"By embracing this partnership while remaining vigilant about ethical considerations, we can harness the potential of AI to enrich our narratives rather than diminish them." (said the AI)
So thought provoking, P.Q. I have SO MANY QUESTIONS that would take until the New Year for you to finally say, “Hey, you know that ChapGPT could probably help you out with these ridiculously low rent questions, dummy!” But, do you know if Substack is actively scanning our work product for any AI generated content that is OVERLY reliant upon the AI of it all? I read that two independent firms, unbeknownst to the other as being hired by the powers that be at Medium to investigate the case for true authenticity of content in their respective readership communities and both firms, independently found that the content of Medium was about 49% AI generated. Now, the CEO had a very important distinction between these two findings in that they were hired to scan the about 4.200 or so randomly selected articles and then detect what portion was actually AI generated, right? Well, the CEO was not buying the findings because the firms themselves did not set a standard quota for the actual AMOUNT of AI generated content that was being published by Medium authors which would make the case for those who used AI generated results as an outline for an article idea or possibly just used an algorithm to enhance an abstract concept that AI helped to make make more palatable for the reader.
However, I find it a tad suspect that both firms independently sampled arbitrary articles of abound 4.200 each, but they had IDENTICAL data results that were uncannily consistent with their respective sources. What is your opinion on this subject? I’d love to know your thoughts?! 😊
Interesting and scary: We may not be around to perceive the full amplitude of AI, but if people cease to think and talk and write for themselves, or even paint and sing; wonder what would unfold for future generations? One child I spoke to , did not know what a dictionary was! Ah the joys of actually looking for meanings of words is lost forever.
In my school the last class was Words are Important: each page on the left hand side had words, whose meanings we had to look for in the Dictionary; on the right hand side of the page were jumbled sentences, where we'd use the words whose meaning we unearthed.
Unless the cycle shifts, and people rediscover the joys of writing letters, then wait for an eternity for the Postman to arrive with a reply; or to sit by the huge telephone, just waiting for that particular Ring:))
Or even go shopping to the friendly cloth merchant and get immersed in yards of lacy, silky, cotton materials, then visit the nearby tailor to get them stitched:)))
On Substack actively scanning, from an Observer article last month: “Substack doesn’t prohibit A.I.-generated content, although the platform has mechanisms in place to detect spam activities like duplicated content and bot activity that often involve A.I. “We don’t proactively monitor or remove content solely based on its A.I. origins, as there are numerous valid, constructive applications for assisted content creation,” said the company in a statement to Observer.” https://observer.com/2024/11/report-popular-substack-authors-use-ai/
On Medium: I’m not sure if AI detection scores are useful at all. People have this deep fear that fully AI generated content will take over the Internet (see ‘Dead Internet Theory’). Detection services play into that fear.
Speaking for myself, I like to let the AI do part of the writing, it’s fun. There are many prompting techniques to get good output. I try to get good at it. Nobody has all the answers, because it’s a new and dynamic field. I wonder if my AI writing is good enough to pass as ‘certified human’ in the eyes of those detection geniuses. What’s your guess?
You are the real deal. You’re too genuine - it shines through with every word of your above referenced response to a complex issue - ironically about authenticity! 😊💜
Someone else said it in the comments, but yeah what kinda gave it away was that it felt a little short on landing an idea, if that makes sense. For this new series do you plan on feeding these articles to the LLM model so it can "develop a style" or are you going to let it run from scratch every time?
For me, part of the fun is trying out different styles and see just how good AI output can become when you push the right buttons. So I'll keep experimenting.
In a way, I hate how good AI output sometimes is, but at the same time I want to learn how to make the best of it.
This whole article felt AI written in terms of blandness. The writing is technically fine, though lacks a real viewpoint. The arguments are logically laid out and sound factual, though I have no idea how the author really feels. There’s zero emotion or heart or sense of a unique individual behind the words.
For me as an author that would never be good enough. For me as a reader, AI writing leads to endless retreads of prior works. Also, the training data limits the world experience and tastes of the AI generating the content. I’d rather experience the original work of humans than the bland regurgitation of a semblance of existing writing of dubious quality.
I see what you mean. It's not one thing in particular that stands out, it's that nothing stands out at all.
Although I have found that 'style transfer' techniques can make for much better output. Which makes sense to me: emotion and heart can be faked, humans do it all the time (think sociopaths, actors, liars).
Yes, emotion can be faked. As a writer I think it’s important to have a clear point of view. I can agree or not but the author puts down a line and says this is who I am. Your AI writing is neither exciting nor offensive. It’s informative in a bland museum curator way, though not memorable. It doesn’t make me feel anything and I feel that’s a sad statement. If you want real fans who will stick with you, you need to make them feel something.
You make a good point, and I think better prompts can make a world of difference. Let me illustrate:
ChatGPT, reword this piece of bland AI output into something much more personal in the provocative style of, for example, Piers Morgan: "Ultimately, the future of storytelling will likely involve a blend of human creativity and machine assistance. By embracing this partnership while remaining vigilant about ethical considerations, we can harness the potential of AI to enrich our narratives rather than diminish them."
"Let's face it, the future of storytelling is staring us right in the face - and it's a fusion of human genius and machine meddling. I'm not naive, I know the purists will scream bloody murder at the very thought of AI sullying the sacred art of storytelling. But I say, bring it on. Let's harness the power of these machines to inject some much-needed adrenaline into our narratives, to push the boundaries of what's possible, and to create something truly revolutionary. And to those who claim AI will somehow 'diminish' our stories, I say, get a grip. The only thing that's being diminished is the lazy, uninspired writing that's been peddled as 'art' for far too long. It's time to shake things up, and if AI is the catalyst, then so be it."
Thank you for hacking patterns. I agree that AI can be a partner, there's trust issues, though. I want to believe that AI will help humanity, but the inclination to mistrust AI has been pre-programned for generations.
Very interesting article. It seems we forgot that we can make mistakes in our writing and it's ok to look as though we aren't perfect. I can say that early on I have used Chat to correct any errors that needed correcting. I don't think AI can write my content, or maybe it can. It is honestly something to be really aware of. Thank you for sharing this with us
😢. I would have been an A student back then if that was around. 😉But alas I struggled through school finding my own voice. So B/C student I was. Writing has been the biggest by far challenge for me.
Choosing to look like I’m still learning is going to be the biggest challenge still. 😉
Thanks! They say to err is human, and that's true now more than ever.
From what I hear, some of Substack's top publications as well as most new posts on LinkedIn are now made with AI. As is probably most schoolwork these days.
So thought provoking, P.Q. I have SO MANY QUESTIONS that would take until the New Year for you to finally say, “Hey, you know that ChapGPT could probably help you out with these ridiculously low rent questions, dummy!” But, do you know if Substack is actively scanning our work product for any AI generated content that is OVERLY reliant upon the AI of it all? I read that two independent firms, unbeknownst to the other as being hired by the powers that be at Medium to investigate the case for true authenticity of content in their respective readership communities and both firms, independently found that the content of Medium was about 49% AI generated. Now, the CEO had a very important distinction between these two findings in that they were hired to scan the about 4.200 or so randomly selected articles and then detect what portion was actually AI generated, right? Well, the CEO was not buying the findings because the firms themselves did not set a standard quota for the actual AMOUNT of AI generated content that was being published by Medium authors which would make the case for those who used AI generated results as an outline for an article idea or possibly just used an algorithm to enhance an abstract concept that AI helped to make make more palatable for the reader.
However, I find it a tad suspect that both firms independently sampled arbitrary articles of abound 4.200 each, but they had IDENTICAL data results that were uncannily consistent with their respective sources. What is your opinion on this subject? I’d love to know your thoughts?! 😊
Interesting and scary: We may not be around to perceive the full amplitude of AI, but if people cease to think and talk and write for themselves, or even paint and sing; wonder what would unfold for future generations? One child I spoke to , did not know what a dictionary was! Ah the joys of actually looking for meanings of words is lost forever.
In my school the last class was Words are Important: each page on the left hand side had words, whose meanings we had to look for in the Dictionary; on the right hand side of the page were jumbled sentences, where we'd use the words whose meaning we unearthed.
Unless the cycle shifts, and people rediscover the joys of writing letters, then wait for an eternity for the Postman to arrive with a reply; or to sit by the huge telephone, just waiting for that particular Ring:))
Or even go shopping to the friendly cloth merchant and get immersed in yards of lacy, silky, cotton materials, then visit the nearby tailor to get them stitched:)))
So fascinating, Seema. Thank you for sharing your experience with us! 😊
Many thanks for your enthousiastic reply!
On Substack actively scanning, from an Observer article last month: “Substack doesn’t prohibit A.I.-generated content, although the platform has mechanisms in place to detect spam activities like duplicated content and bot activity that often involve A.I. “We don’t proactively monitor or remove content solely based on its A.I. origins, as there are numerous valid, constructive applications for assisted content creation,” said the company in a statement to Observer.” https://observer.com/2024/11/report-popular-substack-authors-use-ai/
On Medium: I’m not sure if AI detection scores are useful at all. People have this deep fear that fully AI generated content will take over the Internet (see ‘Dead Internet Theory’). Detection services play into that fear.
Speaking for myself, I like to let the AI do part of the writing, it’s fun. There are many prompting techniques to get good output. I try to get good at it. Nobody has all the answers, because it’s a new and dynamic field. I wonder if my AI writing is good enough to pass as ‘certified human’ in the eyes of those detection geniuses. What’s your guess?
You are the real deal. You’re too genuine - it shines through with every word of your above referenced response to a complex issue - ironically about authenticity! 😊💜
That's so kind of you! Being genuine may be the only thing that sets us apart from the robots at some point in the future, better hang on to it ❤
Someone else said it in the comments, but yeah what kinda gave it away was that it felt a little short on landing an idea, if that makes sense. For this new series do you plan on feeding these articles to the LLM model so it can "develop a style" or are you going to let it run from scratch every time?
For me, part of the fun is trying out different styles and see just how good AI output can become when you push the right buttons. So I'll keep experimenting.
In a way, I hate how good AI output sometimes is, but at the same time I want to learn how to make the best of it.
This whole article felt AI written in terms of blandness. The writing is technically fine, though lacks a real viewpoint. The arguments are logically laid out and sound factual, though I have no idea how the author really feels. There’s zero emotion or heart or sense of a unique individual behind the words.
For me as an author that would never be good enough. For me as a reader, AI writing leads to endless retreads of prior works. Also, the training data limits the world experience and tastes of the AI generating the content. I’d rather experience the original work of humans than the bland regurgitation of a semblance of existing writing of dubious quality.
I see what you mean. It's not one thing in particular that stands out, it's that nothing stands out at all.
Although I have found that 'style transfer' techniques can make for much better output. Which makes sense to me: emotion and heart can be faked, humans do it all the time (think sociopaths, actors, liars).
Yes, emotion can be faked. As a writer I think it’s important to have a clear point of view. I can agree or not but the author puts down a line and says this is who I am. Your AI writing is neither exciting nor offensive. It’s informative in a bland museum curator way, though not memorable. It doesn’t make me feel anything and I feel that’s a sad statement. If you want real fans who will stick with you, you need to make them feel something.
You make a good point, and I think better prompts can make a world of difference. Let me illustrate:
ChatGPT, reword this piece of bland AI output into something much more personal in the provocative style of, for example, Piers Morgan: "Ultimately, the future of storytelling will likely involve a blend of human creativity and machine assistance. By embracing this partnership while remaining vigilant about ethical considerations, we can harness the potential of AI to enrich our narratives rather than diminish them."
"Let's face it, the future of storytelling is staring us right in the face - and it's a fusion of human genius and machine meddling. I'm not naive, I know the purists will scream bloody murder at the very thought of AI sullying the sacred art of storytelling. But I say, bring it on. Let's harness the power of these machines to inject some much-needed adrenaline into our narratives, to push the boundaries of what's possible, and to create something truly revolutionary. And to those who claim AI will somehow 'diminish' our stories, I say, get a grip. The only thing that's being diminished is the lazy, uninspired writing that's been peddled as 'art' for far too long. It's time to shake things up, and if AI is the catalyst, then so be it."
Thank you for hacking patterns. I agree that AI can be a partner, there's trust issues, though. I want to believe that AI will help humanity, but the inclination to mistrust AI has been pre-programned for generations.
But when have tech companies ever betrayed our trust? 😅
Very interesting article. It seems we forgot that we can make mistakes in our writing and it's ok to look as though we aren't perfect. I can say that early on I have used Chat to correct any errors that needed correcting. I don't think AI can write my content, or maybe it can. It is honestly something to be really aware of. Thank you for sharing this with us
😢. I would have been an A student back then if that was around. 😉But alas I struggled through school finding my own voice. So B/C student I was. Writing has been the biggest by far challenge for me.
Choosing to look like I’m still learning is going to be the biggest challenge still. 😉
Relatable.
Thanks! They say to err is human, and that's true now more than ever.
From what I hear, some of Substack's top publications as well as most new posts on LinkedIn are now made with AI. As is probably most schoolwork these days.