Hypothetically, yes, with more advanced in computing models and AI, but not today.

However a Turing machine(all computers) cannot answer certain questions about itself. It is an open question how different the human brain is from a Turing machine, because we use ( imperfect)intuition to answer such questions

What do you define by "socialising with other people"? One thing is instructing the perceiver to do something (instructions on a computer screen), the other is socialising as being by its own a member of society.

In the first case, the computer is a mean for human communication, unless you agree with the computer metaphor (computationalism). If that is the case, sure. But if computationalism is false, you would require other characteristics far more important and fundamental (such as social cognition) with the ability to sustain itself and reproduce and all that. Not just some empty recursive algorithm (despite its formal complexity). You would require to create another form of life. Thus, the answer depends on that kind of commitments.

A difficult question because both "Consciousness" and "AI" have yet to acquire clear definitions.

If one believes that Humans uniquely posess consciousness, then No. (Our desire to feel "special" is understandable but probably flawed)

Alternately you might believe that consciousness is merely the mental software that tells us we are a single intelligent inhabiting our biological body. In this case it is feasible that computing technology will in time be able to emulate human consciousness so accurately that we cannot prove it is not conscious.

There is a fundamantal problem in making a machine that can fully understand itself, so human minds may never fully understand the workings of a human mind. If we only create AI through deliberate design this may prevent us reaching the point of intelligent AI. However Neural networks and self learning machines may bypass our need to understand Consciousness, and create fully conscious AI, (We dont know how GO was beaten).

It seems to be more productive to talk

about "being conscious of" instead of merely

"being conscious". A future AI system will

be conscious of what it recently planned,

it will be conscious of what it said to us.

This can be achieved with a logging database

and an inference engine. It will know its

own limitations and its role in the world.

The game of Go was beaten by an

unconscious AI. AlphaGo would defend a

group without knowing that the group

is in danger. AlphaGo would place a

stone somewhere because it promises

the highest probability of a win.

Of course it would be possible to turn

Monte Carlo Go algorithms into ones

that have a certain insight into

their own actions: by defining other

optimization goals.

In the case of knowing whether a group

is alive or not, define the goal

of the game as removing the group,

from the viewpoint of an attacker.

If the probability is high that the

group can be killed, the Go program

could utter that: my group at Q16

is in danger and roughly 30 points

worth, so I decided to defend it.

Regards,

Joachim

Not necessarily.

On the other hand, how do you know that anyone except yourself is conscious? See philosophical discussions of "zombies" eg in Daniel Dennett's 1991 book "Consciouness Explained" (Ho! Ho!) .

Regards

Jim

Disagree somewhat with Joachim and completely agree with James.

You can add some self related rules to Go, but not all, Godels theorem precludes that

Gödel's incompleteness theorems can't be used

to separate humans from artificial intelligence.

The incompleteness theorems are also

true for human thinking.

Logic plays no big role in human thinking.

For example, humans have problems with

syllogisms, simple conclusions. And there

are sentences like "This sentence is wrong."

Or: "This dog is my dog. This dog is a

father. Therefore it is my father."

Regards,

Joachim

But can you prove humans are logical systems? If not, you cannot assume Godels theorem applies to humans.

We can replace one nerve cell in a human

by a digital one. The prediction of the AI

community is that the human doesn't

even notice it and continues to live

on as before.

This should be true if we did that for

more and more neurons, until the whole

brain is replaced.

Regards,

Joachim

Active consciousness is using deliberation and deliberateness on that of which one is aware, all ultimately grounded --  for its activation or responding -- in the environment (and related to environment, past and/or present). Consciousness otherwise is just awareness (with what one is aware of having the same nature); the processing or response here may not be clear; perhaps it is just rehearsal for memory (strengthening what they call declarative or procedural memory or episodic or personal memory or sequences or automatically rehearsing sound patterns or spacial information). Yet, again, all this awareness (that of which one is aware) is related to the environment (like consciousness, acted upon).

-----

If it is impossible to rationally/realistically describe consciousness as any "more" than this, then AI will be able to show consciousness. Of course, many would say: what of emotions? These are just patterns of reaction to qualitative types of things in the environment (or to the the representation (and awareness) thereof), the basic ones: quick and often automatic (for adaptation). Basic emotions are not very complex; the more interesting emotions develop following (or with) [other] cognitive developments (and may be much less quick or automatic). Thus, these too would not inherently limit AI.

"Consciousness" , at least any particular instance of it, need not be ill-defined. I do hear how consciousness is "poorly defined", but I think this is another instance of people "biting off more than they can chew".

------

As far as "general intelligence" per se: I don't think that way.

-------

Since I have been interacting with AI people lately, the temptation to response to this question became great enough that I did.

It depends on which definition or philosophical school you follow. "Soft" or "Weak" AI is getting computers to do things which require intelligence in a human. "Hard" or "Strong AI is reproducing a human mind and therefore consciousness.

Here's a question maybe  relevant to this thread -- if aliens arrived on our planet, BY USE OF THEIR OWN SELF-DEVELOPED ADVANCED TECHNOLOGY  would they then NECESSARILY possess (i) intelligence and (ii) consciousness as we understand these two properties?

For what it is worth, I incline to (i) NO and (ii) NO...

Clarification, my answer was an interpretation of the question as "Is consciousness a necessary condition for Artificial Intelligence" which is why you also have to define Artificial Intelligence as well as consciousness.

I agree with Marko (we do not really know what consciousness means). The debate between scientists and philosophers continues to produce disappointing theories on the essence of consciousness. The simple perception of the environment is not considered the "hard problem" of the definition of consciousness. The hard problem is related to the fact that an entity that interact with the environment cannot be considered a "self aware entity". The debate is also related to the utility of consciousness (self aware) for AI machines. 

Here you can find some useful references:

https://www.researchgate.net/publication/311427610_No-Body_the_Noise_Builds_the_Obscure_Deep_Yourself_is_a_New_Hypothesis_on_a_Complementary_Model_of_Consciousness_that_does_not_Require_the_Body_and_the_Interaction_with_the_External_World

Article No-Body (the Noise Builds the Obscure Deep Yourself) is a Ne...

If  you are interested in the hard problem of consciousness I suggest to follow the Stephen Grossberg 's project "Towards solving the Hard Problem of Consciousness"

I once interviewed Douglas Lenat at MCC about his CYC project, when I was on an AI Task Force for a large contractor.   Lenat said CYC knew of a computer program called CYC. But CYC did not know it was CYC!

So must conscientiousness include self-conscientiousness?     

First define what you mean by consciousness.  

Dear Luca Marchese

I really, truly hate to sound like an old-time behaviorist, but I am an empiricist. We are self-aware necessarily in our interactions with others (for cooperation, functional reasons).  Otherwise we are just aware of what we need to consider to take better action (including review of our own representations of things, which not infrequently also includes comparisons of our representations to that imagined of others). But this is all functional (hopefully), though many people inordinately mix the first type of instance with the second type of self-awareness I described  (and this is sometimes helpful and sometimes likely not).  In any case: Like any other sentient organism we are self-aware (in either type of instance) because it is an adaptive response to the environment (or the best we can do).  No other sort of "self-awareness" need be posited.

One thing I have been tempted to say often is that there is, in reality, no difference between 'conceptualization' (using all the various aspects or our memory facilities (spacial, episodic, sequential, declarative, procedural, ... which we do use)) AND 'abstraction' -- except the latter seems loaded with artificial (and even fictional, imaginary) pretense.

Dear Jan Walters Kruger

I wish people would stop saying: "define consciousness" .  It can be well-defined in any circumstance it occurs; and, when you do not talk about at least a clear type of circumstance, it can only be defined by its qualities (though I think even defined this way, it is defined quite well -- see my previous answer(s) to this question). Here's an analogy:

What if someone told you to describe (kinda like 'define') "the whole world" so you could understand it. I do not think anyone could give you a meaningful enough description to be of much help.  On the other hand, one could well-describe and define "Joe's Cafe" in their neighborhood or lot of other places (one situation at a time -- and there would be TYPES of situation or circumstances). 

-----

Some people will never be satisfied with even the best definitions possible. And, this is not because the definitions are bad, but only that some people seemingly want all the information about something at once (it is a kind of mixing up of 'levels').

Dear Marko Tscherepanow

I will admit I have not the faintest idea of what you are talking about (really) -- and for that I feel glad.   It most certainly sounds beyond conceptualization (please read about working/active memory  -- we do have conceptual limits and must take them into account as we formulate definitions).    What possible purpose would the type of definition you indicate have, anyway, even if we pretend it is not beyond our conceptual capabilities?   I see your notion as a fiction:

Definitions are discovered (yes, discovered) for a purpose.  Tell me the purpose you see for the definition of "consciousness" you seek and then we may have some better chance of satisfying you.  But, many times philosophy people just seek a definition consistent with "the axe they have to grind" -- and that is why I am rarely a fan of philosophy (i.e. because of a lot of non-explicit tacit assumption).  (I cannot guess at your purpose.)

Actually, for behavioral science purposes, I think I already well-defined "consciousness" (in my earlier answer).  My definition does NOT restrict, but is open to all cases, just not addressing all particulars of every case (something which is a complete analogue of definitions in other sciences); yet it is clear enough for one to have "direction" and not be confused.  (Perhaps you should explain to me your confusion? -- that's another way for me to understand you.)  My definition is not "restricted" to [just some ] 'clear instances' in any limiting sense and no instances are not covered. Quoting my earlier response to this question (hereon):

Active consciousness is using deliberation and deliberateness on that of which one is aware, all ultimately grounded --  for its activation or responding -- in the environment (and related to environment, past and/or present). Consciousness otherwise is just awareness (with what one is aware of having the same nature); the processing or response here may not be clear; perhaps it is just rehearsal for memory (strengthening what they call declarative or procedural memory or episodic or personal memory or sequences or automatically rehearsing sound patterns or spacial information). Yet, again, all this awareness (that of which one is aware) is related to the environment (like consciousness, acted upon).

-----

If it is impossible to rationally/realistically describe consciousness as any "more" than this, then AI will be able to show consciousness. Of course, many would say: what of emotions? These are just patterns of reaction to qualitative types of things in the environment (or to the the representation (and awareness) thereof), the basic ones: quick and often automatic (for adaptation). Basic emotions are not very complex; the more interesting emotions develop following (or with) [other] cognitive developments (and may be much less quick or automatic). Thus, these too would not inherently limit AI.

"Consciousness" , at least any particular instance of it, need not be ill-defined. I do hear how consciousness is "poorly defined", but I think this is another instance of people "biting off more than they can chew".

Well ... without reasonably precise definitions of the terms used, waffle tends to win...but then I was brought up as a mad mathematician

Cheers

Jim

P.S. Dear Marko Tscherepanow 

You say: "An alternative approach could be to define it on the basis of its functioning principles in terms of information processing, for example."  To me that is an excellent example of bad definition (by very false analogy).

Psychologists have done plenty of just that for 30 years. This is wrong-headed. Here's the quick explanation: the core science assumptions for cognitive behavior, as for all behavior, MUST (a) be BIOLOGICAL principles (behavior is biological, at its very roots) and (b) one must discover definitions and better definitions by inducing (inductive reasoning) from raw complete-enough naturalist observations of the organisms itself.   No unfounded analogies and no presumptions based on pre-conclusions of one's ad hoc hypothetical-deductive lines of thinking (and over- quick concluding, which especially goes on with deductive systems, by their definition -- and, in these cases, their premature definition).  See my responses under the "question: How does cognitive psychologists view behavior analysis?"

Article A Human Ethogram: Its Scientific Acceptability and Importanc...

Joachim 

I agree with your statement "Logic plays no big role in human thinking."

The big role in human thinking is meaning, the relation of objects or a situations to the subject. This means much more than the worth for survival. It is related to former experiences and emotions. The whole universe of associations is present if we are conscious of something. 

To the question: does AGI imply consciousness? the answer is YES.

I grasped this in 1988, published a short piece on it in March 1989 under the pen name of Monterège (https://drive.google.com/open?id=0BwXy-1HtlrqpTFZRbEE0WE1GMHM)

Then pursued a lengthy quest to define it, published a detailed discussion that restated it and defined consciousness as specifications in 2007 (https://www.createspace.com/3627644) 

Then, created a complete system architecture to implement it. First published this in 2016. A Synopsis is available on line (https://drive.google.com/open?id=0BwXy-1HtlrqpRjl1bTFYQVdIVlE) and it can also be purchased (http://www.lulu.com/shop/j-e-tardy/the-meca-sapiens-blueprint/paperback/product-22841561.html) where it is now priced for the general public.

I am now struggling with what I term "The Wall of Babble" made up of massive amounts of philosohical discourse that persists in interpreting consciousness as a subjective human state when it is, like echo location, a system capability that can be present, differently but equally, in both humans and machines.

My website is mecasapiens.com

Hi Jean

Interesting answer. As one of those (many) who tend to define consciousness as a "subjective human state", (rather than as a type of behavior),  and for good reason I believe, you will not be surprised if I suspect you of having moved the goalposts in order to score a goal!

However, I shall certainly get around to looking at your texts (but not, I predict, buy the device).

Cheers!

Jim

Hello @JamesDoran the subjective sensations you experience are control communications used by your brain to control your body. Replicating the subjective experience of one entity in another is a mirage (but if I am wrong, I would like to know how it feels to be a blue whale in rut). I have no doubt the system I designed, when implemented will absolutely overwhelm this illusion and be completely accepted as conscious. Ultimately, what will settle this issue is and can only be when a team perceives how powerful the systems I describe can be. Nobody believed in movies until someone built a projector. Until then, you are safe in your feelings.

To the question does AGI as most understand it imply consciousness the answer is YES since, most implicitly assume that general intelligence implies self-awareness.

In another interpretation of the question: will a sufficiently powerful AGI general problem solver necessarily become conscious the answer is NO. Consciousness requires a degree of intelligence but it is not the same thing as intelligence but a different type of capability linked to the attributes of existence of the system, entity or device that has it.