Computers that talk to each other in a secret language, which no one knows, it must not griezeliger. Yet, that is now possible: during an experiment, communicated two AI systems with each other in a coded language which nobody can crack. Writes in The Morning.
It is an invention of Google Brain, the research programme in which Google is experimenting with artificial intelligence (AI). The researchers use deep learning, which means that computers themselves things learning on the basis of algorithms.
The AI-systems of Alice, Bob and Eve had in the experiment, each with its own task. Alice had a secret message to send to Bob, Bob had to take that message to decrypt, and Eve had the two try 'to listen'. It took a while, but after 15.000 attempts could Bob Alices encrypted message smoothly convert in text while Eve the code could not decipher.
According to the researchers used Alice a lot of unexpected techniques to achieve its goal. Their study indicates that robots in the future can talk to each other in a way that we can’t decipher.
The research is preliminary not more than an interesting experiment. Computers have in this area is still a long way to go. But developments in artificial intelligence, going so fast that leading experts such as Stephen Hawking, Elon Musk and Bill Gates have often been warned of the potential hazards.
Source(s): The Morning
No comments:
Post a Comment