Small update that hopefully will help keep the content in this community a little bit more accurate. We have a lot of bubble dev veterans that really know what they are talking about. Their advice is truly valuable, as it is the time the choose to spend contributing to the space.
I have now created a new moderation flag called “AI-slop/Misinformation”. If you come across a post that seems copy & pasted from an AI without any checking for veracity or accuracy, please use this new flag.
This new flag is not a statement against the use of AI. In fact, even Bubble’s support bot is an AI trained on our docs and best practices. This new flag is just one more little way that we can keep inaccurate advice from misleading newcomers that are trying to learn Bubble.
Thank you all for your help flagging things btw! Super helpful. Nothing makes my morning like a long review queue (kidding )
PS: shoutout to @georgecollier for suggesting the change
Thank you for the update, [insert community name here].
The new “AI-slop/Misinformation” flag will greatly assist in identifying posts authored by [insert expert name here], confidently asserting facts about Bubble features released in [insert year here].
Please note: advice may appear polished, authoritative, and formatted in bullet points for clarity, while being completely untested in a real application.
Appreciation to the Bubble veterans for correcting the record, and condolences to the moderators reviewing the queue at [insert time of day here] with [insert beverage here].
———
Had to do it! This post was just asking for it. You kind of walked right into this one. But seriously, thanks. This was a good idea.
I had thought some of the AI-written answers might be because of a language barrier. I.E. the poster used AI because they could get the answer written in English. Maybe not.
I was trying to use AI yesterday to solve a problem in Bubble. Spent quite a few hours over the whole day to figure it out. AI was useless. I finally figured it out, but it wasn’t with the help of AI.
Was also thinking yesterday though about all the people who learned Bubble before AI ever became a thing. Those people are awesome, and I’m sure it took a lot of trial and error.
Anyway, I usually never try to answer a question on the forum if it’s something I haven’t worked with before. I think the whole AI frustration is because it’s obvious sometimes the poster has never had that issue. Maybe?
What I’ve generally seen with the AI answers is they are either completely wrong, mostly wrong, or partially wrong, which is not good for the people following the advice, and then it just lives there in the post for others to stumble across, or another LLM to stumble across.
Most of the time, when it’s AI, I don’t even want to read through it to make sure it’s right. They are usually very long answers. At least I can just mark it as AI now.
Yeah, it’s the blatant cases where people evidently don’t actually care about the quality of information they’re putting out and do it because they think it raises their reputation.
Nothing wrong with AI for translation etc. As long as the substance of what you’re saying is correct
For people learning Bubble, using AI to solve problems at the first instance will likely rob you of the learning experiences that will make you good at Bubble.
When I started developing with Bubble, I saw AI as a light at the end of the tunnel. And it really was, but not every light in the darkness means salvation, hahaha.
)
This post was just asking for it. You kind of walked right into this one. 
But seriously, thanks. This was a good idea.
