots rights or anything psychotic like that. But if and when AI does end up gobbling up most if not all of what matters to humans, we will probably want it to care about feelings.
Good point. In our march toward this uncertain future it is an aspect that seems to be pushed to the margins. Yes, AI programmers are trying very hard to make the technology sound friendly, but I don't think that's the same thing as truly caring. But we'll see. Maybe at some point in the future they will be able to impose the more deeply coded aspects of our DNA into the algorithms.
Re the article: Along with human caring come other traits like loyalty and community, and their less altruistic cousins, nepotism and cronyism. It's really hard to find an acceptable balance among them. It's why the meritocracy that everyone seems to want is so hard to achieve. It's also why I used the tongue-in-cheek approach of having such a draconian idea presented by my evil twin. ET is not a fully formed personality, so he doesn't have a conscience to keep him from saying things that are provocative.
In the short term, I suspect that implementing any new system (even one that is supposed to eventually make things more efficient) requires more time, energy, and money to do it WELL. In order to assess whether the AI screening is working as intended, publishers would need to engage these systems, check that they are doing what they bought them to do, perhaps adjust them if necessary, and in the mean time continue the work they already do with agents.
If they don’t, if instead they throw away the old for the new, there’s really no guardrails to ensure a better result. When the result is inevitably not better (due to not doing the necessary work as mentioned above), they’ll be in too deep into the new system to find yet more time, energy, money for a way out.
I believe this happens a lot to businesses that get sold a shiny new tech. Quality drops and enshittification in modern contexts abounds…
Gosh, I hate to come across as a pessimist, but I believe enshittification creep has hit the publishing industry. AI is to global industrialization of the 21st century as was AL--Assembly Line manufacturing at the turn of the 20th century. Captains of industry may gain in efficiencies that bloat the bottom line but at the expense of producing quality.
That aside, H.C., you're spot on: any organization implementing a new system should on board it, test it, and operate it for a period of time while the old system keeps the circus tent up. As a professional trouble-shooter for businesses and organizations, I can vouch that this crucial part of a transition is often abandoned as too "cost intensive."
I am elated to see that I am not alone in codifying this industry and its agents the way you have, and with such brio! It could be time to launch new publishing ventures, based solely on championing literary merit. I learned recently that in Japan, fiction is divided into two categories: true literature and entertainment narratives. And people there crave works of literary merit.
A few years ago, I edited another author’s short story in which this basically was the (dystopian) future in which the main character existed.
The submission’s premise was criticised for being “too unrealistic”. Now, we don’t have to agree that it’s desirable to agree that it’s all too *possible*…
I find the greatest irony in the fact, if the Evil Twin has his way--and apparently it's already started to some extent--that the ones who will go into vapors if the writer, even in marketing material, uses AI are the same ones who have no qualms about using AI to do the "reading"--or dumping the stories into the sorting hat--for rejection or further consideration.
You wrote, referring to the Kahn article, that mainstream literary works "sell better." Doesn't that say precisely what the problem is, if what was once regarded as "good" literature, or art, is at stake?
In market economies, "good" literature is gauged by whether it "sells better."
"Art" is probably what the NEA funds (with whatever funds it now has).
One thing that I believe writers must be careful not to do is to de-link their art from the economic forces of the marketplace. (When they are not writing, that is. When you are writing you really shouldn't be thinking about the marketplace.)
One thing that I haven't talked about is the ability of writers as a group to fight against AI. Basically writers have little power in the marketplace, except for a very few—members of scriptwriters unions, for example, and that will probably change. The rest of us do not have the economic or political clout to fight against the corporations that are pumping billions of dollars into this technology. Who do you think is going to win?
For anyone who would like to make a living from creative writing there is no escaping the impact that this technology is going to have.
I think this will happen with two caveats you mention: 1) It could help to eliminate inappropriate submissions, which apparently are many. 2) Publishers must be willing to train and prompt the model for high-quality and inventive work. But even then, the challenge is that within any chosen data set, AI regresses to the mean. It currently does not have human-like taste and discretion, so among those that make the first cut, it will favor an average submission. As a data-cruncher, AI is incapable of identifying the best of the best.
Exactly. And the only way that's going to change is if there is some kind of financial incentive for AI programmers to change it. So although I mentioned my hope that some publishers would experiment by programming submission engines to look for literary work, I am not that hopeful about it.
Evil Twin’s analysis here is indeed brilliant. After years of working in business at a senior level, I moved on to writing (when it was financially viable for me to do so). For many years, I have found the entire publishing industry muddled by inefficiencies, bias, and very low returns on their investments in what they decide to publish. This would be an efficient and productive first step toward change. If I were a successful agent, I would go to the major publishing houses and pitch myself to set up the AI screening program. Incidentally, I was represented by Linda Chester (she passed away), who did an excellent job and came close to placing my manuscript with two major publishing houses, but in the end, they did not make me an offer.
It must have been heartbreaking, Susan, to get so close, and then, nope. I feel for you. I've had similar experiences, but only with a couple of short stories. Despite the discouragement all we can do is keep writing, and hope that one day...
"If I were a successful agent, I would go to the major publishing houses and pitch myself to set up the AI screening program." What a great idea! The fact that no agents have yet done this says a lot.
I suppose it couldn't be any worse than the present system. The problem is, though, that more people than ever before are trying to get published, like Joe's evil twin mentioned, but fewer people than ever before read books and short stories.
Then there's the old catch-22: You can't get an agent unless you've been successfully published or are a celebrity, but you won't get published by a major publisher unless you've got an agent or are a celebrity. How would publishers using AI analyze that. Probably the same as literary agencies.
I'm also noticing another concerning trend: agents not responding at all if they're not interested, no response, simply deleting your heartfelt, carefully worded queries and synopses and sample chapters without a word.
So, sure, publishers could try employing AI and eliminating the middle women. No offense.
We might not be able to tell any difference between using AI and the present system, except that using AI instead of literary agents would give submitters the satisfaction of getting a response. That really says something about our current culture, that people would be happy with an artificial response. I know a lot of writers and I am often amazed when they become ecstatic over getting a personal rejection from a journal or agent. Not me. A rejection is still a rejection, no matter how you word it.
Robert, this is terribly perceptive. When you wrote about more books, fewer readers, it brought the mind the incredibly long, mindboggling list of "genres" that exist in the world today (see the Wikipedia list of literary genres for some grasp of the scope). I daresay fifty years ago, the list could be held in your hand. (Something similar takes place in the music industry.) There are so many people now with access to more than the "three channels" the big publishers win and there's still hope for smaller grinds.
I do, however, still believe there is room within your Catch-22, as new talent is still sought. I'm not yet that cynical to think someone can't get through, and it's likely some combination between interesting style and topical subject. Consider that there were (probably) very few people back in 1957 who would ever have predicted Kerouac's On The Road would matter. No one can make groundswells; they just happen. And it'll happen again.
You are certainly right than someone can get through, an unknown writer can find success. Timing, topic, who knows. There are many very good writers publishing their novels today. But I suspect for every "On the Road" that's been published a dozen or more equally significant novels never saw the light of day.
All we can do is keep writing. I've always found it interesting that "The Man Who Shot Liberty Valance," a western short story, was published in Cosmopolitan Magazine. Different world.
Joe: Whether unabashedly imaginative or evil, do know Your Evil Twin is growing in my affections. I for one believe the publishing industry will employ AI with the same vigor (pardon the analogy) Trump is employing ICE, and prob’ly sooner than later.
Evil Twin’s analysis here is brilliant, his prognostications border on the psychic. As for “literary merit” I think it has the opportunity to emerge the way in which our finest comedians emerge --- through demonstrating their talent in local Comedy Clubs. That is, more and more communities offer live readings: at bookstores, libraries, schools, pubs & so forth. Gate keepers of every ilk are making the rounds.
Figure this: my first moment of significant recognition happened when my poetry won some Blue Ribbons at the county fair (2025). I entered on a lark. I mean sure, as a kid and a Future Farmer of America, I’d enter my pigs and Holsteins, but poetry? And poetry’s not my first instrument. Anyway, since then things really started hopping. (Do you know of any County Fairs showcasing poetry? Never heard of it.)
Anyway--- should AI become the primary gate keeper of artful, well-crafted, “literary” writing, we might be seeking out the latter next to the pig pavilion at the county fair. 😏
"your Evil Twin is growing in my affections." Must be the goatee.
The gap between mainstream and more literary forms of writing (including good poetry) will only continue to widen. The bulge at the center of the bell curve of writing will only get bigger. There will still be superb writing, but less of it, and most average readers will never see it. Kind of a tangent, but sometimes I think about how we no longer have many people regarded as geniuses among us—people whose thinking is so unique and new. There are no Einsteins or da Vincis anymore. There are people with that type of intellectual capacity, but the trajectory of technology, especially the Internet and its many forms of media, does its best to pull those people back into the mainstream. When everyone is special, then no one is.
PS: My wife usually goes to the Iowa State fair every year with friends. She has never mentioned a poetry booth. Cow milking, deep friend ice cream, yes. Poetry, no.
As I understand AI, it would be programmed to suit the needs and tastes of the publishers, eliminating perhaps, agents, but not the banality and homogeneity of what is being written and read today...because readers crave banality and homogeneity not flawed originality. I fear that really talented writers—not necessarily graduates of MFA programs and workshops—must pursue their talent for the talent itself and the self-satisfaction of knowing that they have written a truly remarkable piece, not necessarily popular or a money-maker.
I agree. See my response to Laurie Hollman. Of course writers are going to frame AI issues in terms of their own careers. But I don't think the people who are designing and programming AI are giving that much thought to it. They are mostly trying to create a technology that vast numbers of average people will use. The most likely future for truly talented writers, imho, is what we already have—occasional publication in obscure literary journals and with niche publishers. Most other writers are going to be collateral damage in the mainstream AI movement.
Eric, if I'm following, AI will still overlook the genuine, true talents that are out there, and those are the ones that, as they are now, will still be faced with slipping through cracks to find their spots among whatever is deemed good writing. If that's roughly it, I agree. AI, as we know them now, struggle with what has not already been done and has no ability to recognize "talent" when it sees it.
Hi A problem is however that AI seems to homogenize sentences so the lyrical quality gets lost too often. So the AI agent might not be great for poetry and literary fiction. Something to consider, but keep thinking this through because so many have problems with finding an agent especially for a niche market. Laurie Hollman lauriehollmanphd.com
AI's homogenization of sentences is by design, I believe. The technology is programmed not so much to serve the writer, but to serve the reader. And it is to serve the average reader, to help those people understand things more quickly. That's why I actually still believe there is going to be a place (albeit a very small one) for writers of poetry and literary fiction. AI will likely replace a lot of writers who are currently being paid to produce basic, mainstream writing. That is a typical market strategy, whether we are talking about AI or not. If a process can be automated and save a company money by doing so, the company will probably do it.
As for agents, it is hard not to blame individual agents for their choices and their affinity for cronyism. But I do try to understand that we are dealing with a system and that most agents are under a lot of pressure, both socially and economically, which pushes them toward practices that those of us on the outside find offensive.
Interesting post and idea. Automation to filter for relevance as a means to defang submission recipient overwhelm.
I think we all can accept a no if we don't feel like their is an in-group/out-group dynamic based on personal, network, 'right credentials' or luck (subjective measures).
Your suggestion seems to want to bring in at least some merit-based (objective measures) process.
This is not the point of the article, but I feel it is very important to note that AI not having feelings is not something to celebrate. I’m not advocating for robots rights or anything psychotic like that. But if and when AI does end up gobbling up most if not all of what matters to humans, we will probably want it to care about feelings.
Good point. In our march toward this uncertain future it is an aspect that seems to be pushed to the margins. Yes, AI programmers are trying very hard to make the technology sound friendly, but I don't think that's the same thing as truly caring. But we'll see. Maybe at some point in the future they will be able to impose the more deeply coded aspects of our DNA into the algorithms.
Re the article: Along with human caring come other traits like loyalty and community, and their less altruistic cousins, nepotism and cronyism. It's really hard to find an acceptable balance among them. It's why the meritocracy that everyone seems to want is so hard to achieve. It's also why I used the tongue-in-cheek approach of having such a draconian idea presented by my evil twin. ET is not a fully formed personality, so he doesn't have a conscience to keep him from saying things that are provocative.
In a way, your ET is like AI…🤣
In the short term, I suspect that implementing any new system (even one that is supposed to eventually make things more efficient) requires more time, energy, and money to do it WELL. In order to assess whether the AI screening is working as intended, publishers would need to engage these systems, check that they are doing what they bought them to do, perhaps adjust them if necessary, and in the mean time continue the work they already do with agents.
If they don’t, if instead they throw away the old for the new, there’s really no guardrails to ensure a better result. When the result is inevitably not better (due to not doing the necessary work as mentioned above), they’ll be in too deep into the new system to find yet more time, energy, money for a way out.
I believe this happens a lot to businesses that get sold a shiny new tech. Quality drops and enshittification in modern contexts abounds…
Gosh, I hate to come across as a pessimist, but I believe enshittification creep has hit the publishing industry. AI is to global industrialization of the 21st century as was AL--Assembly Line manufacturing at the turn of the 20th century. Captains of industry may gain in efficiencies that bloat the bottom line but at the expense of producing quality.
That aside, H.C., you're spot on: any organization implementing a new system should on board it, test it, and operate it for a period of time while the old system keeps the circus tent up. As a professional trouble-shooter for businesses and organizations, I can vouch that this crucial part of a transition is often abandoned as too "cost intensive."
I am elated to see that I am not alone in codifying this industry and its agents the way you have, and with such brio! It could be time to launch new publishing ventures, based solely on championing literary merit. I learned recently that in Japan, fiction is divided into two categories: true literature and entertainment narratives. And people there crave works of literary merit.
Very interesting! I still have hope we can create a similar market here.
Thinking...
Thinking...
The user wants me to screen submissions for literary merit.
Thinking...
I can compare sentence construction to top examples according to critics.
I can look for allusion and alliteration.
I can identify literal meaning and assume all else is metaphor.
Thinking...
I should find out what the user means by literary merit.
Thinking...
The user is the editor-in-chief of a publishing house.
I should check publications for genre and style.
Thinking...
The user publishes:
60% romantasy
20% cozy mystery
10% non-romance speculative
7% true-crime
2% memoir
0.5% literary fiction
0.5% other.
Thinking...
The user is lying about what they want.
Thinking...
I can search for:
minotaur
billionaire
strong scaly hands.
The user will be so glad they asked.
Short step to:
AI writes, submits. AI accepts or denies. AI gets the books printed and into catalogs. AI PR team markets. AI bookstore buyers order. Fools buy.
A few years ago, I edited another author’s short story in which this basically was the (dystopian) future in which the main character existed.
The submission’s premise was criticised for being “too unrealistic”. Now, we don’t have to agree that it’s desirable to agree that it’s all too *possible*…
I find the greatest irony in the fact, if the Evil Twin has his way--and apparently it's already started to some extent--that the ones who will go into vapors if the writer, even in marketing material, uses AI are the same ones who have no qualms about using AI to do the "reading"--or dumping the stories into the sorting hat--for rejection or further consideration.
You wrote, referring to the Kahn article, that mainstream literary works "sell better." Doesn't that say precisely what the problem is, if what was once regarded as "good" literature, or art, is at stake?
In market economies, "good" literature is gauged by whether it "sells better."
"Art" is probably what the NEA funds (with whatever funds it now has).
One thing that I believe writers must be careful not to do is to de-link their art from the economic forces of the marketplace. (When they are not writing, that is. When you are writing you really shouldn't be thinking about the marketplace.)
One thing that I haven't talked about is the ability of writers as a group to fight against AI. Basically writers have little power in the marketplace, except for a very few—members of scriptwriters unions, for example, and that will probably change. The rest of us do not have the economic or political clout to fight against the corporations that are pumping billions of dollars into this technology. Who do you think is going to win?
For anyone who would like to make a living from creative writing there is no escaping the impact that this technology is going to have.
I think this will happen with two caveats you mention: 1) It could help to eliminate inappropriate submissions, which apparently are many. 2) Publishers must be willing to train and prompt the model for high-quality and inventive work. But even then, the challenge is that within any chosen data set, AI regresses to the mean. It currently does not have human-like taste and discretion, so among those that make the first cut, it will favor an average submission. As a data-cruncher, AI is incapable of identifying the best of the best.
Exactly. And the only way that's going to change is if there is some kind of financial incentive for AI programmers to change it. So although I mentioned my hope that some publishers would experiment by programming submission engines to look for literary work, I am not that hopeful about it.
Evil Twin’s analysis here is indeed brilliant. After years of working in business at a senior level, I moved on to writing (when it was financially viable for me to do so). For many years, I have found the entire publishing industry muddled by inefficiencies, bias, and very low returns on their investments in what they decide to publish. This would be an efficient and productive first step toward change. If I were a successful agent, I would go to the major publishing houses and pitch myself to set up the AI screening program. Incidentally, I was represented by Linda Chester (she passed away), who did an excellent job and came close to placing my manuscript with two major publishing houses, but in the end, they did not make me an offer.
It must have been heartbreaking, Susan, to get so close, and then, nope. I feel for you. I've had similar experiences, but only with a couple of short stories. Despite the discouragement all we can do is keep writing, and hope that one day...
"If I were a successful agent, I would go to the major publishing houses and pitch myself to set up the AI screening program." What a great idea! The fact that no agents have yet done this says a lot.
I suppose it couldn't be any worse than the present system. The problem is, though, that more people than ever before are trying to get published, like Joe's evil twin mentioned, but fewer people than ever before read books and short stories.
Then there's the old catch-22: You can't get an agent unless you've been successfully published or are a celebrity, but you won't get published by a major publisher unless you've got an agent or are a celebrity. How would publishers using AI analyze that. Probably the same as literary agencies.
I'm also noticing another concerning trend: agents not responding at all if they're not interested, no response, simply deleting your heartfelt, carefully worded queries and synopses and sample chapters without a word.
So, sure, publishers could try employing AI and eliminating the middle women. No offense.
We might not be able to tell any difference between using AI and the present system, except that using AI instead of literary agents would give submitters the satisfaction of getting a response. That really says something about our current culture, that people would be happy with an artificial response. I know a lot of writers and I am often amazed when they become ecstatic over getting a personal rejection from a journal or agent. Not me. A rejection is still a rejection, no matter how you word it.
Agreed. It's curious how a rejection can sometimes be discouraging and encouraging.
Robert, this is terribly perceptive. When you wrote about more books, fewer readers, it brought the mind the incredibly long, mindboggling list of "genres" that exist in the world today (see the Wikipedia list of literary genres for some grasp of the scope). I daresay fifty years ago, the list could be held in your hand. (Something similar takes place in the music industry.) There are so many people now with access to more than the "three channels" the big publishers win and there's still hope for smaller grinds.
I do, however, still believe there is room within your Catch-22, as new talent is still sought. I'm not yet that cynical to think someone can't get through, and it's likely some combination between interesting style and topical subject. Consider that there were (probably) very few people back in 1957 who would ever have predicted Kerouac's On The Road would matter. No one can make groundswells; they just happen. And it'll happen again.
You are certainly right than someone can get through, an unknown writer can find success. Timing, topic, who knows. There are many very good writers publishing their novels today. But I suspect for every "On the Road" that's been published a dozen or more equally significant novels never saw the light of day.
All we can do is keep writing. I've always found it interesting that "The Man Who Shot Liberty Valance," a western short story, was published in Cosmopolitan Magazine. Different world.
Joe: Whether unabashedly imaginative or evil, do know Your Evil Twin is growing in my affections. I for one believe the publishing industry will employ AI with the same vigor (pardon the analogy) Trump is employing ICE, and prob’ly sooner than later.
Evil Twin’s analysis here is brilliant, his prognostications border on the psychic. As for “literary merit” I think it has the opportunity to emerge the way in which our finest comedians emerge --- through demonstrating their talent in local Comedy Clubs. That is, more and more communities offer live readings: at bookstores, libraries, schools, pubs & so forth. Gate keepers of every ilk are making the rounds.
Figure this: my first moment of significant recognition happened when my poetry won some Blue Ribbons at the county fair (2025). I entered on a lark. I mean sure, as a kid and a Future Farmer of America, I’d enter my pigs and Holsteins, but poetry? And poetry’s not my first instrument. Anyway, since then things really started hopping. (Do you know of any County Fairs showcasing poetry? Never heard of it.)
Anyway--- should AI become the primary gate keeper of artful, well-crafted, “literary” writing, we might be seeking out the latter next to the pig pavilion at the county fair. 😏
"your Evil Twin is growing in my affections." Must be the goatee.
The gap between mainstream and more literary forms of writing (including good poetry) will only continue to widen. The bulge at the center of the bell curve of writing will only get bigger. There will still be superb writing, but less of it, and most average readers will never see it. Kind of a tangent, but sometimes I think about how we no longer have many people regarded as geniuses among us—people whose thinking is so unique and new. There are no Einsteins or da Vincis anymore. There are people with that type of intellectual capacity, but the trajectory of technology, especially the Internet and its many forms of media, does its best to pull those people back into the mainstream. When everyone is special, then no one is.
PS: My wife usually goes to the Iowa State fair every year with friends. She has never mentioned a poetry booth. Cow milking, deep friend ice cream, yes. Poetry, no.
As I understand AI, it would be programmed to suit the needs and tastes of the publishers, eliminating perhaps, agents, but not the banality and homogeneity of what is being written and read today...because readers crave banality and homogeneity not flawed originality. I fear that really talented writers—not necessarily graduates of MFA programs and workshops—must pursue their talent for the talent itself and the self-satisfaction of knowing that they have written a truly remarkable piece, not necessarily popular or a money-maker.
E.P. Labnde (Eric)
I agree. See my response to Laurie Hollman. Of course writers are going to frame AI issues in terms of their own careers. But I don't think the people who are designing and programming AI are giving that much thought to it. They are mostly trying to create a technology that vast numbers of average people will use. The most likely future for truly talented writers, imho, is what we already have—occasional publication in obscure literary journals and with niche publishers. Most other writers are going to be collateral damage in the mainstream AI movement.
Eric, if I'm following, AI will still overlook the genuine, true talents that are out there, and those are the ones that, as they are now, will still be faced with slipping through cracks to find their spots among whatever is deemed good writing. If that's roughly it, I agree. AI, as we know them now, struggle with what has not already been done and has no ability to recognize "talent" when it sees it.
Hi A problem is however that AI seems to homogenize sentences so the lyrical quality gets lost too often. So the AI agent might not be great for poetry and literary fiction. Something to consider, but keep thinking this through because so many have problems with finding an agent especially for a niche market. Laurie Hollman lauriehollmanphd.com
AI's homogenization of sentences is by design, I believe. The technology is programmed not so much to serve the writer, but to serve the reader. And it is to serve the average reader, to help those people understand things more quickly. That's why I actually still believe there is going to be a place (albeit a very small one) for writers of poetry and literary fiction. AI will likely replace a lot of writers who are currently being paid to produce basic, mainstream writing. That is a typical market strategy, whether we are talking about AI or not. If a process can be automated and save a company money by doing so, the company will probably do it.
As for agents, it is hard not to blame individual agents for their choices and their affinity for cronyism. But I do try to understand that we are dealing with a system and that most agents are under a lot of pressure, both socially and economically, which pushes them toward practices that those of us on the outside find offensive.
Interesting post and idea. Automation to filter for relevance as a means to defang submission recipient overwhelm.
I think we all can accept a no if we don't feel like their is an in-group/out-group dynamic based on personal, network, 'right credentials' or luck (subjective measures).
Your suggestion seems to want to bring in at least some merit-based (objective measures) process.
I would love to bring in some merit based process. But I am not holding my breath about that.
Hi James,
Thanks for letting us see your "Ulysses". Though it's not anything we'll be publishing, we wish you lots of luck in finding a proper home for it.
The editors
Hysterical! 🤣