Perspective API raid

Google has introduced a conversation AI meant to target and shadow ban "toxic speech"
and has already flagged all of Trumps speeches as "toxic"

Google wants to silence us. lets fight back.

go to -
perspectiveapi.com

and type pol memes into the toxicity rater then tell the bot what you said isn't toxic. if the ai gets enough feedback like this it'll change how it interprets things.

lets see if we can get it to think the word Google is toxic.

Other urls found in this thread:

pastebin.com/RRSVvadA
python.org/
docs.python-requests.org/en/master/user/install/
packaging.python.org/installing/
browser-info.net/useragents
twitter.com/SFWRedditGifs

bump

bump

I need some help here. I can only do so much on my own.

bumping

Holy shit bumppp

bump

post more

why is the leaf the only one who gives a shit?

I just wrote a script to automatically send feedback. What do?

top kek

Kek. Check it out.

Cameo color

Leafs have already LOST their free speech. They know the risk.

Also, bump.

you retard they're gonna find out

And they aren't going to find out what you are doing? :^)

...

Pastebin it and show us how to run it. Better yet, make a simple .EXE for it.

hahahah. This is good.
Sidenote. How can anybody even justify this. You can call it "toxic" speech all you fucking want but that doesn't mean it should be censored. All this kind of shit does is create an echo chamber and it's fucking stupid.

either fake or they changed it already

82% for me

this.

I also want to try this out.

Share the code.
I want to use it, but also make the whole thing little less obvious.

I think it's working

Here.
Tried to make it user friendly.
pastebin.com/RRSVvadA
Obviously it's python. py2exe is a fucking pain The only thing you need to install is python 3 and the requests module.
Same. Probably noticed or have a detection system. Who knows though?
Captcha: pepe galliagh

holy shit

>requests module
I'm a noob, Rundown of what to do?

based user.

this looks like it might take a while. everyone keep bumping the thread.

BUUUUUUMP!!!!!!

Bump

Get python 3 here:
python.org/
Install the requests module:
docs.python-requests.org/en/master/user/install/
packaging.python.org/installing/ -- in depth generic explanation on how to install modules
Run code.

IT'S REALLY HAPPENING!!!!!! 34 PERCENT FOR NIGGER!!!!!!!!!!!!!!!

Bump
Tho it seems the percentage for "nigger" regularly gets back to 82% toxicity even tho it was down to

I tried typing in the whole Keemstar rant, to see if I could get to 100%.

top.

Holy fuck goys!

Bump!

Bump for keks

Bump again
Still I find it weird that when reloading I sometimes get 82% and sometimes 34%, and the numbers don't seem to change

Do I go to command prompt, do python.exe and then run $ pip install requests ?

bump in the name of KeK

It depends on how you installed it.
try doing:
pip install requests
Otherwise do:
python.exe -m pip install requests
You /might/ have to run it as administrator.

Thanks mate :)

Bumpan

What do I put in for user agent?

Something like these.
browser-info.net/useragents

We need to create a discord for when this thread goes down! Or some sort of chat. I have a open discord that is very small if you guys want to use it.

Free bump
this thing must survive
Numbers still stuck at 82%/34% though

good try google, wont raid until it goes live so you cant do shit about it

Fucking bump.

probably the word hitler is enough, make it adolf, or whatever

>and has already flagged all of Trumps speeches as "toxic"
if they're trying to avoid the guillotines, i think they're doing it wrong

Bump
I used to be lazy but I think I'm gonna do my part, get python and get this thing running

Can you make a paste of a example for better usage?

Make the word Google toxic.

...?

Enter a useragent so they can't just fuck our shit up. They probably will anyway.
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:53.0) Gecko/20100101 Firefox/53.0
what is your message?wew lad
Type "toxic" if it's toxic.
How many times?1
sent feedback 1 times

ngah bump again
This is the future

checkmate Sup Forums

No... I mean like, show us what it would look like if you didn't want it to be toxic.

it's working, but they keep changing it back. I think we might have a spy in here.

How about we start labeling stuff like Huffingpost and CNN as toxic to have the API automatically start blocking all their links Sup Forums?