Computers and consciousness

[Lullaby Chainer]

In what ways does a computer mind differ from a human mind?

If we were to produce a genome that had genes necessary for a human brain, and then injected this into a suitable host germ cell, would the finished product have a consciousness? Would it be artificial?

What properties of this organism would be responsible for its consciousness?

Can we design a conscious computer with non-biological properties?

Lets look at the popular requirements of a conscious entity.

Wikipedia~
Consciousness is a quality of the mind generally regarded to comprise qualities such as subjectivity, self-awareness, sentience, and sapience.

[a]Subjectivity_ Proceeding from or taking place in a entity's mind rather than the external world

[b]Self-awareness_ Aware of oneself, including one's traits, feelings, and behaviors

[c]Sentience_ The ability to feel or perceive

[d]Sapience_ The ability of an organism or entity to act with judgement



[A] The calculations, acts, data stimuli reception, and data output all happens, or at least starts, within the parameters of what we define a computer(mind) as.

[b] A computer can be programmed to distinguish itself in a mirror using advanced image recognition software. A computer can also be aware of its traits, such as monitoring its heat levels.

[C] This is a tricky one. What qualifies as a feeling? A human finger can touch a hot surface and perceive that it is hot. A computer can do the same thing by placing an appendage on a hot surface, detecting the temperature using a built in thermometer, then perceive and collect the data for conclusion making.

[D] Computers can collect vast amounts of data, argubly our equivalent of knowledge, and calculate and act out conclusions accordingly.

Why can't computers be conscious?

[Zhavric]

I know that most humans want to romanticise our existence. To theists, we're a soul attached to body. To non-theists, we're a barely-tapped super-computer with an attached life-support system (the body). Neither one is accurate.

We aren't nearly so splintered.

The mind without the body is nothing. Consider what would happen to a child who was born in absolute sensory deprivation and was then given stimuli. They'd react crudely at best. It does little good telling an infant "Hey, talk to me. Prove you're intelligent." You won't get any reply.

Why?

Because our personality (intelligence, mind, etc.) is a series of learned experiences. It's not enough to have this huge brain that can make complex calculations. It has to have reference points. Those references come from our bodies. We're hugged and cuddled (hopefully) when we're infants and we begin to learn that the things hugging us are good and things not hugging us are not as good. We learn that being full is better than being hungry. Eventually, we even start to be able to mimick those strange sounds we hear... and in time we figure out that sounds have meanings. It's a long and tedious process as any parent can attest to... and we're very lucky our bodies happen to function that way.

So, what do we do with a computer? How do you teach it? Do you attempt to simulate humanity in it or do you try something outside the box?

Do you want it to have a conversation with you? Remember that every conversation you've ever had is based on previous conversations and things you've learned earlier on. How would you do that with a computer? How would you get it to disagree?

We're still a long way off.

[Samildanach]

If we in theory could decode memories into binary we might be able to produce AI. As far as I can see the only thing that is stopping us is the actual teaching part. This topic has been done to death in another past thread though.

[tman_ndsu08]

In what ways does a computer mind differ from a human mind?

There are obvious structural differences.

Human brains are made from protein.

Computers are made from metal.

Consciousness is a quality of the mind generally regarded to comprise qualities such as subjectivity, self-awareness, sentience, and sapience.

And a computer could one day have all of those qualities.

In that sense, a human wouldn't be able to tell the difference between a computer consciousness and a human one.

[Kite]

I aggree with tman, we're not that far off from making a truly competent AI that can function on its own with minimal pre-programming.

[The Bacon Guy]

In principle, the brain is no different from a computer. The mind exists as highly complex combinations neurone structures in the brain and a computer works by a complex combination of zeros and ones. All the characteristics of consciousness are no more than complicated neurone structures, so in theory, they could be replicated in machine code. The main difference between the mind and a computer is in terms of complexity.

It isn't an overstatement to say that the human brain is the most complex thing in the known universe, whereas computers are extremely simple by comparison. We are nowhere near being able to make a computer as complex as the brain, however, as the principles are the same, it is not inconcievable that we someday will.

[Zhavric]

In principle, the brain is no different from a computer. The mind exists as highly complex combinations neurone structures in the brain and a computer works by a complex combination of zeros and ones. All the characteristics of consciousness are no more than complicated neurone structures, so in theory, they could be replicated in machine code. The main difference between the mind and a computer is in terms of complexity.

It isn't an overstatement to say that the human brain is the most complex thing in the known universe, whereas computers are extremely simple by comparison. We are nowhere near being able to make a computer as complex as the brain, however, as the principles are the same, it is not inconcievable that we someday will.

The problem with your analogy is that the brain isn't just a complex set of zeroes and ones. It's a complex set of zeroes and ones that's hardwired to a human body.

I remember reading a piece of science fiction in college. The premise of the book was that scientists had build an A.I. that could talk, but was having a lot of trouble understanding humans. I bring it up now because there was a chapter in the book where one of the computer programmers tries to "show" the A.I. an image of a lamb. The A.I. replies that it doesn't really see it the same way and spits back the image in its data-set: as a bunch of zeroes and ones that stretch on for about a page.

If you want a computer that is like us, you have to give it input similar to ours. You have to give it five senses, instincts, and built in reactions. Then, you have to give it the ability to use metaphor.

That last bit is really the trickiest as there are a handful of animals that are nearly as smart as us, but they aren't able to use or understand metaphor. There are chimps that can use rudimentary tools, but they can't understand those tools as symbols for other bigger concepts.

How would you get a computer to do that?

EDIT: Think of how much you learned as a child from your parents telling you "No!", yelling at you, or spanking you. How do you approximate that? How do you simulate the negative feelings and make a computer associate certain actions with those feelings?

[The Bacon Guy]

If you want a computer that is like us, you have to give it input similar to ours. You have to give it five senses,

You’re right that I didn't really mention anything about input. However, it seems to me that all this 'human computer' would need for input would be sensory devices to act in place of the body's senses. For example, instead of sight, the computer would have a highly sophisticated sensor that is able to interpret light in the same way as the retina.

That last bit is really the trickiest as there are a handful of animals that are nearly as smart as us, but they aren't able to use or understand metaphor. There are chimps that can use rudimentary tools, but they can't understand those tools as symbols for other bigger concepts.

Understanding a metaphor requires a series of thought processes incorporated with the sensory data of the metaphor. These processes, as with all thought processes are extremely complex and scientists are still unable to understand how they work. However, I would still argue that these thoughts are no more than neurone structures and chemical reactions in the brain. If so, and the mind is completely scientific, this should in theory be possible to replicate.
I think the reason this is so hard to imagine is because we don’t understand how each of these individual processes occur in the brain and it is therefore difficult to imagine programming them into a computer.

How would you get a computer to do that?

I haven’t the faintest idea and neither do any scientists or computer technicians in the world today. I just think that, as the principles of the brain and the computer are the same, it should in theory be possible.

[Zhavric]

I think we're on the same page. I guess whenever I hear anyone talk about making a computer that can think like a human, I automatically think of a human infant with next to no sensory organs.

I actually think the computing part is the "easy" part. Getting the computer to react as a human is the nigh impossible bit.

[Lullaby Chainer]

The problem with your analogy is that the brain isn't just a complex set of zeroes and ones. It's a complex set of zeroes and ones that's hardwired to a human body.

I remember reading a piece of science fiction in college. The premise of the book was that scientists had build an A.I. that could talk, but was having a lot of trouble understanding humans. I bring it up now because there was a chapter in the book where one of the computer programmers tries to "show" the A.I. an image of a lamb. The A.I. replies that it doesn't really see it the same way and spits back the image in its data-set: as a bunch of zeroes and ones that stretch on for about a page.

If you want a computer that is like us, you have to give it input similar to ours. You have to give it five senses, instincts, and built in reactions. Then, you have to give it the ability to use metaphor.

That last bit is really the trickiest as there are a handful of animals that are nearly as smart as us, but they aren't able to use or understand metaphor. There are chimps that can use rudimentary tools, but they can't understand those tools as symbols for other bigger concepts.

How would you get a computer to do that?

EDIT: Think of how much you learned as a child from your parents telling you "No!", yelling at you, or spanking you. How do you approximate that? How do you simulate the negative feelings and make a computer associate certain actions with those feelings?

Is the only difference other than content, the decreasing complexity gap?