I got a complaint that this standard message is too harsh. It is worded in a somewhat cross tone
If you suspect your opponent of using AI, please tell us and not them. When actual cheaters are warned about suspicion, they change their behavior, it makes it harder to detect. Also, accusations of AI use feel like harassment, which is not allowed. Please finish the game the best you can and report it - this provides the best evidence, and we will investigate
You all are good at wordsmithing - can you let me know what the words should be, and Iâll update the template.
If you suspect your opponent of using AI, please inform us directly rather than confronting them. Alerting potential cheaters can lead to altered behavior, complicating our detection efforts. Additionally, accusations of AI use may be perceived as harassment, which is prohibited. We encourage you to complete the game to the best of your ability and submit a report, as this will provide the most effective evidence for our investigation.
I think this has the same problem, of accusing the player of harassment, which ironically, by the overly broad definition implied here, may itself feel like harassment
STOP! Donât even think about accusing your opponent of using AI! Youâre not a detective, and youâll only mess up our investigation. Harassing or falsely accusing others is a serious offense, and we wonât stand for it. Keep playing, gather evidence, and report it to us. Weâll handle it. Accusing someone without proof will get YOU in trouble, not them. Donât try to play vigilante - just play the game and let us do our job!
edit: although to be fair many people are detectives, and do provide great assistance in weeding out AI cheating. Itâs not the detectiving thatâs the problem, itâs the public accusationsâŚ
I dunno, the statement âmay be perceivedâ is not an accusation, itâs an explanation.
âProhibitedâ is an inopportune word though.
Explaining to people that âsaying âhey, youâre cheating with AIââ is perceived harassment is definitely necessary, because many people fail to appreciate this, and not-infrequently they respond saying âoh, yeah I get that sorryâ.
This is a very interesting exercise. You have a good sense of natural, conversational speech, Eugene. ââŚtell us and not them,â short and direct, is far superior to the prolix, bureaucratese of the ChatGPT. âWhen actual cheatersâ is also superior because it is specific and unambiguous. The sentence deals with actual cheaters not âpotentialâ cheaters, because it is the actual cheaters who change their behavior, so it makes no sense to unnecessarily drag in the lawyerly word âpotential.â Following on, âthey change their behavior,â is also better. âCan leadâ has less impact, and âtheyâ and âtheirâ keeps the focus rather than the subjectless âleads to altered behavior.â And âchangeâ is better than âaltered.â
Necessary correction: substitute âmakingâ for âit makes.â
I do prefer the ChatGPTâs âmay be perceived,â which was what I was going to suggest. âProhibitedâ does seem unnecessarily harsh. Something like âwhich might lead to further unnecessary conflictâ would be better, I think.
Again, your final sentence is better, because the ChatGPT is pointlessly wordy (it sounds like typical bureaucratic mush). All that is needed is a tweak to the punctuation there. Rather than the hyphen, put in a proper em-dash (closed-up on both ends), or else substitute a semicolon.
As a side note, our recent experience with the other message that the forum fixed up (the âobjectionable chatâ warning) has got me keen on this exercise.
Of course, it has the ever-present risk that we canât agree on anything in the forum , but having the community debate what messages we send to our fellow community members, and thus actually what policies we have, seems like âwinâ to me.
My attempt at making this message less harsh and a bit more condensed:
If you suspect your opponent of using AI, finish the game the best you can and then report it. Donât alert your opponent to your suspicion. They might change their behaviour and that makes it harder for moderators to investigate your claim.
Thatâs OK as far as it goes, but I think it is important to inform people that accusing people of AI use IS harassment. So itâs not up to their discretion whether they want to alert someone of their suspicion: even if they think that would be OK, it is still not OK to accuse someone, in chat, of AI use.
I donât think itâs OK not to tell them about this - itâs important to know that this is the case.
It is in fact the more important reason for not doing this: it simply is not OK to hassle people in chat even if you think they cheated.
If you suspect your opponent of using AI, finish the game the best you can and then report it. Donât alert your opponent to your suspicion. They might change their behaviour and that makes it harder for moderators to investigate your claim.
Do realise that you should never accuse your opponent of cheating on the basis of a suspicion.
If you suspect your opponent of using AI, finish the game the best you can and then report it. Donât alert your opponent to your suspicion. Firstly, accusing someone of AI use in chat is harassment and so not allowed. Secondly, they might change their behaviour and that makes it harder for moderators to investigate your claim.
[No idea if this is any different to the original really. Maybe better because it starts with what we want people to actually do, finish and report?]
Is accusing someone of murder harassment? itâs certainly a worse charge [citation needed]. Let us suppose that your friend went into a public restroom, then a stranger went in, then the stranger came out, you went in, and saw your friend lying on the floor stabbed to death. If you go back out of the bathroom and call out after the stranger âYou murdered my friend!â, is that harassment? I would say no. The question is not whether or not you accused someone of murder, itâs whether or not you have good reason to believe that accusation is true. You could argue itâs better to go to the police, both to protect yourself and to avoid tipping off the murderer that there was a witness, but none of that is a question of harassment
I donât think any action should be taken against someone who says in chat after a game something like âhope you like your Katago wins. reportedâ. Itâs not even particularly aggressive, just emphatically stating their position (their opponent and the mods are of course free to disagree with said assessment without necessarily concluding that the accusation was malicious: people can be honestly mistaken)
If I got this message, Iâd probably compliment my opponent on the Holy Grail reference. I certainly wouldnât bother reporting them for such a light-hearted phrasing
