My interactions on the forum have been by-n-large very productive. Have made wonderful acquitances and relationships.
What frustrates is some of the posts on forum are spam. I understand that there is a process to stop these; however that relies on actual humans reading and detecting spam and reporting it as such.
Is there a way to actually just have a automated process that can tell me based on my preferences that I should not be reading that post?
There is actually an automated system that attempts to spot these things and deal with them, but as ever it’s a cat and mouse game and we all really apricate the time and effort of our members in spotting these and flagging them where appropriate.
Note that a company can make a single thread about their Ai related service so long as it’s OpenAI related, but they cannot spam links to that on every thread.
Clear spam is usually easy to spot and flagging it helps everyone.
Thanks !
Now it’s becoming a little bit more insidious; but clearly easy to spot to the trained eye.
Day 1 : poster 1 : A simple problem # this could be legit
Day 3: poster 4: The poster seemingly has a solution (but high level and hides details)
Day 4: poster 1: Asks for clarification
Day 5: poster 4: Pitches company in the third person
But yeah, I get it that it is a cat-n-mouse game.
EDIT 1: Would it be violating TOS if I developed a script for me to parse individual new forum posts for potential spam so that I don’t read it. (this would be for me)
The Forum has an API to do that, might be worth looking into.
Hitting the website might cause more issues than it solves.
It would be a challenge, because AI itself thought that the arbitrary response it created with no foundation in reality and tell-tale signs of AI cliche was an informative answer to a question. The posting looks internally to be of low perplexity, which could mean well written and informative human also.
Detection by AI model analysis of language and not a deeper algorithm also informed by pattern and network metadata to connect sockpuppets would likely determine the AI-generated useless posting to be of high value even with the same forum message context that led to the spam creation.
Developing scripts or non-impactful scraping is not a terms of use violation, just like even posting negativity about OpenAI quality or company operations as informative guidance has no codified policy violation anywhere in any TOS language.
It would be the actions the script takes, and if they are beneficial through out-of-band communication or methods, or if they rather just add to the noise.
I have no intention of imposing my views on others.
However it is just that there are majority of the posts that I would NOT rather read; either because the kind of intellectual curiosity is not present for me (i.e. a billing issue ) or a spam or a post turning into spam.
On the other hand, there are other articles that are so insightful to me that I will gladly “drink from the fire hose”. Also I fear that I miss some insightful posts that come in when I sleep or doing something else.
So now you have my complete motivation in words. 
I understand your frustration dealing with the non dev related posts, we are talking to OpenAI about how we best handle that, but clearly there is a need for an open forum method of communication for ChatGPT users.
For now, you can select which categories show up in your alerts in the preferences section on the forum in your account.
There are category specific option below this screen shot.
