The latest winners of the Torring Award again warn of the risks of artificial intelligence

Two corridors who are corridors Today he received the Torring Award for this year To create basic training techniques for artificial intelligence training, the spotlight is used to shine anxiety over the risk of impulsive models of artificial intelligence for public consumption.

A researcher at the University of Massachusetts Andrew Barto and former Deep Mend’s research scientist Richard Soton warned that artificial intelligence companies do not test comprehensive products before their launch, which is similar to development by “building a bridge and testing it by making people use it”, ” according to Financial times.

The TURING Award, which is often referred to as the “Nobel Computing Award”, holds a $ 1 million prize and was jointly awarded to Barto and Sutton to develop “reinforcement learning” – a method of automated learning that trains AI systems to make improved decisions through experience and error. First Vice President in Google Jeff describes this technique As “lynchpin of progress in artificial intelligence” and remained a “central column of artificial intelligence” Google Alpha Before that.

“The launch of programs for millions of people who do not have guarantees is not a good engineering practice,” Parto said Financial times. “Engineering practice has evolved to try to reduce the negative consequences of technology, and I do not see the development of companies that develop.”