Will there be moderators in the future AI landscape?

In ten to twenty years I think Discourse (the software used on this forum) will integrate AI and actual people won’t need to moderate…

Btw, the first person to turn this thread into a mod-bash session gets flagged.

3 Likes

Maybe, I don’t know how fast AI will develop. Got to keep a leash on those things when it happens!!

Your mind will merge with machines. So we could all have a party in VR and meet up with our avatars.

2 Likes

No no no. This is scaring me! I ain’t being no machine!

It is not so easy to code it, an AI would need to understand phrases and rate it as offensive, inappropriate etc…

1 Like

Reddit has auto mods.

Joke’s on you. We’ve had an AI mod for six months now. We trained it to automatically flag and delete all sorts of things and it’s really increased productivity and given us more free time.

4 Likes


Go to 1:50

Yeah but I meant a full time AI who has super complex understanding of the meaning of posts

It’d also have to fit into the abstraction space that the software runs on… and there are a lot of dynamics in that.

Java has very strict public/private nature to it’s classes and it’s functions operate like modules a lot of the time… and that breaks some uniformity yielding abstraction.

I mean where would the AI sit? Inside the server software itself? Or along side it? That’s another contemplation of abstraction.

You might find that the server applications language is totally defunct in handling “intelligent” code… then you’d have to basically have ai sitting piggy back… and that kind of integration would be very complicated… and prone to bugs… even just in the timing and error detection routines in the server software itself.

It’d all be possible… but the technical challenges of phrase recognition is only a small part of that.

I mean ideally you just set the AI with a self-learning aglorithm… human interface for verification… and then the capacity for real mods to be able to doctor the records.

I’d use a flag and then polling system… perhaps make sure that the flags come from diversified and verified users.

then 3 flags could yield the addition to the filter… while also preventing spammers or hackers from hijacking the filtration system to render the board unuseable.

@Nomad… probably not with discourse… but I’d guarantee sometime somewhere there will be an AI moderated website.

1 Like

It also sends us reminder PMs about flagged posts. :smirk: :roll_eyes: :face_with_raised_eyebrow:

2 Likes

In other words you want to clone @rhubot into an AI über-Mod. :scream_cat:

3 Likes