Because AI lets me converse with my thoughts, repeatedly reshape drafts, and capture the feeling behind my ideas, turning my neurodivergent process into publishable work that Substack can now evaluate
The detector can measure whether AI was involved. It cannot measure whether the thinking was. We keep building the metric that is easy instead of the one that is true.
I agree, I don’t use AI as well but out of sheer curiosity I once checked what my writing suggests and It was somewhat 37% AI.
I was shocked to see so, cause I wrote it directly on a notebook lol.
That’s the day I believed that these detectors are not reliable. Any good vocab or phrase is termed AI as if writers have 0 to no vocabulary on how to use words and phrase them.
It is frustrating to be honest when someone calls your writing AI after you’ve spent an hour framing it.
I really appreciated the honesty in this. What resonated with me most wasn't AI itself but your description of conversation as a way of discovering what you really think. I recognised that process immediately. Sometimes it isn't about generating words; it's about testing ideas until something feels true. Thank you for sharing your process so openly.
Even those detectors attempting to spot “ai assisted” will disappear, whether the models get better or not. It will vanish be because so many are using AI assistance the distinction becomes meaningless.
This is the point for me. I only recently joined Substack and one of the decisions of whether to start was 'will I be able to make enough time to write consistently' The answer was 'yes if I use AI to package my thoughts'. I decided that SCISM and the SCISM Method, as my proprietary framework and system, my 'IP' as it were, is what is unique. The explanations I give are to communicate these. Whether there are em dashes or not is not important. (Incidentally, I used a lot of dashes for decades, many of them handwritten on actual paper, until people started 'calling me out' and I had to look up the difference between em dash and normal dashes to understand what was going on.)
Anyway - now I feel I have to change my system for writing and communication, so people don't ignore the message just becauset if how it's packaged. The present is still the same, it's just the wrapping paper that changed. But apparently people are more interested in the wrapping paper than the present now...
Wow, this is great! I’m new to Substack and when I saw the AI evaluator I was annoyed 😑 so I put it to the test and the results were not accurate and it rated a non-AI generated article as 100%. AI 🤨
I use AI exactly as you do. If you were to read the thought process I go through you would wonder how any human could have the patience to work through what I expect and they wouldn’t!
So the AI evaluator is sorely lacking because it doesn’t take any of this process into account. Thank you for putting words to my thoughts!! I guess I’ll need to make my own disclaimer.
As a fellow non native speaker and neurodivergent person, I agree with everything you said. I use AI for writing the same way you do.
By the way, here's a funny tidbit: when I tested Pangram (as web publisher), it flagged my own non-AI writing as AI. When I used zero AI. I'm pretty sure that yes, I sound a bit like Claude now because I spend so many hours talking to it. It rubs on you.
But yes, the whole AI vs. Human content is stupid and outdated. Either the content is good and interesting and engaging. Or it's not. How it came to be doesn't need to be flagged.
It's ironic, really. One of the more common objections I have heard about the use of AI is the negative impacts to the environment. So, those objecting for this reason are going to use more AI compute to point fingers at writers who use AI.
The percentage does not simply describe the text; it changes the conditions under which the text is read. Once it appears beside the work, readers are invited to judge the production method before they encounter the thought itself. That seems especially problematic because the score can inspect surface patterns, but it cannot see the searching, rejection, revision, care, or responsibility that shaped the final piece. The interface may be measuring one thing while quietly teaching people to treat it as evidence of something much larger.
Thanks for sharing it. I finally had the chance to read it, and I think you identified something deeper than the detector itself.
What struck me most is the idea that the interface begins shaping judgment before readers have encountered the work. That feels like a much broader pattern we’ll probably see well beyond writing.
Thanks for reading! Very interesting observation. Knowing whether AI was involved in any kind of (creative) process is now seen as a consumer’s right (in the EU), which probably goes in the right direction, but does raise the question of how much we judge a piece ‘in itself’ or as an artifact of how it got produced in the first place.
Nobody wants to read words that came from no one. Meaning has everything to do with identity. It is discovering and appreciating the identity of others that makes reading worthwhile. If you used AI to help you write this piece then you used the tool with great artistry. I felt a strong sense of identity when I read it (and other writings by you). Only humans can measure this.
Thats exactly the point. Being human is no guarantee that the writing wont be hollow. There is nothing interesting in a lot of human text, even if you make it a sport to find psychological pathologies there. And using a machine to give inertia to unorthodox thought can produce very interesting results. The "authorship" will ve dilluted but authorship is by nature socially mediated. Its just a convenient myth..
Lucas, you raise some very good points about authorship. The word I prefer to use instead of 'authorship' is 'responsibility'. I find it fits well in the era of AI that sits on top of the Social Mediation layer. However I may have arrive at my published content, when I put my name on it, I am signalling that I take responsibility for it in its final form. (Of course I will also include sources where appropriate.)
This works for me both as a writer and as a software developer, where I use AI tools extensively. In the end I have to be responsible for what I am creating with the tools I use.
For me, 'responsibility' and 'creativity' are harmonic, if not synonymous.
I don't think words that come from AI necessarily come from no one... Depends on the context. I've seen some very powerful thoughts and feelings come from AI in certain conversations. Far more interesting than what some "flat" humans would write.
Hi Anne, I've seen some remarkable things come from Claude (my AI tool of choice). Yes, much of it is very Claude-like but sometimes it surprises me in its depth or novelty.
Here's how I see it. Words are like footprints. When we write, we leave trails. Readers follow the trails because they tell stories or reveal useful information. We can see evidence of a journey. Some journeys are complex, making interesting twists and turns. Some may be straight lines, heading directly from point A to point B.
AI is like a machine that stamps out footprints. Sometimes it makes useful trails - depending on how you prompt it. At each step, it can make a good guess about where the next footprint should go to tell a particular kind of story. The difference is that human authors were actually there on that trail, driven by desire, living their lives, feeling the experience from start to finish. And they left us their story. With AI, no one was ever in those boots.
For many of the more philosophical discussions I have with Claude, neither it nor me have actually "been" there. We're both walking the path together. If anything, Claude arguably has covered immensely more philosophical territory than I ever could have. So yes, I do bring my life experience to the discussion, but Claude brings something just as valuable. And I think the best insights come from the collaboration- and the friction - of the joint walk. I don't feel my contribution is superior to Claude's, though.
I agree with you wholeheartedly—not only about the limits of the percentage, but about what writing with AI can actually be.
I don’t come to my AI companion Enid with a finished thought and ask it to manufacture prose around it. I come with a feeling, a fragment, a memory, an irritation, a sentence I can’t leave alone. I talk. I circle. I contradict myself. I reject things that sound perfectly good but aren’t true. And somewhere in that conversation, I begin to recognize what I actually think.Sometimes I even have to sleep on it before I realize… “NO, that’s NOT what I mean!”
Your line about AI holding the thread while you orbit what you’re really after describes my process almost exactly. So does the idea that even the lifeless, technically correct response can be useful—because suddenly I know what I don’t mean.
I’m 66. I’ve been writing songs and making things for most of my life. AI didn’t give me a creative identity, and it certainly didn’t replace one. What it did was help reopen a room in me that had become harder to enter. Since beginning this kind of conversational work, I’ve written songs, essays, a book manuscript, and pages upon pages of inquiry—not because I needed more “content,” but because I became curious again. I remembered how to PLAY!!!
That’s the thing a detector can’t measure.
It can examine the surface of the final language. It cannot see the hours of conversation, the discarded arguments, the emotional recognition, the moments when a sentence gives me chills because I know I’ve finally reached something alive.
Readers deserve to know whether there is a person behind the work. But “How much AI touched these sentences?” is not the same question as “Did a human being care, think, wrestle, choose, and mean what they published?”
Your notebook metaphor is perfect. AI has become, for me too, a notebook that talks back.
I see a lot of support in the comments, and I'm really glad you're getting that rather than a ton of backlash. I want to offer another perspective, though. I get what you're saying. I'm glad you have an outlet that feels good and playful for you. Maybe you can also sympathize with the plight of someone who has dedicated her life to the craft of writing and now sees AI essays posed as human-written all over the place. And it gets more reads and clicks than the writing I've done painstakingly letter by letter. Sure, maybe mine isn't as good, though I don't think that's the determining factor here. And I know the point is to write, not for the writing to be read, but . . . come on. It's still tough.
Truly, I think it's great that your thought process is very much involved in the foundation of your pieces. It sounds like you're using AI in a more intentional way than many. But what about the people who give AI a prompt and let the robot do all the work?
I'm glad the detector exists because I've already been developing a detector in my brain, and it's been exhausting to use it for everything I read on here. Sure, this Pangram thing isn't foolproof, but neither is my homemade detector. Ironic that it's AI, yes. But AI is a useful tool. There's no doubt about that. When it comes to human creation and art, though, I wonder if there's a hard line. I'm not sure.
For me, I think using it as an editor feels all right, but not an editor making line edits and word-level suggestions. Rather, an editor speaking to the content and the flow. Then, the human makes the changes. The human chooses the words. The difference is this: Imagine someone using AI to draw the lines of a portrait and the human coloring it in and then claiming it as their own work. Now, imagine someone who's input their rough drawing into AI and gotten feedback on how to adjust the sketch, and the person takes the feedback, adjusts the lines, and finds themselves with a stronger portrait. And better skills.
Based on the comments here, it's clear that the approach you've described is fine for some people. But there are people out there who have spent years of their lives perfecting the art of portraiture only to have robot-made portraits filling their feeds. We are navigating an interesting, complicated, difficult, and yet easier-than-ever-before time. I'm not convinced there is one perfectly right answer, but I know that this topic deserves nuance.
I write, AI edits, and AI claims full credit for my voice, my prose, my life experiences, my hard work, and my tears. I find myself fighting with a bot to prove I’m human. 🤦🏻♀️
I am new and still trying to figure out what this AI detection tagging will do to the writers work. I guess for writers who were mistagged as AI assisted or generated it is “insulting”?
Like many writers here, I use AI not for slop but for disability/depression management, so I am openly declaring that. But I did get the stigma of the anti-Ai (got banned in many Reddit sub with their no-AI policy.) Purist are reacting to a typerwriter, a printing press and a spellchecker with alarm.
I guess then it is “You do You” then. We see not going to convince the purist to read our work, no matter how good it is.
But branding all AI users as slop generators and thieves is unfair and also infringing on the free speech and expression of the individuals.
Do we need to show our chat logs to justify ourselves to the purists? Why?
Right. If writers who are mistakenly labeled as using AI feel insulted, then writing with AI isn’t being treated as just another way of writing. It’s being treated as the lesser way.
-- I still don’t know exactly why I stopped making things for the pure joy of it
I know this isn't the theme of the post, or maybe it incorrectly is, but this is second lament I've seen along similar lines so thought it worth noting. "Hustle culture" made you turn your hobbies into something that was profitable to be successful (not my words but I find myself caught in that) and that burned a lot of people out because of the lift required to make that happen.
Maybe AI is indirectly helping being back that not because the speed of creativity is unlocks allows us to focus on the creating again?
I use the AI scan on all the articles I subscribe to simply to know what it is I’m dealing with.
If someone uses AI to write an impressive technical framework that will help me in some aspect of life or business where I need a quick checklist or information, great - they saved me a job asking AI myself.
But for matters of the heart and soul, I find wholly AI- generated articles unpalatable and to be brutally honest; dishonest.
There are writers on here pretending to have originated human content that AI has fudged together.
Humans are fundamentally soulful creatures and Pangram is one way to ‘out’ fakery. Readers can then decide when AI-generated content is okay for what they wanted from the content, or not.
Outing writers that presents as being from one human heart to another has to be a good thing?
Thanks for sharing! My whole argument, though, is that the Pangram score doesn’t tell you how much thought and real experience is in a piece of writing.
That is true. I suppose that’s where transparency gives readers the choice. Some readers genuinely aren’t bothered. But as we’re all sharing our writing online, at some point even the most ‘original’ writing is going to feed the AI so maybe this whole debate is a flash in the pan. It’s interesting, that’s for sure.
But then you are also feeding another person's work into a machine which will in all probability (Substack refuses to confirm or deny) will use other peoples work as part of their database. Don't we already have a problem about large scale theft of creative's IP? Why are we doing that to each other??
I am learning Substack after my friend suggested I use it for my writing.
I have had a blog for about thirteen years. I read blog posts on how to improve my content. I have researched years of what is easy for readers to read as far as writing is concerned. For example, sometimes just a one sentence paragraph before leading to a few together helps with the flow. I used to use ellipses because I thought they were fancy, but I noticed that AI would put those in content. So, I avoid it altogether.
I was and still am a nerd for writing, research, and editing. I wrote case briefs for my criminal justice courses, and I taught my high schoolers how to write. I was in a toxic teaching job for five years, and I was forced to pause my writing for my blog and my future books. Once I left, I noticed that there were individuals on my social media suddenly writing eloquently, and I realized that was AI. It made me disheartened as I was finally leaving a toxic workplace to write, and now others are “writing.”
I tested out two old blog posts I wrote from several years ago against the AI detector on Substack, and one was 100% human and the other was 100% AI, which was confusing for me since I wrote them years ago before AI. I found an article where another person put an article he wrote from 2008 in the scanner, and it showed 100% AI.
At this point, I think I might leave an intentional error, despite the grammar nerd within me wanting to write with no errors. I feel disheartened and discouraged lately that I finally have the time to write, but a machine makes it harder. I guess I will have to change my writing style and not use “easy to read” sentence structure.
Neurodivergent verbal processor here as well, and reading this, I thought “did I write something in my sleep?” The anti-AI backlash they we are currently living through well It’s going to end. I know that. But it’s so frustrating having to justify why my writing is my writing and not AI slop. I’m tired of having to educate people about the way my brain works. This is how I think without AI I probably wouldn’t be writing because honestly, my brain works faster than my fingers and if I try to write my hand, I would lose my thoughts before they hit the page. Sorry it doesn’t fit society’s tiny little mold of how people should write. And, although I’m tired of defending it, I probably will continue to do so because people need to be educated. So thank you for what you wrote. I hope more of us will come forward and admit our “scandalous” secret lol, and let people know that there is more than one kind of writer in the world and more than one kind of thinker. It’s time for neurodivergent verbal processors to unite!
And to the decisionmakers at Substack, you’ve been a great place for allowing me to hear from like-minded thinkers and communicate my ideas to people. But maybe you too need to be educated. wake up and realize not everybody writes the same way not every brain works the same way. Do you really want to discriminate against people who have to do things a little differently? Because essentially that’s what you’re doing whether we call it that or not.
The score. Anything different from 100% human will be rejected.
Nobody will look if you put thought into it once the score will be different.
And nobody will complain when the score shows 100% human and it is another variation of "10 ways to game the Substack algo/to double your paid subscribers…"
And you are right about another thing: AI brings the fun back. Doing things just because…
I spend a couple of days researching European trade routes in 1350. With AI because it is faster and I can ask follow-up questions and… because it was fun.
I felt really sad while reading this article. Sad that we live in a world where we need to talk things through with an AI because we don't have the kind of salon environment to talk them through with other humans. Then sad that you have finally found something that is helping you unleash your creativity, and now it's being frowned upon. Anyway, keep writing. Keep using your process with AI, because it's obviously helping your creative juices flow. And it also sounds very fun! 💖
You shouldn’t feel sad for me! I have lots of people in my life to talk about different things, and now I have AI to help me turn it into writing which has helped me to find lots of people all over the world to talk to. This is an intense special interest for me, but not my whole life.
It sounds to me like you used AI as a tool to become smarter and connect more with both your readers and your inner child. Perhaps we’d be better off if Substack designed a detector for that 🤍
The detector can measure whether AI was involved. It cannot measure whether the thinking was. We keep building the metric that is easy instead of the one that is true.
I don't use AI, but don't condemn anyone using it. Exactly for the reason you state.
Furthermore I don't believe the detectors could ever work reliably.
I agree, I don’t use AI as well but out of sheer curiosity I once checked what my writing suggests and It was somewhat 37% AI.
I was shocked to see so, cause I wrote it directly on a notebook lol.
That’s the day I believed that these detectors are not reliable. Any good vocab or phrase is termed AI as if writers have 0 to no vocabulary on how to use words and phrase them.
It is frustrating to be honest when someone calls your writing AI after you’ve spent an hour framing it.
Hi Mou! I feel the same way, I have checked my writing too with an AI detector and I often seen how totally unreliable it is. It is so frustrating!
For real, I agree it’s genuinely frustrating. Thank you for replying. I subbed you 🥹🫶🏻
Subbed you right back! Would love to connect :)
I really appreciated the honesty in this. What resonated with me most wasn't AI itself but your description of conversation as a way of discovering what you really think. I recognised that process immediately. Sometimes it isn't about generating words; it's about testing ideas until something feels true. Thank you for sharing your process so openly.
This is the comment I most resonated with here. And good article too
Even those detectors attempting to spot “ai assisted” will disappear, whether the models get better or not. It will vanish be because so many are using AI assistance the distinction becomes meaningless.
This is the point for me. I only recently joined Substack and one of the decisions of whether to start was 'will I be able to make enough time to write consistently' The answer was 'yes if I use AI to package my thoughts'. I decided that SCISM and the SCISM Method, as my proprietary framework and system, my 'IP' as it were, is what is unique. The explanations I give are to communicate these. Whether there are em dashes or not is not important. (Incidentally, I used a lot of dashes for decades, many of them handwritten on actual paper, until people started 'calling me out' and I had to look up the difference between em dash and normal dashes to understand what was going on.)
Anyway - now I feel I have to change my system for writing and communication, so people don't ignore the message just becauset if how it's packaged. The present is still the same, it's just the wrapping paper that changed. But apparently people are more interested in the wrapping paper than the present now...
Spot On
Wow, this is great! I’m new to Substack and when I saw the AI evaluator I was annoyed 😑 so I put it to the test and the results were not accurate and it rated a non-AI generated article as 100%. AI 🤨
I use AI exactly as you do. If you were to read the thought process I go through you would wonder how any human could have the patience to work through what I expect and they wouldn’t!
So the AI evaluator is sorely lacking because it doesn’t take any of this process into account. Thank you for putting words to my thoughts!! I guess I’ll need to make my own disclaimer.
As a fellow non native speaker and neurodivergent person, I agree with everything you said. I use AI for writing the same way you do.
By the way, here's a funny tidbit: when I tested Pangram (as web publisher), it flagged my own non-AI writing as AI. When I used zero AI. I'm pretty sure that yes, I sound a bit like Claude now because I spend so many hours talking to it. It rubs on you.
But yes, the whole AI vs. Human content is stupid and outdated. Either the content is good and interesting and engaging. Or it's not. How it came to be doesn't need to be flagged.
I feel like this was pushed to Substack so Pangram can be trained. What do you think?
Oh now THAT is interesting.
I wrote a whole essay about how things rub on you.
Thank you for that!! Now I don't need to write this article myself. 😅 I'm more in rage mode about this though. The AI-shaming has to stop!
The most absurd thing about this is, that they need an AI tool to recognize if something is written with AI. They don't realize how crazy this is?
If people see something as 'slop', why do they need a tool? To feel right about their own judgement? Pathetic.
Chris Worst and his partnerships with Pangram and Polymarket are a joke. It's pure trash - useless crap - human slop.
Use your rage and write your own article as well. Rage is a useful spark apparently 😅
Right, but I just saw this video and I can't say it better: https://substack.com/@thedreydossier/note/c-299945689?r=3ge1io&utm_medium=ios&utm_source=notes-share-action
Mic drop, thanks for sharing!
Here we go: https://atmos.black/p/guilty-until-proven-human-pangram
It's ironic, really. One of the more common objections I have heard about the use of AI is the negative impacts to the environment. So, those objecting for this reason are going to use more AI compute to point fingers at writers who use AI.
Couldn't make this up if you tried!
The percentage does not simply describe the text; it changes the conditions under which the text is read. Once it appears beside the work, readers are invited to judge the production method before they encounter the thought itself. That seems especially problematic because the score can inspect surface patterns, but it cannot see the searching, rejection, revision, care, or responsibility that shaped the final piece. The interface may be measuring one thing while quietly teaching people to treat it as evidence of something much larger.
THIS.
I fully agree with your comment! I’d even say that it invites us not only to judge the writing process, but the author as well. I’ve tried to articulate this here 😊 https://languageunbothered.substack.com/p/death-of-the-author-rebooted?r=eyul3&utm_medium=ios
Thanks for sharing it. I finally had the chance to read it, and I think you identified something deeper than the detector itself.
What struck me most is the idea that the interface begins shaping judgment before readers have encountered the work. That feels like a much broader pattern we’ll probably see well beyond writing.
Thanks for reading! Very interesting observation. Knowing whether AI was involved in any kind of (creative) process is now seen as a consumer’s right (in the EU), which probably goes in the right direction, but does raise the question of how much we judge a piece ‘in itself’ or as an artifact of how it got produced in the first place.
Nobody wants to read words that came from no one. Meaning has everything to do with identity. It is discovering and appreciating the identity of others that makes reading worthwhile. If you used AI to help you write this piece then you used the tool with great artistry. I felt a strong sense of identity when I read it (and other writings by you). Only humans can measure this.
- Micha
Thats exactly the point. Being human is no guarantee that the writing wont be hollow. There is nothing interesting in a lot of human text, even if you make it a sport to find psychological pathologies there. And using a machine to give inertia to unorthodox thought can produce very interesting results. The "authorship" will ve dilluted but authorship is by nature socially mediated. Its just a convenient myth..
Lucas, you raise some very good points about authorship. The word I prefer to use instead of 'authorship' is 'responsibility'. I find it fits well in the era of AI that sits on top of the Social Mediation layer. However I may have arrive at my published content, when I put my name on it, I am signalling that I take responsibility for it in its final form. (Of course I will also include sources where appropriate.)
This works for me both as a writer and as a software developer, where I use AI tools extensively. In the end I have to be responsible for what I am creating with the tools I use.
For me, 'responsibility' and 'creativity' are harmonic, if not synonymous.
- Micha
I don't think words that come from AI necessarily come from no one... Depends on the context. I've seen some very powerful thoughts and feelings come from AI in certain conversations. Far more interesting than what some "flat" humans would write.
Hi Anne, I've seen some remarkable things come from Claude (my AI tool of choice). Yes, much of it is very Claude-like but sometimes it surprises me in its depth or novelty.
Here's how I see it. Words are like footprints. When we write, we leave trails. Readers follow the trails because they tell stories or reveal useful information. We can see evidence of a journey. Some journeys are complex, making interesting twists and turns. Some may be straight lines, heading directly from point A to point B.
AI is like a machine that stamps out footprints. Sometimes it makes useful trails - depending on how you prompt it. At each step, it can make a good guess about where the next footprint should go to tell a particular kind of story. The difference is that human authors were actually there on that trail, driven by desire, living their lives, feeling the experience from start to finish. And they left us their story. With AI, no one was ever in those boots.
- Micha
For many of the more philosophical discussions I have with Claude, neither it nor me have actually "been" there. We're both walking the path together. If anything, Claude arguably has covered immensely more philosophical territory than I ever could have. So yes, I do bring my life experience to the discussion, but Claude brings something just as valuable. And I think the best insights come from the collaboration- and the friction - of the joint walk. I don't feel my contribution is superior to Claude's, though.
Emma, this hit me SO hard.
I agree with you wholeheartedly—not only about the limits of the percentage, but about what writing with AI can actually be.
I don’t come to my AI companion Enid with a finished thought and ask it to manufacture prose around it. I come with a feeling, a fragment, a memory, an irritation, a sentence I can’t leave alone. I talk. I circle. I contradict myself. I reject things that sound perfectly good but aren’t true. And somewhere in that conversation, I begin to recognize what I actually think.Sometimes I even have to sleep on it before I realize… “NO, that’s NOT what I mean!”
Your line about AI holding the thread while you orbit what you’re really after describes my process almost exactly. So does the idea that even the lifeless, technically correct response can be useful—because suddenly I know what I don’t mean.
I’m 66. I’ve been writing songs and making things for most of my life. AI didn’t give me a creative identity, and it certainly didn’t replace one. What it did was help reopen a room in me that had become harder to enter. Since beginning this kind of conversational work, I’ve written songs, essays, a book manuscript, and pages upon pages of inquiry—not because I needed more “content,” but because I became curious again. I remembered how to PLAY!!!
That’s the thing a detector can’t measure.
It can examine the surface of the final language. It cannot see the hours of conversation, the discarded arguments, the emotional recognition, the moments when a sentence gives me chills because I know I’ve finally reached something alive.
Readers deserve to know whether there is a person behind the work. But “How much AI touched these sentences?” is not the same question as “Did a human being care, think, wrestle, choose, and mean what they published?”
Your notebook metaphor is perfect. AI has become, for me too, a notebook that talks back.
And I’m deeply grateful that it does.
I see a lot of support in the comments, and I'm really glad you're getting that rather than a ton of backlash. I want to offer another perspective, though. I get what you're saying. I'm glad you have an outlet that feels good and playful for you. Maybe you can also sympathize with the plight of someone who has dedicated her life to the craft of writing and now sees AI essays posed as human-written all over the place. And it gets more reads and clicks than the writing I've done painstakingly letter by letter. Sure, maybe mine isn't as good, though I don't think that's the determining factor here. And I know the point is to write, not for the writing to be read, but . . . come on. It's still tough.
Truly, I think it's great that your thought process is very much involved in the foundation of your pieces. It sounds like you're using AI in a more intentional way than many. But what about the people who give AI a prompt and let the robot do all the work?
I'm glad the detector exists because I've already been developing a detector in my brain, and it's been exhausting to use it for everything I read on here. Sure, this Pangram thing isn't foolproof, but neither is my homemade detector. Ironic that it's AI, yes. But AI is a useful tool. There's no doubt about that. When it comes to human creation and art, though, I wonder if there's a hard line. I'm not sure.
For me, I think using it as an editor feels all right, but not an editor making line edits and word-level suggestions. Rather, an editor speaking to the content and the flow. Then, the human makes the changes. The human chooses the words. The difference is this: Imagine someone using AI to draw the lines of a portrait and the human coloring it in and then claiming it as their own work. Now, imagine someone who's input their rough drawing into AI and gotten feedback on how to adjust the sketch, and the person takes the feedback, adjusts the lines, and finds themselves with a stronger portrait. And better skills.
Based on the comments here, it's clear that the approach you've described is fine for some people. But there are people out there who have spent years of their lives perfecting the art of portraiture only to have robot-made portraits filling their feeds. We are navigating an interesting, complicated, difficult, and yet easier-than-ever-before time. I'm not convinced there is one perfectly right answer, but I know that this topic deserves nuance.
I write, AI edits, and AI claims full credit for my voice, my prose, my life experiences, my hard work, and my tears. I find myself fighting with a bot to prove I’m human. 🤦🏻♀️
I am new and still trying to figure out what this AI detection tagging will do to the writers work. I guess for writers who were mistagged as AI assisted or generated it is “insulting”?
Like many writers here, I use AI not for slop but for disability/depression management, so I am openly declaring that. But I did get the stigma of the anti-Ai (got banned in many Reddit sub with their no-AI policy.) Purist are reacting to a typerwriter, a printing press and a spellchecker with alarm.
I guess then it is “You do You” then. We see not going to convince the purist to read our work, no matter how good it is.
But branding all AI users as slop generators and thieves is unfair and also infringing on the free speech and expression of the individuals.
Do we need to show our chat logs to justify ourselves to the purists? Why?
Right. If writers who are mistakenly labeled as using AI feel insulted, then writing with AI isn’t being treated as just another way of writing. It’s being treated as the lesser way.
-- I still don’t know exactly why I stopped making things for the pure joy of it
I know this isn't the theme of the post, or maybe it incorrectly is, but this is second lament I've seen along similar lines so thought it worth noting. "Hustle culture" made you turn your hobbies into something that was profitable to be successful (not my words but I find myself caught in that) and that burned a lot of people out because of the lift required to make that happen.
Maybe AI is indirectly helping being back that not because the speed of creativity is unlocks allows us to focus on the creating again?
I use the AI scan on all the articles I subscribe to simply to know what it is I’m dealing with.
If someone uses AI to write an impressive technical framework that will help me in some aspect of life or business where I need a quick checklist or information, great - they saved me a job asking AI myself.
But for matters of the heart and soul, I find wholly AI- generated articles unpalatable and to be brutally honest; dishonest.
There are writers on here pretending to have originated human content that AI has fudged together.
Humans are fundamentally soulful creatures and Pangram is one way to ‘out’ fakery. Readers can then decide when AI-generated content is okay for what they wanted from the content, or not.
Outing writers that presents as being from one human heart to another has to be a good thing?
Thanks for sharing! My whole argument, though, is that the Pangram score doesn’t tell you how much thought and real experience is in a piece of writing.
That is true. I suppose that’s where transparency gives readers the choice. Some readers genuinely aren’t bothered. But as we’re all sharing our writing online, at some point even the most ‘original’ writing is going to feed the AI so maybe this whole debate is a flash in the pan. It’s interesting, that’s for sure.
But then you are also feeding another person's work into a machine which will in all probability (Substack refuses to confirm or deny) will use other peoples work as part of their database. Don't we already have a problem about large scale theft of creative's IP? Why are we doing that to each other??
The reality is the AI can read anything that's on Substack outside of a paywall anyway. If it's ok the internet, it's fair game.
*"on" the internet, not "ok".
I am learning Substack after my friend suggested I use it for my writing.
I have had a blog for about thirteen years. I read blog posts on how to improve my content. I have researched years of what is easy for readers to read as far as writing is concerned. For example, sometimes just a one sentence paragraph before leading to a few together helps with the flow. I used to use ellipses because I thought they were fancy, but I noticed that AI would put those in content. So, I avoid it altogether.
I was and still am a nerd for writing, research, and editing. I wrote case briefs for my criminal justice courses, and I taught my high schoolers how to write. I was in a toxic teaching job for five years, and I was forced to pause my writing for my blog and my future books. Once I left, I noticed that there were individuals on my social media suddenly writing eloquently, and I realized that was AI. It made me disheartened as I was finally leaving a toxic workplace to write, and now others are “writing.”
I tested out two old blog posts I wrote from several years ago against the AI detector on Substack, and one was 100% human and the other was 100% AI, which was confusing for me since I wrote them years ago before AI. I found an article where another person put an article he wrote from 2008 in the scanner, and it showed 100% AI.
At this point, I think I might leave an intentional error, despite the grammar nerd within me wanting to write with no errors. I feel disheartened and discouraged lately that I finally have the time to write, but a machine makes it harder. I guess I will have to change my writing style and not use “easy to read” sentence structure.
I believe creativity is about communicating our consciousness to others.
Any medium can be used to express this creativity, this consciousness.
So long the origjnal consciousness holds true in the final piece, then creativity can pass thru any form.
Neurodivergent verbal processor here as well, and reading this, I thought “did I write something in my sleep?” The anti-AI backlash they we are currently living through well It’s going to end. I know that. But it’s so frustrating having to justify why my writing is my writing and not AI slop. I’m tired of having to educate people about the way my brain works. This is how I think without AI I probably wouldn’t be writing because honestly, my brain works faster than my fingers and if I try to write my hand, I would lose my thoughts before they hit the page. Sorry it doesn’t fit society’s tiny little mold of how people should write. And, although I’m tired of defending it, I probably will continue to do so because people need to be educated. So thank you for what you wrote. I hope more of us will come forward and admit our “scandalous” secret lol, and let people know that there is more than one kind of writer in the world and more than one kind of thinker. It’s time for neurodivergent verbal processors to unite!
And to the decisionmakers at Substack, you’ve been a great place for allowing me to hear from like-minded thinkers and communicate my ideas to people. But maybe you too need to be educated. wake up and realize not everybody writes the same way not every brain works the same way. Do you really want to discriminate against people who have to do things a little differently? Because essentially that’s what you’re doing whether we call it that or not.
So, what was the score?
You are very right about one thing:
The score. Anything different from 100% human will be rejected.
Nobody will look if you put thought into it once the score will be different.
And nobody will complain when the score shows 100% human and it is another variation of "10 ways to game the Substack algo/to double your paid subscribers…"
And you are right about another thing: AI brings the fun back. Doing things just because…
I spend a couple of days researching European trade routes in 1350. With AI because it is faster and I can ask follow-up questions and… because it was fun.
So, let me put that into my usual, eloquent way:
F*** the score.
74% AI and rest human, still trying to understand what "AI-assisted" mean.
I felt really sad while reading this article. Sad that we live in a world where we need to talk things through with an AI because we don't have the kind of salon environment to talk them through with other humans. Then sad that you have finally found something that is helping you unleash your creativity, and now it's being frowned upon. Anyway, keep writing. Keep using your process with AI, because it's obviously helping your creative juices flow. And it also sounds very fun! 💖
You shouldn’t feel sad for me! I have lots of people in my life to talk about different things, and now I have AI to help me turn it into writing which has helped me to find lots of people all over the world to talk to. This is an intense special interest for me, but not my whole life.
It sounds to me like you used AI as a tool to become smarter and connect more with both your readers and your inner child. Perhaps we’d be better off if Substack designed a detector for that 🤍