Nov 28, 2017 03:52 PM EST
Does autonomous driving car make better decision than human driving car?
Google has said it will not have a "trolley dilemma," a major ethical problem AI will bring.
Google said in an interview with the Asia-Pacific press conference held in Tokyo, Japan on Monday.
The trolley dilemma is a major thought experiment in the field of ethics as a question of whether or not an AI algorithm should have a direct impact on human life, what to judge based on, and whether such an algorithm can be left to the algorithm.
For example, if you are a driver on a broken train, you are in a situation where you decide whether you can manipulate the line transformer and make a choice to sacrifice one worker on another line to save the five workers working on the track.
Jeff Dean, Google Senior Fellow, commented on the trolley dilemma: "I do not code that artificial intelligence is coded and two people are on the left and three people on the right make the right use." Purpose. "
"Autonomous driving cars have more advantages than people driving," he said. "Like humans, artificial intelligence can see more in the middle without letters. Ultimately better results than people. "
Google has also vigorously rejected the notion that artificial intelligence can be biased. "Artificial intelligence is causing problems in many areas," said Jeff Dean. "Google is also paying attention to making sure that artificial intelligence is well represented in the training data."
"For example, it is not appropriate to use data collected in the United States in Pakistan," he added.
"It is not necessary to eliminate all biases," he added. "It is necessary to bias cancer cells in tests that detect cancer in a human body."
It has a positive prospect that new jobs will be created about the point that artificial intelligence robs people's jobs.
"Automating some jobs through automation is not a new phenomenon," Jeff Dean said. "Technology has evolved over two decades, replacing human labor." "Technology has created interesting jobs for the time being," he predicted. "Artificial intelligence and machine learning will create jobs that I do not think are possible."
"The fear of artificial intelligence is a horrible fear of the future," he said. "There is nothing worrying about the current system."
Meanwhile, at the press conference, a variety of announcements were made about products that Google has combined with AI. Isaac Rwynolds, Product Manager for Google Pixel Camera, "Artificial Intelligence and Hardware," Linne Ha, Google Research Program Manager, "Artificial Intelligence and Natural Language Processing," and Pravir Gupta, Google Assistant Team Engineer The chief executive announced 'Google Assistant'.
In addition, examples of various applications through artificial intelligence machine capabilities were also disclosed. Kaz sato Google cloud platform developer is Cloud Machine Run, Lily Peng Google Product Manager is' Machine Learning and Medicine ', Victor Anton Dr. Victoria University of Wellington' Machine Learning and Environment Protection 'in the United States.
Join the Conversation