By continuing to use the site, you agree to our use of cookies and to abide by our Terms and Conditions. We in turn value your personal details in accordance with our Privacy Policy.
Please log in or register. Registered visitors get fewer ads.
Following my post at the bottom of the championship spending thread, I'd like to propose to Clive that he bans AI-generated content from the forum. It's just misinformation and it wastes everyone's time, while filling threads with slop.
What is AI-generated content? I've used LLM to generate images of the kits without sponsors and other kit designs. They're AI generated – are they banned?
What about an AI-enhanced post? Every major writing tool has some LLM bolted onto it now. Word's editor suggests rewrites, Google Docs has "help me write", and Grammarly's been running on machine learning for years. If someone runs their post through one of those before hitting submit, there's no way for a mod to know and no way for them to prove they didn't.
If the real complaint is wrong facts and thread-flooding, mod on that instead. A wrong fact gets corrected or pulled regardless of how it was written. Six near-identical posts in an hour are spam regardless of what produced them. You don't need to know if AI was involved to act on either, and you don't end up accusing someone of cheating because their English happens to read a bit different.
Ban "AI-generated content" and you've written a rule nobody can enforce fairly.
This is my wheelhouse right now - I've been working with Govt depts, banks, pharma, and third sector on how AI is impacting the work in orgs and it's a really interesting discussion. A blanket "no AI" rule looks decisive on paper and then doesn't survive contact with the low-risk uses nobody meant to catch: drafting aids, image generation, anything without a factual or decision-critical payload attached, e.g. grammar or spelling.
I had this with an org recently; their contract explicitly states no AI and I asked what AI - they knew what 'felt' wrong but couldn't explain it.
I mentioned this on a different thread. I'd rather have a discussion with errors. Recently, I was trawling through youtube looking some folky music. I came across Amelia - AI generated young lady singing British songs, sea shanties etc. I later found out it's put out by far right groups and even has Tommy Robinson in some. Anyone seen it ? It's weird and there seems to be hundreds of them.
What is AI-generated content? I've used LLM to generate images of the kits without sponsors and other kit designs. They're AI generated – are they banned?
What about an AI-enhanced post? Every major writing tool has some LLM bolted onto it now. Word's editor suggests rewrites, Google Docs has "help me write", and Grammarly's been running on machine learning for years. If someone runs their post through one of those before hitting submit, there's no way for a mod to know and no way for them to prove they didn't.
If the real complaint is wrong facts and thread-flooding, mod on that instead. A wrong fact gets corrected or pulled regardless of how it was written. Six near-identical posts in an hour are spam regardless of what produced them. You don't need to know if AI was involved to act on either, and you don't end up accusing someone of cheating because their English happens to read a bit different.
Ban "AI-generated content" and you've written a rule nobody can enforce fairly.
This is my wheelhouse right now - I've been working with Govt depts, banks, pharma, and third sector on how AI is impacting the work in orgs and it's a really interesting discussion. A blanket "no AI" rule looks decisive on paper and then doesn't survive contact with the low-risk uses nobody meant to catch: drafting aids, image generation, anything without a factual or decision-critical payload attached, e.g. grammar or spelling.
I had this with an org recently; their contract explicitly states no AI and I asked what AI - they knew what 'felt' wrong but couldn't explain it.
I think you're trying to bring a nuance that doesn't exist in this use case. Sure, in sectors where machine learning is important, or where AI generated content can be useful (again, coding) then this makes sense. But this is a small message board for talking about QPR, so obviously the issue is not enhanced grammar and spelling tools - if people want to use those, then that's obviously fine. That's not AI-generated content, however, since the core content was written by one of the people on this board and then edited with an LLM.
This issue is the large surge in posts going "I asked ChatGPT etc. and it said this...", where blocks of text are copied directly from LLM's without checking. This behaviour should be instantly banned. Text that is obviously from an LLM should also be banned ("it's not this, it's that"; long bulleted lists etc., reams of em-dashes).
In terms of misinformation policing, this is both harder to accomplish and also misses the point somewhat. Pulling every post that contains misinformation of any kind on this board would be impossible, since frankly we have a lot of "ITK" posters, people posting rumours etc. This makes it impossible to verify what is and isn't misinformation. I actually think this rule is harder to enforce than a "no-AI" rule, as a result. However, the AI slop posts are producing actually verifiable misinformation and it is slowly filling the board with crap. Crap that isn't even any person's opinion, but that of a machine asked to generate information based on training datasets that contain very little about QPR.
I think it's obvious that this could be managed case-by-case. Want to generate a kit design? Sure. Want to post a 900 word post about club economics with all the facts wrong? Not so much.
AI horse may have bolted to be fair… but at least check and edit what chatgpt sends you before posting and make it your own…. What’s the point otherwise?
I wanna hear from fellow Rs not a tech bro… now what’s Claude predicting for tonight?
"Travel is fatal to prejudice, bigotry and narrow-mindedness, and many of our people need it sorely on these accounts. Broad, wholesome, charitable views of men and things cannot be acquired by vegetating in one little corner of the Earth all one's lifetime." (Mark Twain)
Find me on twitter @derbyhoop and now on Bluesky
AI horse may have bolted to be fair… but at least check and edit what chatgpt sends you before posting and make it your own…. What’s the point otherwise?
I wanna hear from fellow Rs not a tech bro… now what’s Claude predicting for tonight?
I have a Claude betting agent that I built quite successfully during the World Cup and now using for the champo (the pricing is less tight than the prem)
It's not a bet as the real prices don't beat the expected value - but the data says both teams to score, over 2.5, Smyth shot on target
Feel like I’m living in a parallel universe. Am I the only one who doesn’t understand a word of this thread?
I asked claude to summarise it. It was my paid account though, so be careful about bias; it has taken a very positive view of my contributions:
TwoHalves, short version.
"LLM" means large language model. That's the tech behind ChatGPT, Claude and the rest. When people say "AI" on here, that's what they mean.
EuroRanger has asked Clive to ban AI-generated content on the board. What's set him off is the growing number of posts along the lines of "I asked ChatGPT and it said…", where someone pastes a wall of machine-written text straight into a thread without checking any of it.
The main threads of argument:
Where's the line? E17hoop points out that AI is now baked into Word, Google Docs and Grammarly, so almost any post could be "AI-assisted". He's also used it to mock up kit designs. EuroRanger's answer is that spellcheck isn't the issue: it's whole posts written by a machine and copied over. Could you even enforce it? derbyhoop reckons good luck policing it. E17hoop's worry is you end up accusing someone of cheating because their writing reads a bit oddly. EuroRanger thinks it's easier to spot than people assume, and that you'd handle it case by case. Ban the tool or ban the behaviour? E17hoop says moderate the actual problems, wrong facts and thread-flooding, and it doesn't matter what produced them. EuroRanger counters that policing misinformation generally is impossible on a board full of ITK rumours, whereas AI slop is verifiably wrong and isn't even anybody's genuine opinion.
For a ban: it fills threads with rubbish, the facts are often wrong, there's very little about QPR in the training data anyway, and it's not a fellow fan's view, it's a machine's.
Against a ban: nobody can define "AI content" precisely, nobody can prove it either way, innocent uses get caught, and the honest fix is to moderate bad posts on their merits.
Where most people seem to agree: nobody is defending unchecked copy-paste. vanrrrr's line sums it up, at least read and edit the thing and make it your own.
Also worth knowing: colinallcars brought up AI-generated folk singers being pushed by far-right accounts, QPROslo wants the whole lot switched off, and robith cheerfully posted his AI betting model's picks for tonight midway through the ban debate.
Feel like I’m living in a parallel universe. Am I the only one who doesn’t understand a word of this thread?
Love it....
My granddaughter is doing coding at Brunel uni... and at the weekend I was asking how she was getting in, having finished her first year. ..... did I not understand a word of what she had done and was going to do...
I'm an electrical engineer so I have I think a little grasp on technology... but this is absolutely mindblowing....what she is doing
I remember my old Granddad bless him... a miner all his life getting into my old Ford escort Mexico... where I fitted an 8 track player and him looking at it with utter bewilderment...ifeel the same
3
LLM Ban on LFW (n/t) on 14:07 - Sep 2 with 2649 views
My granddaughter is doing coding at Brunel uni... and at the weekend I was asking how she was getting in, having finished her first year. ..... did I not understand a word of what she had done and was going to do...
I'm an electrical engineer so I have I think a little grasp on technology... but this is absolutely mindblowing....what she is doing
I remember my old Granddad bless him... a miner all his life getting into my old Ford escort Mexico... where I fitted an 8 track player and him looking at it with utter bewilderment...ifeel the same
I posted about the marked differences between the PREMIER and CHAMPIONSHIP fair play rules.
Got them from the official websites.
Then posted my own ideas about the transfer window and the rules. The championship figures werent totslly up to date but near as damn it. Was not quite the end of the window, so thats where the minor inaccuracies were.
Posted as i felt it was an intetesting comparison with the reality being how unfair it could be.
If you are saying i shouldnt post anything with particular regard to this assessment of the ridiculousness of the divide and as serms there is a consensus, i wont post about certain topics anymore.
Talking and giving out inaccurate information. Who would have thought of such a thing on a football fans forum !!!
Will have to stop doing that. Sorry.
Thought that was one of the main qualifications to post on a football forum myself....
Great, question and I've give this a lot of thought and come to the conclusion that...
The Loft For Words (LFW) forum should ban AI-generated replies to protect its authentic fan culture. Football message boards thrive on raw human emotion, dark humour, and the shared lived experiences of QPR fans. AI-generated text is inherently sterile, overly polite, and neutral, which completely clashes with the passionate, opinionated nature of standard LFW debate.
If users suspect they are interacting with or arguing against an automated tool rather than a fellow supporter, trust will erode, and community engagement will sharply decline.Instead of a blanket ban on all technology, the forum should implement a targeted policy.
While using AI to draft opinions, match reactions, or arguments in the comment section should be strictly forbidden, exceptions can be made for factual research. For example, users could be allowed to paste AI-sourced historical data or statistics, provided the information is explicitly cited.
This targeted approach prevents low-effort spam and preserves the uniquely cynical, human voice that makes LFW a dedicated home for QPR fans.If you are planning to post this on the forum, let me know if you would like me to adjust the tone to fit the LFW board or if you need help drafting a direct message to the moderators.
Great, question and I've give this a lot of thought and come to the conclusion that...
The Loft For Words (LFW) forum should ban AI-generated replies to protect its authentic fan culture. Football message boards thrive on raw human emotion, dark humour, and the shared lived experiences of QPR fans. AI-generated text is inherently sterile, overly polite, and neutral, which completely clashes with the passionate, opinionated nature of standard LFW debate.
If users suspect they are interacting with or arguing against an automated tool rather than a fellow supporter, trust will erode, and community engagement will sharply decline.Instead of a blanket ban on all technology, the forum should implement a targeted policy.
While using AI to draft opinions, match reactions, or arguments in the comment section should be strictly forbidden, exceptions can be made for factual research. For example, users could be allowed to paste AI-sourced historical data or statistics, provided the information is explicitly cited.
This targeted approach prevents low-effort spam and preserves the uniquely cynical, human voice that makes LFW a dedicated home for QPR fans.If you are planning to post this on the forum, let me know if you would like me to adjust the tone to fit the LFW board or if you need help drafting a direct message to the moderators.
Yeah, that's what I think about it.
personally i do put a lot of thought and effort into my posts
sometimes they are not liked or agreed with and thats fine.
But the club is clearly using AI to source out new players with the right numbers and with the chances of securing the best players wirh the best chances of being very low cost. it helps them so for a further explanation as to rhe chasm in the transfer fair play between the premier.and so far their policy is working
anyhow my AI says the club are refunding my seaaon ticket and paying me per game to attend.
I think you're trying to bring a nuance that doesn't exist in this use case. Sure, in sectors where machine learning is important, or where AI generated content can be useful (again, coding) then this makes sense. But this is a small message board for talking about QPR, so obviously the issue is not enhanced grammar and spelling tools - if people want to use those, then that's obviously fine. That's not AI-generated content, however, since the core content was written by one of the people on this board and then edited with an LLM.
This issue is the large surge in posts going "I asked ChatGPT etc. and it said this...", where blocks of text are copied directly from LLM's without checking. This behaviour should be instantly banned. Text that is obviously from an LLM should also be banned ("it's not this, it's that"; long bulleted lists etc., reams of em-dashes).
In terms of misinformation policing, this is both harder to accomplish and also misses the point somewhat. Pulling every post that contains misinformation of any kind on this board would be impossible, since frankly we have a lot of "ITK" posters, people posting rumours etc. This makes it impossible to verify what is and isn't misinformation. I actually think this rule is harder to enforce than a "no-AI" rule, as a result. However, the AI slop posts are producing actually verifiable misinformation and it is slowly filling the board with crap. Crap that isn't even any person's opinion, but that of a machine asked to generate information based on training datasets that contain very little about QPR.
I think it's obvious that this could be managed case-by-case. Want to generate a kit design? Sure. Want to post a 900 word post about club economics with all the facts wrong? Not so much.
I don't disagree - and having moderated other boards it's a thankless task. As you say, misinformation has always been a pain to police, and has always been a case by case activity. What's more annoying now is people sticking up 400 words of crap rather than 40.
Great, question and I've give this a lot of thought and come to the conclusion that...
The Loft For Words (LFW) forum should ban AI-generated replies to protect its authentic fan culture. Football message boards thrive on raw human emotion, dark humour, and the shared lived experiences of QPR fans. AI-generated text is inherently sterile, overly polite, and neutral, which completely clashes with the passionate, opinionated nature of standard LFW debate.
If users suspect they are interacting with or arguing against an automated tool rather than a fellow supporter, trust will erode, and community engagement will sharply decline.Instead of a blanket ban on all technology, the forum should implement a targeted policy.
While using AI to draft opinions, match reactions, or arguments in the comment section should be strictly forbidden, exceptions can be made for factual research. For example, users could be allowed to paste AI-sourced historical data or statistics, provided the information is explicitly cited.
This targeted approach prevents low-effort spam and preserves the uniquely cynical, human voice that makes LFW a dedicated home for QPR fans.If you are planning to post this on the forum, let me know if you would like me to adjust the tone to fit the LFW board or if you need help drafting a direct message to the moderators.
Yeah, that's what I think about it.
Quillbot calls 11% of this AI. GPTZero says it's 100% AI. Pangram says this is 100% AI.
I asked claude to summarise it. It was my paid account though, so be careful about bias; it has taken a very positive view of my contributions:
TwoHalves, short version.
"LLM" means large language model. That's the tech behind ChatGPT, Claude and the rest. When people say "AI" on here, that's what they mean.
EuroRanger has asked Clive to ban AI-generated content on the board. What's set him off is the growing number of posts along the lines of "I asked ChatGPT and it said…", where someone pastes a wall of machine-written text straight into a thread without checking any of it.
The main threads of argument:
Where's the line? E17hoop points out that AI is now baked into Word, Google Docs and Grammarly, so almost any post could be "AI-assisted". He's also used it to mock up kit designs. EuroRanger's answer is that spellcheck isn't the issue: it's whole posts written by a machine and copied over. Could you even enforce it? derbyhoop reckons good luck policing it. E17hoop's worry is you end up accusing someone of cheating because their writing reads a bit oddly. EuroRanger thinks it's easier to spot than people assume, and that you'd handle it case by case. Ban the tool or ban the behaviour? E17hoop says moderate the actual problems, wrong facts and thread-flooding, and it doesn't matter what produced them. EuroRanger counters that policing misinformation generally is impossible on a board full of ITK rumours, whereas AI slop is verifiably wrong and isn't even anybody's genuine opinion.
For a ban: it fills threads with rubbish, the facts are often wrong, there's very little about QPR in the training data anyway, and it's not a fellow fan's view, it's a machine's.
Against a ban: nobody can define "AI content" precisely, nobody can prove it either way, innocent uses get caught, and the honest fix is to moderate bad posts on their merits.
Where most people seem to agree: nobody is defending unchecked copy-paste. vanrrrr's line sums it up, at least read and edit the thing and make it your own.
Also worth knowing: colinallcars brought up AI-generated folk singers being pushed by far-right accounts, QPROslo wants the whole lot switched off, and robith cheerfully posted his AI betting model's picks for tonight midway through the ban debate.
This reads like the weekly intro to Soap - one for the kids!
I left pretty much all social media because of bots and AI slop. Really don’t want to leave this as well (I’ll be left exclusively with QPR WhatsApp channel as engagement).
By continuing to use the site, you agree to our use of cookies and to abide by our Terms and Conditions.