adorablesergal

Go out and make something!

๐Ÿ‡จ๐Ÿ‡ฆ Mom, ๐Ÿ•น๏ธ gamedev, ๐Ÿ”ญ amateur astronomer, ๐ŸŽจ artist, idTech4 engine appreciator, DOOM 3 shotgun enjoyer

โ˜ข๏ธ 2D Lead, @dnf2001rp โ˜ข๏ธ

๐Ÿš€ I run @nasa-unofficial ๐ŸŒ 
๐Ÿ”ž I draw porn, yo. ๐Ÿ”ž

โค๏ธ๐Ÿงก๐Ÿค I was here ๐Ÿค๐Ÿฉท๐Ÿ’œ

ย 


๐ŸŽจ Cara.app
cara.app/adorablesergal
๐Ÿ“บ Picarto
picarto.tv/adorablesergal
โŒจ๏ธ GitLab
gitlab.com/adorablesergal
โŒจ๏ธ Github
github.com/adorablesergal
๐Ÿ’ฐ SubscribeStar
subscribestar.adult/adorablesergal

And it does present a lot of ground for discussion about the future of content moderation.


I've been feeling some undercurrents in my own circles that eschew massive communities in favour of small, more tight-knit groups of friends and trusted colleagues. In these communities, content moderation is handled by the community members themselves, often by merely having a set of shared values that serves as an automatic filter.

But most people aren't like that. I don't think the general populace has quite gotten the message yet despite gestures broadly at the world. They still want that idea of Big Internet, chasing the biggest "Town Square" they can find, having influence. They want Numbers. They like Numbers.

How the fuck do you moderate all of that?

Like, as long as these services exist, as long as those pressures are there to be a star, people are going to want to be in the Big Town Square where the parties are at. No one wants to shout into the void, which is why Web 1.0 had people setting up site counters and guestbooks.

And while you and I might not use Bluesky (I have no plans to, at least for now), we live in a world where sites like Bluesky will draw people in.

As an artist, Numbers are valuable, and until Capitalism is dead and buried in the ground, I have to care about Numbers to know which things I draw are popular or not, and thus which subjects or styles I should focus on to gain audience and money.

However, strolling into that Town Square to set up shop, I would have to deal with a lot of unpleasant people who should be thrown out of the Town Square, but there's so many of them, and most of them blend into the crowds so easily, trained via years of trial and error to not tip off automatic moderation tools or even human moderators. They're still assholes. They're still dangerous. They're still part of the swell of Shitty Ideology that can and will go to Bad Places if we allow it to.

Market values

Handing things over to AI is inescapable, maybe even necessary. It's not like Earth's population is shrinking. More and more people are getting online. Not everyone is nice. There's still plenty of shitty ideas out there, and ways to inflict harm, but quelling that is going to come at a tremendous cost.

We're at the cusp, I think, of the largest imposition of a core value system on the planet that anyone's ever seen, and it's largely going to be unseen. Even if it is seen, the very fact that it's associated with AI subtracts the human element in such a way that there will be a mistaken belief that the machine is unbiased (no matter how much we tell people there is bias inherent in all data sets).

What is that future world going to look like?


You must log in to comment.

in reply to @adorablesergal's post:

Yeah, honestly I don't see anything about training data specifically being sent, though that might be something more on Hive.ai's end, which is why I simply focused my discussion on AI-driven content moderation. Scraping is going to happen no matter what, that's the age we live in, but I'm far, far more worried about the automated systems we'll put into place to impose value systems on the world at a scale no one's ever seen before.

Pinned Tags