I’ve had a few odd experiences on the forum recently and just quickly wanted to ask about how to proceed going forward. The forum Terms of Service states that:
So I addition to flagging clearly spammy content (e.g. containing links) I have been flagging content where especially new users make a first post that reads like it was generated with Chat GPT, even it does not have links and might actually contain “good” information. For instance this one (edit: has been deleted), where I then received a PM from a moderator stating that I should “Feel free to add a better answer to that thread…”
Additionally there has been cases where I took the bait (again, my bad) and have replied instead of flagging (i.e. calling out Chat GPT content with proof from a user I long suspected if doing this). Which was also met with a moderator PM urging me to delete these posts now that the author had deleted their Chat GPT content and following ad hominem rant. Which is fair enough I suppose, but I do wonder if that author ever even received a reprimande for violating the Terms of Service.
Anywho, I don’t mean to stir up any shit. But I am genuinely asking how we are to proceed as forum users, when there seems to be some discrepancy between the rules and the response from moderators. If you don’t want us to flag certain things, please let us know so (and update the Terms of Service/FAQ), so we can stop wasting our time and energy.
Personally, I didn’t see that thread, but I would have flagged it as well.
IMO, if we allow this stuff as valid answers, before long this forum will be filled with them - perhaps some from well-meaning people who don’t actually know the answer themselves so they asked ShatGPT and then decided to post the answer, or, perhaps by bots whose goal might be to catalog places where this kind of content is accepted, for future spam use…
Could you explain what you are flagging up here? In the context of a legal document in English, as opposed to (that rare thing) a statement written in formal logic, I do not see any ambiguity arising from the use of “and”. A clause “A, B and C” equates to “A ∧ B ∧ C”.
I’m explaining what “and” means, and how the moderator behavoir is correct both in terms of ethical and rules.
Rules state they must be machine generated AND useless, which is not the case.
Machine generated content can and often is useful.
Furthermore if the goal is helping users, then It should be alright to use AI as long as It’s helpful to other people.
I use AI everyday and It is very useful, but I also understand that other people really do may dislike AI and that’s fine. Do not limit others fur your own limits, that’s all.
If the goal is helping, AI should be welcomed as It helps.
If it’s useless It will get removed.
That is my opinion but I am curious to hear your opinion,
Looking at that specific example, it seems to me that the advice given is good (I’m happy for you to demonstrate otherwise). In which case, does it matter how it arose? If the response is expunged, will the OP be better off with no advice? Surely the issue is not whether the advice comes from a person or AI, but whether it answers the original need and is correct? After all there are plenty of unhelpful or inaccurate posts made by people (I hold my hand up here) - should they be erased too?
No they need to be culled because AIs can not be trained on their own output, this contamination makes it harder to get useful answers from this repository of Rhino knowledge in the future. Also it seems pretty obvious the entire thread was spam.
Again, I’m not here to prove/demonstrate anything, nor pass judgement. And I literally wrote:
I’m simply asking the moderators if users should flag posts they suspect is LLM-generated (edit: and is intentionally misleading “recipients as to the source of the material”).
No, I’m sorry you are misreading that: the rules are stating what posts must NOT be… NOT A and NOT B and NOT C. So if machine-generated OR “useless” don’t post.
In logical terms the paragraph (read with common-sense, a requirement for the interpretation of any legal document) says something like this:
(¬ spam)
∧
(¬ (machine-generated ∨ randomly-generated) )
∧
( ¬ contain
( ( unethical ∨ unwanted ) commercial content designed to
( drive traffic to third party sites
∨
boost the search engine rankings of third party [sic] sites,
∨
to further unlawful acts (such as phishing)
∨
mislead recipients as to the source of the material (such as spoofing)
)
)
)
)
⇔
OK to post
I’m personally looking forward to the chatbot takeover when the mods get overwhelmed by nonsense…
“How do I fillet this corner?”
“Hi, you can fillet this corner using the fillet function in Rhino 8 SR10”
“That’s great thank you. Is the fillet function?”
“You can use the fillet fillet function by posting SysInfo here, and we can know your problem”
“Can you know my fillet problem?”
“Yes, by using the FilletSrf function in Rhino 8 SR12, you will be able to make your fillet”
“Thanks”.
Fortunately, they aren’t all using inappropriate pictures as profiles images yet. YouTube is absolutely infested with it.
I mean this is someone’s job, I wouldn’t be surprised at all. However cynical you might be, it’s not enough. New user asks not-really-Rhino-related question with a lot of unnecessary detail and user name of “MaryAnn” despite not really sounding like English is their first language. Gets obviously AI slop answer from another new user.
