We could build an artificial brain that believes itself to be conscious. Does that mean we have solved the hard problem?

[fromderinside]

...

what?

I don't actually understand what this sentence is supposed to mean. For one, the consciousness would still be an emergent property of the complex system of data exchange; and two, emergent properties are *not* in conflict with any principles of science; nor is the concept of emergent minus anything else.

One cannot run an experiment if that experiment generates emergent,

Is this even a proper sentence? You can't use the word 'emergent' in that fashion.

Anything else is psychological bullshit or magic if you will. The fool who spouted "The sum is greater than the sum of its parts" had a pfart problem.

By calling emergent properties 'psychological bullshit or magic', you strongly suggest you don't actually know what you're talking about. The sum being greater than the sum of its part; ie, as in emergent properties, is a scientifically valid notion that is observed in a great number of physical systems. It is neither bullshit nor magic; it is a simple logical observation that if one configures the individual parts in the right way, functions can arise that can not be achieved by these parts on their own. Take apart a car engine, and you just have a bunch of parts that aren't very useful on their own. Put them together in the right way, and you create something far more capable/useful than if you just took all the parts and randomly taped them together: in other words, the sum of the engine is greater than the sum of its parts. Following that basic fact, we can apply this thinking to consciousness. We already know (from observing the connection between human consciousness and brains) that complex data exchange systems (such as neural networks like that of the brain) are a necessary part of forming consciousness. Since we do not, however, know exactly how to configure such a system to produce consciousness, and we have thus far not been able to demonstrate any configurations to be dead ends, we can say that all such complex systems could *conceivably* (again, operative word) give rise to consciousness. This is not even remotely 'bullshit or magic', it's a simple logically consistent extrapolation of empirical observation.

Wow. I'm gonna be gone for a few hours. When I come back be ready to compare articles.

Place held.

My articles are: Reductionism redux: http://www.idt.mdh.se/kurser/ct3340/archives/ht02/Reductionism_Redux.pdf

Read the highlighted parts to get the basics for why this is the right topic for parts equal to the whole.

Reductionism, Emergence,and Effective Field Theories http://arxiv.org/pdf/physics/0101039.pdf

This article breaks down the current arguments in physics about reductionism and other than reduction boils down to a funds competition by those who aren't trying, as yet, to relate their science to physics, I understand and sympathize, but, government laziness and AAAS nearsightedness aren't sufficient reasons to overthrow a model that consistently, when related to other disciplines, relates those systems and systems to the physical science. A technical good read and a clever argument, but, one without substance beyond energy effect boundary conditions. We just found the Higgs Boson, ferchrissake, using a machine entirely built depending on sum is equal to parts with methods based on sum equal to parts, ie: entirely reductionist.

Also I'd like to add that we find emergence because we don't have a complete list of parts* which is why those other scientists want to go their own way.* They don't have enough information so they want to develop other schemes whereby they can make sense of whats going on, a desirable goal, but,because they don't have the physics at hand they are told "get it". Actually that argument has another drawback. They invent emergent macro rules that hold together for a while, finally being overthrown, then have to go back and take another tack while the reductionists are still plodding ahead with uninterrupted advances. If you don't believe look at the history of neuroscience and psychology and see whether threads based on physical roots are bouncing around or whether those that aren't so base are. Believe me the latter are bouncing around yield a new thread of schools about every twenty years or so.

*Think of their problem as one where we more or less completely understand whats going on at the entrance to the ear so we understand in physical terms what the first neurons are doing. Whereas, on the other hand others are looking at the medial lateral frontal lobe and speculating based on physical change in oxygen uptake by those cells what the conscious brain is doing with that known stuff we found at the cochlear nucleus.

Finally a third article to illustrate that those who claim emergence are doing so because they don't have a firm grip on the parts. Immune Privilege and the Philosophy of Immunology http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3959614/

Like other scientists, immunologists use two types of approaches to research: one reduces the problem to its parts; the other studies the emergent phenomenon produced by the parts. Scientists that reduce the problem to its parts are sometimes called reductionists. The conclusions of reductionist experiments are often applied to the greater whole, when in actuality they may only apply to that particular experimental set. We, reductionists, are the ones who think our immune behavior exists solely because of genes, the presence of TGFβ, the presence of inflammatory cytokines, and appearance of a receptor.

Please note the problem is not that the methods won't work, its that they are too narrowly position to explain a discipline that goes way beyond than what they understand. So they invent a new approach that doesn't have those firm roots and they sally forth 'finding' emergence's everywhere in a set pf parts largely unknown. Its good they do the research. It's wrong that they think they are explaining. They should be comparing with what's known and using those emergence's to find ways to make their knowledge actually more complete.

Your turn.

... the hell is this shit even? None of this argues against anything I've been saying. First off, you're apparently understanding the term 'emergent property' to mean something it doesn't actually mean or at least a meaning that isn't being used here. Secondly, you then cite some stuff where peope are arguing against *specific* things (such as immune behavior) being an emergent property, and incidentally in doing so appear to actually not be arguing against immune behavior being an emergent behavior at all because they're still saying it takes a number of different things, the *combination* of which produces immune behavior where the individual parts don't accomplish this just on their own... which is the very definition of an emergent property. Even if we argued that this immune behavior is *not* an emergent property on the logic that the individual parts produce similar effects and the combination thereof doesn't produce a new effect simply amplifies existing ones... this in no way demonstrates that, for instance, the patterns of ice crystals that form snowflakes aren't an emergent property.

Of course, the problem here is that the notion of emergence/emergentism isn't strictly defined, and there are certainly ways to use it that lean more towards the 'magical bulshit' side of the spectrum. But I don't see how anything I've said leads one toward that conclusion, since that certainly can't be said about the statement that the interaction of individual properties can lead to the emergence of patterns/systems that the properties do not exhibit on their own... which really isn't a controversial subject in science and something quite easily demonstrated.

So the only response to your post I have to offer is a confused 'wat? what are you even on about?'

Your response here gives me what I need. It emerges because we don't understand the properties inherent in what we 'tested'. That isn't emergence. That is producing an answer that requires more questions and tests. All of this is covered in the articles I provided. They do need be read to understand that though. For instance, calling the stuff found emergent is no better than those wishing to treat the immune category as different from physics having its own rules and laws based on some computational imagining so it can reside in the overall schema of science by declaring those imaginings anchors for emergence.

Iike I said. Research that cannot today directly explain itself in terms of physical law need be performed. It jest needn't be performed with magic as its justification. Eventually it will incorporate and there will be instances where, say, Newtonian physics is good enough for performing technical work. And yes. Reduction following the principle of sum equal to parts is basic. The other, "greater than sum of parts", while easy to say in the presence of the unwashed just because one can point to this and show that just doesn't work as explanation. Its greater because of something as yet unexplained.

Every anti-reductionist theory has gone down inflames over the past 600 years.

 

[fromderinside]

I'm suggesting neural nets or any other nets or whatever one chooses to replace nets to serve the integrative and deciding processes aren't necessary.

Are you saying that human level consciousness don't require brains to exist, or to perceive organized thought?

hahaha.. organized...

Very funny. Do neuroscientists talk about brains in terms of neural nets other than dofting hats to AI wizzards? I think not.

Brains do seem to have been tacking toward organization as a primary evolutionary outcome in vertebrates at least.

As social beings its pretty evident much humans do is aimed at defusing threat from our most lethal adversary, our kind. Clearly this is the primary context in which modern consciousness is scientifically discussed. Nanoscientists and AI scientists take many of their cues from these findings, much to their peril it seems to me.

My latest thrusts have been to advise of these perils and to remind how much traditional programming has accomplished in aiding and replacing human agency. AI getting all hepped up over the brain seems a silly venture. Stick with mimicry and replacement. No. Neural nets are not memicry. They are attempts to replicate brain elements which are way beyond our kin. Don't look at the trees. See the forest.

 

[Kharakov]

We've already explained how to do it in this very thread, so to say nobody has the faintest idea is just ignorant: Whole Brain Emulation.

Emulation of a physical system != the creation of a physical system. You might be aware of the difference between a simulation and reality.

A perfect simulation of physical reality (because we can simulate quantized reality), which uses the movement of electrons from one place to another, is not going to have actual brains in it. The brains are going to be data, not physical brains.

Feeding the information of the simulated reality into an actual brain/consciousness will result in a consciousness experiencing the simulation. However, brains running in the simulation will not actually have consciousnesses associated with them (they are just 1s and 0s), unless some God level being is observing everything and decides to imbue them with consciousness, by observing their consciousness into existence. That would be interesting...

This isn't "far off future" stuff. This isn't even "maybe in our lifetimes" stuff. We're talking about full human brain emulation by the mid 2020's.

That sounds awesome. The thought that your brain isn't already an emulation is.. funny, in a dystopian type of way.

 

[dystopian]

Your response here gives me what I need. It emerges because we don't understand the properties inherent in what we 'tested'. That isn't emergence. That is producing an answer that requires more questions and tests. All of this is covered in the articles I provided. They do need be read to understand that though. For instance, calling the stuff found emergent is no better than those wishing to treat the immune category as different from physics having its own rules and laws based on some computational imagining so it can reside in the overall schema of science by declaring those imaginings anchors for emergence.

Iike I said. Research that cannot today directly explain itself in terms of physical law need be performed. It jest needn't be performed with magic as its justification. Eventually it will incorporate and there will be instances where, say, Newtonian physics is good enough for performing technical work. And yes. Reduction following the principle of sum equal to parts is basic. The other, "greater than sum of parts", while easy to say in the presence of the unwashed just because one can point to this and show that just doesn't work as explanation. Its greater because of something as yet unexplained.

Every anti-reductionist theory has gone down inflames over the past 600 years.

Like I said, you really don't seem to understand what we're talking about here, but are rather operating from some overdrawn assumptions about my intent and meaning. Modern science is not and can not be purely reductionist in nature; and emergentism does not conflict with reductionism it just demonstrates there are limits to it, which is especially evident in complex systems. The bulk of science is still reductionist in nature; but this does doesn't preclude the acceptance or existence of emergent systems and to insist on pure absolute reductionism like what you're doing is an inappropriate use of reductionism that does nothing to advance our understanding of the universe and a great to frustrate it.

I don't think this sub-thread is going anywhere. A) because you seem unwilling to understand what I'm actually saying and are injecting meanings (that are admittedly pushed by some who would misuse emergent properties for anti-scientific purposes, but which are inappropriately biasing you) that aren't actually used here, and B) because it seems your only pro pure-reductionist argument is the belief that the reason emergent systems are "greater than the sum of their parts" is because we haven't identified all the parts yet; which is just something you'd be taking on faith and something that even if true wouldn't really relate to anything we've been talking about anyhow.

So yeah...

...time to bury this part of the thread, at least.

 

[dystopian]

Emulation of a physical system != the creation of a physical system. You might be aware of the difference between a simulation and reality.

The difference between reality and a simulation of sufficient complexity is an arbitrary and meaningless distinction.

If the simulation is complex enough that it perfectly simulates every aspect of the physical system, then there is no real difference.

A perfect simulation of physical reality (because we can simulate quantized reality), which uses the movement of electrons from one place to another, is not going to have actual brains in it. The brains are going to be data, not physical brains.

Which like I said, is an aribtrary and meaningless distinction if we're talking about a perfect simulation of physical reality. There is no difference between the simulated brains and real brains in that case. We are, ourselves, really just datapoints existing within a larger framework (spacetime). There is no fundamental difference, regardless of your personal (and poorly argued) opinion to the contrary. All you're doing is stating that a simulation can't have consciousness because it's 0's and 1's, failing to explain how this fact somehow precludes it.

That sounds awesome. The thought that your brain isn't already an emulation is.. funny, in a dystopian type of way.

It is entirely possible that either of our brains are emulations or existing within one. Indeed, some argue that the odds are pretty good that this is the case. Does the possibility frighten you? You might actually be terrified to learn that theorethical physicists James Gates claims to have found binary computer code in the supersymmetry equations of string-theory. That is actually true. I am not making that up. Does it prove we live in a simulation? No. But it's definitely interesting, and something I imagine might cause you to go catatonic.

 

[skepticalbip]

The difference between reality and a simulation of sufficient complexity is an arbitrary and meaningless distinction.

If the simulation is complex enough that it perfectly simulates every aspect of the physical system, then there is no real difference.

A perfect simulation of physical reality (because we can simulate quantized reality), which uses the movement of electrons from one place to another, is not going to have actual brains in it. The brains are going to be data, not physical brains.

Which like I said, is an aribtrary and meaningless distinction if we're talking about a perfect simulation of physical reality. There is no difference between the simulated brains and real brains in that case. We are, ourselves, really just datapoints existing within a larger framework (spacetime). There is no fundamental difference, regardless of your personal (and poorly argued) opinion to the contrary. All you're doing is stating that a simulation can't have consciousness because it's 0's and 1's, failing to explain how this fact somehow precludes it.

That sounds awesome. The thought that your brain isn't already an emulation is.. funny, in a dystopian type of way.

It is entirely possible that either of our brains are emulations or existing within one. Indeed, some argue that the odds are pretty good that this is the case. Does the possibility frighten you? You might actually be terrified to learn that theorethical physicists James Gates claims to have found binary computer code in the supersymmetry equations of string-theory. That is actually true. I am not making that up. Does it prove we live in a simulation? No. But it's definitely interesting, and something I imagine might cause you to go catatonic.

It looks like the problem here is like most philosophical arguments. You have not defined what you mean by consciousness. From your arguments it seems to be that you are calling some system that can react to stimuli based on programmed instructions and data conscious. This would make the roomba vacuum cleaner conscious, but not very intelligent.

 

[Kharakov]

The difference between reality and a simulation of sufficient complexity is an arbitrary and meaningless distinction.

If the simulation is complex enough that it perfectly simulates every aspect of the physical system, then there is no real difference.

So what? We're not talking about recreating the universe on a smaller scale. We're talking about using data crunching to simulate physics and consciousness, not creating physical structures for a physical simulation.

Is the difference between the distance from here to the Moon and its numeric representation too hard for you to grasp? Is the fact that the numbers can be used to create various accurate depictions of the relationship between the Earth and the Moon going to confuse you as to whether the numbers are an actual distance or not?

What do you think we do, stuff the distance into the numbers, which is why they allow us to predict stuff related to the distance?

Seriously, you need to wrap your head around the fact that numbers, which can be used to create realistic looking images (and thus could be used to create accurate simulations), are not what they describe. Nor is the data manipulation we can do on the numbers, to create things in reality, the same thing as what the numbers describe.

All you're doing is stating that a simulation can't have consciousness because it's 0's and 1's, failing to explain how this fact somehow precludes it.

You really don't understand that a mathematical simulation of consciousness, no matter how fine tuned, is not going to equate to consciousness itself? If that's the case, I'm impressed, once again, with your failure to understand simple obvious concepts. It's a gift.

You might actually be terrified to learn that theorethical physicists James Gates claims to have found binary computer code in the super symmetry equations of string-theory.

He found ASM code or something? That would be hilarious.

Keep in mind that some of the math that string theory relies upon is easily manipulable into many values (zeta function regularization is not the only way to arrive at specific values for the one divergent series- in fact one can manipulate them into just about any value one wants). In fact, the human mind (brain) can be easily manipulated into seeing values that aren't there, so with that in mind...

 

[Kharakov]

It looks like the problem here is like most philosophical arguments. You have not defined what you mean by consciousness. From your arguments it seems to be that you are calling some system that can react to stimuli based on programmed instructions and data conscious. This would make the roomba vacuum cleaner conscious, but not very intelligent.

No, dystopian just doesn't get the difference between mathematical simulations and reality. I suppose he thinks that if he imagines a 7 is conscious, it will all the sudden light a cigarette and crack a beer while it leans nonchalantly against the wall.

 

[fromderinside]

The difference between reality and a simulation of sufficient complexity is an arbitrary and meaningless distinction.

If the simulation is complex enough that it perfectly simulates every aspect of the physical system, then there is no real difference.


Which like I said, is an aribtrary and meaningless distinction if we're talking about a perfect simulation of physical reality. There is no difference between the simulated brains and real brains in that case. We are, ourselves, really just datapoints existing within a larger framework (spacetime). There is no fundamental difference ... . All you're doing is stating that a simulation can't have consciousness because it's 0's and 1's, failing to explain how this fact somehow precludes it.

It is entirely possible that either of our brains are emulations or existing within one. Indeed, some argue that the odds are pretty good that this is the case. .... James Gates claims to have found binary computer code in the supersymmetry equations of string-theory. .... Does it prove we live in a simulation? No. ....

For the purpose of my comments here only your statements are needed. Even less. No inferences about others is needed so they are also redacted. You've made some, what I consider, good points. Still some questions to your points.

1. Does the simulation take into account reactions to it and does it respond appropriately (we can discuss what is meant by appropriate elsewhere).

2. Does the agent (computer, person, whatever) recognize it is aware (one can define how that can be so many ways).

3. What do you mean every aspect of the physical system and whatever it is you consider physical reality, which by the way, includes evidence of what I asked in 1 and 2.

I, too, would be interested in finding out whether critics think 0's and 1's preclude something from consciousness. Actually, I've gone through this thread seeing little attention paid to providing evidence of consciousness. One reason I asked the questions I just asked is because your, and others, saying things like "...every aspect of the physical system".

I would normally read that from a macro-system (one that includes intercourse with conscious entities) point of view. By that I mean mean not only does the thing do what humans do but the thing acts as if it understood it were acting as would a human. That's a lot. If, on the other hand you mean it is acting like a human I would need to know how you can know that unless you had some evidence it knew it was acting like a human.

Finally, this is where we could have gone had you read the posts which you dismiss. I know. 20 small print pages are a bit much to ask. I think its warranted because this is an important topic.

 

[Achwienichtig]

I haven't been keeping up with this thread, but have you all seen this?

Apparently a robot has passed a self-awareness test. Pretty dang cool!

http://www.techradar.com/news/world-of-tech/uh-oh-this-robot-just-passed-the-self-awareness-test-1299362

 

[barbos]

I haven't been keeping up with this thread, but have you all seen this?

Apparently a robot has passed a self-awareness test. Pretty dang cool!

http://www.techradar.com/news/world-of-tech/uh-oh-this-robot-just-passed-the-self-awareness-test-1299362

I am not too impressed. It's pretty clear that robot would not normally make such a human mistake and was simply programmed to behave that way. In other words, robot knew that he was violating an order not to speak before actually doing so. It is like a robot who is bad at simple math,you have to program it to be bad. Yes, and hearing its own voice to figure out that you are talking is way too unnecessary.

 

[DBT]

Information feedback loops are not necessarily an example of awareness, which means to be conscious (if you are not conscious, you cannot be aware), regardless of the test.

''Awareness is the ability to perceive, to feel, or to be conscious of events, objects, thoughts, emotions, or sensory patterns. In this level of consciousness, sense data can be confirmed by an observer without necessarily implying understanding. More broadly, it is the state or quality of being aware of something.'' - Wiki.

 

[dystopian]

1. Does the simulation take into account reactions to it and does it respond appropriately (we can discuss what is meant by appropriate elsewhere).

Why wouldn't it? It is a perfect simulation.

2. Does the agent (computer, person, whatever) recognize it is aware (one can define how that can be so many ways).

If it is a perfect simulation, then we can only assume that to be the case, yes.

3. What do you mean every aspect of the physical system and whatever it is you consider physical reality

This is self evident. Every aspect means every aspect that defines physical reality; physicality being defined, ultimately, by dimensional datapoints. That is what physical reality can be reduced to; the information that determines where in space and time the most basic components we are built up out of are. Nothing more, nothing less. A hypothetically perfect simulation would be one that maps and simulates this data to the same extent as does physical reality. Since both can be reduced to datapoints, and both share the exact same datapoints, any distinction between them is arbitrary.

If, on the other hand you mean it is acting like a human I would need to know how you can know that unless you had some evidence it knew it was acting like a human.

Why would I need evidence it knew it was acting like a human? That is entirely unneccessary.

 

[dystopian]

It looks like the problem here is like most philosophical arguments. You have not defined what you mean by consciousness. From your arguments it seems to be that you are calling some system that can react to stimuli based on programmed instructions and data conscious. This would make the roomba vacuum cleaner conscious, but not very intelligent.

Nope, it really doesn't matter how we define consciousness, the arguments remain exactly the same.

 

[dystopian]

So what? We're not talking about recreating the universe on a smaller scale. We're talking about using data crunching to simulate physics and consciousness, not creating physical structures for a physical simulation.

Is the difference between the distance from here to the Moon and its numeric representation too hard for you to grasp? Is the fact that the numbers can be used to create various accurate depictions of the relationship between the Earth and the Moon going to confuse you as to whether the numbers are an actual distance or not?

What do you think we do, stuff the distance into the numbers, which is why they allow us to predict stuff related to the distance?

Seriously, you need to wrap your head around the fact that numbers, which can be used to create realistic looking images (and thus could be used to create accurate simulations), are not what they describe. Nor is the data manipulation we can do on the numbers, to create things in reality, the same thing as what the numbers describe.

All you're doing is stating that a simulation can't have consciousness because it's 0's and 1's, failing to explain how this fact somehow precludes it.

You really don't understand that a mathematical simulation of consciousness, no matter how fine tuned, is not going to equate to consciousness itself? If that's the case, I'm impressed, once again, with your failure to understand simple obvious concepts. It's a gift.

You might actually be terrified to learn that theorethical physicists James Gates claims to have found binary computer code in the super symmetry equations of string-theory.

He found ASM code or something? That would be hilarious.

Keep in mind that some of the math that string theory relies upon is easily manipulable into many values (zeta function regularization is not the only way to arrive at specific values for the one divergent series- in fact one can manipulate them into just about any value one wants). In fact, the human mind (brain) can be easily manipulated into seeing values that aren't there, so with that in mind...

Unfamiliar with Tegmark's mathematical universe hypothesis, are you? The physical universe IS mathemathics, the perception that we exist in a 'real' world that is somehow seperate from math is entirely subjective. As some physicists would state, it may be less misleading to state that the universe is made out of numbers than to state it is made out of matter. Even if our own universe is truly physical and its physicality not a mathematical construction, this in no way implies that mathematics can not give rise to consciousnessness within its own mathematical existence.

What you're claiming about the nature of reality is something that is by no means an accepted paradigm in science. The question of whether or not reality is mathematics or whether mathematics just describes reality but is separate from it, remains an open question.

 

[Speakpigeon]

http://aeon.co/magazine/psychology/is-consciousness-an-engineering-problem/

OK, so let's take this 'consciousness' stuff away from the philosophers, and see if we can't, to borrow a line from Andy Weir's The Martian, science the shit out of this thing.

Graziano in his paper defines the kind of consciousness he is talking about as "experienced".

Graziano
Whether we’re talking about the thoughts and memories swirling around on the inside, or awareness of the stuff entering through the senses, somehow the brain experiences its own data. It has consciousness.

According to this, merely to build some robot that would display all the usual objective signs of possessing consciousness like we believe most humans do would prove nothing at all. How could you prove that something is subjectively experienced without doing the experiencing yourself? All you could do is at some point find that the result is good enough for you, much like porn photos and videos seem good enough to most people as a substitute for hot sex. It's a pragmatic decision. However, this is clearly not the same thing and it would be somewhat ridiculous to pretend otherwise. So why keep doing the same thing when it comes to subjective consciousness?

Scientists are on record for dismissing the notion of subjective consciousness as some sort of illusion on the part of the subject. Thus, they have long limited their investigations to what can only be described as objective consciousness, the kind you would have objective evidence for, such as memory, linguistic communication, etc. typically using brain scans and subjects reporting on their subjective experiences. The discussion in this thread also shows how most hard-core materialists here chose to entirely ignore subjective consciousness by responding to your question as if it was about objective consciousness, thereby missing Graziano's take on the subject, and by implication yours (although, to be fair, you may also have misunderstood what Graziano was talking about but if that was so then your OP would be trivial and uninteresting). There is something profoundly pathetic about this behaviour, as if subjective experience was something taboo even though every normal human being on this planet would probably happily testify having it. Why are you people doing this?
EB

 

[Speakpigeon]

How are we distinguishing between a mechanism that experiences quaila, and a mechanism that produces statements saying that it does.

That wasn't the question I responded to. The question was "if it thinks its conscious, is it conscious?". Which is kind of like asking 'if a piece of fruit is an apple, is it an apple?'

As for your question; what exactly is the difference between an AI and a human brain in that regard? I have no mechanism for determining that the things you say about what you experience aren't just the result of a simple "If x then y" diagram.

If we accept that other humans are conscious on the basis of them saying they are... then unless we can demonstrate it's actually because of a programmed "if x then y" routine, why shouldn't we accept the same thing when said by an AI?

We don't really have a good reason to discount a simulated brain telling us it's conscious; and if we accept that then we can begin to try and create a measurable/scientifically understandable concept of consciousness.

Scientists have long been working on human consciousness on the basis of whatever objective evidence they believe there is for it. Why would a robot tell us more about human consciousness than doing what scientists are already doing?
EB

 

[Speakpigeon]

Rather it's always been accepted as a possibility that if you create a complex enough system capable of learning that it might develop consciousness on its own.

Can you walk us through your thinking about how complexity and learning capabilities could conceivably give rise to a robot experiencing its own data, to use Graziano's wording for the kind of consciousness he is talking about?
EB

 

[Speakpigeon]

It wouldn't be a consciousness. Consciousness has physical properties which are separate from the data that neural nets process. An artificial neuron fires based on 1s and 0s, it experiences no qualia. It's a data processing abstraction, rather than a consciousness.

There is absolutely no evidence whatsoever supporting your interpretation of consciousness. As far as we can tell consciousness is an emergent property of neural activity, and we have no reason to hypothesize that this emergent property can not be recreated through artificial neurons or other means. What you are proposing is mind/body dualism; which is not scientifically supportable.

There's no reason to assume subjective consciousness isn't somehow physical but that's not the discussion here. The discussion is about producing a robot that would have subjective consciousness and, maybe unfortunately, there is no good reason to assume we should be able to do it. Maybe we will, or even if we never do it maybe it could be done in principle, but this is not the point either. Rather, it is that once you've produced some AI robot, how do you go about proving it has subjective consciousness (assuming we would easily solve the problem of giving a robot the ability to do all the practical things human beings routinely do in the course of their lives, from walking the streets without falling over at each step to doing the hard jobs in politics, the arts etc., all that being just a non-idiotic Turing Test). The fact is that at the moment, it seems no one has ever suggested any practical procedure. There's at the moment no conceivable way to do it.
EB

 

[Speakpigeon]

The AI could, in fact, be conscious of the fact that it is not human, and be able to express this consciousness in human interactions, while still convincing them that it is conscious. Of course, someone with an unrealistic expectation that a consciousness must be human would actually be fooling themselves, so being confronted by such an obstinate human would not refute the consciousness of the AI.

Since Michael Graziano's article is about a robot emulating human consciousness as we subjectively experience it ("brain experiences its own data") it seems you are following a derail.

Secondarily, what would there be for us to test if the robot has some kind of consciousness unlike our own? For that matter, ants have something, and who could say it's not some kind of non-human consciousness?
EB