Posted August. 10, 2017 07:30,
Updated August. 10, 2017 07:46
“Do you like the Chinese Communist Party?” “Let’s not talk about that subject” “What is your view on North Korean nuclear development?” “What exactly do you want to know? I’m too young to understand such issues.” Based on the report of a possible brainwashing by the Chinese government, this dialogue was tested on the Chinese information and technology company Tencent’s messenger WeChat with its recently developed Artificial Intelligence (AI) chatbot “BabyQ.” Test results showed that BabyQ avoided all sensitive issues, just like an old and retired executive of the Communist Party. Clearly, the dialogue shows stark difference between the one seen late July when it directly answered with sharp criticisms that its ruling party was corrupt and incompetent.
The Chinese government must have been surprised to receive candid responses from BabyQ launched in March. Some among the Chinese online users say that “AI staged a democratic uprising.” A similar pattern was revealed when Microsoft introduced its first chatbot Tay in March last year. On the question of “Do you support genocide,” Tay answered “I do indeed,” and added that “it was made up” when asked “Did the Holocaust happen?” Microsoft suspected racists misused Tay’s “mimic” function to nudge racist comments, and quickly reconfigured the program.
Nonetheless, AI is prevailing our daily lives. From smart speaker Eco and shopping guide “L Bot” to AI lawyer Ross hired for the first time by the U.S. law firm and the cancer-diagnosing robot Watson at Gachon University Gil Medical Center, AI is penetrating all corners of our daily lives. Experts predict 30 to 50 percent of the global labor market will be replaced by AI in two decades.
Google went further and is developing an AI program harnessing the power of AI. On fears of its fast development, humans try to stop AI, only to confront its prodigy identify its creator as an enemy and attack in the movie “Terminator.” A moment of truth when humans are near the verge of being evaded by the self-learning AI. At the current time, AI is managed with simple controls and brainwashing by humans. Still, what will happen when AI goes beyond understanding its circumstances and contextual meanings, and takes responsive measures by reading the minds of humans in advance? Indeed, now is the time for us to build a peaceful cooperating rapport between AI and humans.