Echo JS 0.11.0

<~>
MaxArt 22 days ago. 
Lately, I've the suspicion that some of the articles posted here are nothing more than an explanation straight out from ChatGPT and nothing more, in order to create engagement.

I'm pretty sure AI-generated content is already forbidden on EchoJS (rule "No link to summaries, only link to original content"), but this could become more and more a problem in the near future.

What tools do we have to detect (fully or mostly) AI-generated content?

tracker1 22 days ago. link 2 points ▲ ▼
I don't personally have any tools to check for this... I don't care for the relatively Jr level content myself, and often call out errors or at least oddness in the articles that I do read with a comment.  I don't always read every article though, just a quick pass to clear obvious spam.

I could favor nuking the very jr and simple content... just wouldn't want to actually discourage anyone genuinely writing learner/beginner content.
MaxArt 21 days ago. link 2 points ▲ ▼
That's a good point. AI-assisted writing is indeed a thing.

I wonder if there will ever be a tool so smart that could realiably detect how much of the content is genuinely human-made in an article.
tracker1 20 days ago. link 2 points ▲ ▼
That would definitely be a nice tool... I feel that this will only get more difficult in time though.  Right now it's easy enough to prompt a few popular LLMs to see if you get similar text, but man, what will the cost be in the end.

Both in generated content as well as bot detection and anti-bot tooling.  I'd hate to be github.

I can only say I fully expect that we may wee a return to more niche communities that are self-regulated, similar to BBSes of old.  In only that the popular platforms for social media are already inundated with bot activity.