The real reason artificial intelligence can be dangerous
El main danger of artificial intelligence
Several experts have already warned of this possibility about the real danger of artificial intelligence. Stuart Russel warns del principal danger of AI.
The professor at the University of California at Berkeley, Stuart russel, is the author of the work Human Compatible: AI and the Control Problem and who is an expert in the advances of machine learning with artificial intelligence, has explained the reason for his enormous concern about why the AI can be dangerous.
What he explains in his new book is that what should concern us is not that robots with artificial intelligence become aware and rebel against human beings. Rather, the machines become so effective in accomplishing the goals we assign to them and signal that we may end up being inadvertently annihilated. How? setting the tasks in the wrong and / or wrong way.
According to the professor, the general idea about the danger of AI it is completely wrong due to the influence of Hollywood movies. These, in general, always consist of the machine that becomes aware of itself and then begins to hate humans and ends up rebelling against them. The professor refutes this view because he explains that robots do not have human feelings, so it is wrong to worry about something like this happening.
Russell has explained that an evil conscience is not the reason to worry, the danger of artificial intelligence it is the potential of their ability to achieve a misplaced or specified objective that should really concern us.
An example of that effectiveness
The expert took the opportunity to explain to the media an example of a situation like the one he exposes.
If we had a very powerful system of IA which has the ability to control the planet's climate and we want to give you indications to return the levels of CO2 (carbon dioxide) in our atmosphere.
Artificial intelligence determines and concludes that to do this, human beings should be eliminated, because they are the main cause of the massive production of CO2.
Russel made it clear that the important thing now is for human beings to regain control.
We have created artificial intelligence that is highly capable of performing tasks in a matter of milliseconds. Artificial intelligence capable of self-optimization and learning by itself. As long as we are aware of its dangers and direct these profits to ethical tasks there will be no problem. But…
Will we be able to get it on the right track?
And you, what do you think it is the main danger of artificial intelligence?