I’d say your moat is your personal connection with your audience. People read your Substack because they want to hear what you have been thinking (as you have provided value to them in the past)
AI is a huge threat to writers - especially us content marketers. Because of the poor economy (huge inflation), all of my clients left the agency I work with. They went to AI. Every one of them. Why? Because it's cheaper. However, Google doesn't exactly like AI. No, it doesn't say it outright, but it does hint at it. I have one client that came back (but it's a small agency that doesn't have a lot of work). Maybe they figured out that AI wasn't cuttin' the mustard. ;)
Yes, it is certainly better here on Substack than on Medium in regards to AI content. I think this is reflective of the model - on Medium you can make a few bucks a week on sheer volume by mass-generating 10+ AI stories a day in publications that were already there, and that don't filter. Up until now, there hasn't been any sort of penalty for doing this.
Here on Substack, there is no incentive to do this - you can't make a penny by generating massive amounts of shitty click-baity AI content.
Like other authors who have commented here, even with 11,000+ followers on Medium, i am slowly migrating all my stuff over here to Substack.
Hello Alberto, I found you initially on Medium. 100% agree. Also, as a Medium writer myself (well I'm slowly disengaging from there), I noticed that Medium is lately flooded with AI-writing (especially after the MPP changes). What's your view on Medium-vs-Substack regarding AI-writing?
Support human writers and human writing. I don't think the gates of Substack are explicitly closed to AI generated articles, they just don't surface because the platform is not dominated by an algorithm that decides what appears on your feed. There's no way to SEO-hack your way to the top with AI garbage here. (Notes is different but I'm not sure I care as much about notes as about newsletters.)
I think "dislikes" are toxicity-inducing. If you don't like something, leave it alone. I think it's better. If no one likes it or reads it, then it may as well not exist anyway. That's fine to me.
But how do you know there isn't much AI-generated content here? Perhaps it's become so good that it's unrecognizable from human writing. Even if not, it surely seems to be headed that way, no?
It's *definitely* not that good yet (might be soon, though, I don't rule out that possibility). But I also mean that I kinda know the people who I encounter here and know they're not using AI. Of course that's anecdotal because I don't know everyone in here. I'm curious to know if that's everyone's experience or not!
> "How much AI-generated bullshit have you seen on Substack?"
Lots. At least in the tech-related segment of Substack every other image is some AI generated robot-touching-finger-with-human, or blue-glowy-brain, or room-full-of-enigmatic-baubles, or similar kind of klischeed nonsense. But I agree that writing seems mostly fine though.
I get that. Still, I think, it is a bit ironic that the shelter/castle/garden of human creativity shields one form of expression, but another is fair game. So much so that it is not even acknowledged that there is "AI-generated bullshit" in here - it's not recognized as it's not the kind we care about, apparently.
> "The threat is the unscrupulous who publish AI-generated [images], passing them as theirs because they don’t care enough about the beauty of human creativity."
Substack is the online shelter of written human creativity. ArtStation and DeviantArt should have been the digital bastions of visual human creativity. They failed to be.
The question is not how a piece of content was created, but rather did a consumer enjoy that content.
The current wave of hysteria that the newness of AI is creating will pass, because AI won't always be new. And when the newness of AI passes, the same old question that's been ruling the content world since the beginning will still be there. Did the consumer enjoy the content? Did they find it useful in some manner?
It seems near certain that AI development is going to continue, and Gen AI tools will continue to improve, whether that's a good idea or not, and no matter what anybody says about it. Facing that reality can help guide us on to a constructive course.
And so we human creators are, if you will, like the wild animals on the savannah who have to deal with an environment being altered by climate change. Some of the animals will successfully adapt to the changing conditions and thrive, and some will fail to adapt and die. The question each of us Substackers might be asking ourselves is, which one are we going to be?
Such discussion used to be just theory for me, an excuse to blowhard. But I've since rolled up my sleeves, accepted the challenge, and am now daily engaged in seeing what I can learn about the new environment that awaits us. Here's my latest attempt from yesterday. 100% AI content. I had the original idea, and did the video editing. Text, images and sound all courtesy of ChatGPT.
Not fully agree with this "The question is not how a piece of content was created, but rather did a consumer enjoy that content" but even if we accepted the premise, you seem to believe those two things are orthogonal. They aren't: there's an important part of enjoying creativity that's downstream from who created it and how it was created.
„there's an important part of enjoying creativity that's downstream from who created it and how it was created.“ -> behind-the-scenes-shots of Tom Cruise doing the motorcycle cliff jump. Nobody would care if it wasn’t for a real person taking risks
I’d say your moat is your personal connection with your audience. People read your Substack because they want to hear what you have been thinking (as you have provided value to them in the past)
AI is a huge threat to writers - especially us content marketers. Because of the poor economy (huge inflation), all of my clients left the agency I work with. They went to AI. Every one of them. Why? Because it's cheaper. However, Google doesn't exactly like AI. No, it doesn't say it outright, but it does hint at it. I have one client that came back (but it's a small agency that doesn't have a lot of work). Maybe they figured out that AI wasn't cuttin' the mustard. ;)
I am using AI to improve my writing on SubStack by asking Claude and Pi to simulate a personal editor.
Seems rash to throw out the baby with the bath water in this setting.
100% ai generated content does make me want to vomit though, lol.
Just beautiful Alberto!
Yes, it is certainly better here on Substack than on Medium in regards to AI content. I think this is reflective of the model - on Medium you can make a few bucks a week on sheer volume by mass-generating 10+ AI stories a day in publications that were already there, and that don't filter. Up until now, there hasn't been any sort of penalty for doing this.
Here on Substack, there is no incentive to do this - you can't make a penny by generating massive amounts of shitty click-baity AI content.
Like other authors who have commented here, even with 11,000+ followers on Medium, i am slowly migrating all my stuff over here to Substack.
Hello Alberto, I found you initially on Medium. 100% agree. Also, as a Medium writer myself (well I'm slowly disengaging from there), I noticed that Medium is lately flooded with AI-writing (especially after the MPP changes). What's your view on Medium-vs-Substack regarding AI-writing?
Hi Alberto,
I wish I could write well in your language but I cannot because mine is French.
I can understand yours: 99%!
And I like to vibrate in your wonderful world.
Best Regards
Jacques L.
Very well said. Are there any proactive steps that creators and readers can take to prevent AI-generated garbage slowly making its way onto Substack?
Support human writers and human writing. I don't think the gates of Substack are explicitly closed to AI generated articles, they just don't surface because the platform is not dominated by an algorithm that decides what appears on your feed. There's no way to SEO-hack your way to the top with AI garbage here. (Notes is different but I'm not sure I care as much about notes as about newsletters.)
Seems so simple when you put it that way, but perhaps thats all it is.
Would the platform benefit from some sort of 'thumbs down' that readers could use for articles which are obviously garbage? Might help as well.
I think "dislikes" are toxicity-inducing. If you don't like something, leave it alone. I think it's better. If no one likes it or reads it, then it may as well not exist anyway. That's fine to me.
Substack is amazing right now for that reason- the toxicity just isn’t here. I wouldn’t want to introduce anything to change that.
Alberto writes, "That’s false for one simple reason, best illustrated with a question: How much AI-generated bullshit have you seen on Substack?"
My blog is 95%+ AI generated, a service I provide for free to anyone who wishes to go hysterical. :-)
AI itself is the demon.
Sad, but true
People seem hesitant to face reality.
But how do you know there isn't much AI-generated content here? Perhaps it's become so good that it's unrecognizable from human writing. Even if not, it surely seems to be headed that way, no?
It's *definitely* not that good yet (might be soon, though, I don't rule out that possibility). But I also mean that I kinda know the people who I encounter here and know they're not using AI. Of course that's anecdotal because I don't know everyone in here. I'm curious to know if that's everyone's experience or not!
Hear! hear!
> "How much AI-generated bullshit have you seen on Substack?"
Lots. At least in the tech-related segment of Substack every other image is some AI generated robot-touching-finger-with-human, or blue-glowy-brain, or room-full-of-enigmatic-baubles, or similar kind of klischeed nonsense. But I agree that writing seems mostly fine though.
Oh, I mean the writing. After all, Substack is not a platform for visual artists, painters, or designers.
I get that. Still, I think, it is a bit ironic that the shelter/castle/garden of human creativity shields one form of expression, but another is fair game. So much so that it is not even acknowledged that there is "AI-generated bullshit" in here - it's not recognized as it's not the kind we care about, apparently.
> "The threat is the unscrupulous who publish AI-generated [images], passing them as theirs because they don’t care enough about the beauty of human creativity."
Substack is the online shelter of written human creativity. ArtStation and DeviantArt should have been the digital bastions of visual human creativity. They failed to be.
The question is not how a piece of content was created, but rather did a consumer enjoy that content.
The current wave of hysteria that the newness of AI is creating will pass, because AI won't always be new. And when the newness of AI passes, the same old question that's been ruling the content world since the beginning will still be there. Did the consumer enjoy the content? Did they find it useful in some manner?
It seems near certain that AI development is going to continue, and Gen AI tools will continue to improve, whether that's a good idea or not, and no matter what anybody says about it. Facing that reality can help guide us on to a constructive course.
And so we human creators are, if you will, like the wild animals on the savannah who have to deal with an environment being altered by climate change. Some of the animals will successfully adapt to the changing conditions and thrive, and some will fail to adapt and die. The question each of us Substackers might be asking ourselves is, which one are we going to be?
Such discussion used to be just theory for me, an excuse to blowhard. But I've since rolled up my sleeves, accepted the challenge, and am now daily engaged in seeing what I can learn about the new environment that awaits us. Here's my latest attempt from yesterday. 100% AI content. I had the original idea, and did the video editing. Text, images and sound all courtesy of ChatGPT.
https://hippytoons.com/p/a-hippytoons-academic-explores-the
It's not incredibly great, but neither is it crap. And there's no much yet to learn.
Not fully agree with this "The question is not how a piece of content was created, but rather did a consumer enjoy that content" but even if we accepted the premise, you seem to believe those two things are orthogonal. They aren't: there's an important part of enjoying creativity that's downstream from who created it and how it was created.
„there's an important part of enjoying creativity that's downstream from who created it and how it was created.“ -> behind-the-scenes-shots of Tom Cruise doing the motorcycle cliff jump. Nobody would care if it wasn’t for a real person taking risks