So what have we established so far?
1. Consiousness (the action of my mind) is redicible to the physical facts of what happens in my body and brain when I express my inner experiences. This is, presumably, what Kurzweil would refer to as 'sentience'.
2. Consiousness is purely subjective and it is in principle impossible to reduce my conscious experiences, the facts of my mental actions, to mere physical facts: there must be something more than just the physcial facts to explain this strange inner life that only I can experience.
Position 1 is termed the 'materialist' perspective and is usually held by scientists. They explain that just as two properties that at first seem distinct are in fact identical (eg. water and H2O), mental states and brain states are in fact identical. Don't be fooled by advertisements for pain killers which tell you that the drug, once taken, homes in directly to the part of your body that is throbbing (usually indicated by a fancy visual of a red spotlight on the affected area). The pain is actually a certain brain state and the drug's work really takes place there.
Position 2 is termed the 'dualist' position and is usually held by people with a leaning towards a religious or supernatural explanaton for the origins of consciousness. 'Substance dualists' argue that the mind and body are two disctinct substances and can exist independently of eachother but can interact with eachother. Property dualists argue that there is only phyiscal stuff in the universe, but this stuff has two distinct properties, mental and physical properties. Our mental properties are extra properties and exist in addition to our physical properties.
Before we look at the arguments and counter-arguments presented by believers in each position, consider the following variations on an intelligence test for machines.
The Turing Test
In the 1950s, Alan Turing devised an intelligence test to establish whether a machine is displaying any conscious mental activity. There are different versions of the test, but here's a summary: there are three participants, A, B and C. B & C are in separate compartments and A is the interrogator looking into both. A communicates with B & C by means of a telegraph apparatus. It is up to A to pose questions to B & C in an attempt to find out who is the human and who is machine. It's up to B & C to convince A that each of them is indeed human. According to the terms of the test, if A cannot tell the difference between B & C more than 50% of the time, then the machine is said to be intelligent. As yet, we haven't been able to create a machine or computer that has passed the Turing test. Why? Largely because, as Kurweil might suggest, our computers have not mastered emotion...yet!
The Chinese Room
In 1980, John Searle presented a thought-experiment to prove that machines or computers don't have minds or conscious mental states, they simply imitate intelligent behaviour or understanding. In short, computers, however complex their computational outcomes, are simply fulfilling a program that we have created for them. Here's how the experiment works. Imagine that a person who has no knowledge of Chinese is locked in a room in which there are a set of cards with strange symbols on them. These symbols are in fact Chinese characters, but to this person they are meaningless squiggles. Suppose now that he is given other cards with similar symbols on them plus instructions (in his own language) about how to shuffle the symbols and hand them back in response. The first set of cards were a story in Chinese and the second set were questions posed to him in Chinese about that story. The instructions for the symbol shuffling are his 'program', so to speak, to respond correctly in Chinese to these questions. The people posing the questions outside the room are Chinese. When the answers come back from the person in the locked room, they might be fooled into thinking that he understands Chinese and is capable of following the story. But clearly he doesn't understand a word of Chinese. He's simply shuffling symbols according to his instructions. Searle's point is twofold: first, that computers, however complex, work in exactly the same way - they mechanically shuffle symbols according their program without understanding any of the symbols and their meaning. They merely provide a simulation of understanding. Second, the main reason for this is because they're made of the wrong stuff - they are not biological, organic beings that have evolved naturally.
No comments:
Post a Comment