[Cspace]

Today, probably not... But what about the future?


- What if a virus can "evolve" on its own to spread itself, adapting to get through security programs even after they are built to stop it? Natural selection with the successful versions spreading and the unsuccessful ones disappearing?

- What if an AI can learn, use information, and solve problems without any human intervention?

- What if a program can code automatically?

- What if programs with differing purposes can learn to interface with each other without a human coding them specifically to do so?

- What if a program seeks to preserve itself? Learning to deal with viruses or bugs on its own, or maybe even (at the extreme) learning how to prevent a human from shutting it down?


What would you think?

[Aaron]

- What if a program seeks to preserve itself? Learning to deal with viruses or bugs on its own, or maybe even (at the extreme) learning how to prevent a human from shutting it down?


What would you think?

I think that is a self-reference, Cspace .

[Cspace]

- What if a program seeks to preserve itself? Learning to deal with viruses or bugs on its own, or maybe even (at the extreme) learning how to prevent a human from shutting it down?


What would you think?

I think that is a self-reference, Cspace .

[Zoo]

Well... that's a tricky question. We haven't even really determined what it takes to be "alive". If we consider viruses to be alive, then it's not any sort of consciousness, or even an ability to live on its own. It seems they exist solely to reproduce. They're just a bit of well organized protein, a bit of genetic material. They can even be "revived" after they're crystallized (they are inert in this state).

Comparing the two, a virus and a program, both require a host (a living cell and a computer, respectively). Both are, at the core, simple (a bit of protein and RNA [usually], vs. uhh... well, I understand the principle, just don't know the terms...).

So... seems reproduction of some sort is the important thing. If a program can somehow make copies of itself without human intervention (whether they spread or not) then I would sort of call that alive. The difference is (so far as I know) living things are composed of matter. Is a program? I don't know, myself. Nor do I know if being composed of matter is a requirement for being alive. Interesting thought. Before computers that wouldn't have been a question in anyone's mind....

Consciousness is a different story. That's kind of an easy one at first glance. If a program is somehow self-aware, then it would be easy to call it alive. The question then becomes how to determine if it indeed is self-aware.

There's problem enough with this in animals. Some are easily determined with the mirror test, where you put a dot on the animal's forehead or do something else even wilder (one group dyed cotton top tamarins' hair bright blue, and other colors, I believe), but that the animal can't see without a mirror. (I'm assuming here that the animal does not know that someone did that... I'm gonna have to reread those....) Show the animal a mirror. Animals that we consider self-aware will use the mirror to reach up and touch the mark on themselves, or somehow otherwise indicate this... as they have done this with dolphins too. Otherwise they will reach out and touch the mirror image. (It's usually assumed that the animal that touches the mirror thinks it's a conspecific.)

But there are problems with that approach, I think. Mostly anecdotal, but I have been working with animals of about 50 to 60 species, inverts, amphibians, reptiles, birds and mammals, for 6 years. There are some issues with bias in this case (I do love quite a lot of those animals). It's hard not to ascribe many characteristics to many of those animals. There would be many people who would immediately yell "anthropomorphism!" but I would dare to say many of these people don't know the variety of animals I do. (And I say this considering the places I hear it most often are in discussions of articles involving animals, where many of the participants are only casually exposed to animals.)

Yes, there are many animals that operate on instinct. Alligators, all the amphibians we work with, the birds of prey. Their goals are eat, sleep, reproduce and don't die. They, for the most part, seem to see humans as a food source (except the amphibians, they're asleep when their food is given), and something to fear. They get used to us, but most never come to "like" us (and I use like loosely, tolerate might be more appropriate). For animals that imprint (birds of prey, for example), if they see humans early enough, they take them to be something like parents. It's an instinctive... not even a behavior, just a process.

Then we have the parrots, hornbill, the iguanas, everglades rat snake, the baby bearded dragon. These have different characteristics that could be argued either way.

Lots of people have trouble with the green iguana. (Interestingly, being the only one besides a keeper to be bitten, I'm apparently the only one that likes him, but that's beside the point.) I don't. It seems he recognizes me when I come in, not even close enough to smell me. He still gives me hell when I take him out, but give him two minutes and he's good to go. Somehow he knows I like him, or so I believe (and others have said the same without prompting). When I stand and look at him through the glass, if I reach up and wiggle my fingers in front of his face like I'm petting him, he will often close his eyes (like he does when I actually am). Seems he remembers what that is.

The rat snake and the baby bearded dragon (the adult one is old, so he mostly just sleeps) are both much of the time watching what's going on around them. Granted, they would do something of this sort in the wild. On the other hand, they are neither of them fearful. If anything they seem to be curious. There's no way to prove that though. But curiosity doesn't seem to be an instinctive behavior for the animals I mentioned in the first section. (I do believe it is an instinct, however. Kids don't need to be taught to be curious. If anything it's the opposite... it seems they are taught (maybe inadvertently, maybe purposefully, depends on who you ask...) to not be curious.)

The hornbill (and the parrots, actually) seems to show affection by regurgitating bits of food. (Yes, it's absolutely disgusting, in the case of parrots. He only gave me cricket heads, which aren't as bad as mushed up zupreem.) Although I have to assume it's affection. (There's always that caveat.)

The parrots call attention to themselves. They learn that doing certain things gets people to come talk to or play with them.

On top of that, it seems some of them are very smart, as humans like to measure it (i.e. based on what is smart in humans, which of course has a flawed basic assumption; that human smart is the only smart). You've likely heard of Alex the African grey (my mom has, so some of you have to know already ). He could tell you shapes and colors, and could tell you if two objects were same or different. This was tested at least double-blind (I can't recall if they actually went as far as triple blind off the top of my head) to prevent things like the horse that could "do arithmetic" which was actually him picking up subtle (subconscious) clues from his trainer. This means the person actually giving the test did not know what answers to expect, and so could not cue him.

And that seems like a lot of rambling.... Distilled, my point is we don't really know what alive is, nor do we really know what conscious is, so applying these terms to programs is, at present, a judgment call, rather than something that has a definitive answer.

[Neraphym]

For me, what it means to be alive is to have metabolism and be able to reproduce, by ANY means and at least at one point in its life. So viruses and embryos are alive by my standards. The problem with computers evolving is that they are digital. The copying phase of their reproduction would need to be 100% exact, otherwise the program would fail completely. Evolution will not function with computers; they will need intelligent designers. Of course, computers can, in the future, be able to intelligently alter their own code, but they will obviously have to come from a human.

I think it would be ethically wrong to make a computer experience feelings. I only apply morals and ethics to things that can suffer, and often to the extent that they can suffer. Personally, I feel that I am not acting with malice when I kick a rock or turn off a computer. I would feel very different if I kicked a squirrel or 'turned off' a turtle. They can feel. To give a computer feelings is to give that computer rights. They would no longer be morally acceptable to keep as slaves like we do now. Think about that.

[Zziggywolf5]

Ooh, boy. This is a complicated question.
Well, (let's see if I remember what classifies something as alive) reproduction is the main point. Anything alive, cell, animal, etc, has to be able to reproduce without another organism of a different species. This is, if I remember, the main reason viruses are not considered alive. Now, if we want to follow this strictly, then we have to decide if the program is using the computer to reproduce. This is only complicated by the fact that the program requires a computer for all processes. Personally, I don't think this should be strictly followed because, well, no programs would ever be deemed alive. Plus, none of this specification was conceived with this idea in mind.

I think "alive" also means it requires a metabolism. Utilizing energy, signals, etc. I really don't know enough about computers or computer programs to even speculate on this.

Stimulus response is considered a sign of life. I think computer programs are already on the threshold of this.

[[Sentience-based response may be later, if I remember.]]

[Res]

No. You can give a machine all that it needs to be considered living. You can program it to respond, you can design it with sensors so it can 'feel', but if you take an axe to a robotic arm, or a chuck of metal held up and controlled by wires, you are not damaging living cells of a living organism.

The only way I can even begin to consider living computer programs is severely biotechnical engineering. Even then, a human mind must have more control over the program.

Right now, I could take my computer apart, leave it, poke and prod just about every part of it, and reassemble it, because it is a machine. I have a hard time getting to "Man and machine", and will even go so far as to say even someone with mechanical prosthetics is not "Man and machine", but a human with a replacement that is augmented by a machine.

There is a line between where living programs come in and normal life. Programs are merely tools used by machines to make them easier to use as tools by us.

Computer with programs that resemble life are the equivalent of a power drill. Something to be used.

[?FaLLeN?]

What if the Will Smith movie I Robot came true?

[Machine 12]

I believe computer programs can be considered alive. We haven't yet made any standardized qualifications of "alive" anyway. Personally, I feel computers could be sentient and, probably by default, alive.

[Cspace]

I guess some of the question may be whether something can be fully conscious, but not alive.

Many would say "no" because a machine isn't biological, and it may even be more of a stretch when it comes to a program which doesn't have any physical existence. If a program is fully aware, able to learn, and able to make independent decisions based on its own motivations, would its embodiment actually matter?

I don't know whether I can say "yes" or "no" to that, but I think there's a difference between "alive" and "like us". Personally, I think there are other ways for life to exist, but it may be difficult to determine when a program is actually conscious as opposed to it being a convincing simulation. There is theoretically a point where a program can be conscious, but how would we really know for sure when that point is reached?

[Aaron]

There is theoretically a point where a program can be conscious, but how would we really know for sure when that point is reached?

That really would be hard to discern.

[?FaLLeN?]

can my fart be considered alive?

[Machine 12]

can my fart be considered alive?

That is disgusting.

[Ruckus Fox]

How can we define anything alive if we, ourselves, cannot fully define the meaning of "to live."

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

being manifested by growth through metabolism, reproduction, and the power of adaptation to environment through changes originating internally.

Well...as can be expected, A.I. currently do not possess, and probably never will, a lust for glucose. However, can you not consider the energy itself that powers, that operates, that -drives- this machine to do what it does...a "glucose" of sorts? True, it is not broken down from its original state. Not chemically. The electricity (or plasma) from your socket runs through your power cord and into your computer, evenly distributing it amongst the system. Almost like a circulatory system. It's your call there. Can you suggest that the machine itself, if it wanted to, could breathe? It's expected that by 2050, an average $1000 computer will have more thought process and problem solving skills to beat out the entire human race. Sounds exciting.

A.I. being able to reproduce is a slippery notion. As all things do, A.I. will eventually lead to money, then laundering, and then sex. Who's to say sexual machines won't be produced? Flawless in appearance. Perfected in design. Sounds ideal to me. It is at that point when humans will decide whether or not to slap the idea of god in the face. As far as data goes, you can fully consider the genetic make-up of a computer to be reproductive. Try it for yourself. Right click on any icon on your desktop and select "copy." You have just made this data reproduce. An exact cloning, the exact make-up. Data can, and will reproduce. But it cannot bore a new without an original copy.

A.I. can adapt. If you've ever had a crappy old Windows 98, let me tell you. It is not very fun. I had several viruses on the damn thing. Rooted so deep that every time I ran a scan to root them out, it would give me an error. However, it was still there. Your computer can adapt to these changes. It may look at it, then skip over it the next day. Indeed, our home computers are capable of learning. How do you think a personal disc becomes corrupt? Adaption. In this case, adapting poor qualities of system errors. But your computer will realize this and give you one of the fun "close this application" boxes. If you have a firewall, many of the new ones help new computer users to opt for learning more about the programs and deciding which ones to allow and which ones to block. Surely, a highly developed A.I. will not need much adaptation, but they could learn some from the humans. As I said, computers have a learning curve, it's all there.

In my final synopsis, I have to conclude that yes. I think that A.I. has the possibility of becoming "alive." However, we are far from that moment in time.

--Ruckus