German company develops AI dissection of language

Source: Xinhua| 2018-10-29 20:59:09|Editor: xuxin
Video PlayerClose

BERLIN, Oct. 29 (Xinhua) -- Digital gadgets and smartphone applications that can answer simple questions or perform rudimentary tasks such as setting an alarm clock via a spoken command have been offered by companies like Google, Amazon or Apple for several years. The algorithms that allow communication between humans and machines, however, have recently taken a great step forward.

As one of Germany's few successful AI companies, Precire Technologies has developed a language analysis tool that generates psychological profiles to support application processes for new employees, for example. The program called "Precire" learns and recognizes how certain character traits and psychological characteristics affect language and, as a result, can tell which characteristics a tested person exhibits.

"We try to identify patterns that represent psychological characteristics. We do not try to understand content-related aspects, Dirk Gratzel, founder and CEO of Precire Technologies told Xinhua, "when Precire analyzes a speech sample, the program does not know whether you are talking about the weather, artificial intelligence or your workplace. This is comparable to a hair that is examined through a scanning electron microscope. When you see it, first you do not realize that it is a hair. Instead, you will be able to see what the hair consists of," Gratzel illustrates.

However, the majority of Germans are suspicious of the use of artificial intelligence. According to a survey conducted by the German YouGov Institute in September, only 15 percent of respondents believe that the benefits of AI technology outweigh the risks, while 26 percent rate the risks as higher than the possible benefits. "Such a system does not make you more exposed, it makes you more understandable," Gratzel said.

Although Precire is nothing but a highly efficient tool for recognizing psychological traits, it will nevertheless be "extremely important for the way people communicate with machines in the future," Gratzel muses. He expects that future robots and programs which communicate with humans will "recognize who we are, what we are thinking about and what is important to us at that moment".

"I expect that in five years you will no longer be able to know whether your conversation partner on the telephone is a human being or a machine," said Gratzel. Such use of machines would usually be applied to specific situations, such as a telephone call with a bank or a telecommunications provider.

According to the Precire founder, the symbiosis of humans and machines will provide us with an enormous leap forward in development, especially because machines were superior to humans in certain areas. "If we succeed in absorbing machines' abilities within ourselves, we can, for example, improve the processing of visual stimulations in our brain enormously or increase our average intelligence quotient by a factor of ten within two generations."

Physicist Stephen Hawking warned of the dangers that can originate from artificial intelligence during a lecture at the Oxford Union debating society. The renowned scientist predicted several times that the development of artificial intelligence could be the most important, but also the final event in the history of mankind. Tech billionaire Elon Musk even described the "competition for supremacy in AI at national levels as a probable trigger of the Third World War" in September 2018.

"I do not share these dystopian points of view. I am an optimist," states Gratzel. "Every technical evolution has brought us enormous progress as a species. I think it will be similar with artificial intelligence. But it is true that the more powerful our technological achievements are, the greater the potential damage they can cause."

TOP STORIES
EDITOR’S CHOICE
MOST VIEWED
EXPLORE XINHUANET
010020070750000000000000011100001375671241