A thought experiment due to John Searle attacking the Strong AI Postulate.
- A person in a locked room carries on a dialogue with us by way of Chinese written on paper passed back and forth under the door.
- The person in the room responds according to instructions stored in a vast library of rule-books, and does not understand Chinese.
Since the person doesn't understand the language and the rule-books obviously lack understanding, Searle claims that there is no real language knowledge involved. Searle likens dialogue with a computer to this situation, and hopes that it makes it clear why he says that computers are not aware.
The scenario has been widely debated, but proponents of Strong AI point out that the system room + person could be said to possess knowledge of Chinese, in just the same way as the neurons in a human brain (which themselves lack knowledge about Chinese) can form a system that can know the language.