Hacker News Reader

Comment by shagie | Hacker News Reader

My crystal ball of unfounded speculation on this says...

It's that in the GPT hunting/banning back in December and January they probably hit a lot of false positives. SE lacks a good way for people to complain about moderation (the ‘send feedback’ is a rather poor tool to do so) and things got escalated to where lawyers started getting contacted. SE as a global company can’t attempt to hide behind a “we’re only in the US” and so is now swinging the moderation rules to the other extreme (its always one extreme or the other) since they don’t have sufficient CMs or lawyers to properly address the issues.

So now, six months later, the EU digital services act is kicking in hard with the unanswered/unresolved moderation challenges that should have been done in a timely manner.

View on HN

Yeah... that's not it. I used to handle escalation tickets on SO and we were handling tickets from the EU years before I joined. Also, that's not how any of this works.

It would have then helped for SE corporate to expand a bit more on what constituted misdiagnosing ChatGPT content, how that was determined and how it was determined that certain countries were prone to false positives.

The disconnect between regular users capable of flagging (though the "what is an acceptable flag" has likewise gotten more and more restricted too - see also "not an answer" is not the right flag if there was any attempt to answer, even if it was the wrong answer to the wrong question), moderators, and corporate makes trying to have a community run site difficult - especially as the tools that one would use to manage, moderate, and curate it get more and more restricted.

The curious part of the "is this it?" with the digital services act is that it was signed into law 6 months ago (November 2022) and has provisions that have a 6 month window and concerns moderation of users and improperly removing their content.

Anyways, that was my unfounded speculation.

I am still disappointed that incorrect information written by a human isn't more severely moderated or curated on SO and SE sites. If I'm going to get wrong information, might as well get it from a source I need to double check anyways and is in a more consumable format. The value of SO was that "good answers are voted up and rise to the top" which is much less commonly the case anymore.

Reply on news.ycombinator.com