Controlling Super-Intelligent Artificial Intelligence Is Almost Impossible, According To Scientists

Controlling Super-Intelligent Artificial Intelligence Is Almost Impossible, According To Scientists


Scientists from the Max Planck Institute theoretically investigated whether a super-intelligent artificial intelligence could be controlled. Unable to reach a clear conclusion as a result of the researchers, the scientists concluded that it is almost impossible to control such artificial intelligence.

Artificial intelligence, especially in the last few years, has begun to reach the deepest points of our lives. Now we can get help from artificial intelligence for anything in our usual life cycle. There are also those who think outliers when a certain segment defends artificial intelligence to the end. Even according to some conspiracy theories, artificial intelligence will capture our lives, and perhaps the end of humanity will come because of artificial intelligence. It seems that this theory also attracts the attention of scientists.

A group of scientists working at the Max Planck Institute began to investigate whether a real super-intelligent artificial intelligence can be controlled. Research on a theoretical basis has concluded that such a thing is almost impossible. Frankly, if you're afraid of artificial intelligence, we can say that the next part of our news may make you even more nervous.

What's super smart AI?

Can artificial intelligence be controlled?

Artificial intelligence technologies used today are getting a little better every day. Machine learning techniques used to improve such technologies are making artificial intelligence a little smarter every day. The phenomenon that we call super-intelligent artificial intelligence does not actually have a full meaning. That's because the human brain and our perceptions aren't ready for something like this right now. That's exactly why the scientists ' research remains in theory. Because in order to develop and control super-intelligent artificial intelligence, it is necessary to have a more powerful algorithm (protection algorithm) than it.

What scientists are wondering is whether super-intelligent artificial intelligence can be controlled. But in order to experiment with this, it is necessary to develop super-intelligent artificial intelligence, and a more powerful technology or mind is needed that can control or stop this artificial intelligence. It is believed that a truly super-intelligent artificial intelligence cannot be managed, since the results of this type of artificial intelligence are unknown, and it is also unclear whether it is more powerful than this artificial intelligence.

It may not be possible to identify all scenarios and run the "do no harm" command as a result of these scenarios

According to theoretical research by scientists, super-intelligent artificial intelligence is a versatile algorithm. This, in turn, suggests that the artificial intelligence in question means that the number of possible scenarios is infinite. In addition, such an artificial intelligence has the ability to mobilize various resources to achieve a goal. It's this ability that makes things even more impenetrable.

Computer scientist Alan Turing pursued an interesting theory in 1936. This theory questioned whether a computer program would stop when it failed to achieve results or enter into an infinite loop. Having reached a clear conclusion through mathematical studies, Turing had shown that such a question could not be answered, at least for some software. For a super-intelligent artificial intelligence, you can be given a command not to end humanity. Even artificial intelligence can implement this command. But mathematically, how and how to implement this command is incalculable. That's one of the reasons why super-smart AI can't be controlled. Explaining this, Iyad Rahwan says that the inability to make mathematical calculations for such an artificial intelligence will also override the protection algorithm.

On the other hand, restricting super-smart AI can provide a frightening afterlife, according to the researchers. Scientists who also give examples of this say that this artificial intelligence may not use a certain part of the internet or may be blocked from certain networks. But when this idea comes to the fore, the concept of super-intelligent artificial intelligence has no logic. So that prevents the research from connecting somewhere.

The scariest explanation on the subject was made by Manuel Cebrian. Cebrian, who continues his work at the Max Planck Institute, said that what they are thinking about may seem meaningless, but even today there are artificial intelligence that can perform some important tasks. In addition, the programmers who developed these algorithms do not fully understand how artificial intelligence learns to perform these important tasks.

For More Blogs ;