You said Medium doesn’t allow AI behind the paywall but also a huge amount of the work there is AI. Can you clarify what you mean? They are checking some writing but not all of it? I thought you had to pay to read anything on medium. But maybe you mean they’re just not paying the AI writers? I don’t know medium as well so that’s on me but I just wanna be clear. It’s all really difficult and disheartening her writers everywhere and I appreciate your articles For every bit of attention they bring to this new world.
Medium's policy says they allow AI writing but (a) it must be disclosed and (b) it may not be put behind the paywall. So anyone using AI is supposed to disclose and they are not permitted to earn money for those stories. However, people are not disclosing use of AI and they are putting their AI stories behind the paywall.
Medium is just a small team and don't have the resources to check all content because the volume going up is staggering. They are trying to catch AI, at least so they say, but it's no small job. And they have tools in place to allow readers to report an AI generated piece behind the paywall and do take them down. I report AI writing almost daily there.
But the AI writers who don't disclose and put content behind the paywall aren't always caught. I would say the few who get caught are some tiny minority. A lot of them get away with it and it depletes the pay pot for human writers. Did that help?
OK, as soon as you read that, it seems super clear but I appreciate you explaining it because I couldn’t wrap my head around it. And of course they’re not gonna disclose it because they just wanna make the money. That’s gross. I All humans could be more honest, but I suppose we just don’t work that way. Sigh.
I am so surprised by those figures! That is scary that it is a such a high percentage of Ai posts on the internet. I have been noticing it too with images. Many smaller-medium sized companies are now using Ai and no one can seem to tell. As an artist I spot them straight away, but they are comical, either too cartoony with bizzare hands or else far too perfects to be real. I guess education of the public is key, but soon these Ai generated images and texts will be too good for anyone to be able to tell, so what then!?
It is terrible, for sure. According to people who work in the industry, it will eventually start to degrade. Once they exhaust the human work they've been training on, they will have to start training on AI. And once that happens, it will begin to degrade. I hope they are right. Can't happen too soon for me
That is interesting, I hope that is true. If anything, it will never be capable of complex story yelling, because it doesn’t have the comprehension of thoughts and feelings. Like all things new and shiny, I think the excitement will soon wear off and people will be craving real human connection.
"Here’s the irony —the algorithm doesn’t know readers are chastising the writer. The algorithm just sees engagement. So it sees all that engagement and pushes the post out to more people. Who also engage.
So it goes, as Vonnegut liked to say."
One of the writers I follow on Notes yesterday that they are giving up publishing there because the algorithm is horrendously biased against "quality" content in favor of stuff like this. And they do have a point, since I don't get the kind of traction on my posts there I would like to have- and since I mainly use Notes to promote things I have written on my newsletter...
In the 20th century, the American writer Edward Abbey wrote resentfully about how mechanical and urban culture were expanding into previously pristine rural and wilderness areas and ruining them. He advocated using industrial sabotage as a means of counteracting this. (This is the subject of his novel "The Monkey Wrench Gang"). I am starting to believe the only way human writers can fairly compete with AI is to "sabotage" in ways similar to Abbey's- though more legal. (e.g. driving the companies to the brink of bankruptcy by boycotting them and calling them out, a la Tesla).
Sam Altman only got where he is now because he thwarted an attempted coup against his leadership and drove away the people who opposed him. The key is to bring those people back, everywhere...
I must now read The Monkey Wrench Gang. I cannot tell you how many times my husband has suggested we become corporate saboteurs. We can start with insurance companies, then utilities, then, well, we'll take a poll.
Right? That's the part that ate my brain. Why are we so damn gullible? Shouldn't we know better by now? God, I'm kvetching at you in two separate places lol.
I'd also note that people who post about "real writing" often do so as engagement bait. The same is true for those who "want to hear from the smaller writers," imploring them to share their work and promising to follow the interesting ones.
The other thing flaring up again is the call for Substack to bundle subscriptions or adopt a revenue model similar to Medium's. They might not realize that a flood of AI slop and people gaming the system via volume would be sure to follow in short order.
Omg, good point on the engagement bait, Kevin. And yeah - absolutely agree - if Substack adopted a revenue model like Medium's it would be flooded with AI instantly, just like Medium is. There's already a lot here. It's just not in my pocket like it is on Medium
Absolutely, we must become far less credulous as readers. I mean, seriously, Keanu Reeves?? That should've been a red flag right there. I remember years ago when "Keanu Reeves" wanted to friend me on Facebook.
Yea to being better readers. The good story here, I believe, is ai is requiring us to step up at being better humans. To develop our craft of humaning, and while painful now it will pay well over time 💜
I don’t know how I would recognize AI unless it would be to suspect the several-posts-a-day or every-day post contributors.. At this point it’s difficult to even read something from a new writer because I’m so busy reading and supporting those writers I admire and writing my own once-a-week post. At least I know they’re not AI.
That's it for me too, Patricia. Volume. When I stumble across a new writer I always go to their profile first. If I see multiple posts/day I test for AI. lol
You say you test for AI when you see multiple posts. I'm trying to find a scheduling program that will post for me on IG, (posts and photos I've made) although I'm less and less fond of anything associated with FB. How do you test for AI? That could be a new lesson you teach us, unless you've already done so.
Volume is the crux of the issue -- like a Catch 22. The algorithms are meant to better surface "content of interest" because there is too much content... and no one has enough time to read it... so we need a means of culling it down to what would really be of interest... if the algorithm is not the solution -- because the algorithm cannot discern a positive interaction from a negative interaction -- then it's arguable what could be used... besides you yourself opting to read less.
I only subscribe to 23 publications, 2 of which are my husband's, and I can't even keep up with reading all their new posts! Let alone going out and reading new people. That's where I'm quite forgiving that barely anyone seems to read my work. It's sad that AI is taking away potential readers... but the ones who are able to spot it are probably already packed to the gills with other great writing!
I write a daily post. I wouldn’t have the faintest idea how to use AI. I’ve been writing every day for 70 years (I’m 85). It just takes practice. I could write two but I just don’t want to.
Just because it shows up on the internet don’t make it true. Discerning readers are harder and harder to find. It’s bizarre what’s going on right now. As a writer and a reader, I take my responsibility to be authentic seriously. Competing with AI is not sustainable. We must keep on keeping on, writing our best work, supporting other writers who match authenticity with authenticity.
Thank for shining your light on the truth once again, Lynda.
Right? You'd think by 2025 we'd know not everything online is true. But, but, but, there was a picture. Yes, but the picture was ai generated. What a mess we've created
I think I can spot AI writing, but sounds like I need to be even more astute! I saw that article about the Keanu Reeves’s video, it felt like AI because the writing was hyperbolic to the point of being cheesy, and was easy to debunk because the video itself was nowhere to be found. But others might not be so obvious. Thank you for writing about this!.
After I read the snopes piece I looked at the video. I didn't watch the whole thing, just a couple of seconds. It was so horribly bad I'm not surprised he didn't include it. The bad AI writing was better than the even worse AI video
Absolutely great. I think what we need is to develop criteria for what we're reading but also, restrain from reading things just to react. I think it's like, developing restrain for rage bait.
You've given me an idea for a prayer. In order to resist rage bait RRB, I'm going to pay a prayer I heard half of growing up. My folks often said " Jesus, Mary, and Joseph" when they were exasperated with us six kids. It always seemed to me that they must have said a prayer silently after that outburst, otherwise they'd be breaking a commandment by taking the Lord's name in vain. So my prayer will now be: Jesus, Mary, and Joseph, save me from rage bait and the snares of the devil.
Thank you for this - I had no idea how prevalent AI has become. What's scarier is the plan to use AI to teach children. We have outsourced much of our lives, fully embracing our "purpose" as consumers and data products. RESIST!
Thank you for bringing Vonnegut into my morning. The day just got better!
And you followed his rule, too. When something is nice for the love of god, take a minute to say if this isn't nice, I don't know what is. lol
Very well said! In the age of algorithms, the best form of protest is lack of engagement.
That's it exactly, really well said. The best form of protest is lack of engagement -- I love that.
Exactly. But unfortunately our society keep falling to the same trap and make stupid people famous.
You said Medium doesn’t allow AI behind the paywall but also a huge amount of the work there is AI. Can you clarify what you mean? They are checking some writing but not all of it? I thought you had to pay to read anything on medium. But maybe you mean they’re just not paying the AI writers? I don’t know medium as well so that’s on me but I just wanna be clear. It’s all really difficult and disheartening her writers everywhere and I appreciate your articles For every bit of attention they bring to this new world.
Medium's policy says they allow AI writing but (a) it must be disclosed and (b) it may not be put behind the paywall. So anyone using AI is supposed to disclose and they are not permitted to earn money for those stories. However, people are not disclosing use of AI and they are putting their AI stories behind the paywall.
Medium is just a small team and don't have the resources to check all content because the volume going up is staggering. They are trying to catch AI, at least so they say, but it's no small job. And they have tools in place to allow readers to report an AI generated piece behind the paywall and do take them down. I report AI writing almost daily there.
But the AI writers who don't disclose and put content behind the paywall aren't always caught. I would say the few who get caught are some tiny minority. A lot of them get away with it and it depletes the pay pot for human writers. Did that help?
OK, as soon as you read that, it seems super clear but I appreciate you explaining it because I couldn’t wrap my head around it. And of course they’re not gonna disclose it because they just wanna make the money. That’s gross. I All humans could be more honest, but I suppose we just don’t work that way. Sigh.
I am so surprised by those figures! That is scary that it is a such a high percentage of Ai posts on the internet. I have been noticing it too with images. Many smaller-medium sized companies are now using Ai and no one can seem to tell. As an artist I spot them straight away, but they are comical, either too cartoony with bizzare hands or else far too perfects to be real. I guess education of the public is key, but soon these Ai generated images and texts will be too good for anyone to be able to tell, so what then!?
It is terrible, for sure. According to people who work in the industry, it will eventually start to degrade. Once they exhaust the human work they've been training on, they will have to start training on AI. And once that happens, it will begin to degrade. I hope they are right. Can't happen too soon for me
Like some strange version of Darwinian natural selection... But for both of us... AI trained on AI and humans trained on AI....
That is interesting, I hope that is true. If anything, it will never be capable of complex story yelling, because it doesn’t have the comprehension of thoughts and feelings. Like all things new and shiny, I think the excitement will soon wear off and people will be craving real human connection.
I agree, Ashley, I hope we're right that the AI excitement will wear off. And I love your typo: "story yelling," lol. I'm going to use that soon.
😂 oh my, my predictive text in my phone is unhinged! It gives my friends a good laugh daily though!
"Here’s the irony —the algorithm doesn’t know readers are chastising the writer. The algorithm just sees engagement. So it sees all that engagement and pushes the post out to more people. Who also engage.
So it goes, as Vonnegut liked to say."
One of the writers I follow on Notes yesterday that they are giving up publishing there because the algorithm is horrendously biased against "quality" content in favor of stuff like this. And they do have a point, since I don't get the kind of traction on my posts there I would like to have- and since I mainly use Notes to promote things I have written on my newsletter...
In the 20th century, the American writer Edward Abbey wrote resentfully about how mechanical and urban culture were expanding into previously pristine rural and wilderness areas and ruining them. He advocated using industrial sabotage as a means of counteracting this. (This is the subject of his novel "The Monkey Wrench Gang"). I am starting to believe the only way human writers can fairly compete with AI is to "sabotage" in ways similar to Abbey's- though more legal. (e.g. driving the companies to the brink of bankruptcy by boycotting them and calling them out, a la Tesla).
Sam Altman only got where he is now because he thwarted an attempted coup against his leadership and drove away the people who opposed him. The key is to bring those people back, everywhere...
I love the idea of bringing those people back. How to do that, I'm not sure. But it's getting really stupid and why? Corporate profit. No other reason
Hayduke Lives!
I must now read The Monkey Wrench Gang. I cannot tell you how many times my husband has suggested we become corporate saboteurs. We can start with insurance companies, then utilities, then, well, we'll take a poll.
100% agree is what I think. You're right. There's no excuse for restacking a piece that can be debunked in 30 seconds.
Right? That's the part that ate my brain. Why are we so damn gullible? Shouldn't we know better by now? God, I'm kvetching at you in two separate places lol.
Well said!
I'd also note that people who post about "real writing" often do so as engagement bait. The same is true for those who "want to hear from the smaller writers," imploring them to share their work and promising to follow the interesting ones.
The other thing flaring up again is the call for Substack to bundle subscriptions or adopt a revenue model similar to Medium's. They might not realize that a flood of AI slop and people gaming the system via volume would be sure to follow in short order.
Omg, good point on the engagement bait, Kevin. And yeah - absolutely agree - if Substack adopted a revenue model like Medium's it would be flooded with AI instantly, just like Medium is. There's already a lot here. It's just not in my pocket like it is on Medium
Absolutely, we must become far less credulous as readers. I mean, seriously, Keanu Reeves?? That should've been a red flag right there. I remember years ago when "Keanu Reeves" wanted to friend me on Facebook.
Omg Jan, I am so jealous. Keanu never wanted to friend me and I'm Canadian like him. lmao.
He must have a thing for Muricans :-)
Yea to being better readers. The good story here, I believe, is ai is requiring us to step up at being better humans. To develop our craft of humaning, and while painful now it will pay well over time 💜
I agree, Kev. We need to step up at humaning. Great way to put it
I don’t know how I would recognize AI unless it would be to suspect the several-posts-a-day or every-day post contributors.. At this point it’s difficult to even read something from a new writer because I’m so busy reading and supporting those writers I admire and writing my own once-a-week post. At least I know they’re not AI.
That's it for me too, Patricia. Volume. When I stumble across a new writer I always go to their profile first. If I see multiple posts/day I test for AI. lol
This is very helpful, thank you Linda.
You say you test for AI when you see multiple posts. I'm trying to find a scheduling program that will post for me on IG, (posts and photos I've made) although I'm less and less fond of anything associated with FB. How do you test for AI? That could be a new lesson you teach us, unless you've already done so.
Volume is the crux of the issue -- like a Catch 22. The algorithms are meant to better surface "content of interest" because there is too much content... and no one has enough time to read it... so we need a means of culling it down to what would really be of interest... if the algorithm is not the solution -- because the algorithm cannot discern a positive interaction from a negative interaction -- then it's arguable what could be used... besides you yourself opting to read less.
I only subscribe to 23 publications, 2 of which are my husband's, and I can't even keep up with reading all their new posts! Let alone going out and reading new people. That's where I'm quite forgiving that barely anyone seems to read my work. It's sad that AI is taking away potential readers... but the ones who are able to spot it are probably already packed to the gills with other great writing!
I'd fail at the volume hurdle. I've only got time to write one post a week. No wonder I'm only at 28 subs after 2 months
I write a daily post. I wouldn’t have the faintest idea how to use AI. I’ve been writing every day for 70 years (I’m 85). It just takes practice. I could write two but I just don’t want to.
Just because it shows up on the internet don’t make it true. Discerning readers are harder and harder to find. It’s bizarre what’s going on right now. As a writer and a reader, I take my responsibility to be authentic seriously. Competing with AI is not sustainable. We must keep on keeping on, writing our best work, supporting other writers who match authenticity with authenticity.
Thank for shining your light on the truth once again, Lynda.
Right? You'd think by 2025 we'd know not everything online is true. But, but, but, there was a picture. Yes, but the picture was ai generated. What a mess we've created
I think I can spot AI writing, but sounds like I need to be even more astute! I saw that article about the Keanu Reeves’s video, it felt like AI because the writing was hyperbolic to the point of being cheesy, and was easy to debunk because the video itself was nowhere to be found. But others might not be so obvious. Thank you for writing about this!.
After I read the snopes piece I looked at the video. I didn't watch the whole thing, just a couple of seconds. It was so horribly bad I'm not surprised he didn't include it. The bad AI writing was better than the even worse AI video
Absolutely great. I think what we need is to develop criteria for what we're reading but also, restrain from reading things just to react. I think it's like, developing restrain for rage bait.
Oh man, I love that Carlos. Yes, we need to learn to resist rage bait. It's not easy. But falling for it doesn't go anywhere good
Oh no, it's not easy at all and I still fall for it! But I think we owe it to ourselves to keep trying! 😅
I have mixed feelings with all the AI debate, but I think we owe it to ourselves to keep developing our critical thinking as we engage with it.
It pains me that we're so easy to distract from doing things we love because we're reacting to something that bothers us.
You've given me an idea for a prayer. In order to resist rage bait RRB, I'm going to pay a prayer I heard half of growing up. My folks often said " Jesus, Mary, and Joseph" when they were exasperated with us six kids. It always seemed to me that they must have said a prayer silently after that outburst, otherwise they'd be breaking a commandment by taking the Lord's name in vain. So my prayer will now be: Jesus, Mary, and Joseph, save me from rage bait and the snares of the devil.
Thank you for this - I had no idea how prevalent AI has become. What's scarier is the plan to use AI to teach children. We have outsourced much of our lives, fully embracing our "purpose" as consumers and data products. RESIST!
Right? Omg, I'm so glad my kid is done school. I truly am.
Thanks for sharing so many valuable points!
Poo-tee-weet?