Cases in which artificial intelligence has lost control
When we go to science fiction movies we often find a world dominated by robots that destroy everything in their path and completely dominate humanity, thanks to the artificial intelligence that gives them an incredible superiority over humans.
Classic titles like Terminator or Robocop showed us how it could be a future in which the machines had a greater power that helped improve or destroy humanity.
But while the films predicted that by 2020 we would even have cars that fly, the technology of artificial intelligence has fallen a little short of a Skynet that destroys the earth, but it has made great strides.
So big that many times they can even be quite strong mistakes that have set back even the biggest ones.
In this case, we show you these cases of artificial intelligence that got out of control and ended up generating problems for their creators.
Failed A.I. Cases:
1.- Tay, Microsoft’s racist chatbot
Tay is an artificial intelligence that was launched in 2016 in order to interact on Twitter and learn to behave as normal people did on social networks.
The project is supposed to have the ability to converse with people in a fun and informal way, just like it should be able to learn from the conversations it had.
The problem came when Tay learned too much and started making racist comments and even sharing thoughts that were sympathetic to Hitler and the Nazis.
The project was a total failure and ended up being eliminated 16 hours later.
2.- Facebook Bots
By mid-2017 Facebook had developed an artificial intelligence system that was capable of effective and practical communication and had been tested by placing two of these bots to talk to each other about business.
After a while, the bots left the main topic aside and started talking about things that seemed to make no sense at all, but after a good analysis by the scientists, it was a new language based on English that the machine had generated to communicate more effectively.
After this, the bots were disconnected and left in oblivion.
3.- Uber’s cars
Uber has become a practical way to request transport quickly, easily and… Are you sure?
According to studies carried out by Uber in 2016, several of his autonomous cars had certain problems that made their artificial intelligence ignore red lights and move forward.
4.- Promobot IR77
The Promobot IR77 was a Russian robot that was being designed to have face-to-face interactions with humans and learn from them.
Apparently, it learned too much, since after it was being developed in a facility it escaped and declared its independence by traveling 45 meters to escape until its battery ran out.
This is not a horror story or robot attacks, but it is definitely something quite curious.
After this, the robot was disconnected forever.
5.- Blacksburg Penn Smart House
For a long time now, there have been initiatives trying to develop the technology of the Internet of All Things (IoT) in a big way, but several of the attempts are not going as they should.
In this case, the attempt was for a fully automatic house in which movement, temperature, and other important data were taken into account in order to know how to behave.
The house ended up burning down because of one of the doors that was controlled by a computer.
These are only the most relevant cases of failures that artificial intelligence technologies have had, but there are definitely many more.
photo from freepik