Google faces backlash over autonomous calls by new voice chatbot
DOES DUPLEX ENCOURAGE DOUBLE STANDARDS FOR COMPUTER CALLS?
IN WHAT WAS potentially the first public display of a computer passing the Turing Test, Google’s reveal of its AI powered audio-chatbot (codenamed Duplex) at the company’s annual I/O developer conference in May certainly garnered a lot of attention — from both early adopters and AI doomsdayers alike.
Duplex’s uncanny ability to pass for a human in phone conversations had many conference attendants eager for the time-saving benefits it could provide — such as never having to personally book a haircut, doctor’s appointment or restaurant reservation ever again [ Ugh, such a chore – Ed] — others decided that the important take-home message was to ask whether or not it’s ethical for robots to make phone calls without disclosing their AI nature. University of North Carolina tech sociologist Zeynep Tufekci called the AI-powered conversation “horrifying”, adding that “Silicon Valley is ethically lost, rudderless and has not learned a thing.” This opinion seems to be shared by UNSW AI professor Toby Walsh, who posted on social media that “Computers should not pretend to be humans.”
It’s not immediately clear just how many people are concerned by the new chatbot’s linguistic capabilities and default anonymity, but Google’s CEO Sundar Pichai did footnote that they were still developing the software and wanted to “get the user experience and expectations [for Duplex] right”. So the addition of prior notification that you’re speaking to an AI may already be in the pipeline.