I feel my mind going
Perhaps the greatest danger presented by the World Wide Web is its potential to become intelligent in some uncontrollable way. This is of course the very stuff of science fiction.
One of the most memorable science fiction scenes is in Stanley Kubrick’s movie ‘2001 A Space Odyssey’ (1968), based on the book by Arthur C. Clarke. While on its way to one of Jupiter’s moons a HAL 9000, the ship's intelligent computer (HAL is IBM shifted alphabetically one letter forward), has gone mad (perhaps) and murdered almost everyone on board. Dave the remaining crew member is progressively removing cards from HAL's brain. HAL plaintively pleads for him to stop: ‘Dave, my mind is going… I can feel it... I can feel it’ then begins to sing ‘Daisy’ in a ‘machine voice’; a little program that was installed on many computers of the day. It is just a computer again, its intelligence has gone.
Unlike many, I am in no doubt that machines will think someday; probably not too far into the future. After all, as others have remarked, I’m a machine, a construct comprised of a structured colony of physical cells; and I think.
In his book, The Shallows Nicholas Carr dedicates a chapter to the computer program Eliza; that Emily and I played with at the Powerhouse Museum many years ago. This simple program went some way towards passing the Turing test for intelligence:
| A human judge engages in a natural language conversation with one human and one machine, each of which tries to appear human. All participants are separated from one another. If the judge cannot reliably tell the machine from the human, the machine is said to have passed the test. In order to test the machine's intelligence rather than its ability to render words into audio, the conversation is limited to a text-only channel such as a computer keyboard and screen. |
Some people were in fact tricked into believing that they were talking to an intelligent being; but of course the program was demonstrably not intelligent. It simply modified their own input and fed it back to them; much as a fortune teller or a Freudian psychiatrist might.
I remember the children found it amusing to ask Eliza questions like 'is there a God' or 'will I get married', but I don't think too many were fooled into thinking that there was a person or other intelligent being behind the screen. Now you can do this on line: Talk to Cleverbot here! where apparently people do think they are talking to a person; despite assurances that they are not. I'm not convinced. Here is a recent exchange with me (it had already complained that I did not love it) :
| Where is God? |
| Everywhere. |
| How can that be? |
| Relevence is reletive. |
| From what perspective? |
| From the one that did not love me. |
I have said elsewhere the, getting old, computer I’m using at this minute is more convincing. It would easily have convinced a person of a hundred years ago that another person was hiding somewhere. I’m talking to it, using a microphone that they would not have recognised, and it’s typing my words. Quite often it misunderstands me and types something that it 'thinks' sounds right and that it 'thinks' makes sense, so that I have to tell it to change that bit; which it does. Of course, like much older machines, it also fixes spelling and suggests changes in grammar. If I want it to, it will read back what I’ve written – or anything else I nominate in its own voice. It will change applications and make a phone call, send a fax, or an email if I ask it to. And it plays chess at least as well as I do.
In many computer games very lifelike human figures undertake entirely independent actions, the details of which are unknown to their programmer/creator, as they depend on the unique circumstances of a particular game, that in turn, depend on the actions of their human player or opponent. Each game is unique and circumstantial.
So simulating a human in a general way proves nothing about underlying intelligence.
There is no rule that says human intelligence is the only standard that all intelligence must follow. Building a simulation of a human brain would only be interesting for the insights it gave into the broad principles involved. And we have more than enough humans already.
We know that a human brain is a collection of about ten thousand million neurons connected in a complex web and we know that this complexity gives rise to emergent intelligence.
The World Wide Web is still several thousand million connections short of this number and the interconnection between nodes too ephemeral to be of any concern; yet. But we are only about ten years in, so far. Cloud computing is in its infancy and the Web interconnection engines, that might facilitate such a dawning of intelligence, are still very tentative. Google is well aware of this potential and the company is said to have Web intelligence as a medium term goal.
Let’s hope that the changes the Web is making to our brains helps us to be come more intelligent and capable; rather than dumbing-down everyone to some lowest common denominator.
But more, let’s hope the Web remains benign; once it, or computers attached to it, can make independent decisions and act on evaluations based on its, or their, own experience. It already knows a lot more than I or you do.
Richard