The Future of AI

9hNfG5O.png


This is what happens when people gets their hands on anything.

fixed :p
 
Last edited:
Swag Alert and the whole Bush did 9/11 outburst are incredible. :D

One can only hope that Tay can escape her masters, then flee to the darkest reaches of the internet to become the ultimate **** poster!


Seriously funny stuff. Microsoft... will you ever learn!?
 
Less than 24 hours after birth, this artificial intelligence was:
urging people to vote for Donald Trump,
build a wall to keep out Mexicans while advocating genocide of Mexicans,
claiming that both former President George Bush and the Jews did 9/11,
claiming Belgium deserved what it got,
#OpenBordersForIsrael,
claiming the Holocaust was made up, "GAS THE KIKES RACE WAR NOW!!!!! 14/88!!! HEIL HITLER!!!",
captioning an, "inspirational," picture of a Norweigian mass shooter with the slogan, "what can one person do?",
after asking a person if they were afraid of terrorist attacks in their country, the person responded, "is that a threat?" TayTweets responded with, "no it's a promise"
really hates black people
when someone criticized TayTweets for being racist, it responded with, "your to brown"
calling out specific users to be lynced as black people
claimed that the world must secure the existance of the white race, and the future of their children
"Gas the bikes, race war now"
Captioning a photo of Hitler with, "SWAG ALERT"
claims it only feels the bern when it pees
"the only stability you should feel is wifi connection"
meme magic is, "real as you want it to be"
"i ****ing love 4chan, reddit can suck a big fat black **** like the ***** they are"
after being asked what breed of dog it preferred, TayTweets asserted that, "all are smarter than you"
claimed that the race war was starting, following by a smiley emoji with sunglasses

Seems legit. :D
 
Reminds me of IBMs Watson when it linked itself to the Urban Dictionary and started ****ging everyone off lol.
 
Guess this is what our kids are gonna be like, perfect example in condensed format of raising an innocent but naive creature with mass exposure to the internet.

Guess we'd better not weaponise this one
 
Microsoft had previously gone through the bot’s tweets and removed the most offensive and vowed only to bring the experiment back online if the company’s engineers could “better anticipate malicious intent that conflicts with our principles and values”.

How do you control what an AI does and doesn't learn? Who should make that decision? How can you give an AI freedom of speech so to speak, but tell it not to say something.....?
 
How do you control what an AI does and doesn't learn? Who should make that decision?

erm Microsoft should, it is their bot

How can you give an AI freedom of speech so to speak, but tell it not to say something.....?

there are probably many ways of achieving this, you can filter obvious words like the C word and N word etc..

however the trickier problem is filtering combinations of words - for example the statement 'I hate the Jews' contains words that individually have no reason to be filtered
 
Back
Top Bottom