- Blind mode tutorial
lichess.org
Donate

The Decline of Quality Chess Content on Lichess: How AI is Flooding Our Blogs

#2

What do you think? Have you noticed this trend?

What do you think? Have you noticed this trend?
#3

Not sure blogs were the best idea to begin with... there was lots of low quality BS before AI ever showed up. But you're right, it's a worrying trend.

Not sure blogs were the best idea to begin with... there was lots of low quality BS before AI ever showed up. But you're right, it's a worrying trend.
#4

I was almost an adult when the Internet made its appearance in my country and I had to learn to navigate it as a strange new land with unknown rules. I've learned about spam, clickbait and so on very early on because, guess what, the Internet was full of that even back then, with no software or laws or rules to help you either. The only thing that made it better was that people haven't managed to fully monetize the Internet, so spam was not an industry yet.

So my brain learned to recognize patterns, as in chess, of what spam is or is not. You might find it funny to know that most "modern design" pages I see now trigger that sense of spam or that I have difficulty finding download buttons if they are too large, bright or colorful, because I am looking for that inconspicuous "Download" text and so on.

Why am I telling you this? Because the same pattern repeated again and again and again. And the reactions were identical, too. People complaining about the death of an era. Every death is a rebirth on the Internet. Your brain will learn to detect AI slop or - that would be the happier outcome, maybe - AIs will get so good that you won't be able to tell the difference.

Clickbait only works because you are willing to be baited. It is very annoying in any case, I agree, but when people learn en masse to recognize and reject clickbait, it will just become another genre that certain people will enjoy and most people will stop bothering with. Spam will always be there, AI generated or not. It's more annoying when it's generated by human beings because you know they could have done better, but were too dumb or disinterested to do it.

As for the Lichess blog and forum... I've seen a lot of people, including senior Lichess developers, complaining about the onslaught of bad content. I will let you know that I check the chess blogs ordered by date from time to time and choose what to read, as well as look in the Timeline to check what my friends liked. I see something worth reading every 5-10 posts, and the others I rarely even need to open to see I would not enjoy them. Same for the forums.

Oh, and most of it is just normal people, children or dumbasses that just can't figure out what they are doing is public or subject to a little bit of self-censorship. Why would people even spam Lichess? Everything is free! No one pays for anything. Attention-seeking aside, there is no profit in spamming.

So while agree with some of your points, I would have to say that your blog feels just as inaccurate and spammy as the things it accuses. I am sure it was not your intention, but what is an emotional blog about the terrible deluge of spam if not clickbait?

I was almost an adult when the Internet made its appearance in my country and I had to learn to navigate it as a strange new land with unknown rules. I've learned about spam, clickbait and so on very early on because, guess what, the Internet was full of that even back then, with no software or laws or rules to help you either. The only thing that made it better was that people haven't managed to fully monetize the Internet, so spam was not an industry yet. So my brain learned to recognize patterns, as in chess, of what spam is or is not. You might find it funny to know that most "modern design" pages I see now trigger that sense of spam or that I have difficulty finding download buttons if they are too large, bright or colorful, because I am looking for that inconspicuous "Download" text and so on. Why am I telling you this? Because the same pattern repeated again and again and again. And the reactions were identical, too. People complaining about the death of an era. Every death is a rebirth on the Internet. Your brain will learn to detect AI slop or - that would be the happier outcome, maybe - AIs will get so good that you won't be able to tell the difference. Clickbait only works because you are willing to be baited. It is very annoying in any case, I agree, but when people learn en masse to recognize and reject clickbait, it will just become another genre that certain people will enjoy and most people will stop bothering with. Spam will always be there, AI generated or not. It's more annoying when it's generated by human beings because you know they could have done better, but were too dumb or disinterested to do it. As for the Lichess blog and forum... I've seen a lot of people, including senior Lichess developers, complaining about the onslaught of bad content. I will let you know that I check the chess blogs *ordered by date* from time to time and choose what to read, as well as look in the Timeline to check what my friends liked. I see something worth reading every 5-10 posts, and the others I rarely even need to open to see I would not enjoy them. Same for the forums. Oh, and most of it is just normal people, children or dumbasses that just can't figure out what they are doing is public or subject to a little bit of self-censorship. Why would people even spam Lichess? Everything is free! No one pays for anything. Attention-seeking aside, there is no profit in spamming. So while agree with some of your points, I would have to say that your blog feels just as inaccurate and spammy as the things it accuses. I am sure it was not your intention, but what is an emotional blog about the terrible deluge of spam if not clickbait?
#5

you wrote in your article: "Demand citations—if a post makes bold claims, ask for proof."

AI can give sources.
The problem isnt so much AI-help when writing stuff, but the users of AI and their attention seeking. AI on its own does not write articles (as of today) in lichess forum. You need some users to deliberate "farm" attention.
If some professional likes to improve the things he want to say (and english is a foreign language to him), I can not see any harm from AI help. Rather the AI improves and reduces the amount of words needed to express the ideas of the professional.

But ye, there is some point, where users of AI get lazy, do not read through what the AI has produced (according to the input of the user) and it is then thrown into forums and articles without the user having read critically through the AI improvements. Nor was it ever the intention of the user to write a good article. So in some way, its trolling, and trolling is on the rise, and AI enables trolls to spam more articles with Pseudo information.

I have not written this with an AI, but I do copy the above and let AI sum up what I just said. I read through it, and yes, it says in less words what I said above, so I think it is ok to use these AI generated words:

chat gpt 4.1: "AI itself isn’t the core problem—it’s how people use it. While AI can provide sources and help non-native speakers communicate ideas more clearly, issues arise when users post AI-generated content without reviewing or caring about quality, often just to seek attention. This leads to low-effort, pseudo-information flooding forums. The real issue is not AI, but users who misuse it for trolling or spamming, rather than meaningful contribution."

you wrote in your article: "Demand citations—if a post makes bold claims, ask for proof." AI can give sources. The problem isnt so much AI-help when writing stuff, but the users of AI and their attention seeking. AI on its own does not write articles (as of today) in lichess forum. You need some users to deliberate "farm" attention. If some professional likes to improve the things he want to say (and english is a foreign language to him), I can not see any harm from AI help. Rather the AI improves and reduces the amount of words needed to express the ideas of the professional. But ye, there is some point, where users of AI get lazy, do not read through what the AI has produced (according to the input of the user) and it is then thrown into forums and articles without the user having read critically through the AI improvements. Nor was it ever the intention of the user to write a good article. So in some way, its trolling, and trolling is on the rise, and AI enables trolls to spam more articles with Pseudo information. I have not written this with an AI, but I do copy the above and let AI sum up what I just said. I read through it, and yes, it says in less words what I said above, so I think it is ok to use these AI generated words: chat gpt 4.1: "AI itself isn’t the core problem—it’s how people use it. While AI can provide sources and help non-native speakers communicate ideas more clearly, issues arise when users post AI-generated content without reviewing or caring about quality, often just to seek attention. This leads to low-effort, pseudo-information flooding forums. The real issue is not AI, but users who misuse it for trolling or spamming, rather than meaningful contribution."
#6

Besides, your own article is AI-improved. You can see this by Chat GPT typical use of "—" (thinking dash, without a space between words). Example from your article: "The rise of AI-generated articles—shallow, repetitive, and often factually dubious—has"

By comparison, a human would write: "The rise of AI-generated articles - shallow, repetitive, and often factually dubious - has"

I would not even know how to get a longer "—" (minus sign) on my keyboard. It is a typical feature of Chat-GPT.

Besides, your own article is AI-improved. You can see this by Chat GPT typical use of "—" (thinking dash, without a space between words). Example from your article: "The rise of AI-generated articles—shallow, repetitive, and often factually dubious—has" By comparison, a human would write: "The rise of AI-generated articles - shallow, repetitive, and often factually dubious - has" I would not even know how to get a longer "—" (minus sign) on my keyboard. It is a typical feature of Chat-GPT.
#7

Thank you for pointing that out, and I’m glad you noticed this distinctive feature! It's a good reminder of how AI can be recognized by these small yet unique markers in its writing.

Thank you for pointing that out, and I’m glad you noticed this distinctive feature! It's a good reminder of how AI can be recognized by these small yet unique markers in its writing.
#8

I notice that this article uses an AI generated image as its header... what is your intent?

Also, your previous post uses AI generated images, and ends with this...:
"In conclusion, modern AI tools have revolutionized chess analysis, making it more accessible and impactful than ever. By combining self-reflection, engine-assisted reviews, and targeted training, you can transform weaknesses into strengths and accelerate your growth. What’s the biggest challenge you face in analyzing your games? Share your thoughts, and I’ll offer tailored guidance to help you excel!"

If you are against the use of AI I fully support that (and this article was a good start) but I hope you can take your own advice as well

I notice that this article uses an AI generated image as its header... what is your intent? Also, your previous post uses AI generated images, and ends with this...: "In conclusion, modern AI tools have revolutionized chess analysis, making it more accessible and impactful than ever. By combining self-reflection, engine-assisted reviews, and targeted training, you can transform weaknesses into strengths and accelerate your growth. What’s the biggest challenge you face in analyzing your games? Share your thoughts, and I’ll offer tailored guidance to help you excel!" If you are against the use of AI I fully support that (and this article was a good start) but I hope you can take your own advice as well
#9

@insipidlight said in #8:

I notice that this article uses an AI generated image as its header... what is your intent?

Also, your previous post uses AI generated images, and ends with this...:
"In conclusion, modern AI tools have revolutionized chess analysis, making it more accessible and impactful than ever. By combining self-reflection, engine-assisted reviews, and targeted training, you can transform weaknesses into strengths and accelerate your growth. What’s the biggest challenge you face in analyzing your games? Share your thoughts, and I’ll offer tailored guidance to help you excel!"

If you are against the use of AI I fully support that (and this article was a good start) but I hope you can take your own advice as well
Maybe the author was intending to be ironic! :)

@insipidlight said in #8: > I notice that this article uses an AI generated image as its header... what is your intent? > > Also, your previous post uses AI generated images, and ends with this...: > "In conclusion, modern AI tools have revolutionized chess analysis, making it more accessible and impactful than ever. By combining self-reflection, engine-assisted reviews, and targeted training, you can transform weaknesses into strengths and accelerate your growth. What’s the biggest challenge you face in analyzing your games? Share your thoughts, and I’ll offer tailored guidance to help you excel!" > > If you are against the use of AI I fully support that (and this article was a good start) but I hope you can take your own advice as well Maybe the author was intending to be ironic! :)
#10

@Munich said in #6:

Besides, your own article is AI-improved. You can see this by Chat GPT typical use of "—" (thinking dash, without a space between words). Example from your article: "The rise of AI-generated articles—shallow, repetitive, and often factually dubious—has"

By comparison, a human would write: "The rise of AI-generated articles - shallow, repetitive, and often factually dubious - has"

I would not even know how to get a longer "—" (minus sign) on my keyboard. It is a typical feature of Chat-GPT.

I use the longer em-dash all the time. Google docs and a few other word processors that I use will convert a few - in a row to a longer one, like -- turns into — once you hit spacebar. I wouldn't use that as a litmus test of AI generation, personally. Maybe if it's combined with several other things that raise red flags, but on its own it really isn't enough. Chat-GPT is just mimicking what other people have written, after all, so it learned to use that dash from people actually using it online.

I'm raising this point because in the art field and among crafters (I follow a lot of knitting and crochet content), people are becoming hypervigilant about AI generated images, and it has led to artists being essentially bullied off the internet because someone said their art was AI generated when it wasn't. Sometimes we think it's glaringly obvious when something is AI, and sometimes that thing is really just photoshopped or badly drawn.

@Munich said in #6: > Besides, your own article is AI-improved. You can see this by Chat GPT typical use of "—" (thinking dash, without a space between words). Example from your article: "The rise of AI-generated articles—shallow, repetitive, and often factually dubious—has" > > By comparison, a human would write: "The rise of AI-generated articles - shallow, repetitive, and often factually dubious - has" > > I would not even know how to get a longer "—" (minus sign) on my keyboard. It is a typical feature of Chat-GPT. I use the longer em-dash all the time. Google docs and a few other word processors that I use will convert a few - in a row to a longer one, like -- turns into — once you hit spacebar. I wouldn't use that as a litmus test of AI generation, personally. Maybe if it's combined with several other things that raise red flags, but on its own it really isn't enough. Chat-GPT is just mimicking what other people have written, after all, so it learned to use that dash from people actually using it online. I'm raising this point because in the art field and among crafters (I follow a lot of knitting and crochet content), people are becoming hypervigilant about AI generated images, and it has led to artists being essentially bullied off the internet because someone said their art was AI generated when it wasn't. Sometimes we think it's glaringly obvious when something is AI, and sometimes that thing is really just photoshopped or badly drawn.