The words "safety", "hateful", and "harm" are triggering me, but realistically this isn't much different than the algorithms that search engines or parental control filters already use, with a fresh coat of Buzzword Paint slapped on top and sold as a new SaaS.
The fact that it's adaptive as well as automated is the concerning part. With static automated censorship it was relatively straightforward to communicate concepts via workaround, and with manual adaptive censorship there's an element of "can't censor us all" that lets a community spread the word inside itself if it's important enough to them.
Overcoming both played a key part in the genesis of GG back in the day. But this AI, if it's good enough at it's job, will be able to wipe information from communities so quickly and effectively you'll have no idea anything ever happened no matter how hard you look.
youtube already using ai bots to scan people's videos and age restrict them. scan footage. i know channels that when they upload a video (lets say a video game character in a skimpy bikini) and it is private so no one can see, the video will get auto aged restricted by youtube. Although... seems REAL girls in skimpy bikinis seem to be okay.
Tired of internet autists ruining all of their AIs by corrupting them with based takes, pornography and violence, Microsoft have themselves decided to program a new AI to specifically look for based takes, pornography and violence. What could possibly go wrong?
Because they operate on keywords, the autists are just going to find alternative ways and keywords or terms to fuck with the AI even more. When the censors get a hold of a keyword filter system, they always break it because they keep adding more and more shit until it becomes impossible to even use whatever crap they're selling.
This is why I defend Elon Musk when it comes to Twitter and even fixing stuff, the poor guy is probably having to spend half his time with the back end staff unfucking all of the keyword filter bullshit they have on 'right wing accounts' designe to fuck with peoples' access and reach through shadowbans etc. it all runs on the same sort of idea.
Exactly, you could could replace famous brand names for swearwords and write to each other in a form of code. The SJWs would then have no other choice but to block these words whenever they were used especially if the normies start noticing what's going on. This is why keyword filters barely ever work, the only time they do work is when it's just on your own private client and you don't want to see certain shit which I completely get.
The left does this all the time when they want to re-write language but when it comes to censorship it's a great way to get around their bullshit. An example of this would be when everybody started calling covid 'the coof' because they knew the keyword was being aggressively tracked.
“New [AI] models are able to understand content and cultural context so much better,” they explained of the company’s decision to invest in Azure. “They are multilingual from the start … and they provide clear and understandable explanations
The words "safety", "hateful", and "harm" are triggering me, but realistically this isn't much different than the algorithms that search engines or parental control filters already use, with a fresh coat of Buzzword Paint slapped on top and sold as a new SaaS.
The fact that it's adaptive as well as automated is the concerning part. With static automated censorship it was relatively straightforward to communicate concepts via workaround, and with manual adaptive censorship there's an element of "can't censor us all" that lets a community spread the word inside itself if it's important enough to them.
Overcoming both played a key part in the genesis of GG back in the day. But this AI, if it's good enough at it's job, will be able to wipe information from communities so quickly and effectively you'll have no idea anything ever happened no matter how hard you look.
This will ensure that the next generation can NEVER discuss vaccine harms nor lies that lead us into false wars (basically every war since ww2)
Unfortunately right now the far left on other websites LOVE censorship.
We need to show them it impacts their everyday life too :
Want an 8th month abortion? Banned from discussing it if you live in red states.
Want to smoke marijuana in bed before your mom wakes up? This technology will ban you from finding a new source using codewords online.
Want to organize a nightly riot in portland? Big tech will stop it.
Lol lies didn't get us into ww2? All the lying began after the great holocaust origin story of our wretched modern system.
youtube already using ai bots to scan people's videos and age restrict them. scan footage. i know channels that when they upload a video (lets say a video game character in a skimpy bikini) and it is private so no one can see, the video will get auto aged restricted by youtube. Although... seems REAL girls in skimpy bikinis seem to be okay.
Tired of internet autists ruining all of their AIs by corrupting them with based takes, pornography and violence, Microsoft have themselves decided to program a new AI to specifically look for based takes, pornography and violence. What could possibly go wrong?
Because they operate on keywords, the autists are just going to find alternative ways and keywords or terms to fuck with the AI even more. When the censors get a hold of a keyword filter system, they always break it because they keep adding more and more shit until it becomes impossible to even use whatever crap they're selling.
This is why I defend Elon Musk when it comes to Twitter and even fixing stuff, the poor guy is probably having to spend half his time with the back end staff unfucking all of the keyword filter bullshit they have on 'right wing accounts' designe to fuck with peoples' access and reach through shadowbans etc. it all runs on the same sort of idea.
Such as that time company names like Yahoo and Google were used as pejoratives.
Well, more so than normal I mean 🙃
We're saving those for the AIpocalypse. Which isn't now, but may be soon.
Exactly, you could could replace famous brand names for swearwords and write to each other in a form of code. The SJWs would then have no other choice but to block these words whenever they were used especially if the normies start noticing what's going on. This is why keyword filters barely ever work, the only time they do work is when it's just on your own private client and you don't want to see certain shit which I completely get.
The left does this all the time when they want to re-write language but when it comes to censorship it's a great way to get around their bullshit. An example of this would be when everybody started calling covid 'the coof' because they knew the keyword was being aggressively tracked.
No they fucking aren't, stop lying.
nah fuck off i don't need your nannybot.
Scunthorpe.
all gorillas pictures banned