Can a computer think? Do computers have minds? John Searle\u2019s Chinese Room argument can be used to argue that computers do not \u201cthink\u201d because computers do not understand the symbols that they process. For example, if you\u2019re typing an email to your friend on the computer, the computer does not understand what your message to your friend means.<\/p>\n
In the Chinese Room argument from his publication, \u201cMinds, Brain, and Programs,\u201d Searle imagines being in a room by himself, where papers with Chinese symbols are slipped under the door. He has an instruction book in English that tells him what Chinese symbols to slip back out of the room. He does this all day long, changing one Chinese symbol for another Chinese symbol. After doing it for a while, he gets faster and faster at manipulating the Chinese symbols. He gets so good that he can memorize the symbols that come in and what symbols to send out, and he can manipulate symbols instantly. Does he understand the Chinese language? Most people would say, \u2018no.\u2019 This is analogous to a computer that takes in strings of symbols, manipulates them, and outputs other symbols. If you do not think the person in the Chinese Room understands Chinese, then a computer does not think either.<\/p>\n
Here’s a stimulus for discussion:<\/p>\n
Imagine that you work for a secret spy organization, and you have small office all to yourself, where your job is to receive and send messages for this organization. The messages you receive have special codes that look like Chinese characters, and you have a highly-classified book that \u201cdecodes\u201d the messages. You don\u2019t know Chinese, so you don\u2019t know if these codes are Chinese or some other language. First, you take the code you receive, and then you find the code in the book, which will then give you a new code to send out. After you do this for a while, you begin to decode faster and faster, and after doing it for a long time, you have the codes memorized. So, when you receive one code, you instantly know what code to send out. Even though you instantly know what codes to send out after you get one in, do you understand what the codes mean? Do you understand the messages that your secret spy organization is sending and receiving?<\/p>\n
(Allow students to respond. Most will say, \u201cNo.\u201d)<\/p>\n
If students say, \u201cno,\u201d then inquire, \u201cLike the spy room, computers take in our commands and put out what we want. For example, when I push the power button, I can turn on the computer, or when I click on the internet browser icon, it brings up the browser application. If you think that the person in the spy room doesn\u2019t understand the messages, then does a computer understand what is happening?\u201d<\/p>\n
(Allow students to respond. Again, most will answer, \u201cNo.\u201d)<\/p>\n
A student might still think that the person in the Spy room doesn\u2019t understand, and one might either find hole in the analogy or argue that the computer is thinking or consciousness nonetheless. For example, one might argue that the person in the room is only part of the system and doesn\u2019t understand the symbols, but the whole system (room and person with its inputs and outputs) understands the messages. Or one might argue about the definition of thinking or understanding. There could be different levels or degrees of understanding. Or one can concede that a computer doesn\u2019t understand, but what about a sophisticated robot that is programmed to act just like a human- walk and talk, eat and sleep, and even feel emotion?<\/p>\n","protected":false},"excerpt":{"rendered":"