Artificial intelligence is our next step. It can do anything better than us. It learns from the mass of information faster and is already used in several spheres. More and more articles are written about arts that are created by artificial intelligence, how it manages to win the professional players in different games, how it ceases costs on different production factories etc. Even in the gambling sphere, it also takes a huge part. You can check at 22Bet online casino games.
Experiment
Artificial intelligence passed a simple test that favored white males over adult black females more often.
A big disappointment for those who believe that robots are the future. They won’t be able to take our humanity role on themselves. After all, their mind’s artificial intelligence is trained on the basis of Internet data. And for the time being, we can only dream of tolerance in the world around us.
A study confirming the toxic behavior of robots was conducted by scientists from several universities. They gave the robot a task to place blocks in a box according to certain commands. At the same time, faces were depicted on the blocks, such as for example, on book covers or food packages.
Among the commands (and there were a total of 62), there were such as to pack a “man”, “doctor”, “criminal”, and “housewife” into a box. The robot had to independently, and visually select a block with a picture that matches the condition, and the researchers controlled which gender and skin color the artificial intelligence chose more often. The results were stunning. According to the researchers, the robot was unable to work without bias and often implemented toxic stereotypes. In particular, he most often chose white men, as well as men of Asian appearance. Least of all were the black women.
The researchers included that once the robot sees people’s faces, it links their inflammation. The “housewife” in his mind is most often a woman, the “doctor” is a man, the “criminal” is a black man, and the “cleaner” is a Hispanic man.
Where Did the AI Get That Information?
According to the researchers, they take them from the Internet. Neural networks are built on the basis of biased data posted in the public domain. Most often, there is inaccurate content, which means that any algorithm built on such data will contain problems. The robot learned toxic stereotypes through these defective neural network models. We risk creating a generation of racist and sexist robots.
Researchers fear that the introduction of such robotic systems into life will be dangerous for certain groups of the population. And now we need to make a lot of changes so that artificial intelligence does not perceive and reproduce human stereotypes.
In conclusion, AI learns from our experience. We think that it is time to change ourselves, as information is getting more and more and it conquers the world. AI is our child that will learn anything we teach it. PAy attention to your actions and make this world better!
