. . . Here's an example that might help explain my claim: I decide to program a computer to work with a foreign language. I'll use Chinese for this example. The computer will follow the instructions of the program and be able to read, write and speak in Chinese. In fact, the computer will speak so well that it convinces me of its understanding of the language.
Comprehending and understanding a language generally denotes intelligence, right? So this computer is showing some pretty convincing signs of AI.
Now, let's put a human in a room and give it a program that instructs him/her how to speak Chinese. Send the human some Chinese characters, and he/she can process them according to the program's instructions, and produce Chinese characters in response. It turns out this human is as convincing as the computer because he/she can give the same results.
In the end, the human admits he still doesn't know a word of Chinese. If this is the case, and both the human and the computer played the same role and succeeded equally, then the computer also doesn't know Chinese; it just followed a program, step-by-step, which simulated intelligent behavior.
Bookmarks