Thank you. I am going to begin expanding the list now. I have already found more today. Please let me know if you see anyone suspicious. Of course, you can always check it too. Just copy/paste into a draft like it was your own post and then you can scan the text like it was yours.
Quite the detailed piece and I appreciate the thorough investigation. Not completely surprised by the outcome, but that doesn't mean im not triggered about it. The fact that people are rewarded for this kind of "work" or think it's ok is what triggers me. The lying afterwards and sheer denial is what really gets me. Do some of these people really think there's absolutely no problem with this? I'll admit there was a time where I thought AI could shortcut my work in this way and allow me to scale my business faster but I've recently become disgusted by that approach. I remember how it felt, soulless, empty, and at a time, I believed there was nothing wrong with it. Yes, writing a piece yourself takes longer, but the feeling of satisfaction afterwards is the real reward. What is life if we can't honestly feel that within ourselves? Carey's conversation was especially interesting, she got increasingly more frustrated as the conversations kept going and less "proper".
Nice work on the research. At first, I was a bit uncomfortable because it felt like a witch hunt but I can see you've given people the opportunity to respond. I was also thinking that some people might have a different first language to English or there's other things to consider like dyslexia. But the more I moved through your piece it's quite evident what is going on. It's a shame, I subscribe to a few people who are open about their use of AI and for one or two they have an insightful approach. Most importantly, they're transparent. But there are a lot of posts where it's obviously AI. There's going to be a lot of people who game the system, adding random typos here or there that are completely unaligned with the rest of their text. Plus, some people are pumping out 3 posts a week, perfect prose and about 2000+ words! Now, I might not be as proficient as others, but that is another eyebrow raiser. A very thought provoking piece.
Thank you, Lloyd. I appreciate that a lot. I’m working on the sequel now. And two of the people on the list actually responded positively, acknowledged their AI use and we had positive exchanges. So now instead of condemning them, I’m encouraging others to subscribe to them.
If you have any thoughts or suggestions as I continue the series then do let me know. I’m planning to explore it in multiple angles. My dad had a good idea: I’m going to compare/contrast the Science category bestsellers with the “Health and Wellness” bestsellers and see what we find. It’s not hard to guess which will likely have more hidden AI use.
Every time I dictate my ideas to AI apps to write a Substack post for me they all respond, "no way are we generating this shit and you're now banned from using our products ever again and you do realize ideas like this are why you have no friends, right?"
But seriously, I just ran my first and only post through an AI checker. It says that 100% of my writing was written by AI!
That’s a whole new level of bullshit, because I wrote my 1 min read, all by my self. I did however, run it through a grammar check in Microsoft Word before posting.
I suspect many of the AI checkers, like the one that got its evaluation of my note 100% wrong, are completely unreliable.
Everything I write comes back as 100% human written. Most of the stuff on here is human written. Pangram is very effective. The only people who try to malign Pangram are people trying to hide their AI use. And it is often people like you who hide behind a pseudonym.
I have been writing under my own name on the internet since I was a teenager in the 20th century. What I do as a writer writing my own words under my own name is not the same thing as an anonymous person using an LLM to try and dupe people into thinking they’re writing themselves.
You’ve been using your real name on the internet since the 90’s?!?
Well, good for you!
As a solo parent who can’t afford to lose my income over sharing some of my controversial views, I took my employer’s advice and use a pseudonym when posting anything to social media that could *gasp!* cause a stir.
Cancelling the average citizen for acknowledging biological reality, is still very much a thing, you know.
I do find it fascinating that even this reply, when run through a detection scan, is suspected to be partially written by AI as well.
That only reinforces my suspicion that AI “scanners” are poppycock, and those that use them to judge the legitimacy of other Substackers’ writing shouldn’t be taken seriously.
Which begs the question: What is the psychological payoff of using unreliable software to invalidate the hard work of other writers?
The reason why I turn it off is ideological, not because I rely on AI.
There are still false positives, and if I had signed up with pangram being part of the deal, I would have been fine with it (as you can see I don't draft my stuff with AI). What I object to is it being forced on me without having any say in the matter. Which is why I'm not opting in.
This isn’t rolling over. This is being transparent with your audience. As my list makes clear, this is a genuine problem on Substack. The execs who added it aren’t stupid. And they know the technology is reliable.
OMG thank you. I have had to unfollow soooo many people - not listed in here - who also write pro-Israel articles and who I admire because their writing is unbearable AI slop!
This thing's got legs Mr. S. It actually gets worse. Some big name authors have disabled their detection, like Salmon Rushdie. We're used to lying, cheating, and hypocrisy from our political pundits. But our best-selling novelists? Ouch
You shouldn’t ignore those of us who know firsthand how dangerous and unreliable this technology is. I’ve worked for 3 of the big AI companies training their models. If you’d worked on those projects with me then you wouldn’t be so optimistic. When we learn how the sausage is made then we become less enthusiastic about taking a bite.
But this whole series has nothing to do with tech optimism. It has to do with ethics. People who use LLMs to write should be open and honest about it. They shouldn’t try and conceal it. Don’t you agree that people should be transparent here instead of disabling Pangram?
Wow! This is an incredibly researched article. You found the plagiarists (or at the very least 47 of the plagiarists). Good work.
Thank you. I am going to begin expanding the list now. I have already found more today. Please let me know if you see anyone suspicious. Of course, you can always check it too. Just copy/paste into a draft like it was your own post and then you can scan the text like it was yours.
Quite the detailed piece and I appreciate the thorough investigation. Not completely surprised by the outcome, but that doesn't mean im not triggered about it. The fact that people are rewarded for this kind of "work" or think it's ok is what triggers me. The lying afterwards and sheer denial is what really gets me. Do some of these people really think there's absolutely no problem with this? I'll admit there was a time where I thought AI could shortcut my work in this way and allow me to scale my business faster but I've recently become disgusted by that approach. I remember how it felt, soulless, empty, and at a time, I believed there was nothing wrong with it. Yes, writing a piece yourself takes longer, but the feeling of satisfaction afterwards is the real reward. What is life if we can't honestly feel that within ourselves? Carey's conversation was especially interesting, she got increasingly more frustrated as the conversations kept going and less "proper".
Nice work on the research. At first, I was a bit uncomfortable because it felt like a witch hunt but I can see you've given people the opportunity to respond. I was also thinking that some people might have a different first language to English or there's other things to consider like dyslexia. But the more I moved through your piece it's quite evident what is going on. It's a shame, I subscribe to a few people who are open about their use of AI and for one or two they have an insightful approach. Most importantly, they're transparent. But there are a lot of posts where it's obviously AI. There's going to be a lot of people who game the system, adding random typos here or there that are completely unaligned with the rest of their text. Plus, some people are pumping out 3 posts a week, perfect prose and about 2000+ words! Now, I might not be as proficient as others, but that is another eyebrow raiser. A very thought provoking piece.
Thank you, Lloyd. I appreciate that a lot. I’m working on the sequel now. And two of the people on the list actually responded positively, acknowledged their AI use and we had positive exchanges. So now instead of condemning them, I’m encouraging others to subscribe to them.
If you have any thoughts or suggestions as I continue the series then do let me know. I’m planning to explore it in multiple angles. My dad had a good idea: I’m going to compare/contrast the Science category bestsellers with the “Health and Wellness” bestsellers and see what we find. It’s not hard to guess which will likely have more hidden AI use.
Nothing from me. I'm just a fan of good investigative journalism. You're on it!
Every time I dictate my ideas to AI apps to write a Substack post for me they all respond, "no way are we generating this shit and you're now banned from using our products ever again and you do realize ideas like this are why you have no friends, right?"
So, you didn’t use AI.
Would it have killed you to use spell check?
I kid, I kid!
But seriously, I just ran my first and only post through an AI checker. It says that 100% of my writing was written by AI!
That’s a whole new level of bullshit, because I wrote my 1 min read, all by my self. I did however, run it through a grammar check in Microsoft Word before posting.
I suspect many of the AI checkers, like the one that got its evaluation of my note 100% wrong, are completely unreliable.
Write your post directly in the post window and see what happens.
Well, at least Pangram gave me some credit. Regardless, I did not use AI.
I’m skeptical you are telling the truth.
I don’t trust people who hide behind pseudonyms.
Everything I write comes back as 100% human written. Most of the stuff on here is human written. Pangram is very effective. The only people who try to malign Pangram are people trying to hide their AI use. And it is often people like you who hide behind a pseudonym.
I have been writing under my own name on the internet since I was a teenager in the 20th century. What I do as a writer writing my own words under my own name is not the same thing as an anonymous person using an LLM to try and dupe people into thinking they’re writing themselves.
Oh my Gosh…
You’ve been using your real name on the internet since the 90’s?!?
Well, good for you!
As a solo parent who can’t afford to lose my income over sharing some of my controversial views, I took my employer’s advice and use a pseudonym when posting anything to social media that could *gasp!* cause a stir.
Cancelling the average citizen for acknowledging biological reality, is still very much a thing, you know.
I do find it fascinating that even this reply, when run through a detection scan, is suspected to be partially written by AI as well.
That only reinforces my suspicion that AI “scanners” are poppycock, and those that use them to judge the legitimacy of other Substackers’ writing shouldn’t be taken seriously.
Which begs the question: What is the psychological payoff of using unreliable software to invalidate the hard work of other writers?
The reason why I turn it off is ideological, not because I rely on AI.
There are still false positives, and if I had signed up with pangram being part of the deal, I would have been fine with it (as you can see I don't draft my stuff with AI). What I object to is it being forced on me without having any say in the matter. Which is why I'm not opting in.
Yisrael, can you understand why I have lost faith and trust in you because you do this?
It’s your prerogative to do so.
It’s my prerogative not to agree to roll over at Substack’s whims. I’ve responded here:
https://substack.com/profile/109090086-yisrael-m-eliashiv/note/c-337710579
This isn’t rolling over. This is being transparent with your audience. As my list makes clear, this is a genuine problem on Substack. The execs who added it aren’t stupid. And they know the technology is reliable.
The technology is not reliable the way you think: https://unlimitedhangout.com/2026/09/reports/the-pangram-paradox/
OMG thank you. I have had to unfollow soooo many people - not listed in here - who also write pro-Israel articles and who I admire because their writing is unbearable AI slop!
Who? I am expanding this list with sequel pieces and intend to grow the list as long as I can. Feel free to DM me if you prefer.
People are giving this dude $$ to read his stuff? Isn't AI free? 😀
That’s why this is such a scandal. People are charging readers for material that is actually low quality and unoriginal.
This thing's got legs Mr. S. It actually gets worse. Some big name authors have disabled their detection, like Salmon Rushdie. We're used to lying, cheating, and hypocrisy from our political pundits. But our best-selling novelists? Ouch
That's what's up 💯❤️🔥
You shouldn’t ignore those of us who know firsthand how dangerous and unreliable this technology is. I’ve worked for 3 of the big AI companies training their models. If you’d worked on those projects with me then you wouldn’t be so optimistic. When we learn how the sausage is made then we become less enthusiastic about taking a bite.
But this whole series has nothing to do with tech optimism. It has to do with ethics. People who use LLMs to write should be open and honest about it. They shouldn’t try and conceal it. Don’t you agree that people should be transparent here instead of disabling Pangram?
Which corner is that?
What are some of your favorite Substacks? I’m curious to learn more about your corner.