Do Books Think?

I don't buy into the AI concerns. When I was taking my computer science degree, I asked one of my profs (a huge AI fan) the following question. Keep in mind that he thought the Turing test was an adequate test for the existence of AI.

"Let's say that I have a book that lists all possible responses to any question or comment that anyone could make. Let us also say that I appended an index to the book that allows me to look up the appropriate response and read it off to my opposite within a few seconds, so that time lag would not factor into the Turing problem.  

Is that book, sitting quietly closed on the desk, is it artificially intelligent?"

He couldn't say yes - it was clearly just a book. But he also didn't want to say no, because that would negate his darling. So, he just said, "I think there's more to it than that." When I asked what that might be, he laughed and said, "Well, I don't know, but something."

Insofar as AI is a database lookup, there is no way it can be artificially or otherwise intelligent. A database is just a large digital e-book. Books, it is commonly agreed, are not examples of artificial intelligence.

You might say it is the program (the procedures) combined with the book, but isn't that the same as just reading a book: procedure (reading) plus database (book)? Even when a computer is able to modify its own programming, it is following a predetermined procedure operating on a pre-existing database (its own programming).

No matter how I think about AI, it just isn't obvious to me how AI is any different from reading a book. Which means every printed book, every written anything, is artificially intelligent. But every printed book is just the product of a real intelligence, a person, in many cases, a dead one. The book can't do anything apart from the procedure of being read.

Does that mean it is the procedure which is artificially intelligent? The act of reading is artificial intelligence in action? I can't wrap my head around how the action itself is intelligence, apart from the actor. Even the ancient Greeks created little robotic machines, in which springs and ropes wound up gears so that the "robot" would move in a specific direction for a bit, then in another specific direction for a bit. Were those robots artificially intelligent? I don't see how. Everything we've done since then is just a glorified Greek robot.

Can we create procedures which will kill us all? We've done that lots of times: just consider the weapons of mass destruction we have created. Is poison gas, along with the procedure for disseminating it, an artificial intelligence? Is a machine gun? By the rules given above, the answer would have to be "YES!". So, in that sense, I suppose artificial intelligence is dangerous.

So, when people tell me I should be concerned about AI, it seems that they are telling me I should be concerned about my fellow man. Yes. Of course. That's not news.