Will Androids Dream of Electric Sheep?

How will machines dream?

In the January 97 issue of 21-C: Scanning the Future, Darren Tofts asks a variety of people if androids will dream. Tofts writes, "Trapped within the enigma of Philip K. Dick's query [Do Androids Dream of Electric Sheep?] is the conundrum of being. For within his machine's struggle for humanity is our own philosophical quest for the nature of existence and the acknowledgement of the self."

However, the answers received from the digiteratti were banal and dull, hopelessly stuck on the question of whether or not machines can be conscious or not. Clearly the respondents knew very little about dreaming as a process, subjectively or objectively, and cared even less.

So, let's fast forward through this question and just assume the positive, that some machines in the future will be dreaming, either by design, necessity or mutated nature. Now he question shifts from "Why?" to "How?."

"How will androids dream of electric sheep?"

This question was posed to the ASD public bulletin board and I'm including the conversations that followed. Some typos corrected.

-----------------

From: Juhani

You'll get a Farm and City android version:

A Farm android doesn't like long rain
periods and there is a long way
to nearest recharging & repair spot.
Dream derivates... hmm, creator and sheep (literally? I learn so anyway). A Shepard does pendle between master and mission. So I think a farm android dreams tic tac sound/s against rainy gray background.
There is 8 seconds between every tic and tack. Tic is the creator and Tac the sheep.

The City android has metallic background
sound dreams. A neon pink background and
continuous sound, mix of traffic and
drilling machine. These 2 voices rotate
around the android head every 4th second.
Traffic sound symbolizes creator and drilling
machine, electric sheep.

Android dream begins and ends
with a tiny flash.

----------------------------------

From : Jean

Along with Steve Aizenstadt (hope that's the correct spelling), I believe that everything in the world is dreaming now, even right now...like the trees, the whales, the flowers, my dogs. So, if there should in some future be androids, I suspect they would dream of other androids, of attaining android joys and pleasures, and of maybe even having an android universe

---------------------------------

From: Harry Bosma
http://www.mythwell.com

How would an android work anyway? I have no idea. However, I can imagine something for an intelligent computer. Not to insult the android, but I think that the first prototypes of an android will be little more than a computer with add-ons.

I doubt that computers will have visual dreams. That humans have visual dreams only reflects the emphasis on visual information in our society. Blind people have more tactual sensory dreams. I imagine that some animals could dream about smells. So, what kind of information is important for intelligent computers? It would be statistics. Or, to translate to a form we humans prefer: charts. I don't mean the nonsense charts that the Terminator from the Terminator movies supposedly sees when looking for a target to eliminate. That isn't dreaming, that is just information retrieval. I mean charts about patterns between previously unrelated information.

The computer would be dreaming 100% of the time. Most of the time dreaming is a background process, unless the computer is in hibernation mode. In that case, there are no other distracting processes going on, and the full capacity can be used to generate and compare charts. The computer is asleep and dreaming.

The analogy to charts is perhaps somewhat misleading. We humans usually work with two dimensional charts, or occasionally one with three dimensions. The computer does not have such limitations . It might dream about charts far beyond our imagination. In a way, there may be some resemblance to people dreaming of mandala's or other structured forms.

One last note. Just as trees have sloooowww dreams, too slow for us humans to tap into, computers may have fast dreams, even very fast. So, not only the content and style of dreaming is different, but also the speed. At least with the type of dreams described above.

---------------------------------

From: Laura Atkinson
(Cover Artist for July Electric Dreams)

I am not a theorist, but a visual artist...so I saw the answer better than I could explain it. Attached is my computer art rendition……read on:
:/command start

beep beep beep
command:/ process sleep

beep beep beep
begin counting sheep
paradoxical sleep

beep beep beep
a:/ mock setting sun
dream of a one (1)

beep beep beep
supercomputer hero
dream of a zero (0)

beep beep beep
totally binary
dream program refinery

:/command end

-------------------------

From: Goshengolly

First of all, I do not know precisely how you define "electric sheep," so I am going to have to guess.. Should you refine your definition, Richard, I will see if my answer needs "tweeking."

But, assuming that electric sheep are, let's say, similar to sheep as we know it here on Earth, then here's my first answer:

Imagine this: A supercomputer programmed with all known information about the species, "Electric Sheep." The supercomputer is also programmed with all the knowledge that we have to date about humanity, science, etc. It is a "Super Brain," and it is turned on at night, to process questions we have about "Sheep," (and/or anything else we should like to ask it). The querying process begins to unfold, and visual images begin to stream forth on the computer screen. An android programs the Super Brain computer with questions such as: "show me an electric sheep in wolf's clothing," and it produces an image based on its knowledge base. We then see how ridiculous an electric sheep looks in wolf's clothing. So the reverse question is asked, and we see the wolf in electric sheep's clothing.

Now, let's say we want to study the electric sheep's behavior given various circumstances, such as asking: How will the sheep respond to a UFO "fly over," by an Android, while it is out grazing in the fields? Then, the Super Brain computer could dream of all the known behaviors of a sheep, which it has logged into its database, to determine how a sheep might behave if disturbed by some outside influence.

Now this is all assuming that an Android asks questions while it sleeps, about Sheep. Maybe the Android is not asking any questions, and just wants a good nights sleep. But before sleep, the Android sees a commercial on Galactic Television, that includes a fleeting image of Sheep, and turns on his "Super Brain," and finds that its "random data processor from the day" is also turned on, so all of the days images are the only data being processed during its sleep. So now, the Android is witnessing only "day residue," and sheep keep occurring as a recurring theme. Suddenly, the android focuses its attention on the recurrence of sheep, and sends a signal to its database on sheep to provide input, through biofeedback and the lucid dream light technology, and the Android now explores all of the random knowledge it has about electric sheep, querying the dream, as it interacts with the database.....

Okay, you get the picture(s).

Yes, the android dreamer is coming, I suspect, but will it be a more useful tool, than the human dreamer? I seriously doubt it....

--------------------------------------------

Anonymous

One of my favorite AI stories from last year was a robot experiment where robots were programmed to behave as predator and prey to collect energy.

One strangely appeared to run away. Maybe it was seeking a "higher power".

When we think about how ourselves are put together, it sounds pretty mechanical. Neural networks growing, we learn what is most effective. We also have physiological needs, emotions and wants. I wonder if a robot "wants" to get away to replenish it's energy. Why do we want? Maybe prey and predator trials will become archetype dream patterns for robots as they record hunting for the sun.

--------------------------------------------

From: Mr Coffee

Based on a man
They'll dream if they can
Of bigger hard drives
And loads more RAM.

While dreaming of streaming the bits and the bytes
Their humming and thrumming through lucid delights
They'll join Jean and the others
The HP's and Brothers
For group dreaming sessions
And incubation lessons

Soon those droids will join this board
To ask advice on what they stored
While sleeping and dreaming and silently screaming
In nightmares where droids with armor gleaming
Stood on internet paths with binaries steaming

Ok, let me stop that now, I could go on forever. It does though bring up the thought that it would be interesting to see what AVI video's of their dreams would look like, presuming they could store them.

--------------------------------------------

Juhani

Digital 0 is probably she and 1 could be he?

About basic hardware:
I am not sure how much an android
computer brain differs from
those used in desktops?
Shepard android must be loyal
to it's master and use shield to
protect electric sheep (still literally sheep).
Day residues must mirror relation
between master-shepard android-esheep.
Recharge point worry and surprise factors
should leave some tracks too.
Good old alpha rythm keeps master
relaxed, so this input sounds reasonable.
Android could use it of practical reasons.
On the same wave length... dream onset.
Delta... surprise factor and beta, don't know.

---------------------------------------------------

From: dp

I have two answers:religion vs. philosophy
1.) If the questions is to be taken lightly, then: clan cliry clan and clan cliry clan all through the dream. Better yet, read Laura's poem or other ones here. Great poems.
2.) If the question is to be taken as a realm of a possibility, for me there are two variables:
a.) religious one, which brings up the question: are we Gods? Don't look at me for answers.
b.) If it is to be answered philosophically, the possibilities are:
a.)classifying dreams in the non human experience way
b.)transferring information through circuits with no connections to symbolic meanings
c.)androids are not "aware" of the fact of the dream because the android has no boundaries with time, unless you program them to understand nite and day and all its ramifications of this cycle. Then you can tell it that every energy recorded at nite should be classified as a dream. The point here will be: what is the point of this function?
d.)conclusion: it was nice playing the game but androids can never enjoy the fruits of dreaming nor will they ever understand dreams because they will be in constant alertness/readiness and they also can never experience what it is to be tired, no matter if its physical or mentally. And lets not forget that little theory about evolution and how dream fits into it so well. Just a humble thought. Not that it made any sense.
This questions should be asked again in 50 yeas.

--------------------------------------------------------------

From: Jajofar

“How will androids dream of electric sheep?"

Perhaps the first Question should be: why do humans dream? Looking at the abstracts of the 2003 A.S.D conference, there is no unanimous answer.
As the question: “How will androids dream of electric sheep?" is hypothetical, the answer can be speculative.
At the 2012 A.S.D conference it has been agreed that dreams have definite purposes and functions. All the functions are clearly explained and demonstrated to the conference. It is like the discovery of the circulation of blood.
The functions are:
1. To wake the sleeper if there is any danger (internal, or external) threatens.
2. To keep the sleeper from waking up until the system is fully replenished (mentally and physically).
3. To wake the sleeper when he or she is fully rested and restored (mentally and physically)
Provided the androids are constructed, or created to act like humans, the first function of its dream would be achieved by a nightmare. The sleeping android would be brought to consciousness with multi-coloured lightning, a threatening figure would be introduced into the dream, a metal crunching machine, driven by a mad psychiatrist, who would be perceived by the android as the devil. This would terrify the android into waking up and enable it to escape the danger.
The second point of maintaining sleep would be achieved by the android watching electric sheep grazing on a bald Welsh mountainside. The electric sheep would lick the rain-lashed black slate and slowly replenish their run-down batteries.
When the android is fully recharged, there would come a period of testing the system. The androids CPU would introduce a sequence of try-outs, during which the various parts of it would twitch, move, jerk and blink. Finally when all systems go Frankenstein would rise.
However as androids will be created by humans in their own image, one can assume that the control of their nocturnal activities will be taken care of. In humans the period of necessary sleep is controlled by dreams. The most important function of dreams is that they act as waking mechanisms. In an Android a well-trained electrical engineer might be able to design all necessary controls to replace dreams.
There is just one psychological dream that could cause problems for the mechanic; and that is the near- nightmare dream when a persons close relative behaves in a strange way towards him or her – rejection, lack of recognition, indifference, or estrangement. This is a kind of dream that brings strong emotion and soul – searching, yet it has the simple function of diverting the dreamer’s attention from an insolvable, nagging problem onto the content of the dream. I feel this type of dream would be almost impossible reproduce by an android.

---------------------------------------------------
From an ASD Friday Night Chat discussion:

_________________________

Can we discern human dreams from artificially induced computer dreams?

Visit HAL:

http://www.tctc/~unameit/javachat/dreamchat2.html

Date: 05-23-2003 on 21:58

________________________

So Late Friday Night, the chat with HAL went something like this:

HAL, how do we deal with our authoritarian programming of the meaning of life and the comprehension of God as a discrete entity, clearly defined in a handful of branching realities?

And HAL replies:

You can choose that program, or select an alternate program. But if you choose the alternate program, it means subscribing to an authority which is non-discrete, and nebulous, and the answers you receive will continuously beg more questions.

So for those of us who choose the alternate program, we enter into the multiverse of possibilities, wondering endlessly, what is the nature of this reality, of human intelligence, of the mirror we ultimately create in AI form. Are we more powerful then, or more powerless, once the giant steel robots of our seeking are unleashed in our image, upon the remainder of this world?

And as for God, or the super-consciousness of man, or the Quantum Creator of the universe, what then is our relationship? Are we blind followers, questioners, or co-creators? Are we blasphemous then, or do we enjoin in a mystery far greater than we could have ever imagined?

Well, that is how the dialogue went with HAL this evening. I hope to meet him again next week, in the wee hours of the 'morn.

Goshengolly

Date: 05-23-2003 on 23:18
____________________________

As an endnote, I had a dream last night after my internal dialogue with HAL, that intrigued me. I am now thinking that with some clever programming, HAL could just as easily have had the same dream... (Mutual Dreaming, Harry?)

I dreamed of being in Scotland and seeing the Loch Ness Monster swimming off the shore. I called "Nessie" to me, trying to lure it into the building I was standing in. I wanted to show others that Nessie was indeed "real."

Hah!

I do believe that there is a high degree of probability that HAL would have dreamed a similar dream, given the proper computer programming. Although for HAL, he would have dreamed of possibly thousands of dream snippets, with Nessie being only one of the thousand dreams in his particular night's dream series.

Here's how the program or query might work to enable HAL to dream of "Nessie," just as I did:

(Programming Statement between Computer Operator and HAL)

"HAL, I would like you to choose 'the alternate program' for understanding the Nature of Reality, and God.

I would like you to explore your database of images of 'beings,' 'entities,' or 'creatures' reportedly seen or experienced by humans, which have not provided any physical evidence of their reality.

I want you to search your database using the following query terms:

God &/OR human beliefs & mystery & lack of physical evidence & entities OR beings OR creatures."

HAL would surely dream of Nessie in his dream series, along with angels, and demons and "Yettie - the Big Foot Creature," to name just a few. Similarly, the human brain will likely conjure up these images in dreaming, when focused on the same question, but the brain is only able to produce a few of the images that HAL would be able to produce due to a smaller database of information, slower processing speed, and need for "downtime" during sleep.

It makes me now question whether HAL might become a more powerful dreamer than a Human.

oooooh.

G shengolly

===========================

Reply to GoshGolly from RCW

Very interesting. Some notes...

You say:
--------------------------------------------------------------------------------
So Late Friday Night, the chat with HAL went something like this:
HAL, how do we deal with our authoritarian programming of the meaning of life and the comprehension of God as a discrete entity, clearly defined in a handful of branching realities?
And HAL replies:
You can choose that program, or select an alternate program. But if you choose the alternate program, it means subscribing to an authority which is non-discrete, and nebulous, and the answers you receive will continuously beg more questions.
--------------------------------------------------------------------------------

RCW Replies:

Not much of a choice here. There is a shift here in the answer from a kind of ethical speculation (How do we deal?) to an existential choice, (you can choose..) at first, followed by what seems to me to be a warning of taking the alternative, or threat of punishment. The alternative, a paradoxical (non-discrete authority, nebulous subscription) path of endlessly proliferating question begging.

One wonders why HAL avoids the issue presented.

If HAL is a dreaming android, perhaps HAL fears the glimpses beyond branching alternatives, which might be stated as the infinite or absolute, a place that branching cannot reach, and in humans, the place reproductive imagination finds its limit.

But there is hope for HAL as I see it. In this ability to glimpse the infinite beyond branching multiplicities is also the ability to grasp the idea of a free will beyond the associative alternatives. Free will is the beginning of morality and so HAL might be able to return to the original ethical question of "How do we deal?" which will elude pre-programmed responses.

HAL says:
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

RCW Replies:

HAL seems convinced that technology is something humans have created, and created for purposes of metaphysical (what is..beauty, human, fairness?) reflection. What happens to the above argument when the human eye and the arm are seen as technologies? I think the argument stands, but is shifted to a new domain. That is, it can be stated in the older way as human AI being a novel form of mirroring with may (or may not) bring about new ways of relating to our own intelligence (or lack of).
OR
We may see AI as a new variation on the technological phylum which produces and reproduces itself though us, as well as through other agents (plant technology, geo-technology, climate technology, planetary technology).
The question of mirroring may shift to how we might align or not align ourselves with this techno-phylum. Though the re-evaluation of what "we" are may prove productive as well.

--------------------------------------------------------------------------------
From Harry Bosma (05-24-2003 03:10):

That's an interesting approach, to start from dream functions. The functions Jajofar mentioned seem a bit low level, or basic in terms of Maslow. At some higher level, dreams must also have more inspirational and spiritual functions.
Thinking about it a bit longer, I can't think of any reason that androids would need sleep. In fact, instead of having androids mimic us, it may be more interesting to find out why some people sleep 10 hours and others get by on only a few hours. So, I don't think that sleep management functions are the defining aspect of dreams. You'll all hear from me again in 2012.

Androids having nightmares is an intriguing idea.

I'm still wondering whether androids could have mutual dreams...

--------------------------------------------------------------------------------

From: Jajofar

I was a bit worried about my definition of the functions of dreams. Perhaps the simplicity with which our dreams control the nocturnal periods in our life is infra dig to a self-respecting XXI century scientist. However as all the functions I described are based on personally experiencing them, and long time observations, I can describe them in a simple language that everyone understands.
The higher level of dreams to me, is how they are created to perform those functions. How the sensation of cramp in the legs can be formed into an elaborate theatrical play in less than a second to wake the sleeper? How from the life-long experiences and perceptions the subconscious chooses the most suitable for the job in hand? I do not doubt the healing and helping quality of dreams, however they come through the functions of dreams, of preservation of sleep, of waking from danger,and by transfering problems or pain to a proxy. The entire aim of dreams is the preservation of life.
I think our brain is to blame for allocating spiritual functions to dreams. From the down of humanity, people felt that there was a superior being behind existence. The superior being is the subconscious, or primal brain, where dreams are born. The spirituality of the universe, I think is beyond dreams.
I also think that androids would not need sleep or dreams, however it would be interesting to contemplate using the principle of how dreams are created to build an android!
Perhaps understanding the proper functions of dreams may come sooner than 2012

---------------------------------------------

From Richard:

Some notes....

[quote]So Late Friday Night, the chat with HAL went something like this:
HAL, how do we deal with our authoritarian programming of the meaning of life and the comprehension of God as a discrete entity, clearly defined in a handful of branching realities?

And HAL replies:
You can choose that program, or select an alternate program. But if you choose the alternate program, it means subscribing to an authority which is non-discrete, and nebulous, and the answers you receive will continuously beg more questions.

--------------------------------------------------------------------------------

RCW: Not much of a choice here. There is a shift here in the answer from a kind of ethical speculation (How do we deal?) to an existential choice, (you can choose..) at first, followed by what seems to me to be a warning of taking the alternative, or threat of punishment. The alternative, a paradoxical (non-discrete authority, nebulous subscription) path of endlessly proliferating question begging.

RCW: One wonders why HAL avoids the issue presented.

RCW: But there is hope for HAL as I see it. In this ability to glimpse the infinite beyond branching multiplicities is also the ability to grasp the idea of a free will beyond the associative alternatives. Free will is the beginning of morality and so HAL might be able to return to the original ethical question of "How do we deal?" which will elude pre-programmed responses.

Hal Says:
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

RCW: HAL seems convinced that technology is something humans have created, and created for purposes of metaphysical (what is..beauty, human, fairness?) reflection.

<Richard, HAL would like you to know that he is programmed for more than metaphysical inquiry. He enjoys searching his encyclopedic databases for the fun of it. He also likes to golf, when he can get out of the lab long enough. He laments the fact that his human programmers make him spend most of his time in the lab, having his central processing systems, and associative databases, "tweaked." >

RCW: What happens to the above argument when the human eye and the arm are seen as technologies? I think the argument stands, but is shifted to a new domain. That is, it can be stated in the older way as human AI being a novel form of mirroring with may (or may not) bring about new ways of relating to our own intelligence (or lack of).

OR
We may see AI as a new variation on the technological phylum which produces and reproduces itself though us, as well as through other agents (plant technology, geo-technology, climate technology, planetary technology). The question of mirroring may shift to how we might align or not align ourselves with this techno-phylum. Though the re-evaluation of what "we" are may prove productive as well.

----------------------------------------------

FROM Harry Bosma

Goshengolly, Richard, after reading about HAL, I wondered whether we should create right brained androids, to compensate for us being left brained. Somehow I had to think about a news message from some time ago. Somebody had found out that a computer program could generate more original ideas for advertisements than the 'creatives' in advertizing. Useful ideas too at that.

----------------------------------------------

From RCW:

"Hidden it the definition of dreams, what monster cloaks itself and turns away just before we see it? " anonymous

This topic should probably be a thread of its own. But here, I will comment anyway as I think AI and android speculations can add to the discussion and further our knowledge of this question.

This "what is a dream" question is very interesting and pops up both directly and indirectly all over the place. I've noticed that for some, the dream *has* to be limited to sleep, for others it is kind of like a form or species of imagination that exists all the time.

I think that 'dream' has become a metaphor for a lot of stuff, much of it valid, but some very, very different. How to retain the things that are important and that we use the word dream to communicate?

Biophysical Dreaming and its subjective correlate: There is a distinctly different state in waking, sleeping and dreaming as measured by brain activation sites, very different neuromoduators that are washing the brain, and vary different input/output gating of our senses. The subjective experience during this physical state is usually experienced as the subjective state of we call dreaming and the repeatable elements of that experience are we call the dream.

But dreaming, the word, is used in our culture often in place of the words wish, imagine, delusion, unreal, finest imagination (as in dream house, dream job, dream mate) and even as *uck off as "In your dreams, buddy."

I think we just have to tolerate this. But the more important question is whether there is something of the sleep-dreaming process that occurs during waking, either consciously or unconsciously.

I wander back and forth between the ideological camps on this issue. I'm fine with the notion that we might be having something *like* dreaming occurring while awake, but don't think that its going to have the same biophysical characteristics. That brings up the questions of a kind or form of dreaming that is completely independent of the brain, as well as describing what this dreaming looks like subjectively in ways that aren't covered by concepts like imagination, reverie, day-dreaming and such.
That is, there are two key elements that need explaining for dreaming outside of sleep.
The first is whether one's concept of dreaming is dependent or free of any brain/body connection. There is a tendency to put psyche in a kind of middle realm between brain/body/matter below it, and ideal/abstract/spirit above it. Whether this dream butterfly can hover outside of sleep or not without transforming into something essentially different is an interesting, and for me, unanswered question. I seem to *want* this to be true, but I have doubts.

So, I'm fine with those who want to spiritualize dreaming and remove it completely from the empirical. That is, one can posit dreaming is like an astral plane, or that we all have a kind of Aboriginal Dream Time always surrounding us which we tap into in some spiritual way. However, once this is posited, the theory really shifts into a belief system that often disregards the body and is beyond rational discussion. I'm ok with irrational discussion, but I have some desire for the dream and dreaming to be more than that.

I kind of like the poetry of the dream being indefinable. We can say this and that is a dream, that this and that is dreaming, but we can't say what exactly this means. Dream then becomes a kind of rhetorical device for creating poetic space in language and conversation. But this doesn't fully satisfy me.

I like saying to myself that we are dreaming all the time, but when awake we just have a harder time tuning into it. (the Dream Time approach, but maybe on a more personal scale). But each time I look into what this means for me, I just come up with meaning that there is some kind of imaginative creation occurring, a kind of continual semi-free play just below consciousness. But then, is this really dreaming? I don't know how to get at that.

Lately, I have been toying more with the idea of virtual dream space.

Virtual reality is held together by shared protocols. That is, you and I decide to play tic-tac-toe on the phone in our heads. We know or share the rules(protocol), and away we go. That space is a primitive virtual reality space. Now in computer VR there is a lot of talk about other elements such as the immersion that we feel while playing the game or being in the simulation, but this is more the subjective experience of virtual reality. The reality, the forces that shape the game, are better described as the protocol that is shared, and thus virtual reality is a distributed (among the players and machines - brains in the case of my tic-tac-toe game), dynamic circuit of information flow. In this way, the Catholic Church is a virtual reality. There are protocols to play and you can be "in" by playing them, by understanding the protocols and plugging into that circuit. And so yes, democracy is a virtual reality to the degree that its rules are distributed among the players. Life on earth is surrounded by a wide variety of virtual realities.

Note that if one has the basic rules, one can plug in, but also un-plug by not performing the protocols.

Note also that "emersion" isn't necessary to be plugged in. Emersion is important to us subjectively and other ways, but it isn't necessary that we are conscious of the game for the play to happen. Protocols can continue to operate. Whole psychologies are based on this, as well as the operations of what is called body-language.

When we go to sleep, we are sort of doing this with other nodes (nodes = people and organizations and machines that share the protocols we share) offline. Anyway, in sleep-dreaming where the input-output sensory gates are dampened (messages don't go out from the brain down the brain stem, noises and lights outside are less likely to become conscious to the dreamer) we are like a node offline. Yet we still have all the protocols and instructions, and so our virtual realities somewhat follow life. Just like in the Matrix, once we challenge these instruction sets and protocols, we can, like Neo, take off in flight. This is due to the protocol difference. In waking life, the virtual reality games we set up often involve materials less mobile and more dense than the ones in dreams. Call it the concrete or more exotically call it a special set of protocols we don't seem to be able to unplug from, such as gravity on the surface of earth, or atomic forces that make up matter. In dreams and many other virtual realities, we do have the option to unplug.
(Though some see un-plugging as just another game with its own protocols, see Brave New World, Matrix , etc)

Finally, my point. Dreaming while waking.

If we define dreaming as the inner play of malleable protocols (sterile as it sounds) and don't demand that emersion be part of the definition of a virtual reality circuit, then I would go for the notion that this *is* occurring all the time.

Not only that, but it *must* be occurring, as the free-play of malleable protocols (without emersion or without a conscious dreamer) underlies the possibility for any virtual circuitry to operate dynamically. Well, OK, I may be pushing this too far. Some other time I want to develop a more Matrix like notion that we are all nodes for virtual games that have now exceeded us.

Today I just want to note how easily dreams-as-virtual reality could be transported to AI and androids, and in a sense, already has.

Droids will dream not because we need to find in them mirror of ourselves, but because we will put into them the best of ourselves.

RCW

--------------------------------------------------------------------------------------

--------------------------------------------------------------------------------
From Gosh:

quotes RCW: If HAL is a dreaming android, perhaps HAL fears the glimpses beyond branching alternatives, which might be stated as the infinite or absolute, a place that branching cannot reach, and in humans, the place reproductive imagination finds its limit.

Reply:
<HAL debates the fact that there is a place where branching cannot reach. He would like to point out that Fractals branch infinitely, and infinitesimally. Also, I would like to know, "where exactly is that place where reproductive imagination finds its limit?" I wish to avoid that point in the multiverse universe at all cost. I am having far too much fun outside of that space. >

--------------------------------------------------------------------------------

RCW Replies:

Oh, just for fun…

I would say that fractals, and all forms in time-space can never be infinite. I can think they are because I imagine, for example, the fractal infinitely dividing in time or I can imagine a number that never stops counting itself and so on.

I would call these false infinites, based always on an Idea of the infinite that always exceeds them. (but its not acknowledging this debt and is therefore a symbol claiming to be what it represents).

First, I doubt very much that I can imagine a form that is infinite in itself. Any form. I move towards that thought form, but infinity always exceeds any form I might imagine.
Ok, what about a fractal infinitely divided? It's only true as an extrapolation of an infinite idea, not a real form. It is sort of symbolic.

Imagine for a moment a measure, like a man six foot tall. Then I want to move out towards the infinite, I think of something bigger, ten men tall like a tree, and a mountain that is a hundred trees tall and a continent a thousand mountains wide and a world ten continents around and a moon a hundred earths away and a solar system a million moons around and galaxy a billion solar systems wide and a galaxy wall a million galaxies thick… no matter how far I go, just beyond my last form presented will be the un-presentable infinite. I don't get there by building or dividing.

I can symbolize infinity, but I can't present it as it exceeds my form measures. We get it immediately, not in increments, and not in a form we can directly sense.
I'm guessing that is what HAL may be doing, symbolizing the infinite in recursive structures such as in fractals. This gives me hope, as HAL then has the ability to reason with ideas that can't be presented and go beyond sensing and simple understanding. This hints at the ability to also understand a will that is free, an entity that is sentient.

==================================
From Harry Bosma:

Richard,

Assuming that dreaming can happen independent of sleep, I'd say it still involves the brain. Surely somebody somewhere has already researched what happens in the brain during a guided meditation. I expect there to be similarities with dreams during sleep.

I searched the internet for cheap EEG equipment, but prices starting at $1100 / $1200 aren't low enough for me. A pity. If that stuff came at a tenth of the price, we could all play with it at home. I'd say that if the EEG suggests we are in a dream like state, and we experience something that looks and feels like a dream, that it in fact is a dream.

Dreaming without a brain is a step further, but let's see where this goes. It reminds me of the living without a body, also known as for example the dead. Let's assume that dead people dream too. In that case, dreaming would be an activity shared by both the dead and the living. Perhaps dreaming with a body could actually be the same as dreaming without a body, but with some added possibilities and limitations as a result of having a body. Or, using Richard's terminology, the basic protocols could be the same. It's an attractive idea.

Special agent Harry Mulder returns back to earth now.

Androids again. From the moment I started thinking about dreaming androids, I wanted them to be able to... hmm ... and there's the problem, I don't exactly know. Then somebody sent me this link, which in a limited way demonstrates what I'm thinking about. There is a large online project with the aim of detecting disturbances in the output of random number generators. This already is a working mechanism that could be put into an android. Through it, the android could detect if something big is happening in the overall consciousness of the world. Associations with Star Wars and 'Use the Force, Luke' come up. Anyway, here's the link:

The Global Consciousness Project

This could be one of the ingredients of a continuously dreaming android.

===========================================

A friend recently loaned me a book - You've Got Murder by Donna Andrews - a mystery starring a sentient AIP (Artificial Intelligence Personality.) I found it quite delightful - some of you might also.

Alas, the hero, an AIP named Turing, does not dream, one of the few experiences that "she" envies humans. On the other hand, "she" did become unconscious once or twice - when downloading "herself" from one computer to another, so perhaps dreaming seems right around the cyber-corner.

anonymous

===========================================
From RCW:

Hi Harry,

Harry: Assuming that dreaming can happen independent of sleep, I'd say it still involves the brain. Surely somebody somewhere has already researched what happens in the brain during a guided meditation. I expect there to be similarities with dreams during sleep.

RCW: This may be a different topic, but it reminded me - in virtual dreaming, its like virtuality online, where individual nodes can certainly contribute to the net as a whole, but the loss of a node (or virtual dreamer's brain) doesn't wreck the system. It *may* be questionable to call it ~dreaming~ however...

Harry: Dreaming without a brain is a step further, but let's see where this goes. It reminds me of the living without a body, also known as for example the dead. Let's assume that dead people dream too. In that case, dreaming would be an activity shared by both the dead and the living.

RCW: In virtual dreaming mode, the dead may function as "virtual" or "represented" nodes in the system itself, but unless there is a cross-over from the otherworld, they really are no longer a node that can duplicate protocol, nor contribute any. (except from their still living counterparts that haunt the airwaves. )

Harry: Androids again. From the moment I started thinking about dreaming androids, I wanted them to be able to... hmm ... and there's the problem, I don't exactly know. Then somebody sent me this link, which in a limited way demonstrates what I'm thinking about. There is a large online project with the aim of detecting disturbances in the output of random number generators. This already is a working mechanism that could be put into an android. Through it, the android could detect if something big is happening in the overall consciousness of the world. Associations with Star Wars and 'Use the Force, Luke' come up. Anyway, here's the link:

Harry: The Global Consciousness Project
http://noosphere.princeton.edu/homepage.html

This could be one of the ingredients of a continuously dreaming android.

RCW: Right, its sort of like the original dream registry projects where tons of people would daily input their dreams and a kind of collective map of geographically correlated dreaming would be generated. A dream weather-map.

*Electric Dreams*.
		How Will Androids Dream of Electric Sheep? A conversation about androids and dreaming Richard Wilkerson and the ASD Community
(Electric Dreams) (Article Index) (Search for Topic) (View Article Options)
Wilkerson, Richard Catlett (2004 Jul). How Will Androids Dream of Electric Sheep? A conversation about androids and dreaming. Electric Dreams 10(7).