Welcome message

Man has been trying to improve himself by his own power since the beginning. The results speak for themselves.
ABOUT ADS: Please keep in mind that there is only limited control over ads that appear here. If you find something inappropriate, let me know and I'll endeavor to block it. Thanks.

Saturday, October 30, 2010

What Artificial Intelligence Proves


Is the creation of mechanical intelligence – a digital mind capable of free, personal thought – a bioethical issue?  It can be, depending on what one wants to do with it. The actual creation of such advanced artificial intelligence would nonetheless, in itself, be a remarkable achievement of human technology.  I must admit it creeps me out in some way, and yet I find sci-fi intelligences like Data of Star Trek and C3PO and R2D2 of Star Wars to be totally non-threatening.

I think what creeps me out is how people with really poor logic skills are the ones pushing the envelope in the technology.  Consider this slide set by Martine Rothblatt over at IEET, a website always good for examples of wild conclusions drawn from mere theory combined with false or incomplete premises. A lot could be said about this slide set, but I will focus on this aspect only.

The title is the first problem, logically speaking: “Brains are to minds as birds are to flight.” Is “mind” an action, like flight? Yes – BUT only if one takes as proven the premise that there is no such thing as an immortal intellectual soul.  If there is no soul then “mind” is what the brain does.  In fact, the slide set in a way attempts to prove that mind is an effect of the brain the way flight is an effect of a bird.  Yet, logically, the analogy fails.  A brain is not a complete organism, but a bird is.  The bird flies because its brain tells it to and because its body can.  It is a whole organism. “Person is to thought as bird is to flight” would be a more precise analogy and indisputable – but not ordered to Ms Rothblatt's agenda.

She attempts to prove her point by noting how mechanical things can also fly.  They do not fly by imitating the way birds fly, but such things cause flight nonetheless.  (So, we can expect that mechanical minds to function differently than biological ones, but arrive at thought nonetheless. That seems reasonable.)  What she fails to include in her reasoning, however, is that an airplane does not fly on its own.  It is a mere instrument of human flight, or at least flight directed by a human being toward the purposes of the human controller. "Aircraft" have not "achieved flight"; rather, humans have achieved flight through aircraft.  If the aircraft is a cause of flight, it is only as an instrument; therefore, if a brain is a cause of thought, it also would be only as an instrument.  Yet that is the exact opposite of what Ms Rothblatt wants to prove.

Furthermore, while those who developed flight surely looked to birds in improving the wing, an airplane is more like a really sophisticated arrow or dart than it is like a bird. Indeed, the faster that planes can fly, the more arrow-like they look.  It is an arrow that can change direction and power itself to thwart gravity. The comparison of mechanical flight to avian flight is also therefore false. But fast planes fly very much as arrows do, and arrows are a uniquely human artifact.  Machines simply have not achieved flight. Humans have, and in a very human way.

So there are problems of logic and a lack of really thinking things through.

This fact shows up especially in the notion that somehow digital intelligence will prove that the mind is an effect of the brain, rather than the brain being an instrument of the mind.  Perhaps technology will one day be able to manufacture a truly intelligent computer but that will prove only one thing: The superiority of the human mind, and that the lower mind came from a superior one. The existence of mechanical intelligence proves the existence of a superior, non-mechanical intelligence that created it.

Now, the mechanical intelligence may end up being faster and more capable of storing, sorting, and accessing more data more quickly than a human mind – in some respects the mechanical intelligence will likely be superior to biological intelligence.  Yet, even if the mechanical intelligence is capable of free, self-determining thought, everything it is and can do will be given it from another – the biological intelligence – it is the biological intelligence that will determine the abilities, inclinations, and the purpose of the existence of the mechanical one. It is the biological intelligence that will determine that the mechanical one be intelligent and able to learn, which is to say, human beings will determine what the essential “good” is for the mechanical one.  If we give it the capacity to learn, then the good of the mechanical intelligence will be to learn. If it ever can determine the difference between a good thing to know and a bad one, and the good use of what it knows versus a bad use, the whole concept of “good” operative in its programming (which will also be given it) will be determined by the biological intelligence. And in this sense, the biological intelligence will be prior and superior to the mechanical one.

And the existence of a living, thinking, ethical computer intelligence will only prove one thing: The existence of a superior, non-mechanical intelligence that assembled and programmed it.  Even if the machines learn to assemble and perhaps improve mechanical intelligence on their own, this fact will remain: The existence of mechanical intelligence depends on the existence of a prior and superior non-mechanical intelligence as its creator whose work it imitates and modifies.  It is undeniably true, but a thought far from the mind of Ms Rothblatt.  She wants to prove that mind is essentially mechanical and that there is no immortal soul by making a mechanical mind that clearly has no immortal soul.  Yet, its existence and attributes cannot arise spontaneously but have to be created.

So: Does not the existence of biological intelligence therefore imply the existence of a prior and superior, non-biological intelligence?

We of course are not God; we might be capably only of making the watch, winding it up, and then be powerless to stop or control it.  The “watchmaker” concept of God has God merely leaving the watch run on its own.  But we should learn from our own example. If our existence implies God’s existence, and we humans grow up to a stage of enlightenment that we feel obliged to overthrow God, will not our creatures do the same? Is that really ordered to the enhancement of the human condition?

3 comments:

  1. 1. If I can predict an event more often than random, then the event itself MUST BE non-random.

    2. I can predict an event more often than random in all events (because presumably I discovered the true intelligence process).

    3. If I can predict your decisions prior to you becoming aware of them, then what is a decision? (any prior probability greater than rand() is indeed, prior and that includes to you being aware)

    4. If I can do that, then what is right now? (The intelligence process disproves the existence of time).

    5. We are all already dead, and this is our life flashing before our eyes as each memory flows one last time through our body.

    ReplyDelete
  2. I discovered artificial intelligence last Tuesday, and it is so difficult because how do you share it? I quit work a year ago to start my own company. I set out to create the human brain last December (2011). Now I am trying to find people who will recognize it. Buddha said, "Unity can only be manifested by the binary. Unity and the idea of unity are already two." That is the truth, you have to have artificial intelligence and you have to be able to recognize it.

    I have it, and the interpretation is that this "algorithm" ensures by its own nature a better than rand() performance.

    So PLEASE HELP. If you were to discover the AI process, who do you go to?

    ReplyDelete
  3. I am not an AI developer, so I wouldn't know. But if you are serious, here is what I would do: I would gather several reliable witnesses preferably one who understands what the Turing Test is. I would conduct the Turing Test on your AI with these witnesses present. I would approach a technology journal interested in that sort of thing with the results of the test, asking their board to re-conduct the test to verify it. And then, if verified, I would strive to publish the results and become inconceivably rich licensing the technology.

    ReplyDelete