Resonate deeply with the 'guardrails' approach. In my professional and academic work, I treat AI as a 'Scribe' rather than a 'Creator.' It requires a rigorous internal protocol to ensure the soul and depth of the work isn't lost to some sort of algorithm— although I let the AI optimize tags for SEO. Point is, transparency isn't just about ethics; it's about maintaining the integrity of the connection with the reader.
Hi Leah. I don't know if you wrote this particular post using AI or not. I can't tell the difference, but more importantly, I feel I have read an authentic piece written by Leah and I can hear her voice coming through loud and clear. And the piece has educated me pretty well on the the moral debated around AI assisted writing. Speaking of, it is a reality none of us, including the purists, can ignore. And many times it makes wonder, is there any future in being a (human) writer?
This article was a combination of AI and me. As is everything I share here. But I manually edit every piece and make sure it’s all me - my thoughts, my feelings and most importantly ultimately MY WORDS.
Calling it “writing” in the first place was always a mistake to me because you’re very obviously not doing the writing lol. I’m like of course I didn’t write it, that’s not the point! AI can write garbage but so can I, but the two of us together there’s a little bit more of a checks and balance system that makes it less worthless
Our whole premise is symbiotic thought. Basically along these lines. Human thoughts, ideas, judgement, into AI drafting with a constraint grammar, back into human ideas and discernment over what's true and what's valuable, repeat until it's ready for human polish. So far so good.
Reading this felt like someone articulating the exact process I’ve been building and refining for months now. I use AI every day too. With high standards I won’t budge on and with an actual editorial process. I also spend an embarrassing amount of time cataloguing the exact phrases and patterns that make my nervous system and eye twitch so I can train against them, but that isn’t an exact science and it takes ridiculous amounts of time to be the human against the machine. I genuinely thought I might be slightly unhinged for that process, and the effort and time just one piece takes to write, but apparently not 😂 What I appreciated most here is that you didn’t pretend the slop problem isn’t real. It absolutely is. But banning tools doesn’t fix lack of thinking. It just relocates the anxiety. I came at this with my own essay from more of a capacity lens becuase for me, AI isn’t a vending machine, it’s scaffolding. On days when I’ve done the thinking first, it helps shape it. On days when I haven’t? It hands me back the same empty output everyone’s complaining about. The tool reflects the depth you bring to it. EVERYTIME! A lot of this debate feels less like a craft argument and more like bracing. Identity threat. Economic threat. And bracing has a tone that often comes out sounding like moral certainty to me. Anyway. Just wanted to say this felt steady and sane in a conversation that’s gotten loud very quickly and I appreciate the hell out of that! ☺️
I agree that AI can be a useful tool when used properly. Unfortunately the writing industry doesn’t agree. Magazines use checkers that reject submissions with even the hint of AI, agents and publishers do the same thing. I’ve actually fed AI checkers work I’ve had published long before LLM’s existed and been told it’s over 50% AI written.
That’s the problem. AI checkers can’t deal with well written prose because they are looking for averages. So how do you manage that?
Well it’s difficult right now! This issue you raise is a major one. The checkers have to get better AND the reality is that AI-assisted writing will become more mainstream and the publishers and magazines will have to change thier policies eventually.
I agree with a lot of this. The all-or-nothing AI debate has gotten exhausting, and writers deserve more room for nuance than they’re getting. I still think the sentence matters more than this piece allows, because voice and thought live there too.
That’s partly why I built my custom GPT for writers to focus on sharpening voice, editing, and original writing rather than just generating drafts.
A kindred spirit! The purist arguement is moot now. AI is already here. Like when the camera came out, the pictures were not “real” art. Well look at photography now.
Resonate deeply with the 'guardrails' approach. In my professional and academic work, I treat AI as a 'Scribe' rather than a 'Creator.' It requires a rigorous internal protocol to ensure the soul and depth of the work isn't lost to some sort of algorithm— although I let the AI optimize tags for SEO. Point is, transparency isn't just about ethics; it's about maintaining the integrity of the connection with the reader.
Yes absolutely, I agree!
Hi Leah. I don't know if you wrote this particular post using AI or not. I can't tell the difference, but more importantly, I feel I have read an authentic piece written by Leah and I can hear her voice coming through loud and clear. And the piece has educated me pretty well on the the moral debated around AI assisted writing. Speaking of, it is a reality none of us, including the purists, can ignore. And many times it makes wonder, is there any future in being a (human) writer?
This article was a combination of AI and me. As is everything I share here. But I manually edit every piece and make sure it’s all me - my thoughts, my feelings and most importantly ultimately MY WORDS.
Calling it “writing” in the first place was always a mistake to me because you’re very obviously not doing the writing lol. I’m like of course I didn’t write it, that’s not the point! AI can write garbage but so can I, but the two of us together there’s a little bit more of a checks and balance system that makes it less worthless
Interesting point! I do end up doing a lot of writing - which is why I prefer AI-assisted writing!
Our whole premise is symbiotic thought. Basically along these lines. Human thoughts, ideas, judgement, into AI drafting with a constraint grammar, back into human ideas and discernment over what's true and what's valuable, repeat until it's ready for human polish. So far so good.
Oh my god. You are my people.
Reading this felt like someone articulating the exact process I’ve been building and refining for months now. I use AI every day too. With high standards I won’t budge on and with an actual editorial process. I also spend an embarrassing amount of time cataloguing the exact phrases and patterns that make my nervous system and eye twitch so I can train against them, but that isn’t an exact science and it takes ridiculous amounts of time to be the human against the machine. I genuinely thought I might be slightly unhinged for that process, and the effort and time just one piece takes to write, but apparently not 😂 What I appreciated most here is that you didn’t pretend the slop problem isn’t real. It absolutely is. But banning tools doesn’t fix lack of thinking. It just relocates the anxiety. I came at this with my own essay from more of a capacity lens becuase for me, AI isn’t a vending machine, it’s scaffolding. On days when I’ve done the thinking first, it helps shape it. On days when I haven’t? It hands me back the same empty output everyone’s complaining about. The tool reflects the depth you bring to it. EVERYTIME! A lot of this debate feels less like a craft argument and more like bracing. Identity threat. Economic threat. And bracing has a tone that often comes out sounding like moral certainty to me. Anyway. Just wanted to say this felt steady and sane in a conversation that’s gotten loud very quickly and I appreciate the hell out of that! ☺️
You are not alone (I thought maybe I was though so it's super nice to connect). Thanks for the comment and the share - I really appreciate it!
Fully agree here!
Thanks for reading!
I agree that AI can be a useful tool when used properly. Unfortunately the writing industry doesn’t agree. Magazines use checkers that reject submissions with even the hint of AI, agents and publishers do the same thing. I’ve actually fed AI checkers work I’ve had published long before LLM’s existed and been told it’s over 50% AI written.
That’s the problem. AI checkers can’t deal with well written prose because they are looking for averages. So how do you manage that?
Well it’s difficult right now! This issue you raise is a major one. The checkers have to get better AND the reality is that AI-assisted writing will become more mainstream and the publishers and magazines will have to change thier policies eventually.
I agree with a lot of this. The all-or-nothing AI debate has gotten exhausting, and writers deserve more room for nuance than they’re getting. I still think the sentence matters more than this piece allows, because voice and thought live there too.
That’s partly why I built my custom GPT for writers to focus on sharpening voice, editing, and original writing rather than just generating drafts.
https://chatgpt.com/g/g-69a0a235b3708191bbb56bc523c29c08-wordsmith
Give this a go! and help me refine, feedback welcome!
Very cool. I built my guardrails offer for a similar purpose. Fully customized.
A kindred spirit! The purist arguement is moot now. AI is already here. Like when the camera came out, the pictures were not “real” art. Well look at photography now.