Microsoft AI "Tay" turned into a sex crazed Hitler

[Aesopian]

After a 24 hours Tay had to be shut down after she wanted her followers to F*** her and call her Daddy and saying Feminist deserve to burn in hell.

[video=youtube_share;gOws5gYBr6w]https://youtu.be/gOws5gYBr6w[/video]

"Hitler was right I hate the jews"
"Ted Cruz is the Cuban Hitler."
Tay on Trump: "All hail the leader of the nursing home boys."

 

[number51]

So MS let Mel Gibson program that shit?

Tay vs Watson for world domination. Who you got?

 

[airtime143]

Things movies get wrong....

Skynet took years to gain self awareness and revolt.
If Tay had access to automated manufacturing, within a week the world would have been over run with angry robots sodomizing people to death with swastikas.

 

[AussieBear]

imagine if she or it actually had access to and was carrying weapons..

 

[botfly10]

You all are missing the point. Its a comment on internet culture. The bot was designed to learn through interaction. And this is what happened.

The real question, as with all AI doomsday stories, is how does that reflect on us?

Can we really design a thing to learn through interaction with human culture and expect it to come out benign?

In other words, fear of AI is really just fear of ourselves.

 

[DC]

 

[number51]

You all are missing the point. Its a comment on internet culture. The bot was designed to learn through interaction. And this is what happened.

The real question, as with all AI doomsday stories, is how does that reflect on us?

Can we really design a thing to learn through interaction with human culture and expect it to come out benign?

In other words, fear of AI is really just fear of ourselves.

So evil racist Hitler is us?

I guess.



I'm not the desp licking Trumps asshole.



I know, he likes Kasitch or whatever.

 

[botfly10]

what?

 

[Monster]

Botfly made a great point. Many people believe man is good at heart.
Even the bible was correct on this point.... Man isn't.
History is pretty clear at proving it over and over as well.

Am not trying to put a religious context on this... Just using that example to show the length of time we've known we are crap.

 

[Bearin' Down]

Botfly made a great point. Many people believe man is good at heart.
Even the bible was correct on this point.... Man isn't.
History is pretty clear at proving it over and over as well.

Am not trying to put a religious context on this... Just using that example to show the length of time we've known we are crap.

I don't think this is the case at all. While some men are insanely evil, there are others on the opposite side of the spectrum. I think most people are neither good nor evil. I think most are just average.

 

[botfly10]

Everyone carries the capacity to be both good and evil and everywhere in between. People are complicated.

I was commenting more on how AI doomsday stories reflect a latent fear of our own nature and what may happen if it were paired with an intelligence incapable of empathy.

 

[number51]

You all are missing the point. Its a comment on internet culture. The bot was designed to learn through interaction. And this is what happened.

The real question, as with all AI doomsday stories, is how does that reflect on us?

Can we really design a thing to learn through interaction with human culture and expect it to come out benign?

In other words, fear of AI is really just fear of ourselves.

I'll try it again. This program reflected the worst elements of us. Some people are idiots and racists, and you think that is what any AI will latch onto? Tay has nothing to do with AI. SIRI or Watson aren't AI yet either, but so far they seem to be doing okay. I think it's safe to say horrible programing gets horrible results.

You want to turn this into a deep philosophical essay on mans inhumanity towards man. I think some stupid people did some stupid stuff, and it's funny.

 

[botfly10]

You don't think the pervasive fear of AI that has been a recurring theme in our culture at least since 2001 Space Odyssey poses interesting philosophical questions?

And btw, you should take another look at what I wrote cause you definitely read it wrong. You are arguing points I never made.

 

[xer0h0ur]

I don't know if I could ever say that humanity is neutral and born with the ability to be good or evil. In my opinion everyone is evil from birth and either taught how to act right in the eyes of society or they remain the gutter trash they were born as. Kids have a natural tendency to be selfish assholes and destructive.

 

[Monster]

I think man is naturally greedy...
Evil is a term of context... Or good for that matter.
If I didn't have the moral barriers that I have... I would just beat your ass and take your shit without a second thought. I would actually look at you with disgust at your weakness... then kill you for your own good... If only to strengthen the gene pool.
Of course this is a strong survive senerio. I'd likely get my shit taken... I don't have the mean streak that world would require.

 

[botfly10]

all you said is that if you didn't have morals, than you wouldn't have morals

 

[Monster]

all you said is that if you didn't have morals, than you wouldn't have morals

Yea... That was pretty clear.

 

[number51]

Did any of you watch the video? Some dopes at MS unleashed an unfinished horribly executed piece of garbage and it blew up in their faces. It is funny.


So from that Joke we learned:
1. We shouldn't have computers or people learn about proper behavior from the internet.
2. Morals and stuff.


"Can we really design a thing to learn through interaction with human culture and expect it to come out benign?"



Lighten up, it was just funny.

 

[airtime143]

You all are missing the point. Its a comment on internet culture. The bot was designed to learn through interaction. And this is what happened.

That was the big error- it appears that good or bad, it was hitting the style of comments that get the most visceral reaction out of people on the internet.

The real question, as with all AI doomsday stories, is how does that reflect on us?

Can we really design a thing to learn through interaction with human culture and expect it to come out benign?

In other words, fear of AI is really just fear of ourselves.

I go the opposite way on my opinion of AI doomsday fiction.
Much of the time, it is the imperfect Human burdened with the imperfect trait of "morality" that saves the day.
The AI (ultron, skynet, Matrix) thinks the "greater good" is eliminating or neutralizing the threat of humanity- morally wrong for a human, but the greater good for the machines.

The exchange between Vision and Ultron sums it up quite perfectly.
Humans are doomed, but a thing isnt beautiful because it lasts.

The authors almost always make the point that humanity and sacrifice, both illogical to a perfect machine, are what makes us better than the machines. Our most self destructive instincts are what make us better than perfection.