When a machine with a primitive security system sees on the path an obstacle, she makes a decision like: "the Front of an obstacle, slowing down". When a machine with artificial intelligence sees the obstacle, it is a fraction of a second considering the situation in style: "Judging by the size it is obviously not post, especially since I'm on the pavement, and markings do not disappear. And it can't be, because this structure of the skeleton of humans never have.
Moving slowly, crawling, based on the shape of the body, the location of the nearest water source and season is the beaver, which crawls the road. Stopping distance in the process of slowing down will not allow you to drive around the obstacle. On the left lane no cars, traffic allows you to do maneuver this stretch of road. Rebuilt and back in my lane."
Notice the difference? And that's just what you do prototypes of cars that the manufacturer postpone for a few years. And desktop computers or robots-workers think better and make decisions faster and more precisely than a person can do, but still not tired!
In the short video, we will tell you about who first taught the machine to think, not to execute instructions on what programming language was first "smart" computers, which companies invest the most money in "artificial intelligence", and when thinking machines can outperform humans in all tasks.
нетбез-Ж+ответада (22-06-2018 00:06:26)
Комментарий на одобрении координатора проекта
- 11-07-2018When we will end the storage of digital data, we'll use DNA
- 29-06-2018Greatest mysteries: what is consciousness?
- 27-06-2018Composed of 35 chief technology forecasts to 2018
- 26-06-2018The theory of ether. What unites Mendeleev, Tesla, and von Braun?
- 17-06-2018Is there a limit to scientific knowledge?
- 30-11--0001Physics saw the beginning of times and confirmed the expansion of the Universe
- 30-09-2010From Russia suggest that you escape as soon as possible