John Searle's thought experiment, the Chinese Room Argument, is used to address this the question, "how far can we advance with AI, in this episode of The AI Minute. For more on Artificial Intelligence: https://voicesinai.com https://gigaom.com https://byronreese.com https://amzn.to/2vgENbn... Transcript: What does it mean to understand something? There have been recent stories in the news about artificial intelligences understanding contracts or legalese, or being able to read medical journals and understanding them. The idea of what understanding is best illustrated by a classic thought experiment by John Searle called the Chinese Room Argument. It goes something like this: there is a man who doesn't speak any Chinese, he doesn't know anything about Chinese, and he's in a room with a bunch of special books. People slide questions to him under the door written in Chinese, and he can go through these books, looking up matching symbols, until he gets to a point where the book instructs him to copy down symbols; that again he doesn't know what they mean, and he slides his answer back out under the door. Now, the answer is perfect, but the man doesn't know whether it was about cholera or coffee beans or anything like that. So the question is does the man understand Chinese, and at root what you think about that question should inform how far we can advance with artificial intelligence.
2019年4月18日星期四
How Far Can AI Advance?
John Searle's thought experiment, the Chinese Room Argument, is used to address this the question, "how far can we advance with AI, in this episode of The AI Minute. For more on Artificial Intelligence: https://voicesinai.com https://gigaom.com https://byronreese.com https://amzn.to/2vgENbn... Transcript: What does it mean to understand something? There have been recent stories in the news about artificial intelligences understanding contracts or legalese, or being able to read medical journals and understanding them. The idea of what understanding is best illustrated by a classic thought experiment by John Searle called the Chinese Room Argument. It goes something like this: there is a man who doesn't speak any Chinese, he doesn't know anything about Chinese, and he's in a room with a bunch of special books. People slide questions to him under the door written in Chinese, and he can go through these books, looking up matching symbols, until he gets to a point where the book instructs him to copy down symbols; that again he doesn't know what they mean, and he slides his answer back out under the door. Now, the answer is perfect, but the man doesn't know whether it was about cholera or coffee beans or anything like that. So the question is does the man understand Chinese, and at root what you think about that question should inform how far we can advance with artificial intelligence.
订阅：
博文评论 (Atom)
没有评论:
发表评论