Jump to content
Objectivism Online Forum

Artificial Intelligence

Rate this topic


Prometheus

Recommended Posts

Mr. Speicher, you are essentially saying that consciousness is caused by the brain, but it cannot be reduced to just physical processes, or have i misunderstood what you are saying?

If say physical processes are responsible for only 20% of consciousness, then where does the other 80% come from? Or are you saying, that given this day and age, we just don't know yet, what constitutes the other 80%? (the percentages are just a conceptual tool, no more)

If consciousness cannot be wholly reduced to physical processes, wouldn't the rest have sprung from a void?

I am not a biologist, so if there's some grave error i have committed then please point it out.

dinesh.

Link to comment
Share on other sites

  • Replies 88
  • Created
  • Last Reply

Top Posters In This Topic

...I know nothing about digital algorithms (i.e. a Turing machine) that prevents them from supporting intelligence.

Here's one thing: computers don't have free will. They are deterministic, and unavoidably follow whatever course is necessitated by their programming, environment, etc. They are fundamentally more similar in their nature to a rock (albeit much more complex) than to a human being. Absent volition, there can be no such thing as intelligence or objectivity.

For that reason, I will consider any discussion of AI as a serious possibility to be arbitrary until some evidence is given that we could reproduce free will in inanimate matter. And the only way that might ever happen is after we discover how it arises in ourselves, and we don't even know that yet.

Link to comment
Share on other sites

On a serious note, I think it would be correct  to say that while consciousness cannot be reduced to any biological process, there are biological environments and potentially digital algorithms than give rise to consciousness.  I know nothing about digital algorithms (i.e. a Turing machine) that prevents them from supporting intelligence.

All of the evidence supports consciousness as being dependent on life, and none of the evidence supports consciousness to depend on "digital algorithms." If you think it possible, the burden of proof is on your part. No one has to show what would prevent "digital algorithms" from supporting consciousness, but rather you need to show why this might be so. Absent of such evidence, it is not possible, it is arbitrary.

Link to comment
Share on other sites

Mr. Speicher, you are essentially saying that consciousness is caused by the brain, but it cannot be reduced to just physical processes, or have i misunderstood what you are saying?

Consciousness cannot be reduced to neural processes, period.

If say physical processes are responsible for only 20% of consciousness, then where does the other 80% come from? If consciousness cannot be wholly reduced to physical processes, wouldn't the rest have sprung from a void?

You misunderstand. Our consciousness is 100 % dependent on the brain for its existence, and consciousness is a completely different thing from the brain. Consciousness is ... 100 % itself. It is an irreducible primary. As far as we know there is nothing else in existence like consciousness, and any attempt to reduce it to neural processes will only lead to failure and confusion. Some people like to think of consciousness as an emergent property, an independent property arising from the brain. Personally, I do not think that too helpful, because it really does not explain anything. Consciousness is not neural processes, period.

Link to comment
Share on other sites

You are wrong, and as long as you continue to not understand why, you will continue to think our differences to be "trivial" or "absurd." They are neither.
Don't misunderstand, I absolutely disagree with you on several important issues in regard to conscoiusness. What we agree on is that consciousness has an existence separate from the brain, and that consciousness cannot be reduced to the brain (neither philosophically nor scientifically). I can say that now because I am using the precise definition of reduction that you've given, which you say can also be regarded as the Objectivist definition. That is not at all a trivial or merely semantic issue. It is a matter of clearly defining the concept involved. Discussion of whether this or that could be reduced would be impossible without such a definition. Your insistence that I do not agree with you on this particular point is bizarre, since I have explained myself several times now. Consciousness does not consist of neural processes.

What is very trivial and absurd is your continued misunderstanding over what the word corresponding means. You seem to be insisting that I'm contradicting the fact that consciousness cannot be reduced to the brain when I say that conscious states have a corresponding neural mechanism. But what I told you is that when I say corresponding, I mean what you meant when you said "accompanying." Corresponding does not have to mean "similar or analogous to." For the Random House Dictionary definition see my previous post.

The only way we could still be in disagreement over this is if you are in fact attempting to claim that consciousness, or some aspect of it, has no physical expression in the brain. You've said flat out that you are not claiming this, but some of your statements seem to contradict that. I cannot discern your actual position.

Now, if you disagree with me that the output of the neural circuitry controlled by consciousness is not determined, that is a different matter, and I would be happy to hear why.

To start with, how can consciousness itself be nondeterministic, but its actions in the physical world be determined. Neural action, after all, is precisely the action that results when consciousness does anything, from thinking to moving a muscle. So how can the output of the neural systems controlled by consciousness be determined?

Link to comment
Share on other sites

Here's one thing: computers don't have free will.  They are deterministic, and unavoidably follow whatever course is necessitated by their programming, environment, etc.

I think that’s an incorrect representation, because what you are in fact saying is that the underlying physical components of computers don’t have free will – which is true, as they are entirely causal. On the other hand, the human brain is entirely causal as well. Consciousness is a phenomena that (necessarily) arises from a causal mechanism. It is incorrect to attribute the characteristics of consciousness to the mechanism that supports it.

For that reason, I will consider any discussion of AI as a serious possibility to be arbitrary until some evidence is given that we could reproduce free will in inanimate matter.

You are right in that discussion of artificial intelligence consciousness is arbitrary because we have no idea of what physical conditions cause it to come about. However, it is also an arbitrary claim to say that a biological process can give rise to consciousness while an inorganic one cannot – unless you know something about computers that I don’t.

Link to comment
Share on other sites

All of the evidence supports consciousness as being dependent on life, and none of the evidence supports consciousness to depend on "digital algorithms." If you think it possible, the burden of proof is on your part. No one has to show what would prevent "digital algorithms" from supporting consciousness, but rather you need to show why this might be so. Absent of such evidence, it is not possible, it is arbitrary.

That’s not entirely correct. It is true that there is self-evident evidence that living entities give rise to consciousness, but it is not true that there is no evidence that digital algorithms can do the same. Turing proved that a digital algorithm can simulate any algorithm found in nature – including those of the human brain. At the very least, this proves it possible that it is possible to make an exact simulation of a human brain. Now, a computer running such a simulation would probably be bigger than our galaxy, but we do in fact have some evidence that it is theoretically possible. The challenge is to determine the essential physical process of the brain that give rise to consciousness and reproduce them in digital form. The current state of computer science and psychology may be light-years away from achieving that, but we know that it is at least theoretically possible.

Link to comment
Share on other sites

GC, you might be interested in the article, "Mindless Intelligence" by Christian Beenfeldt in the latest (April) issue of The Intellectual Activist. Briefly, it describes how the Turing Test of computer intelligence is logically unsound and philosophically corrupt. It substitutes the ability of a computer to mimic human conversation for the actual presence of any conscious understanding. The article shows that Turing's ideas about consciousness were thoroughly irrational. As to a Turing Machine, I'm not familiar enough with what it is to comment.

That said,

The challenge is to determine the essential physical process of the brain that give rise to consciousness and reproduce them in digital form.
I agree that this is possible. I also agree with your statements two posts up.
Link to comment
Share on other sites

...However, it is also an arbitrary claim to say that a biological process can give rise to consciousness while an inorganic one cannot – unless you know something about computers that I don’t.

I did not say that an inorganic process cannot give rise to consciousness. I said that it is arbitrary to posit that they can, until some evidence for this is provided.

(I don't take Turing's "proof" to be evidence. It is rationalism, and logically fallacious rationalism to boot. Anyway, he--to my understanding--proved only that all computable functions are "Turing computable"--not that all functions are computable. And if you buy into that kind of "logic," it's supposedly also been "proven" that there are many non-computable functions.

You will never get intelligence, free will, or even consciousness from a tape with instructions to move back and forth and mark or unmark. If we can ever reproduce intelligence, it would have to be...well, I don't even know what it would look like, because we just don't know what gives rise to those things yet, and it would be arbitrary speculation on my part to try to guess at this point. But it will certainly not be a Turing machine.)

Link to comment
Share on other sites

What is rationalistic about the Turing Machine, and which functions not computable?

Ash, how can you say what will not produce intelligence if you admit that you don’t know what gives rise to it in the first place? Again, I think you are attributing some functions of consciousness to the mechanism that enables it, which is a fallacy. It’s true that Turing’s test has been used by philosophers to make tall claims about intelligence itself, but I think the concept of a Turing machine itself is entirely sound.

Amagi, I would probably agree with Christian’s evaluation of the Turing test. At the same time, I think it would be impossible for any non-intelligent algorithm to fool me into confusing it with the real thing. The notion that consciousness can be “simulated” without the existence of a conscious entity is wrong.

Disclaimer: My knowledge of computer science and Turing is limited to a casual skimming though a few websites.

Link to comment
Share on other sites

All of the evidence supports consciousness as being dependent on life, and none of the evidence supports consciousness to depend on "digital algorithms." If you think it possible, the burden of proof is on your part. No one has to show what would prevent "digital algorithms" from supporting consciousness, but rather you need to show why this might be so. Absent of such evidence, it is not possible, it is arbitrary.

That’s not entirely correct.  It is true that there is self-evident evidence that living entities give rise to consciousness, but it is not true that there is no evidence that digital algorithms can do the same.

Okay. That was my point -- provide the evidence and convince me.

Turing proved that a digital algorithm can simulate any algorithm found in nature – including those of the human brain.  At the very least, this proves it possible that it is possible to make an exact simulation of a human brain.

No. Consciousness is not an "algorithm found in nature." Consciousness is guided by volitional choice, and therefore stands apart from any algorithm, digital or otherwise.

You have failed to provide any evidence that "digital algorithms" can "give rise to consciousness." Though, I do remain open to such evidence, when and if you can provide it.

Link to comment
Share on other sites

What is rationalistic about the Turing Machine, and which functions not computable?

One such function is the so-called "halting problem," and there are supposedly an infinite number of other non-computable functions.

...It’s true that Turing’s test has been used by philosophers to make tall claims about intelligence itself, but I think the concept of a Turing machine itself is entirely sound.
Yes, I think you're right that the concept of a Turing machine itself is sound, but I think that the attempt to apply it to everything that exists is rationalism.

Ash, how can you say what will not produce intelligence if you admit that you don’t know what gives rise to it in the first place?  Again, I think you are attributing some functions of consciousness to the mechanism that enables it, which is a fallacy.

Perhaps you're right that I am going too far in making that claim. I'll have to think about it some more. But certainly the claim that they can produce intelligence is arbitrary, even if the claim that they can't is also arbitrary. We just don't know at this point.

Amagi, I would probably agree with Christian’s evaluation of the Turing test.  At the same time, I think it would be impossible for any non-intelligent algorithm to fool me into confusing it with the real thing.  The notion that consciousness can be “simulated” without the existence of a conscious entity is wrong.
Why? I think it's much more plausible that we could create a computer that would simulate consciousness (though this would be the galaxy-sized computer you hypothesized in an earlier post) than one that would actually be conscious. One claim I can at least imagine. The other is arbitrary. How would you tell the difference?

Disclaimer: My knowledge of computer science and Turing is limited to a casual skimming though a few websites.

My knowledge of Turing is limited mostly to what I've learned in my symbolic logic class this semester, so that may explain why I distrust him. ;)

Link to comment
Share on other sites

You have failed to provide any evidence that "digital algorithms" can "give rise to consciousness." Though, I do remain open to such evidence, when and if you can provide it.

I did not say that a “consciousness algorithm” exists in the human brain, or is even possible. As I understand the term, “consciousness” is a concept that refers to the capabilities of a mental faculty. Foremost among these capabilities is perception, cognition and the initiation and direction of mental action.

Given that consciousness is a volitional state, I think it would be incorrect to claim that consciousness can be reduced to or represented by any algorithm, if by “reduced” you mean, “this algorithm is a conscious process.” However, I think it is correct to say that there are certain biological and potentially digital processes/algorithms that support or give rise to consciousness. By “give rise to” I mean “support a process with certain characteristics.”

Again, refer to my concept of a simulated brain. It is theoretically possible to create a digital simulation of every single atom in the brain. Since the behavior or atoms is entirely causal, it can be represented exactly in virtual form. The same physical properties that give rise to consciousness in humans will then be replicated in digital form. Now, I am not saying that such a simulation is feasible, but it is a useful theoretical model. Note that it would be a fallacy to claim that the simulated neurons don’t have free will or an “arbitrary” state, since the real neurons of the human brain do not have such features either – objects of consciousness cannot be reduced the entities that support it.

Link to comment
Share on other sites

I don't take Turing's "proof" to be evidence.  It is rationalism, and logically fallacious rationalism to boot.

I think you are almost being too kind. It is due to such "thinkers" as Daniel C. Dennett (Consciousness Explained) that a portion of the current generation has swallowed whole some of the worst anti-mind irrationalities that have ever been perpetuated by man. The only thing worse than straight philosophical absurdity is when it is clothed in scientific robes. At least the mystics of old attacked the mind directly. Today we witness the insidiousness of consciousness being attacked from within. Self-destruction, so to speak. Very sad.

Link to comment
Share on other sites

Why?  I think it's much more plausible that we could create a computer that would simulate consciousness (though this would be the galaxy-sized computer you hypothesized in an earlier post) than one that would actually be conscious.  One claim I can at least imagine.  The other is arbitrary.  How would you tell the difference?

As I said in an earlier post, consciousness is a mental state capable of “perception, cognition and the initiation and direction of mental action.” Simply put, I don’t think it’s possible to simulate these functions without them actually occurring. I just can’t conceive of how concept-formation can be performed without a conscious entity actually performing it. I will also add that while I am ignorant of theoretical computer science, I do know something about programming, and I can deduce the basic algorithm of any intelligent chat bot within a few responses. (Btw, I once attached a chat bot to my AIM profile with some interesting responses – more than a few people were fooled.)

Regarding the difficulty of creating artificial intelligence, I think there are two opposing factors to consider: that the speed of digital computations is much faster than the biological kind versus the fact that the brain is a neural network whereas digital computers use a linear model. The key question to me is whether intelligence requires a large-scale neural processing model. I don’t think we know enough about consciousness to make a determination of that, but if it does not, true AI may be possible within the next few decades. After all, human intelligence did arise by a natural process, and four billion years of biological evolution may add up to a few weeks of processing with a supercomputer.

Link to comment
Share on other sites

GreedyCapitalist writes:

Again, refer to my concept of a simulated brain. It is theoretically possible to create a digital simulation of every single atom in the brain. Since the behavior or atoms is entirely causal, it can be represented exactly in virtual form. The same physical properties that give rise to consciousness in humans will then be replicated in digital form. Now, I am not saying that such a simulation is feasible, but it is a useful theoretical model.
Exactly. If deterministic matter gives rise to consciousness, but consciousness cannot be reduced to matter, why can't deterministic algorithms give rise to consciousness, even though consciousness cannot be reduced to algorithms?
Link to comment
Share on other sites

Again, refer to my concept of a simulated brain.  It is theoretically possible to create a digital simulation of every single atom in the brain.  Since the behavior or atoms is entirely causal, it can be represented exactly in virtual form.  The same physical properties that give rise to consciousness in humans will then be replicated in digital form.

Real life is not The Matrix, and digital simulations are not the brain.

Link to comment
Share on other sites

why can't deterministic algorithms give rise to consciousness, even though consciousness cannot be reduced to algorithms?

A note on terminology:

Determinism is “the philosophical doctrine that every state of affairs, including every human event, act, and decision is the inevitable consequence of antecedent states of affairs.” (Heritage Dict.) Since living entities are initiators of their own actions, determinism is an erroneous theory, and the term “deterministic” has no valid applications. Actions of all entities are causal, not deterministic. Living entities are unique in that they are the prime movers of their actions.

In answer to your post, I think it would be correct to say that if my example of a simulated brain is valid, (causal) digital algorithms can theoretically give rise to consciousness.

Link to comment
Share on other sites

Real life is not The Matrix, and digital simulations are not the brain.

While your statements are obviously correct, I fail to see how they answer my point.

Link to comment
Share on other sites

GreedyCapitalist, I don't see any reason to reject the term deterministic simply because it is associated with an invalid philosophy. We don't reject the word subjective because subjectivism is philosophic drivel.

To say that some isolated system of matter is deterministic does not imply that the whole of existence is deterministic.

Even long-time Objectivists knowledgeable in the ways of physics use the word:

For a non-technical overview of Little's TEW, which is a causal, deterministic, and local theory, see my three-part article at:

--stephen speicher

The word is useful because causal is not an exact synonym for deterministic in the context you've used it in. You can certainly find, for instance, talk of the causal nondeterministic random nature of quantum mechanics. Whether or not this is logically sound is another matter...

Link to comment
Share on other sites

Since living entities are initiators of their own actions, determinism is an erroneous theory, and the term “deterministic” has no valid applications.  Actions of all entities are causal, not deterministic.

The philosophical/psychological doctrine called "determinism" does deny volition and is indeed self-contradictory. However, you over-generalize when you state that "the term 'deterministic' has no valid applications." The term "deterministic," as it is commonly applied in the physical sciences, is most certainly a valid application.

One definition of "determinism" is "The doctrine that everything that happens is determined by a necessary chain of causation." [Oxford English Dictionary] The physically deterministic behavior of matter means those actions which are caused and necessitated by the entity's nature. That is a perfectly valid use of the term.

Living entities are unique in that they are the prime movers of their actions.

No. All living entities are self-regulatory, but only an entity which possesses a volitional consciousness, like man, can be said to be a "prime mover of their actions."

The self-regulatory actions of life on the vegetative level are fully determined by the physiological/biological processes of the entity. The self-regulatory actions of the lower-level conscious animals are much more complicated but are still biologically "programmed" through the physical mechanisms (sensory-perceptual, etc.) which regulate their behavior, and are automatic.

By contrast, with man's volitional consciousness comes the unique ability to regulate the action of his own consciousness, and it is that fact -- the existence of a primary choice -- which makes man the only "prime mover of [his] actions."

Link to comment
Share on other sites

However, you over-generalize when you state that "the term 'deterministic' has no valid applications.
So is “deterministic” is a synonym for “casual?” Why the need for a separate term them? I find it interesting that Amaqi mentioned that QM is “causal nondeterministic,” since it shows the errors one can fall into when using that term.

The self-regulatory actions of life on the vegetative level are fully determined by the physiological/biological processes of the entity.

So there isn’t a fundamental difference between a computer virus a dog?

Link to comment
Share on other sites

I find it interested that Amaqi mentioned that QM is “causal nondeterministic,” since that shows the errors one can fall into when using that term.
The point was to show that the words denote different concepts. Causal means simply having a cause. Deterministic means that the outcome of a process is fixed by the initial conditions.

[Edit: I removed the rest of this post, primarily because it was confused and superfluous]

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.

×
×
  • Create New...