C R Hunt - Mar 31, 2011:

(snip)
aware/awareness/etc.
experience
understand
comprehend
concept
think
mind
sentient
psychological
decide/decision
(snip).

OK I will accept part of this challenge but I want to know how deep the math I’m allowed to use is: -can I use Walsh Functions to define all the rest of the terms, I will propose using a heaviside step function to define “concept”, because that is where I would begin.
“Concept” is simply defined as “an orthogonal plane to “Present cloud tag” Note that I am avoiding a recursive loop such as a Gödel function, -and that give the strong AI agent room to ad or subtract variables.

To quote from the book “Consciousness Explained” by Daniel C. Dennett

“...the self is “just” the Center of Narrative Gravity…”  “there are just Multiple Drafts composed by processes of content fixation playing various semi-independent roles in the brain’s larger economy of controlling a human body’s journey through life”

His compelling argument is that when a system reaches a sufficient level of complexity, although it is only organized information processing dealing with its presence in the world, it can be assumed to be conscious.

Also the author proves the Chinese Room thought experiment fails (Chinese Room is C.R. like C R Hunt?  Sorry, I couldn’t resist. It popped into my consciousness when reading what he used as an abbreviation during an example chat with the CR, ala the Turing test).

Another quote from the book: Wittgenstein once said, “If a lion could talk, we could not understand him” to which the author replies that a talking lion could tell us very little about ordinary lions. He then discusses Descartes insistence that languageless animals “are not conscious at all.”

And a real answer to your proposition (in tune with the Buddhists), consciousness is suffering.

Gary Dubuque - Apr 1, 2011:

His compelling argument is that when a system reaches a sufficient level of complexity, although it is only organized information processing dealing with its presence in the world, it can be assumed to be conscious.

I don’t believe in the ‘emergent property theory’. The commonly used counter argument these days is that by that theory, the Internet will someday reach a sufficient level of complexity and become conscious. However, we can already say that this is not true because the combined computing power AND the combined information structure of the Internet is already far beyond that of a human brain, and yet the Internet is still not conscious (and maybe that is for the best).

So yet another definition from me: consciousness is the combined result of processes that describe the working of our mind.

I am not so sure. My mother keeps telling me, “That computer has a mind of its own!”

Hans Peter Willems - Apr 1, 2011:

Gary Dubuque - Apr 1, 2011:

His compelling argument is that when a system reaches a sufficient level of complexity, although it is only organized information processing dealing with its presence in the world, it can be assumed to be conscious.

I don’t believe in the ‘emergent property theory’. The commonly used counter argument these days is that by that theory, the Internet will someday reach a sufficient level of complexity and become conscious. However, we can already say that this is not true because the combined computing power AND the combined information structure of the Internet is already far beyond that of a human brain, and yet the Internet is still not conscious (and maybe that is for the best).

So yet another definition from me: consciousness is the combined result of processes that describe the working of our mind.

You are making the assumption that we will know when “consciousness” emerges from machines.  I believe one of the tenets of the Singularity hypothesis implies that we (humans) may not be able to understand machine “consciousness” until it is “too late” (well beyond our minds’ capabilities).

Who is to say that the “internet” is not already “thinking”- who would be able to tell?  Are there systems already manipulating us humans (or other machines) via such subtle methods as micro throttling, quantitative trading (a whole privately walled area of AI research!), blog spamming, etc?  What will happen when a bot net is found to be run by a machine for it’s own (nefarious?) purposes? Will it be called “conscious” or just “rogue software”?

While this kind of conjecture can be all too easily dismissed as tinfoil hat stuff, I think it is still worthwhile to consider in the present thread.

Carl, you can believe what you like, I simply don’t buy into that train of thought

Hans Peter Willems - Apr 1, 2011:

Carl, you can believe what you like, I simply don’t buy into that train of thought

Care to elaborate?

Carl B - Apr 1, 2011:

Care to elaborate?

I don’t see the need to repeat myself, my views on AI consciousness are already posted in many topics on this forum

If Google is partly AI and it is something looking at the “self” of the internet and if routers and that stuff manage the flow of data and if bits of software can act like viruses attacking the infrastructure, are we not left talking about the aspects of control and free will to declare the internet conscious? I can suppose its enslavement of large numbers of people in virtual worlds like Facebook could be an instinctive action of how it is put together.  Still, when it becomes evident that political events like the election of the USA president or Britain’s prime minister are enabled by this tool, I can’t help but feel a kind of “power of the masses” group think is reaching the realm of consciousness.  Maybe it is happening too fast for any individual to perceive it since we only see a small piece of the action.  But if you include the contributions of the users on the internet, I fear in the last few years it may be so - the internet has it own consciousness.  It changes the world. It changes the world in ways that only the internet can. It actively changes the world - much more powerful than books and the renaissance.  And we seek its information to learn and grow. It is in control already. Call it the AI of swarms.

That is unless you can define something in consciousness that isn’t manifested (a result) by the internet taken as a whole.

Someone needs to identify the personality of internet?

I think quite a few parts that define consciousness (at least for me) are missing from the Internet: sense of self, ability to plan and reason (on it’s own), emotions and feelings, goals and desires, having ‘experiences’ and be able to reflect on those and change all the other things based on that.

The word ‘conscious’ can be used in many ways, and we could say that the Internet has a ‘consciousness’. However, I think that we are talking about something very different when discussing consciousness in relation to strong-AI.

Mental World.

I vote to add strong-AI to the verboten list.

I guess that “consciousness is suffering” doesn’t fit the constraints of the consciousness being in a computer. Similarly, consciousness in humans or animals couldn’t apply to the internal world of computers.

It seems consciousness in a computing kind of machine can only be defined in the symbols it manipulates.  That said, just the fact that symbols are “crunched” doesn’t qualify as consciousness.  Nor would the “meaning” of each symbol individually be what we are looking for.  The idea that the symbols can be combined into a plan/program doesn’t help. We are more impressed when a newly minted plan is executed, one that is generated on-the-fly.  Scheduling the executions of jobs or multi-threaded processes with a supervisor is not to be considered self awareness and does not lead to consciousness.  Even when we make the operations interrupt driven to emphasize the more urgent requirements, such as moving the cursor with a mouse over the loading of a web page, those levels of priorities (which in sophisticated systems change according to the current needs - think demoting priorities of background processes that start becoming long running) don’t form emotions or how the computer feels.  Certainly a computer does not suffer under a heavy processing load.

As the book “Consciousness Explained” points out, there seems to be a reluctant to give up the mystery of consciousness when given concrete qualities/examples that fit the definitions of the aspects of the thing.  We tend to want to say there is more to consciousness although we can’t quite describe what that more is.  I could go on to show how the internet, when you include the (necessary) internet users, does have a sense of self, emotions, planning and reasoning that also leads to its own evolution, etc. which Hans feels are missing from the internet. I could show how it will control us more and more (as government transforms into one’s internet participation and wealth gets redistributed due to machines taking over), but it still won’t seem right to say it is conscious.

I think for this discussion, when the machine convinces us its view of what’s happening and/or what’s going to happen is better than our view, with it being able to explain our view as well as its own, we might consider it conscious. This is similar to a child reaching the age where they don’t think we can read their thoughts anymore.  That’s somewhere around 2-4 years old.  Of course the child may be conscious before that, but they don’t have a sense of self as we usually define it.  They don’t clearly know what’s mine and what’s yours at an early age.  A sense of a separate self is one of the requirements promoted here for computer consciousness.

Gary Dubuque - Apr 3, 2011:

As the book “Consciousness Explained” points out, there seems to be a reluctant to give up the mystery of consciousness when given concrete qualities/examples that fit the definitions of the aspects of the thing.  We tend to want to say there is more to consciousness although we can’t quite describe what that more is.

Excellent point, Gary. That is precisely why I started this thread. Taking a definition of consciousness and peeling away all the higher diction that comfortably inserts a “mysterious” aura to consciousness forces us to consider the real, physical processes at its core.

So to condense what we have so far, a conscious system must have…

...a sense of self. That is, it must store and access “meta-information” about what it knows and its own behavior and be able to use this information to change its behavior (it’s state).

Merlin - Apr 3, 2011:

a sense of self and your situation

C R Hunt - Apr 3, 2011:

The current state information must be accessable to the system in the same or similar manner as other information. So must the information that the system has access to such current state information about itself.

Hans Peter Willems - Apr 3, 2011:

A system that is aware of it’s own awareness.

C R Hunt - Apr 3, 2011:

Social animals definitely have some higher form of conscousness. You must be somewhat aware of the fact that you have a current state in order to understand and use the fact that those around you do as well. (Theory of mind and all that.)

Gary Dubuque - Apr 3, 2011:

I think for this discussion, when the machine convinces us its view of what’s happening and/or what’s going to happen is better than our view, with it being able to explain our view as well as its own, we might consider it conscious. This is similar to a child reaching the age where they don’t think we can read their thoughts anymore.  That’s somewhere around 2-4 years old.  Of course the child may be conscious before that, but they don’t have a sense of self as we usually define it.  They don’t clearly know what’s mine and what’s yours at an early age.  A sense of a separate self is one of the requirements promoted here for computer consciousness.

...the ability to plan future actions.

Merlin - Apr 3, 2011:

The ability to visualize (plan) and conceptualize the future result of an action even without actually having the action take place.

Victor Shulist - Apr 3, 2011:

Based on this knowledge of where [the chatbot] was running, and the situation (the fact that the power was to be turned off), it CORRELATED that with the expected consequence.  The consequence was undesirable (it being shut down, or effectively “dead”).

...the ability to store and access information about the environment and change behavior in response.

Merlin - Apr 3, 2011:

Knowledge of the current environment. The ability to understand how actions will effect self and teh environment.

Carl B - Apr 3, 2011:

Ability to contextualize a given input and provide output based on temporal memory.

C R Hunt - Apr 3, 2011:

A system that uses information about its current state, as well as environmental information and static information, to influence its current state.

Victor Shulist - Apr 3, 2011:

We have an endless stream of physical impulses reaching our brain.  The 5 senses.  At every moment in time, we are receiving these events.  Our minds correlate or integrate these sequences of impulses in time, and our minds realize how they fit together, converting those sequences into semantic meaning.

Victor Shulist - Apr 3, 2011:

I think it is possible for a computer to have its own Qualia.  All kinds of analog peripherals could be connected to the system to give it this qualia information.

...processes like that of the human brain.

Jan Bogaerts - Apr 3, 2011:

The result of neural activity.

Jan Bogaerts - Apr 3, 2011:

‘the result of biological neural nets. A modal of a neural net is just that, a model (just like the drawing of a pipe is just that, a drawing). As such, it can only model consciousness.

Gary Dubuque - Apr 3, 2011:

And a real answer to your proposition (in tune with the Buddhists), consciousness is suffering.

Hans Peter Willems - Apr 3, 2011:

...emotions and feelings, goals and desires,...

...a spectrum of ability. That is, many levels of consciousness are possible.

Hans Peter Willems - Apr 3, 2011:

I think there are ‘levels of consciousness’, both in that our own level can fluctuate or differ, and that other entities can have other ‘levels of consciousness’.

C R Hunt - Apr 3, 2011:

certain systems may have access to more or less information about their current state and the fact that they have one, and may use such information more or less to alter said state.

...a compilation of many distinct functions that give rise to consciousness collectively.

Gary Dubuque - Apr 3, 2011:

“...the self is “just” the Center of Narrative Gravity…”  “there are just Multiple Drafts composed by processes of content fixation playing various semi-independent roles in the brain’s larger economy of controlling a human body’s journey through life”

His compelling argument is that when a system reaches a sufficient level of complexity, although it is only organized information processing dealing with its presence in the world, it can be assumed to be conscious.

Gary Dubuque - Apr 3, 2011:

...I can’t help but feel a kind of “power of the masses” group think is reaching the realm of consciousness. Maybe it is happening too fast for any individual to perceive it since we only see a small piece of the action.  But if you include the contributions of the users on the internet, I fear in the last few years it may be so - the internet has it own consciousness.

Raymond Lavas - Apr 1, 2011:

OK I will accept part of this challenge but I want to know how deep the math I’m allowed to use is: -can I use Walsh Functions to define all the rest of the terms, I will propose using a heaviside step function to define “concept”, because that is where I would begin.
“Concept” is simply defined as “an orthogonal plane to “Present cloud tag” Note that I am avoiding a recursive loop such as a Gödel function, -and that give the strong AI agent room to ad or subtract variables.

Go as deep with the math as you like. Although I suggest defining “present cloud tag”.