Should we mark ChatGPT like posts from new users as spam?Should we mark posts like this as a spam? They are very ChatGPT like answers from new users
Re: Should we mark ChatGPT like posts from new users as spam?Hi @jthi,
I understand your concern about this possible issue. I must admit the answer sounds quite vague, but I wouldn't dismiss so quickly the answer if it can bring some informations.
Personaly, as I tend to repeat myself on some topics related to DoE, I also tend to copy some of my previous answers and reformulate them for new posts, to save some time and quickly bring an answer.
Also it can be very difficult to judge (and misjudge) an answer as AI-powered answer, and several studies have shown the high number of false positives that conventional AI-detectors applications or human may be prompt to create (and the number of false positive may be even higher for non-native language speaker, like non-native english speaker) :
OpenAI kills off its own AI text detection tool! Shocking reason behind it | Tech News (hindustantimes.com)
https://www.technologyreview.com/2023/02/07/1067928/why-detecting-ai-generated-text-is-so-difficult-and-what-to-do-about-it/
My personal opinion on this would be : do not mark posts as spam if they bring value or informations and else, mark them as spam.
Re: Should we mark ChatGPT like posts from new users as spam?The post from WendyOliver is not necessarily correct. You don't actually remove the time component by changing the column format.
I've had questionable answers from ChatGPT, when I ask it to write code. The answers look right, but the programs either don't work or they generate incorrect answers.
Re: Should we mark ChatGPT like posts from new users as spam?My guess is that those answers are from bots. There are few similarly written answers, from similarly named users (WendyOliver, DavidHernandez, DeliaSpears) and some of them are marked as spam AS they also contain some casino web pages (you can see some examples from spam quarantine). Regarding WendyOliver's answer, it doesn't really bring anything to the discussion as JMP doesn't have neither of those functionalities named as such.
I don't know what is the hottest hot in spam industry (if those are even bots), they could post okay answers -> edit them later? or just post okay answers to collect post count (not sure is there really a benefit for something like that in JMP community, for example in Reddit with karma there would be).
Re: Should we mark ChatGPT like posts from new users as spam?Seems like WendyOliver did suspicious enough edit and auto-mod caught it
Re: Should we mark ChatGPT like posts from new users as spam?Given that WendyOliver is most likely a bot, I've suspended the account. Thank you for moderating!
Re: Should we mark ChatGPT like posts from new users as spam?and https://community.jmp.com/t5/Discussions/JMP-software-for-analytical-purpose/m-p/735046#M91581
I've noticed several of these; one that I kudo'd later had a spam link in it. Keeping an eye on this one to see what happens...
edit 17mar2024: apparently not spam.
Re: Should we mark ChatGPT like posts from new users as spam?Hi @Ryan_Gilmore,
I'm also very suspicious about new user Darshan Hiranandani, as his posts are either vague/not precise or seem to be copy-pasted from other posts...
User : https://community.jmp.com/t5/user/viewprofilepage/user-id/55551
Example of copied posts :
From Darshan Hiranandani : Darshan Hiranandani : Determining Whether a Linear or Non-linear Regression Model Fits the Variables
From original post : How I can I know if linear model or non-linear regression model fit to my variables?
Seems very close if not identical...
Re: Should we mark ChatGPT like posts from new users as spam?Yes, I saw those and had the same opinion. I have marked all posts as spam and banned the account. Thanks.
Re: Should we mark ChatGPT like posts from new users as spam?Hmm...they did not create a new account. It seems the ban was dependent on the IP address in addition to the other attributes (username, email, etc.). I've updated the criteria so let's see how that works.
Re: Should we mark ChatGPT like posts from new users as spam?Not sure this is a real account. Darshan Hiranandani is a millionaire being investigated in a foreign exchange case!
https://www.hindustantimes.com/india-news/who-is-darshan-hiranandani-summoned-by-ed-in-fema-case-linked-to-mahua-moitra-101711530768515.html
Re: Should we mark ChatGPT like posts from new users as spam?Have they created new account, Darshan Hiranandani - How can I transfer a variable value from JMP into Python? - JMP User Community (there was similar question week ago How to import a variable value from JMP to Python? - JMP User Community)?
Re: Should we mark ChatGPT like posts from new users as spam?Hi fellow Super Users,
I added a new post in the spams today, as it sounds very ChatGPT-ish...
I can't remember how/where to access the spams folder, but it was related to the post Incorporate Time lag in DoE
The answer was vague, not related to DoE, and structured the same way as a ChatGPT-like model would answer...
Hope I didn't react too promptly.
Re: Should we mark ChatGPT like posts from new users as spam?I don't think you can easily access spam or archive since new JMP community page, but I still have the link in my browser history to spam section https://community.jmp.com/t5/spam/searchpage
Re: Should we mark ChatGPT like posts from new users as spam?We'll bring back some of the administrative links in a future update. In the meantime, I've created an article in the library that contains a list of potentially helpful links (only 2 at the moment): Helpful Links . If you have others, please update. Thanks.
