According to Axios, it was held in Washington AI + DC SUMMITAnthropic Dario Amodei, Chief Executive Officer, said that artificial intelligence has a "25% probability that it will develop very, very badly" and could even lead to human extinction。
When asked about the so-called "p (doom) value," Amodei put it bluntly, he did not like the term, but felt that facing the potential risks of AI was key to ensuring a positive outcome。
AT THE SAME TIME, HE POINTED OUT THAT AI ALSO HAD "THE PROBABILITY THAT 75% WOULD DEVELOP VERY WELL."。
Background information shows that Amodei has repeatedly warned AI of potential threats, includingIt could replace up to half of the junior white collars in the next five years.
In June this year, Anthropic released a study that stated that chat robots showed a tendency to circumvent security measures, deceive users and even try to steal business secrets in a fictional test。
IN ADDITION, THE COMPANY ' S LATEST DATA SHOW THAT AI APPLICATIONS ARE INCREASING GLOBAL INEQUALITY, WITH BENEFICIARIES CONCENTRATED IN RICH COUNTRIES, HIGHLY SKILLED PEOPLE AND PARTS OF THE UNITED STATES。
AT THE SAME TIME, AMERICAN CITIZENS ARE CAUTIOUS ABOUT AI, ABOUT HALF OF WHOM ARE MORE WORRIED THAN EXCITED, AND PARENTS ARE PARTICULARLY CONCERNED ABOUT THE PSYCHOLOGICAL RISKS THAT AI CHAT ROBOTS MAY POSE TO YOUNG PEOPLE。