h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

I just read a piece on 'The Straight Dope' about making androids - robots that look like humans. A part of the answer was this: we can already make "self-propelled homunculi that realistically simulate a mature human being" - namely, humans. We make new humans all the time. So why make artificial ones?

I cannot believe that at 'Straight Dope' they didn't think of this. Namely:

Technology progresses. So ... if one day we can make androids that pass for humans, we will at some later day be able to make androids that surpass humans in all respects.

I am serious about this. IF we can make robots that measure up to humans then there is no doubt that potentially we can make robots much, much, MUCH better. Such as:

1. Potentially immortal.
2. Intelligent with the best features of human AND computer intelligence: vast and perfect memories, the ability to think, calculate, and do any other kind of mental processing, very much faster than we.
3. Bodies stronger, able to survive where we can't, able to survive damage and injuries that we can't.

Robots could even be superior to us morally. If they are to pass for human then they will need real emotions ... real feelings, and attitudes, and desires, and everything else belonging to a human mind. They could be MORE than we are. Isaac Asimov thought of the 'three laws of Robotics' meant to keep robots in line ... in practice, robots would have to have an 'ethical structure' very much more complex, but we would probably only accept very ethical robots who would certainly never hurt or murder people ... but these same ethical constraints might also make robots that are truly *virtuous* ... free from 'sin' in a way that almost no human is.

In fact we might make robots so much better than humans that they could end up replacing us ... and we might even accept that. THEY might be the inheritors of our civilisation ... and maybe even the Universe.

Your idea reminded me of that splendid collection of short stories by Stanislaw Lem, 'The Cyberiad':

http://en.wikipedia.org/wiki/The_Cyberiad

Personally, I wish fervently that people would stop trying to build androids, and try instead to improve upon themselves. But that isn't going to happen in this universe.

Hey, I love Lem's work! I wish I could get those stories ... there are only a few of his books in the library.

I don't think people are actually spending much time and money on building androids right now. There's that Asimo robot ... but you can't really call it an android. I think we'll first see the development of human qualities on a computer - or a computer program - that can pass the Turing Test. In my view we're still very very far away from that. But I do think it will happen. And I think it would be useful because teaching a computer how to think will mean we will have to learn a lot more about how humans think.

Learning how to think is good exercise.

The IBM computer that won the Jeopardy! contest has done wonders for algorithmic processing, no doubt.

You can actually buy a t-shirt with the immortal words of one of the losing contestants:

http://www.thinkgeek.com/interests/looflirpa/e82d/

Hi there Dmitri! That was quite a thing when the computer won the Jeopardy. But of course it was the PROGRAM, written by PEOPLE, that did it!

Right you are. The purpose of the game was to let the programmers do their stuff and figure it out. They were very grateful to the experienced contestants - they said they studied all their tapes over and over.