What Makes Us Human? Part 2
Monday, September 06, 2010By Steven H. VanderLeest
Does our ability to remember past events make us human? Animals and computers can do that also. What about our ability to pretend, to play games, or to laugh? Is humor something intrinsically human? Or perhaps our creativity is core to our humanity. Our music, art, or literature (all culture making activities) might be a key part of our being. Some would point to our ability to make and use tools. Although a few instances of animals using simple tools have been observed, such as a tool-making crow, we are far more advanced than any other creature in this respect. Do our emotions set us apart? Some animals exhibit sadness and joy, though humans seem to have a much broader repertoire of emotional responses. Is it altruism – our ability to sacrifice ourselves for the good of someone else? Or perhaps our humanness is in our ability to sense something beyond ourselves – our ability to recognize and worship God.
In the 1950s, Alan Turing proposed a test to determine whether a computer could think, i.e., whether it was artificially intelligent. In the “Turing Test”, a human interrogator communicates with two rooms, A and B, via a computer terminal. A human responder is in one of the rooms and a computer responder in the other, but the interrogator does not know which is which. The interrogator can ask either room any question. The conversation might go like this:
Question to Room A: “What is 2 plus 3?”
Room A response: “5.”
Question to Room B: “What is your favorite color?”
Room B response: “I don’t have a favorite, though I tend to prefer blue shirts.”
Question to Room A: “Are you self-aware, and if so, how do you know?”.
Room A response: “Je pense, donc je suis. Ha ha … do you think you have Descartes in here?
On and on it goes, with as many questions as the interrogator can think to ask. The answers come back from each room via the terminal (so that the interrogator cannot use physical appearance to decide which is the human and which is machine). The computer passes the test if the interrogator is unable to determine which room contains the computer. Turing claimed that passing the test was proof that the machine was artificially intelligent.
Although Turing developed his test to check the creativity and human-like behavior of the computer, I think such a test might be not so much an endorsement of the computer’s ability but rather a rebuke of the human in one of the rooms for lacking creativity – or perhaps it is a lack of creativity and cleverness on the part of the interrogator? This is really a three-way contest between the interrogator, the human responder, and the programmers of the computer. Can the interrogator out-smart the programmers? Can a series of questions be constructed that the computer did not anticipate or answers in such a way that it gives away its true identity? Can the human responder answer in such a way to prove they are the true human in this trial?
Philosopher John Searle suggested a counter-argument to Turing in his “Chinese Turing Test”. He posited a test similar to Turing’s test, but this time it is all done in Chinese. A Chinese interrogator writes out questions in Chinese (either by hand or with a word processor that allows typing Chinese characters) and passes them to each room. Each room responses with answers also written or printed in Chinese.
In Searle’s gambit, we imagine that the computer passes the test and thus according to Turing, the machine is intelligent. Searle then suggests a second round of the test. Before running the second test, take the computer program for the supposedly intelligent computer and write it out as a set of English commands. Certainly the instructions would be numerous, filling many pages in a massive tome. It would contain detailed steps for analyzing characters. Perhaps it would be a series of checks for various slashes and dots and swirls on the page, sending the reader to a variety of other pages in a giant “choose your own adventure” book. Eventually the instructions would have the reader start putting their own slashes and dots and swirls on the answer page. Searle then would have the test run again, this time with a human Chinese speaker in one room (same as before) but a human English speaker along with the English instruction book in the other room (instead of the Chinese-speaking computer). As the test proceeded, every time the English speaker got a question, she would have no idea what the question meant (since it was written in Chinese). But she could follow the instruction book step-by-step. Eventually the book would lead her through constructing glyphs on a page. The symbols would mean nothing to her, but eventually she would have a whole set of them on the page and finally she would come upon an instruction that said to send it back to the interrogator. The Chinese interrogator would have no trouble reading the response and it would make good sense as an answer to his query (since the original program worked flawlessly, so too would the English implementation, though at a much slower pace). Searle then asks us to stop and think about this incredible turn of events. The English speaker has just responded to the Chinese interrogator – in Chinese, even though she does not know Chinese! We cannot possibly say that she is intelligent in Chinese. In fact, the person who wrote the original program (and perhaps the person who translated the program into an English book of instructions) was the one that understood Chinese enough to break it down into a long list of instructions about how to interpret and respond to it. Searle then concludes that just as the English speaker did not know Chinese, neither did the original computer. Rather the computer programmer was the one that knew Chinese. So the only thing that was proven by the test was that the programmer is intelligent. But we already knew that!
If we believe that we humans ourselves are simply controlled by a set of neurons that automatically, instinctively, and deterministically respond to the environmental stimuli presented to our senses by producing programmed behavioral outputs, then Searle’s argument applies to us as well. We are merely the biological machines that run a wetware program. We are not actually intelligent ourselves; rather, we only serve to prove that our Creator is intelligent.
I protest such an interpretation. I am not an automaton that carries out orders dictated by instinct without deviation and without choice. I have the ability to make my own decisions, to choose my own path. I am the master of my own fate. Life is meaningless if I cannot create my own future. Life is a cruel joke if I cannot escape a predetermined fate. But of course this desire for self determination may be pride that was, at least in part, the original sin of humankind. In perhaps the ultimate creative act of the drama of creation, God granted free will to the steward of his creation. While I can envisage a computer program that makes choices based on pre-packaged rules or even randomized selections, I cannot imagine any means to give a computer actual free will. While I apparently have it myself, I do not know how to bestow it on my machine creation. Perhaps that is part of the imago dei .