Turn over moderation duties to AI

[Ben J.]

A perfect role for artificial intelligence; discussion forums moderation. That would be much better than the way it works now, with fallible humans making decisions based on their fallible judgements. I'm sure it would be much more capable of weeding out chatbot posts aswell. It would be both more fair and more predictible, I think.

In time, we might consider it for our judicial systems too.

 

[icanhazmac]

A perfect role for artificial intelligence; discussion forums moderation. That would be much better than the way it works now, with fallible humans making decisions based on their fallible judgements. I'm sure it would be much more capable of weeding out chatbot posts aswell. It would be both more fair and more predictible, I think.

In time, we might consider it for our judicial systems too.

LOL, what? 😉

Chat bot and spam posts, ok.

Moderation... aw hell na!

 

[Ben J.]

LOL, what? 😉
Chat bot and spam posts, ok.
Moderation... aw hell na!

Why not?
What downsides would there be?

Another good thing; it would be real-time. Posts breaking rules would not be first posted and then reported/deleted. Much tidier and less frustration.

 

[za9ra22]

LOL, what? 😉

Chat bot and spam posts, ok.

Moderation... aw hell na!

Not just bots and spam, but also personal attacks, abusive posts, users being belligerent, threads being dragged continuously off topic, and a general demeanor of anything is good as long as it gets a few more clicks... AI can't handle that at all well.

 

[Ben J.]

…and anytime you had a post rejected because it violated some forum rule, you'd get an immediate notification, with a reference to the rule you violated. And there would be no need for the disgusting temporary or permanent banning punishing thing. You could simply adjust and try again.

 

[icanhazmac]

Why not?
What downsides would there be?

Unless I am underestimating the capabilities of AI for this purpose, one word... nuance.

Avoiding politics in a non-political thread:

"and the president is an idiot" - is that the president of a country or the president of a company, one is moderation worthy, the other is not.

This list could go on forever. I think you would find AI to be extremely heavy handed which would probably annoy the community.

…and anytime you had a post rejected because it violated some forum rule, you'd get an immediate notification, with a reference to the rule you violated.

I get that now, and again, nuance. Most rules have gray areas open to interpretation. Some rules may be bent or ignored in promoting healthy conversation.

And there would be no need for the disgusting temporary or permanent banning punishing thing.

LOL, assuming you meant to say "discussing", so no recourse to moderation? No thanks.

Why post at all, why don't we all just use AI to post for us? Hmmm, scratch that, I think some people here do that already and we don't need more of it.

We are getting closer to this:

 

[Ben J.]

LOL, assuming you meant to say "discussing", so no recourse to moderation? No thanks.

I meant what I said.

 

[Scepticalscribe]

An appalling idea, but, unfortunately, a chilling, and telling, symptom of the nature of in the discussion on the role (and place) of tech in our world.

Agree with @Herdfan.

Nuance - and judgment - are key in any such discussion.

More to the point, the OP's original post assumes that AI is "rational" and "objective", when it is nothing of the sort; instead, it will reflect (and project) the biases of those who have created it.

Just because something is a technological creation does not mean that it is free of prejudice and incapable of sustained exhibitions of poor judgment.

 

[za9ra22]

…and anytime you had a post rejected because it violated some forum rule, you'd get an immediate notification, with a reference to the rule you violated. And there would be no need for the disgusting temporary or permanent banning punishing thing. You could simply adjust and try again.

My point, badly made, was that it is not in the interests of the forum (here) to manage behaviors and control posting. More aggravation/frustration/anger equals more clicks. More clicks equals more revenue. If MacRumors wanted to have better moderation, they could achieve it by the simple expedient of letting the moderators they have do their jobs. This appears not to be their aim, as plenty of threads demonstrate.

Our existing moderators are neither stupid nor asleep.

 

[icanhazmac]

I meant what I said.

Ok, that is even worse. You cannot post until you appease the silicon overlord? No thanks. Feel free to start just such a forum and we'll see how far you get. I would hate to be in a nuanced conversation and need to spend a half hour trying to post my way around an overtly strict digital censor.

 

[Ben J.]

We already have automatic moderation.
You know those four letter words that turn into ****.

 

[icanhazmac]

We already have automatic moderation.
You know those four letter words that turn into ****.

And I don't care for that either. Words are just words, we assign different weights to them for some strange reason. To me there is no difference between "poop" and "****", they both have 4 letters and have the same meaning in context. One we teach children to say and one we punish them for saying, we aren't the sharpest knives.

 

[Apple_Robert]

This isn't just a website; it is Arn's business. It would be absurd to put the business at risk by solely relying on AI.

 

[Ben J.]

This isn't just a website; it is Arn's business. It would be absurd to put the business at risk by solely relying on AI.

But, AI is supposed to be even better than us at things.
That's one of the main reasons that people are afraid of it.
And inforcing a set of written rules isn't all that difficult.
And it could, of course, be "super-moderated" by Arn or whoever.

Listen, it was just a thought, and I thought I'd just throw it out there.
Bit disappointed with the one-sided response, though.

 

[Scepticalscribe]

But, AI is supposed to be even better than us at things.

Perhaps.

But, placed in a position of authority without adequate - and exceedingly robust - oversight, AI will simply run the risk of replicating and reinforcing the prejudices, attitudes, and biases, of those who have created it in a more efficient - and inflexible - manner.

Besides, AI has no head for human problems, and no capacity for nuance or subtlety - the sort of areas where good judgment (rather than an exceptionally inflexible application of the rules from which there is no appeal) comes into play.

 

[icanhazmac]

But, AI is supposed to be even better than us at things.

And it certainly is, at some things.

That's one of the main reasons that people are afraid of it.

LOL.

And inforcing a set of written rules isn't all that difficult.

Yes it is, for all the reasons we listed.

And it could, of course, be "super-moderated" by Arn or whoever.

Only after a members experience has been ruined by an overtly strict censor.

Listen, it was just a thought, and I thought I'd just throw it out there.

That is great, it was a good discussion, you didn't think everyone would agree with you, did you?

Bit disappointed with the one-sided response, though.

I'm sure there are folks out there that would agree with you.

My .02 is, if I am posting on a forum it is to engage with other human beings (hopefully, but I fear more and more are bots etc.) and I don't want to speak to other humans through a digital filter/censor. In that endeavor we will, see/read/hear things we might not feel are appropriate, that is the cost of "unfiltered" communication. I will gladly accept the occasional horrible post versus what you propose.

 

[dmr727]

Can't speak for MR, but I moderate on two other boards, and in each case we reach decisions through collective discussion. Unless it's a clear case of spam, no one moderator can make a decision on his/her own. In many cases it's black and white and the mods will be in immediate agreement, but in some situations we'll have dozens of posts between us before a decision is made. Point is, there is enough nuance in moderation to warrant discussion before decisions are made, and even if an AI were capable of such nuance, I wouldn't want a single source making the call.

 

[Pakaku]

But, AI is supposed to be even better than us at things.

Algorithms are not better than us (people invented algorithms, after all,) and that is exactly why machines should never have the final say in any moderating decisions

 

[annk]

Can't speak for MR, but I moderate on two other boards, and in each case we reach decisions through collective discussion. Unless it's a clear case of spam, no one moderator can make a decision on his/her own. In many cases it's black and white and the mods will be in immediate agreement, but in some situations we'll have dozens of posts between us before a decision is made. Point is, there is enough nuance in moderation to warrant discussion before decisions are made, and even if an AI were capable of such nuance, I wouldn't want a single source making the call.

Yes - collective discussion is how it's done here too.

 

[chown33]

For comedy, I regularly visit the website "AI Weirdness":

AI Weirdness

The weird side of artificial intelligence.

www.aiweirdness.com

Flightless birds are pretty funny, but the chocolates are also outstanding. The older posts go back quite a ways, and can be fun to explore. They're also good as a brief illustrated history of bot-versational interactions.

Sometimes the gist of a post is more serious, even if it's framed as "Let's see what kind of weirdness we can get". For example, think about how disconnected from reality the bot in each of these is:

Don't use AI detectors for anything important

I've noted before that because AI detectors produce false positives, it's unethical to use them to detect cheating. Now there's a new study that shows it's even worse. Not only do AI detectors falsely flag human-written text as AI-written, the way in which they do it is biased. This is

www.aiweirdness.com

Optimum tic-tac-toe

ChatGPT text can sound very knowledgeable until the topic is something you know well. Like tic-tac-toe. Once I heard that ChatGPT can play tic-tac-toe I played several games against it and it confidently lost every single one. Part of the problem seemed to be that it couldn't keep track of

www.aiweirdness.com

What does this say?

Large language models like ChatGPT, GPT-4, and Bard are trained to generate answers that merely sound correct, and perhaps nowhere is that more evident than when they rate their own ASCII art. I previously had them rate their ASCII drawings, but it's true that representational art can be...

www.aiweirdness.com

Next, think about whether you'd want that bot being the sole moderator. As an exercise, you could ask a bot to rewrite a post so it doesn't violate any MR rules.

Or as a somewhat larger exercise, setup a public chat room somewhere, and run all moderation questions thru ChatGPT. My guess is it would show some results pretty quickly.

 

[Chuckeee]

But, AI is supposed to be even better than us at things.
That's one of the main reasons that people are afraid of it.

Some people believe AI is even better than us at things.
That's one of the main reasons that people should be afraid of it.

 

[bousozoku]

…and anytime you had a post rejected because it violated some forum rule, you'd get an immediate notification, with a reference to the rule you violated. And there would be no need for the disgusting temporary or permanent banning punishing thing. You could simply adjust and try again.

I'd expect 20-30 people banned in the first 10 minutes.

 

[bousozoku]

Some people believe AI is even better than us at things.
That's one of the main reasons that people should be afraid of it.

I spent a fair amount of my software development career creating automated programmes to make corrections in the way people would, mostly to repair or upgrade source code based on governmental requirements or something similar.

Keeping it simple worked.

In a case where people would expect Artificial Intelligence to act in their favour, it isn't likely to go that way at this stage. I suspect all the touchy-feely decisions that moderators make to be lenient would not be applied and almost every member would be banned in hours, staff included.

It would take thousands of hours to train AI to accurately begin to emulate a human.

 

[Mousse]

Until AI surpasses this guy, it's not ready to moderate.

 

[icanhazmac]

Until AI surpasses this guy, it's not ready to moderate.

Fry with or without the worms? 🤣