Thank you for your post. It was delicious to read, as usual! :)
Agreed with everything. It's complicated, but I, for one, am optimistic about our ability to tell, at least for now.
I dunno but for me about 20-30% of what I see even here on Substack screams of AI. It feels like it's still easy to tell. For now, at least.
Also - I came in the comments mostly to share one thing. When I was a young writer one of my first jobs (and I held it for 3 years!) was for an agency that used one of those screen recording software kind-of-like-Big-Brother and I still shudder thinking about how intrusive it was, how we were all afraid of letting the computer go 'idle' for more than 5 minutes if we were thinking etc.
Just because one manager dude had trust issues and just had to watch the writing process live. Ugh.
Yeah I am really worried about younger writers now. A lot of us have a big body of work to point to. But if you get accused and have nothing out yet. That would be tough. My concern isn’t so much with the obvious slop but more with how folks can train the most advanced models to write things and then edit them to not sound like ai. Not sure if there is anything to be done about that except theory 2
I appreciate your research, approach, and thoughts here. I am also so. tired. As a writer, I am tired of being scared that my work might be flagged as AI because of a few mdashes. As a lit mag editor, I'm tired of worrying one of the essays I loved is AI because the writer used a "not this, but that" phrase in a sentence. It feels like just the beginning ?? and I am already so tired! And grumpy! And I don't want to have to put a "human created" badge on all my shit. And I don't want to pay for an AI detector for my lit mag which is already cost strapped. But. I do appreciate the continued conversations here and am interested in where your talks go and what manifests.
I know. It’s exhausting. And I can tell from the folks I’m talking to that they are exhausted too. It feels like fighting a losing battle. That’s why I’m the end I hope that anyone who would use ai to write, would use it to do something other than write because there isn’t a lot of profitability in writing. Maybe?
I’m tired of AI taking up so much of my time and energy when I want nothing to do with it. A tool that uses AI to detect AI in writing just seems like a perpetual loop of wasted resources. I’m not sure I want to post on Substack anymore. I don’t want to be part of something that normalizes relying on AI to tell us if something was written by AI or not. I understand why we need to talk about it and do something about it, but I’m not sure what. I hope this generative AI explosion will run its course and soon crash. Who really wants this shit besides the guys trying to tell us we need it?
This is, I think, the sentiment a lot of people are feeling right now with this stuff. Unfortunately, a lot of folks are using it. And a lot of editors are worried about it. Whether we will ultimately do anything at all to get involved is still up in the air. Leaning toward, "not my monkey, not my circus" but when so many folks ask us, we feel we should do our due diligence with researching solutions.
I commend the editors, agents, and publishing professionals (and teachers, consultants, etc., really anyone who works with words because it’s infiltrated every business) who are trying their best to navigate this AI invasion.
I run a micro-press (really tiny) and I've placed a clause in the contracts I use where the writer has to verify that they did not use generative AI. Will dishonest people get through? Probably. Will I sue them? Probably not. I don't think we've published anything AI-generated yet, but we might have. Really, I'm just trying not to worry about it much. My worry will not change anything except my blood pressure.
Solution: Close the internet and go back to dead tree writing and publishing. That'll show 'em ;-)
Or we train up an army of aye-ayes. They are monkeys that tap on trees and listen for echoes that signal hollows inside them, which are often infested by grubs (aka AI), then gnaw holes into the wood using their forward-slanting incisors to create a small hole into which they insert their narrow middle fingers (how apt is that in this context) to pull the grubs out.
Have been going down an AI-hole these past few days, and considering the risks I have to say, stealing the jobs of writers has become the least of my concerns. Did you know that according to the higher ups we are all simulated humans in a simulation, and once the singularity happens (optimistic estimates put it at 10 years from now) we will all be herded into a (granted, humane) post simulation zoo for simulated humans?
Anyhoo.. AI writing is very shitty and easily detectable once you’ve read its work. Just saying.
curious what happens to ppl who write by hand and then copy it onto the google doc once they’ve got a draft (in the watch as u go scenario). thinking we should hire gen z to exclusively post Reels about the otherworldly high of writing with pen and paper and let that fix all these issues for us
I was thinking something similar as I do like to write out by hand many of my ideas. I kind of mix writing and drawing/doodling when I’m trying to organize how the piece will go, whether that’s story arch or chapter breakdowns or what have you. Or I’ll additionally type them in my phones notes app and copy/paste that. I’ll even take pictures of my what I write by hand and grab the text from that for a copy/paste into a document (to organize and work with later).
I’ll work on all of that on a Google or Word doc to then copy/paste to Substack, and made adjustments as I see fit.
Thank you for doing this research, esp. w/ levity that makes the medicine go down bit more easily. What really freaks me out is the "11 ft wall, meet 12 ft ladder." I've found that in my work as an adjunct instructor at a community college. Every six months the AI use (cheating) ramps up, and I'm super tired of being a detective and cop. I'm out of solutions and (soon) fucks. But to your piece's focus on creativity, I can only hope the "normalizing" of real human creation being cherished, even sacred, will continue (because I believe with urgent hope that it exists now among most artists/writers).
Thank you for your post. It was delicious to read, as usual! :)
Agreed with everything. It's complicated, but I, for one, am optimistic about our ability to tell, at least for now.
I dunno but for me about 20-30% of what I see even here on Substack screams of AI. It feels like it's still easy to tell. For now, at least.
Also - I came in the comments mostly to share one thing. When I was a young writer one of my first jobs (and I held it for 3 years!) was for an agency that used one of those screen recording software kind-of-like-Big-Brother and I still shudder thinking about how intrusive it was, how we were all afraid of letting the computer go 'idle' for more than 5 minutes if we were thinking etc.
Just because one manager dude had trust issues and just had to watch the writing process live. Ugh.
Yeah I am really worried about younger writers now. A lot of us have a big body of work to point to. But if you get accused and have nothing out yet. That would be tough. My concern isn’t so much with the obvious slop but more with how folks can train the most advanced models to write things and then edit them to not sound like ai. Not sure if there is anything to be done about that except theory 2
I appreciate your research, approach, and thoughts here. I am also so. tired. As a writer, I am tired of being scared that my work might be flagged as AI because of a few mdashes. As a lit mag editor, I'm tired of worrying one of the essays I loved is AI because the writer used a "not this, but that" phrase in a sentence. It feels like just the beginning ?? and I am already so tired! And grumpy! And I don't want to have to put a "human created" badge on all my shit. And I don't want to pay for an AI detector for my lit mag which is already cost strapped. But. I do appreciate the continued conversations here and am interested in where your talks go and what manifests.
I know. It’s exhausting. And I can tell from the folks I’m talking to that they are exhausted too. It feels like fighting a losing battle. That’s why I’m the end I hope that anyone who would use ai to write, would use it to do something other than write because there isn’t a lot of profitability in writing. Maybe?
Love the humor (as usual)
Extra points! 🚨
I’m tired of AI taking up so much of my time and energy when I want nothing to do with it. A tool that uses AI to detect AI in writing just seems like a perpetual loop of wasted resources. I’m not sure I want to post on Substack anymore. I don’t want to be part of something that normalizes relying on AI to tell us if something was written by AI or not. I understand why we need to talk about it and do something about it, but I’m not sure what. I hope this generative AI explosion will run its course and soon crash. Who really wants this shit besides the guys trying to tell us we need it?
This is, I think, the sentiment a lot of people are feeling right now with this stuff. Unfortunately, a lot of folks are using it. And a lot of editors are worried about it. Whether we will ultimately do anything at all to get involved is still up in the air. Leaning toward, "not my monkey, not my circus" but when so many folks ask us, we feel we should do our due diligence with researching solutions.
I commend the editors, agents, and publishing professionals (and teachers, consultants, etc., really anyone who works with words because it’s infiltrated every business) who are trying their best to navigate this AI invasion.
I run a micro-press (really tiny) and I've placed a clause in the contracts I use where the writer has to verify that they did not use generative AI. Will dishonest people get through? Probably. Will I sue them? Probably not. I don't think we've published anything AI-generated yet, but we might have. Really, I'm just trying not to worry about it much. My worry will not change anything except my blood pressure.
I think that’s a good philosophy
A refreshing take!
Solution: Close the internet and go back to dead tree writing and publishing. That'll show 'em ;-)
Or we train up an army of aye-ayes. They are monkeys that tap on trees and listen for echoes that signal hollows inside them, which are often infested by grubs (aka AI), then gnaw holes into the wood using their forward-slanting incisors to create a small hole into which they insert their narrow middle fingers (how apt is that in this context) to pull the grubs out.
Have been going down an AI-hole these past few days, and considering the risks I have to say, stealing the jobs of writers has become the least of my concerns. Did you know that according to the higher ups we are all simulated humans in a simulation, and once the singularity happens (optimistic estimates put it at 10 years from now) we will all be herded into a (granted, humane) post simulation zoo for simulated humans?
Anyhoo.. AI writing is very shitty and easily detectable once you’ve read its work. Just saying.
Wait wait wait…whaaaat? Wait, what if we are already in that zoo?
curious what happens to ppl who write by hand and then copy it onto the google doc once they’ve got a draft (in the watch as u go scenario). thinking we should hire gen z to exclusively post Reels about the otherworldly high of writing with pen and paper and let that fix all these issues for us
Haha whatever we do, getting Gen Z to post reels about it is the only way it will happen
scary! as a gen z not on instagram or tiktok, lol
I was thinking something similar as I do like to write out by hand many of my ideas. I kind of mix writing and drawing/doodling when I’m trying to organize how the piece will go, whether that’s story arch or chapter breakdowns or what have you. Or I’ll additionally type them in my phones notes app and copy/paste that. I’ll even take pictures of my what I write by hand and grab the text from that for a copy/paste into a document (to organize and work with later).
I’ll work on all of that on a Google or Word doc to then copy/paste to Substack, and made adjustments as I see fit.
Thank you for doing this research, esp. w/ levity that makes the medicine go down bit more easily. What really freaks me out is the "11 ft wall, meet 12 ft ladder." I've found that in my work as an adjunct instructor at a community college. Every six months the AI use (cheating) ramps up, and I'm super tired of being a detective and cop. I'm out of solutions and (soon) fucks. But to your piece's focus on creativity, I can only hope the "normalizing" of real human creation being cherished, even sacred, will continue (because I believe with urgent hope that it exists now among most artists/writers).
Bring on the metal detectors — you know who doesn’t use AI to write? Serhiy Zhadan
nice plug. we should catch up soon
Thank you. This is the only thing that’s made me laugh about this crisis. Ever.
It's what I live for
This is starting to feel like a moral panic.
I don't need an AI "detector" to spot bad writing.
It announces itself.
That's kinda the whole point of the end there if you read that far
That's kinda what I did.
not read to the end?