2 new reporter blocks have been added to the AI category that checks if an image is appropriate. One takes the URL of an image and the other takes the name of the costume:
The return value will be “Pass” or “Fail”.
This group is for users who have assisted others by answering or helping with at least 5 situations. Eligible users may request to join the group.
2 new reporter blocks have been added to the AI category that checks if an image is appropriate. One takes the URL of an image and the other takes the name of the costume:
The return value will be “Pass” or “Fail”.
Currently this block only returns the profile image of the current user who is running the project, and it can not get the profile image of any user given a user ID.
Can you please explain the use case for that function?
@info-creaticode can it be made to have an input for their user id so I don’t need to send the pfp along the rest of the message?
A new reporter block has been added to the sensing category, which returns the URL of the user’s avatar image.
If the user has not uploaded an avatar, this block will return an empty string. In that case, you can use the default avatar image: https://play.creaticode.com/static/images/profile_avatar.png
Not a bad idea, not gonna lie. However I feel like it going to Extentions is a bad idea. Maybe we could make a separate category called Community Code so that official Extentions don’t get mixed up.
Yes those things can happen, but they can also occur with new comments or projects, and we can’t really review all of them manually. So we will just have to assume most users are well-intended and delete any user that tries to do harmful things immediately when we find out.
@info-creaticode I originally was going to say either AI or moderator review, but then I realized, for a snippet trying to get a user to run dangerous code or say a snippet that places inappropriate images, the AI can easily be fooled due to its lack of spatial awareness/capabiltiies. You could obfuscate a string, with various methods like say jumbling the letters and dejumbling them, to put strings the AI can’t interpret, and it would be even more effective to put comments that distract the AI like “3D raycasting algorithm” or something or misleading function names. I think human review would be good alongside AI since bad things could happen.
@info-creaticode That was just an idea I didn’t think much about before putting it there. To access things from an extension it would be done the normal way.
@jeffreyrb03-gmail said in Code block presets?:
For both of these, they would need to be moderator reviewed, like in a posting queue system.
We will probably allow the users to submit a new snippet after automatically running it through a moderation AI checker. Then, later, anyone can review them and report issues.
@jeffreyrb03-gmail said in Code block presets?:
For libraries added to a project, their information could be accessed using this perhaps?
Can you please explain this part more? An extension is usually just a list of new blocks. It is not clear why you would need this sensing block to access it.