Where should the line be drawn for what is AI coded vs AS assisted (in some manner) and thus tagged?
XerahS is very openly coded with AI and should be tagged. But what about the recent debate about the use of LLM in cURL?
Comment on Selfhosted & AI
curbstickle@anarchist.nexus 3 weeks agoActually both are pretty good - AIT for it as a more discussion oriented, AIP for a project post.
I like it, I think something like that would be a great idea
Where should the line be drawn for what is AI coded vs AS assisted (in some manner) and thus tagged?
XerahS is very openly coded with AI and should be tagged. But what about the recent debate about the use of LLM in cURL?
So thats one of the questions that will need to go to a poll, but if I were to give my opinion it would be that any use would mean a tag, and the disclosure area is where it can be detailed.
Otherwise we’d end up with a bajillion differing tags, so a disclosure section makes more sense.
Not tagging for assisted would also mean setting arbitrary lines, and would be way too subjective imo.
any use would mean a tag
So when I write code, I regularly ask an LLM how to do things. A few examples from my history is how I could overlap rows when generating PDF files (have a library for this), how I could assign an empty Char and to create a singleton class. A LLM never changes any of my files, I use it as super-search-engine.
Would that constitute an AI tag?
Yes, because it is a bit different than a search engine. It may not be changing the files, but it is telling you the way to do it. It might give you an outdated method/pattern, it might ignore conventions, and most importantly, it doesn’t really understand the problem its solving. Its not finding the optimal answer, its finding the most common.
So the resulting answer may work, but not necessarily be right. In a disclosure (see the new thread) that would be referred to as “hint” for lllm use.
That doesn’t mean the answer you got was necessarily wrong either, just an answer based on an amalgam of the most common for all the code fed into the model, no matter when the code was written.
It would be put into an ai disclosure, so it qualifies for a tag. Make sense?
brucethemoose@lemmy.world 3 weeks ago
Oh, both! Yeah. I didn’t even think of that, but [AIT]/[AIP] as separate tags makes sense.
I guess two tags runs the risk of “overcomplicated rules,” but that’s more of a moderation load question. I got no say in that, heh.