23 replies to this topic

[Xcrypt]

Some time ago I've seen a documentary about some person that wrote an AI application.

He said that there was a law that determined when something is alive. Each unit in his application had all the properties necessary that made them "alive" by definition

Does anyone happen to know the name of that law?

[TheNut]

Are you referring to the Turing Test?

[v71]

Something is alive when it has consciousness about its existence , otherwise is just a machine executing instructions

[rouncer]

v71, but still we are still just machines executing instructions?

[Stainless]

Self awareness is the main thing, but very difficult to quantify. It is easy to write some code that checks to see if it's host is in a good condition, cpu not too hot, not running out of disk space, etc, but to be truely self aware .....

The turing test has been passed by some software, I saw a note about it somewhere but cannot remember where.

However that code is by no means "alive", it's just very cleverly written.

[Xcrypt]

It all depends on how you define "alive". Some scientist made some laws about it, and I'm here to know what those laws are. By no means is this thread meant to be a philosophical discussion of any kind. I'm just here to find out what those laws are so I can try to make a program that suffices to those laws.

Oh, btw, I don't think that guy was referring to the 'Turing Test' - I'm not an expert on this, but I think Turing test is more of an 'intelligence' test. Something that is 'alive' doesn't have to be intelligent by definition.
Even plants are alive, but I've never seen a plant behave intelligently, have you?

[v71]

Rouncer, we aren't machines , our biologic body is a machine, i agree on that but we are more than our brains, many scientist are starting considering the brain more like a receiving antenna , rather than a computer executing instructions.
We are more than our bodies i have experienced this first hand.

[Stainless]

More than our bodies.... now that is moving into philosophy

I'm not convinced. If you could set up a neural net with the same number of synapses etc as a human brain, organised in the same way as the human brain, would you end up with a baby in a box?

Not sure at all.

[rouncer]

Consciousness is something thats beyond my thinking, but im positive the brain is just the greatest machine ever to be, and eventually it will be understood, 10000 years from now.

[Reedbeta]

Xcrypt, there is no widely accepted fixed definition of life. Whatever law you're referring to is probably somebody's pet theory, but without any more specific information there is probably no way to track it down, since any number of scientists, philosophers, etc. have created pet theories about what constitutes "life"! However, Wikipedia has a list of some commonly encountered elements in proposed definitions of life. Maybe that will be helpful.

[alphadog]

The Turing Test is to test for a thinking machine, ie. that has some amount of reasoning ability that makes it indistinguishable from a human. The Turing Test is actually a weak test.

"Alive" is an ambiguous word. If by that you mean self-aware or consciousness, there is no one "law" or "definition" of it. In fact, it is quite a contentious subject.

Personally, I'm more interested in "Artificial Stupidity" studies, in order to better simulate humans.

[geon]

> Something is alive when it has consciousness about its existence

How about bacterias? They are certainly alive, but probably not very aware of anything.

One characteristic of "life" is self replication, but it is not enough in itself. A machine that could replicate itself would not necessarily be "alive".

A more philosophical definition is that for something to be alive, there must also exist a state in which it would be dead.

[Xcrypt]

Reedbeta, on 19 December 2011 - 12:20 AM, said:

Xcrypt, there is no widely accepted fixed definition of life. Whatever law you're referring to is probably somebody's pet theory, but without any more specific information there is probably no way to track it down, since any number of scientists, philosophers, etc. have created pet theories about what constitutes "life"! However, Wikipedia has a list of some commonly encountered elements in proposed definitions of life. Maybe that will be helpful.

Thanks

[touch_the_sky]

editor thing ;(

[touch_the_sky]

"We are more than our bodies i have experienced this first hand."

Sadly no, we are nothing more than our bodies and our wonderful brains. Any OOBE, phone call from a dead relative / friend, divine revelations while in deep meditation / trance, etc. are only tricks of your mind, which has some serious capability to generate really awesome stuff. Think about amazing / unbelievable things which your brain can generate while you're dreaming or a person with a diagnosed mental disfunction. I'd love to hear about your experience though, if you don't mind sharing.

And the topic itself is very interesting indeed. Don't know there exists a particular 'law which determines when something is alive', simulating mere 'aliveness' seems a really simple task for todays technology (especially that, as already mentioned, even bacteria or plants are alive). Even if we limit the scope to games, as opposed to all tech / resources available to mankind. If you for example look at some NPC in a RPG game - it has some 'alive' features of at least a real life spider / insect / etc. - it inhabits a certain area, it has a set of behavioural routines, it reacts to external stimula (defends itself when attacked by player, etc.). Sims anyone?

Xcrypt didn't give us much detail about that AI application, so we can only guess what it did or how complex it was.

The only thing not yet possible is to create a self-conscious AI with an intelligence, emotions, true feelings and subtlety equivalent or better than that of humans (that would of course pass all sorts of Turing Tests with no trickery). Reverse engineering the human brain FTW!

[Xcrypt]

It's not because it has not been done yet that it means it's not yet possible

[teddybot]

A feedback loop may be a requirement for self awareness.
I read 'On Intelligence' and never had these questions again.

"it has not been done" - ah appeal to time.
That one hurts it gives no credit to existing tools, knowledge or intuition.
Time is up- google neuromorphics lab- rat brain in computer behaving same.

Here is a cool idea from a PHd with genetic proof to back it up- Cognitypes.
Illusion is not a choice see the blue dragon illusion (and make one)

Since our fabrication of the world is based on our cognitype it is possible that what makes sense to people who can see that we are circuits is impossible for other cognitypes or cultulreotypes.
Our beliefs are made up from both hard wiring and layers in a hierarchy.
The analogy is that you can not flip the blue drag back to paper and see it correctly even if you try so your reality is a product of lower functions.
To see that we are circuits requires a lower layer seeing something 'flipped out' vs. 'flipped in'
These analogies are easier to understand if you check out 'hierarchical temporal memory' based on our neo-cortex.
numenta.com
Cognitypes are important. It is natures way of making a phenotype compete within itself.
Now who is right? flipped out or flipped in? That is why we have the scientific method.

If you have a puzzle in the room does it not exist because nobody put it together yet?
Its called a puzzle before and after its complete so there are those who know the pieces are in the room and there are those who do not.

"Even when a human was first run inside of a computer he STILL insisted there will never be A.I.!" - Teddybot
teddybot.blogspot.com

[alphadog]

teddybot, on 05 May 2012 - 02:45 PM, said:

Since our fabrication of the world is based on our cognitype it is possible that what makes sense to people who can see that we are circuits is impossible for other cognitypes or cultulreotypes.

I think teddybot would almost pass the Turing Test...almost.

[BlackWingedGames]

This is an area in which I have some expertise (a master's degree in artificial life)
As mentioned, there is a variety of biological definitions of life, and you can find exceptions to all of them in nature. Things like self replication, metabolism, growth, respiration, etc. All definitions fail in certain circumstances. For example, fire fulfils most of the definitions but we don't say it is alive. And things like viruses are 'in-between', exhibiting some characteristics of life but lacking the big one, self reproduction (they have to hijack that functionalty from their hosts).
I have seen (and programmed) many examples of artificial life which simulate many of these aspects.
If you want to go deeper into this, I'd suggest looking into Tom Ray's "Tierra", Chris Langton's "self replicating loops", and Karl Sims' 'creatures'. Those were some of the fantastic examples that fired my imagination and got me into the field.
As for the Turing test, it is pointless and irrelevant. I dont hold it against Turing though, after all, he invented reaction diffusion systems (you can look into them too) :-)

[Schoening]

Haven't read all the posts, but it is baffling to me that those I have read keep comming back to "conciousness" and "aware".

I never considered that Earthworms, trees and eggs aren't "alive". But I guess by the definitions so far they are not.