Ray Kurzweil Wonders, Can Machines Ever Have Souls?

Celery writes "There's an interview with Ray Kurzweil on silicon.com talking up the prospects of gene therapy as a means to reverse human aging, discussing different approaches to developing artificial intelligence, and giving his take on whether super intelligent machines could ever have souls. From the interview: 'The soul is a synonym for consciousness ... and if we were to consider where consciousness comes from we would have to consider it an emerging property. Brain science is instructive there as we look inside the brain, and we've now looked at it in exquisite detail, you don't see anything that can be identified as a soul — there's just a lot of neurons and they're complicated but there's no consciousness to be seen. Therefore it's an emerging property of a very complex system that can reflect on itself. And if you were to create a system that had similar properties, similar level of complexity it would therefore have the same emerging property.'"

Comments

Define soul.

[h4rm0ny]

See subject.

Re:Define soul.

[zeromorph]

Yeah, but why does he use this unscientific and highly religiously charged word? As if consciousness wouldn't be enough of a problematic notion.

We don't know what consciousness is and calling it an emerging property is not really much of a progress.

Re:Define soul.

[chrb]

He didn't choose to use this "unscientific and highly religiously charged word" - he was asked a specific question in an interview - Will super intelligent machines ever have souls? and he responded by saying that the soul was a synonym for consciousness and continued from that point.

Don't blame Kurzweil for an interviewer who uses fuzzy pseudo-religious language.

Re:Define soul.

[cunamara]

Now: demonstrate its existence.

Re:Define soul.

[Roland Piquepaille]

You're a very clever troll, but I'll bite...

here come the knee-jerk atheists

Feel threatened in your religious beliefs much? Don't worry, that phenomenon you see around you whereby people abandon irrational creeds is called progress. It's slow coming, but it's coming.

What is your consciousness? What makes you sentient? They've poked and prodded every orifice of your body and they have still not been able to determine where your consciousness

Have you considered that consciousness is an illusion of a human brain that has become powerful enough to reflect on its own existence? That's why you won't find it in the body or the brain, anymore that you'll find a tummy ache if you look inside your stomach.

this 'thing' in quantum physics called 'the observer'

Nice confusion here. The "observer" in quantum physics doesn't have to be sentient or conscious. A simple camera is enough to skew a quantum physic's experiment.

Yet, most people seem to acknowledge its existence. Even many scientists, atheist or not.

Wrong logic here. It's not because scientists and "most people" acknowledge the existence of consciousness that they all agree it's a metaphysical being. In fact, if I had to guess, I'd say most scientists believe consciousness is a physical brain process that has nothing to do with metaphysics or religion.

Re:Define soul.

[nawcom]

I'm all for science, cause I believe that if our understanding of science was ever to become good enough that we actually could prove/disprove Gods existance, all you atheists would be in for a surprise.

The fact that you have a thought process that says, "Here's a completely unsupported idea - disprove it people! If you can't my unsupported idea is true," shows that you are, in fact, not supportive of science at all. Do you believe in people being abducted by little green men? No? If your logical reason is lack of proof (and personal opinions are not logical reasons), then you are a hypocrite. Please understand what kind of joke makes you, along with anyone else who thinks like this, look like. A hypocrite.

Also, the last time I checked, your god is losing ground on "supposedly" his own words, especially within the last 100 years. Religion dissolving and people losing parts of their core belief system and religious cherry-picking is proof of that. Science is replacing god's place more and more, year after year. If your god really "loves" you then he will wake the fuck up and replace some science with himself. The atheists are waiting.

Re:Define soul.

[MightyMartian]

Quantum physics tells us that in some situations, it's not possible to measure all aspects of a particle at once.

Which has nothing to do with the soul. Why do people trying to support unevidenced ideas always jump immediately to quantum mechanics? It isn't what you think it is, it's not some sort of universal "get my shitty idea instant credibility" card, not even if your Roger Penrose.

Re:Define soul.

[Kazoo the Clown]

When you look at individuals who have suffered brain damage or degenerative brain diseases, you see that their "consciousness" can be fragmented or incomplete in a whole variety of ways (read most anything by Oliver Sachs for some specific examples, or carry on a conversation with someone with Alzheimer's).

Apologists have tried to argue that their "soul" has flown the coop or is halfway-out of them or simply some "connections" to the body are broken or something, but that seems a much more convoluted argument and fails to explain much of the observed phenomena.

Certainly memory is often impaired, which if it were a function of the "soul" would presumably exist somewhere, yet access to it is clearly limited in many individuals. So go ahead, make up a lot of voodoo to explain it so that it includes a "soul," but don't expect the convoluted rain-dance around the evidence to be very convincing.

What's a soul?

[VincenzoRomano]

Do Humans have one?
If so, anything else can.
Unless someone has a proof otherwise.

Pointless...

[Spad]

The religious will argue that a soul is something unique to mankind, embued by whichever creator their faith believes in, making it impossible for machines to ever have soul.

The athiests will argue that there's no such thing as a "soul", only sentience and/or self-awareness.

Others will meander aimlessly between the two.

Re:Pointless...

[ScentCone]

No matter how fine the granularity of the responses of the AI becomes, it's still just a collection of little functions that passed the point of "photorealism" from a conversational perspective. That doesn't mean it's self aware.

So, does that means that YOU are not self-aware? All you are is a collection of complex cellular interactions. You have a finite number of neurons, make a huge but finite number of connections. Those connections behave in subtle ways that are influenced by a finite number of conditions. So, you can't imagine a combination of hardware and sortware that can emulate a neuron? Or a thousand neurons and their interactions?

There's a big difference between the number of neurons you have and an infinite number of them. You want to be careful using the word "never" when all of the pieces of the puzzle are finite, and increasingly well understood. Massively interconnected neural pathways - whether hardware, software, or some future bio-engineered replacement - are no more inconceivable than are tiny microprocessors containing millions of transistors, operating miniscule radios with keyboards that let you read this sentence on a photorealistic display. Those technologies were previously considered outlandish or prohibitively, incomprehensively complex. And that was just earlier in the lives of millions of people who today use such technology before breakfast every morning.

Re:Pointless...

[cowscows]

Interesting analogy, but I don't think you take it far enough. Of course images on a screen won't ever actually become an apple, because images aren't trying to actually be the object, just a picture of an object.

Let's take it further. What if I hook that computer up to a 3d printer that can model an Apple? What if somewhere down the line the technology advances to the point where that 3D printer can assemble the appropriate organic molecules and shape them into an apple? What if it gets to the point where it can do so with such accuracy that you could eat that assembled apple and not be able to discern at all whether it was grown in a machine or grown on a tree? Is it an apple then?

I would agree that an AI that's functionally just a bunch of pre-programmed responses to various inputs is unlikely to be self-aware, no matter how many responses it's capable of and who it might fool. But to argue that that is the only direction in which AI can go seems to be short-sighted.

Do humans have souls?

[Reality Master 201]

Before we talk about computers, let's talk about ourselves. Do humans have souls?

I don't the answer is clear, and I personally lean towards saying that we don't.

I don't get it

[Andr T.]

The brain is a very, very complicated organ that is still being mapped. We don't even know exactly what part of the brain is responsible for what. It changes, it is a complex matter.

We understand how the muscles work. We know that if they act one way or the other, the person's leg will move one way or the other.

We don't understand how the neurons interact with each other. The consciousness is the sum of the work of those cells we don't understand. So,

there's just a lot of neurons and they're complicated but there's no consciousness to be seen.

This seems rather obvious.

And then, you say 'maybe we can give this thing we don't know what is and we don't know for sure how to define for robots'. Ok, maybe. Maybe there's a FSM above us judging our actions. Maybe.

looks like a three step process?

[v1]

Historically, the way we've discovered what part of the brain does what first is by running into someone with an abnormal part of their brain. The visual cortex, hippocampus, etc. So far though, no one has shown up that lacks consciousness.

Now I suppose that would be by a lot of definitions "brain dead", since consciousness is akin to being awake or dreaming, but still we haven't ran into someone that for example, had a brain tumor or took a nailgun to the head that hit a key area that put their lights out for good, on a consistent basis for that area of the brain.

Now not every location in the brain is highly localized. For example, the area of the motor cortex that controls speech is known, roughly, but it varies slightly from person to person. It's likely that consciousness is a highly distributed function of the brain. That's going to make it a lot harder to study.

I think the whole idea of referring to consciousness as an "emergent property" boils down to our not understanding what causes it, multiplied by it seeming to require a highly complex system to support in the first place.

100 years ago if you'd have presented a mathematician with a laptop with Mathematica loaded on it, he'd probably consider it sentient.

My personal take on it is that consciousness is the brain constantly considering a myriad of possibilities, trying to determine their outcome/impact, in an effort to shape future events in a desirable way by adjusting our actions to try to achieve those outcomes. This is a brute force search, and requires the insanely massive parallelism the brain is designed for. Until we can build a system capable of parallelism on that level, we will not have a "conscious" machine. Everything else before that is a fake, trying to cheat that basic requirement by using shortcuts through linear processing. Simple organisms we don't consider sentient behave exactly as we'd expect a linear system to, directly reacting in a predictable way to provided stimulus, with no ability to learn. Learning is the process of tweaking the values used to consider past events, in order to alter present behavior, to achieve a more desirable outcome in the future. Learning and consciousness go hand in hand.

You can see the middleground in a lot of less complex animals. Give a reasonably advanced animal a tool and a reward achievable by proper use of the tool, and they will play with the tool, experimenting with different way to use it until they get lucky and get the reward. Then it quickly becomes easier and easier for them because they've learned to use the tool. That's the "considering the possibilities" done live and with the tool, which may be most of what people consider "thinking" or "consciousness". I believe what "separates us from them" is that we can do this consideration without having the tool in hand. We can imagine future use of the tool and work out in advance what we need to do with it, or to at least select the proper tool in advance. If you give a monkey a toolbox full of tools it may take them some time experimenting to figure out which tool is the right one to loosen the screw to open the box with the banana in it. Maybe this "imagination" is a third ingredient?

Even after we get the parallelism problem solved, there's the matter of the wiring. Evolution has lead brains to be preprogramed to do both the learning and the consideration, and that may turn out to be a tough system to figure out and duplicate. Or it may be pathetically simple. Best guess here is we will get parallelism figured out, then learning, and the last hurdle will be the imagination behavior.

Oh, of course machines can have souls!

[Mateo_LeFou]

We're machines, and *we have souls...