/tech/ - Technology★

>>997556 (OP)

I think I found the website you're looking for http://twitter.com

Sounds like you want Reddit or Voat.

>distributed imageboard software which uses its nodes to power an advanced AI which moderates for us

sounds like a pipe dream

>>997556 (OP)

So basically you want a "votekick" option where everyone in a thread has a say? Yea because that works so well and it's never abused in online games

>we need a self moderating imageboard or moderated by the users

Usenet's Killfile and Scorefiles.

Invented 35 fucking years ago.

Because democracy always works out when the userbase consists of retards and brainwashed NPC drones.

>>997771

>killfiles

>scorefiles

>requires compromising anonymity of all users

jewsnet ~= imageboards

>>997556 (OP)

>self moderating imageboard

>>>/reddit/

>>997774

It can be anonymous if you strip out the IP address. Everything else is basically identical to a tripcode.

But even with the IP address, you were fairly anonymous in the 90's, if you didn't use your real name. Your ISP had logs, but no random faggot could tell who you were. If you wanted to go one step further, there were email->news gateways like anon.penet.fi, which did strip out all the identifying infos. Of course the cianiggers shut it down eventually.

>>997556 (OP)

The only way to really do that is to train an AI to recognize child porn or build a CP database and both methods are not only controversial but in the case with the AI specifically can also potentially read false-positives and permaban innocent users

>>997597

>more than half the people vote on something

>it's abuse of voting rights if I disagree with the outcome

>>997788

In case of neural network based AI, a single purpose-crafted pixel is enough to generate a false result.

>>997597

Votekick does work. You're probably a faggot and got kicked for saving in a casual game

votekick works badly when everyone is anonymous

you can forge votes, and any random assholes who don't even post on this board can also wipe content at will

it could work if there are hidden user IDs but this is not 100% true anonymity

(but then there are IP addresses and they are even worse tool for identification and they can be leaked as well so this is still better than how it works right now)

there could also be made a measure to limit the quantity of new user ID generation, for example you need to mine a certain little amount of XMR to get one, (which goes to fund the hosting expenses) and this doesn't hurt anonymity

and anyways a more robust solution is per-user filtering, I proposed it earlier btw

it will work a bit similar to how browser blocking addons work (such as uBO), filters are created, edited and shared by users, so you can block anything you want but it will be visible to others if they don't subscribe to your filter lists

As >>997771 has brought up, something like Usenet is what you naturally end up with when you engineer a solution that encompasses all of decentralization, moderation and anonymity. The main Usenet federation, the one made up primarily of ISPs, universities and paid services, is barely anonymous and hardly decentralized. It is however also not the only way to run a Usenet services. OStatus is an example of much healthier federation. Take its propagation model and orthodoxy, use them to distribute Usenet-style non-microblog articles and you have a decent product.

You'll naturally end up with swarm fracture when it comes to level of anonymity; different relays/instances will choose whether they want to accept Tor and/or clearnet posts, anonymous and/or registered posts, etc., and also whether they propagate any identifying information for their own posts. Depending on which relay/instance you post on (maybe your own), your ID might look like "anon@sf78oa3ibud.onion", "user-cooldude69@sf78oa3ibud.onion", "anon@shitpostersinc.com", "user-cooldude69@shitpostersinc.com" or "cooldude69@cooldude69ontheinternet.com".

>>997556 (OP)

Millchan.

>>997842

NNTPchan.

8ch.net

>>997556 (OP)

The problem with this is it suppresses unpopular opinions.James Madison recognized this as "tyranny of the majority".

>>997798

See

>>997861

>>997799

I don't think a single pixel can throw off nets working with high resolution images or ones trained to recognize noise. Regardless, I think you would need and obscene amount of CP to recognize it reliably enough for policing a website.

>>997556 (OP)

>be oldfag on selfmoderationchan.org

>it gets linked on GoyBook and CNN talks about it because the newest mass shooter read it once ten years ago

>get permabanned because you were too spicy for the normalfag invasion

Finding a non-retarded mod would be a better idea.

Alternatively, host it on I2P or IPFS or anything else that requires miniscule effort to keep out normalfags.

>>997556 (OP)

>we need a self moderating imageboard

If you don't explain what that means, it makes as much sense as "we need to abolish private property" and "we need a 4 sided triangle".

>or moderated by the users

Like 8chan?

Any user can create a board and moderate it as they see fit, after all.

>>997796

You're ignoring all the issues non-anon forums have shown in practice.

>>997798

>>997808

>defending mob rule

>expecting internet votes to not be manipulated

How fucking new are you two?

You realize you're basically suggesting reddit but with autoban on enough downvotes?

>>997788

>AI to recognize child porn

It is very suspicious that such a system is not already operational. It would take a lot of stress off moderators. I'm thinking that spamming a board with cp is something that TPTB want to keep.

Until chatbots become indistinguishable from humans AI spam detectors could also help moderation and eventually allow for a very small team to run a site or even better a mirror of it, in the way NNTP implements it.

>>997556 (OP)

So, 08chan?

A way this could work:

Any user is allowed to create a list of posts. Each post in the list would have associated with it a series of tags, like "spam", "cp", "programming", "politics", etc. An individual could subscribe to any number of these lists, and give each list a weight. For each posts they encounter, their client would find the post in each list, and take the weighted average of the tags on the post. It would then filter out posts with weights above/below certain values. Eg they might like the ignore all posts with "spam" above .5, or if the wanted to look at programming posts, they might filter for posts with "programming" above .2.

note: this wouldn't be a democracy, and it doesn't require (though could be used alongside) federation of either the posts or the moderation. It could be implemented as a normal image board,

>>997862

Yeah that's how real democracy functions - like absolute shit. That's by design.

>>998019

What

>>997994

I think most people would be too lazy to maintain those lists. And it would quickly devolve into reddit faggotry. Instead of "le upboated xD" they would say "le tagged xD"

>>998050

>complains about reddit faggotry

>posts with flag

kill yourself

>>998086

>complains about flag

>posts with flag

???

>>998089

>is a retarded faggot

!!!

TL;DR, did someone sugget subscriptions to mods?

>>997593

sounds like hell. Leaving moderation to machines is foolish, ignorant and extremely dangerous. Better tools for blocking and muting content you don't like is a good idea and all that you need.

>>998090

>

???