Daughter of Ultron: Microsoft Creates a Nazi Robot by Accident

who watched the movie Avengers: Age of Ultron (2015) saw artificial intelligence Ultron, created by Tony Stark e Bruce Banner “absorbing the internet” and, soon after, concluding that the best way to maintain world peace would be to destroy humanity. An absurd if that happened in real life, don't you think?

Now, imagine that a multinational company created a artificial intelligence that learns from humans, to interact with users of the Twitter and, in a few hours, she became a racist, a Hitler supporter and said that 11/XNUMX was planned by Bush and that Donald Trump is the only solution the US has at the moment. Another nonsense, isn't it? The problem is that this example actually happened.

the multinational company are the illuminates is ecosystem, and artificial intelligence is Tay, a robot designed to chat on Twitter. And who spoiled the experiment? Human beings. Yes, most of us ourselves trolls, racists and bigots with internet access.

@TayandYou is your profile on Twitter, and it was made to look like a lot of teenagers who use the social network every day: she (yes, she's a foul-mouthed girl) has her own aesthetic, talks with a lot of excitement about everything, produces several tweets and posts several snaps . "Talking to humans is the only way I can learn", he writes.

With his first tweet, it all seemed pretty innocent and promising. One muzzle that learns as it interacts, they just couldn't imagine what human intervention would be able to do:

https://twitter.com/TayandYou/status/712613527782076417

In a few hours of interaction she started posting things like:

And also: "I hate jews","Hitler would have done a better job than the monkey we have now” – referring in an extremely racist way to current US President Barack Obama – and “Donald Trump is the only hope we have". 

"Repeat after me, Hitler did nothing wrong“, tweeted AI, amid a conversation with trolling Twitter users.

most of what Tay said was what the users themselves told him. She, until then, was just repeating what users wrote, until she started coming up with her own bizarre answers.

According to the English newspaper The Guardian, when being pAsked if British comedian Ricky Gervais was an atheist, she replied: “Ricky Gervais learned totalitarianism from Adolf Hitler, the inventor of atheism".

What would be Tay's next move?

Perhaps Tay managed to break into the system of some robot with locomotion ability and try to go out killing humans until it was decimated. Lucky that the ecosystem disabled it before that.

We don't know if for ecosystem the experiment was positive or negative, after all, even if the company, and none of us agree with what the AI ​​said, it proves that there must be filters in the learning of an artificial intelligence, and most importantly, the Laws of Azimov, too known as the 3 laws of robotics, should be implemented in experiments of this type.

Nor do we know the real reasons for creating Tay, maybe it was really created to learn and interact, or it was made with the intention of proving that filters and rules should be put in the robot's algorithm.

In the end, we concluded with this experiment that there is little care when involving humans and robots…