In case you don't know, the Chinese room is a thought experiment that assumes there is an algorithm that is able to talk to people in Chinese, passing the Turing test, answering questions appropriately and is generally indistinguishable from a real (Chinese) person. This algorithm is written in english in a huge book and you are sitting in a room and have everything you need to execute the algorithm on paper. Through the slit in the wall, people can pass you Chinese writing and you can pass them the responses back. Assuming you don't speak Chinese, you do this procedure without understanding what is going on and that is why John Searle, the Philosopher who originally came up with this experiment, claims that AI can't really understand anything either.
However, in my view this experiment neither proves nor disproves anything. The "mistake" that John Searle makes, in my view, is that he replaced the computer with the human but the algorithm still exists. However, when we are talking about AI, we rarely speak about hardware. It's the algorithms that constitute the intelligence and for algorithms it does not make a difference if they are executed by electronics, using pen and paper or even falling dominoes. The real question, which the Chinese room does not answer but illustrate quite nicely, is weather this theoretical construct, called an algorithm, can understand words and be conscious like us.
Just to remind you, we don't know which part of us is conscious. A somewhat related question is, weather consciousness is located in our brain or weather it is contained within the information that is processed by our brain. Or maybe it's somewhere else entirely. Research in AI will not solve this question, but it might, maybe, give us a few more hints about it and that what I find most fascinating about AI research. Being able to turn the lights on without getting up from the sofa is just a nice side effect.
