At what point could an AI become human?

[SpaceGhostC2C]

11 hours ago, Tm_Carbon said:

Even if the computer had free thought and had feelings of its own? 

My cat has free thought and feelings of its own. I don't think it's human, though.

 

7 hours ago, straight_stewie said:

A human is an evolved (or created, or both) animal. A computer can never be that.

A computer could be a created animal.

[straight_stewie]

18 minutes ago, SpaceGhostC2C said:

A computer could be a created animal.

Life and sentience are two different things. And life is something that we do know a lot about. Namely, one extremely basic criteria: It must be able to replicate or reproduce.

[dalekphalm]

7 minutes ago, straight_stewie said:

Life and sentience are two different things. And life is something that we do know a lot about. Namely, one extremely basic criteria: It must be able to replicate or reproduce.

If a robot can mine natural resources and use them to build another robot, does that count as replication?

[straight_stewie]

38 minutes ago, dalekphalm said:

If a robot can mine natural resources and use them to build another robot, does that count as replication?

While I suppose that that is one form of replication, there are other criteria that I did not speak of, namely: 

• Organisms that are alive must maintain Homeostasis
• Organisms that are alive are composed of cells
• Organisms that are alive have a metabolism, that is they can consume a resource and turn it into energy.
• Life can grow
• Life can adapt
• Life can respond to stimuli
• Life can reproduce

From these criteria there are 2 that computers definitely could or are already doing (adapting and responding to stimuli), 4 that machines might be able to do or could be construed to do (growing, having a metabolism, maintaining homeostasis, and reproducing), and 1 that computers or machines will likely never do (be composed of cells).

Construed is the key word in that last sentence. If we can construe things to mean something which was unintended or only partially intended, then I can claim that the processor in my computer meets 5 of the 7 criteria:

• It maintains homeostasis, that is, it has state and it can access and regulate that state.
• It is composed of "cells" that could be defined in various ways, such as groups of transistors, logic gates, cores, or even execution units.
• It has a metabolism, that is; the entirety of the system between natural resource production, electricity production, and consumption by the processor is a metabolism. The processor can regulate this metabolism by changing power modes.
• The processor can adapt to changing resource requirements. It can be made to adapt to changing inputs in other ways.
• The processor can respond to stimuli in various ways; External interrupts could be defined as stimuli, or using AI and sensors to respond to faces, touch, smells, or sounds.

Yet no one would claim that my computer is a living being, artificial or otherwise. On a slightly more on topic note, being sentient does not require life. It is conceivable, probably probable, that a computer will one day be made to be sentient. I don't believe that it is realistically conceivable that a computer could ever be made to be a life form.

 

[SpaceGhostC2C]

1 hour ago, straight_stewie said:

Life and sentience are two different things. And life is something that we do know a lot about. Namely, one extremely basic criteria: It must be able to replicate or reproduce.

I don't see how that invalidates computers as created (by human) animals.

1 hour ago, straight_stewie said:

While I suppose that that is one form of replication, there are other criteria that I did not speak of, namely: 

(...)

 

 

Those criteria are quite arbitrary and seem a bit ad hoc, to be frank. Yet even if I stick to them, I don't see why we couldn't build a computer that fulfills all of them. Even the most debatable one (being made of cells?) could be met without twisting the "common understanding" of the word the least: we could potentially make a computer based on, say, mouse cells.

Put differently, we can create, and we already have created, animal species that don't exist. I don't see why we couldn't create a species that can be considered a computer. Most seem to be tackling the question as "will my Celeron ever be alive?", but we could very well wonder "will my aloe vera ever contribute to calculate pi decimals?".

That alone would sidestep all the requirements (that you admit that don't seem very stringent as even non-living things can meet them).

[straight_stewie]

Just now, SpaceGhostC2C said:

Those criteria are quite arbitrary and seem a bit ad hoc, to be frank.

If you look up any credible paper, book, writing, or video on the subject of how to define or determine life, those are the 7 basic criteria that are always given or atleast mentioned. Some choose to disagree, but most agree that those are the basic criteria for life.

 

2 minutes ago, SpaceGhostC2C said:

That alone would sidestep all the requirements (that you admit that don't seem very stringent as even non-living things can meet them).

 

My point when showing that my computer could be construed to be alive was to rebut @dalekphalm statement:

1 hour ago, dalekphalm said:

If a robot can mine natural resources and use them to build another robot, does that count as replication?

I believe I specifically said that my goal with those statements was to show how we could construe any of the definitions to show that anything is alive. It is a well known fact, especially in politics, that you can construe anything to mean anything else, oftentimes without stepping too far away from the intended meaning.

[SpaceGhostC2C]

2 minutes ago, straight_stewie said:

. It is a well known fact, especially in politics, that you can construe anything to mean anything else, oftentimes without stepping too far away from the intended meaning.

But we are not discussing politics: in science, when something is vague enough that it can be made to mean anything, that's simply lack of rigor. 

So basically you admit that your definition had politics-level precision, which is fine, but it doesn't precisely put it in a good light 

[straight_stewie]

Straw man attack, nice job.

4 minutes ago, SpaceGhostC2C said:

So basically you admit that your definition had politics-level precision, which is fine, but it doesn't precisely put it in a good light

 

 

For the record, it's not my definition. It is the most well known and accepted definition of the minimum requirements life.

[SpaceGhostC2C]

1 minute ago, straight_stewie said:

Straw man attack, nice job.

 

For the record, it's not my definition. It is the most well known and accepted definition of the minimum requirements life.

To begin with, as many things in biology, it's not even a definition: it's a description. That's why it has the problems it has. When it says "has cells", it is basically saying that what we have observed so far and called "alive" has at least one cell. And so on. For all the true biological science out there, a huge chunk of it still is glorified cataloging. 

 

Now how can it be a "straw man" when I did not represent your view in a distorted way, or any way, but merely quoted your own words? You gave a definition. You said that that definition could be construed to include anything. Those are facts (read your own post). 

Then I conclude, by accepting your own claim, that the definition is therefore a pretty bad one. Or, as in this case, not a definition at all. 

I took your words. Then derived their logical implications, in my name. Where's the strawman? 

 

But if you are going to cry wolf anyway, let me ptoceed to sketch your position, and this time feel fre to complain:

"I have this great definition of life, although technically you could make anything fit in it, but I know what I actually mean so it's cool"

 

 

Let me know when you are willing to have a rational, rigorous discussion, though. 

[Satisfoxy]

Can you make a masterpiece?

[kelvinhall05]

AI and human aren't the same, and they never will be. I'm not saying that AI will never be as smart as a human, but AI is a computer, we are living organisms.

[MrDynamicMan]

21 hours ago, SamStrecker said:

But for AI to be self aware we are going to have other hurdles outside of technology. People will be moral crazy calling it a human, or people scared of humans become obsolete. 

Well, not really. For a self aware robot to exist it is a technological hurdle, not a semantic one. The hurdle would be the acceptance of self aware bots. 

 

However, that brings up a very interesting question. Definition of self aware conscious. You can't replicate a mirror test - even basic computers can 'recognise' themselves, emotion =! Conciousness. Personality? How can an unfeeling robot have personality? Speaking about itself in the third person? That's probably it, mind you. 

[straight_stewie]

9 hours ago, SpaceGhostC2C said:

it's not even a definition: it's a description.

Please stop talking now. Especially after claiming that I was making shit up. In Merriam Webster: 

 

Quote

Definition of definition

1. an act of determining; 
2. a statement expressing the essential nature of something
3. the action or the power of describing, explaining, or making definite and clear

 

9 hours ago, SpaceGhostC2C said:

Now how can it be a "straw man" when I did not represent your view in a distorted way, or any way, but merely quoted your own words? You gave a definition. You said that that definition could be construed to include anything. Those are facts (read your own post). 

The definition of a straw man attack is:  a common form of argument and is an informal fallacy based on giving the impression of refuting an opponent's argument, while refuting an argument that was not advanced by that opponent. One who engages in this fallacy is said to be "attacking a straw man". But I'll bet my last dollar that when you read that you immediately jump to "But he said something about politics" without even pausing to think that my statement was an example of how often and easily one can construe anything to mean anything.

 

9 hours ago, SpaceGhostC2C said:

"I have this great definition of life, although technically you could make anything fit in it, but I know what I actually mean so it's cool"

This shows your extreme lack of understanding, of education, and of thought. And shows that you are nothing but a troll. When I "showed" that my processor could be made to fit the definition of life (which I must reiterate, is not my definition but the most widely accepted definition of life), I did so under the explicit and specific pretense that I was doing so as an example of how anything could be misconstrued to mean anything as a direct response to @dalekphalm proposition that mining for resources and building a new robot is equivalent to biological procreation. 

 

And finally, since you obviously do not know the definition of construe or misconstrue, or your stupidity would have never showed:

 

Quote

Misconstrue

The act of wrongly interpreting a persons words or actions.

 

[ObscureMammal]

For anyone interested in AI, I would suggest reading a book by Nick Bostrom - Superintelligence: Paths, Dangers, Strategies 

http://www.goodreads.com/book/show/20527133-superintelligence

Spoiler