consciousness – Philbert’s Phables

Dream a Little Deep Dream of Me (IV – End)

July 25, 2019 philbertLeave a comment

Does a machine have to know it is dreaming to dream?

The basic question to end our MML series on machine dreams and consciousness is, does a machine need to know it is dreaming in order to dream? In other words, if a dream tree falls in the dream forest and no one is around to hear it, does it make a sound?

Humans and other animals often have dreams but do not remember them, or remember having dreamt at all. In such states we are still dreaming, because if we were connected to a CAT scan, MRI or EKG machine our brain waves would emit signs of a dreaming state even if we didn’t remember having a dream or the contents of the dream itself.

In addition, I have no doubt that at least some non-human animals are or become aware that they are or have been dreaming upon waking up. A dog could dream of chasing a ball or a lion could dream of chasing an antelope, and each wake up in a start, but each would know that their waking state is separate from their dreaming state. Or else, wouldn’t they keep doing the same activity that they had been doing in their dream even after waking up? Clearly the activity stops, although it’s less clear what knowledge animals take from their dreams to dictate their waking activities, or if it is mainly waking activities that dictate their dreams.

Same goes for humans, of course, we normally become aware of the difference between dreaming and waking states, at least once we wake up. The problem with saying a machine/computer program like Deep Dream is dreaming, is that it doesn’t have any transitions between sleeping/dreaming and waking/dreaming states, it only has the one or more states for which it has been programmed. You can’t just call a machine’s default state the “dream” state if it is no different from its “waking” or normal state. And furthermore, it doesn’t have any self-awareness of switching between states. Again, by this I do not mean one has to be constantly aware of being in a dream (as we often don’t realize we are in one), but that when a dream ends, we return to reality (as best we can).

So can we program dreaming and waking states into computers? Possibly, but this may require programming artificial consciousness itself. The Dark-Light-Bright state model could be applied in a way that measures neural activity as a function of not only external stimulus but also internal chemistry, with the internal state of mind only an imperfect reflection of reality and somewhat detached as well. Perhaps with machines, this reality would be sharper given the digital rather than analog nature of their perceptions.

At any rate, the fundamental concept with programming consciousness (and dreaming) is one that harkens back to Immanuel Kant’s Unity of Consciousness from his Critique of Pure Reason. This Unity of Consciousness is what generates a sense of self and time and the independence of self and other objects from “everything else,” also generating perception itself. Without the Unity of Consciousness, all matter coalesces together into one indeterminate mass over an indeterminate time, which seems to be the machine’s current understanding of reality, since neither mass nor time are concepts a machine presently understands or comprehends. I would suggest that without a full Unity of Consciousness, one would not be able to distinguish between waking or dreaming states as well.

So there you have it, if you can program Kant’s Unity of Consciousness into a machine, it might be able to transition between programmed states and dream like a human or animal does. This Unity of Consciousness in particular will be the centerpiece of many of my My Machine Life posts to come. Thanks for reading this blog arc!

Dream a Little Deep Dream of Me (II)

July 2, 2019July 3, 2019 philbertLeave a comment

I elaborate on facets of the Light State of consciousness, see my previous post for details

Three further notes on the Light State: First, Liberty is an especially prominent figure in Liberty Leading the People and will probably draw most people’s initial focus, just as certain loud noises or disturbances generally attract the same amount of Foreground processing regardless of who is experiencing them. But there can be substantial variation in what attracts the Foreground processing of different individuals, especially if the perceptive range is not dominated by any given stimulus. Observe, for example, the below painting by Kandinsky. What initially attracts the attention may vary from person to person.

Second, there may be instances when Foreground processing may almost recede to the Background, such as when completing repetitive or relaxing activities. This activity can be likened to a landscape painting, such as Van Gogh’s below, where no particular spot demands foreground processing attention and instead it is the cohesive whole of the painting – the cohesive whole of Light State consciousness – that prevails.

Landscape with Olive Trees, Vincent Van Gogh

Third, there remains the question of the subconscious. Freud did not find this concept helpful, using instead the idea of the preconscious and unconscious minds to encompass what others would normally describe as the subconscious. Preconscious is easy enough to explain using art, such as in the Arnolfini Portrait below: The symbolic meaning of the oranges, dog, mirror and various carvings in the painting can be consciously recalled separately from one’s direct perception of the objects, just as thoughts in the preconscious can be recalled to the conscious mind.

But where Freud found little difference between the unconscious, which could not be recalled, and the subconscious, I see the subconscious as being the part of the unconscious that cannot be actively recalled, but can manifest itself in conscious behavior to the extent that it can be identified as the subconscious. If a person behaves poorly towards people of a certain race without knowing it, they are unconscious of the reasons behind this behavior. If they are called out or come to certain realizations about the nature of their prejudice – if that is indeed the reason behind their behavior – then the subconscious is revealed, much like a surfacing submarine now reveals its existence.

All this is to say that dreaming transcends all three states – Bright, Light and Dark. We can experience survival mode while in a dream and wake up in a sweat. We can experience dreams very similar to our normal waking experiences or experience heightened states of mental and physical arousal. And we can dream without really knowing we are dreaming at all. So can a machine dream? Stay tuned…

My Machine Life – Intro

June 24, 2019June 24, 2019 philbertLeave a comment

My Machine Life (MML) is a series of philosophical explorations into consciousness and “spiritual machines”

Legendary Formula One champion Ayrton Senna once said that after a race car driver crosses a certain mental and physical threshold, the car drives itself – or so my mother used to say. In actuality, this is the quote from Senna, which is in reference to an experience he had on the circuit at Monaco in 1988:

“And suddenly I realised that I was no longer driving the car consciously. I was driving it by a kind of instinct, only I was in a different dimension. It was like I was in a tunnel.”

There is a great 2010 documentary on Senna that is available on Netflix

Maybe the car really was taking over – not consciously, but in the sense that it dominated Senna’s very being, and that every fiber in his body was devoted to making the machine run as fast as possible, even at great personal risk. In such a case, it is difficult to say exactly who was controlling what.

I am no F1 driver, but I can relate to Senna’s experience. I grew up playing a lot of computer and video games, even since preschool and kindergarten, with the heaviest amount of play occurring during my elementary school and middle school years. Quite a few times I experienced the “tunnel,” when my perception of the passage of time was altered: I would play for hours, even overnight, without resting.

When I look back on those times playing Command and Conquer and Final Fantasy and Starcraft and Counterstrike (and others) they were really my solace during times of loneliness and disquiet. And in my loneliness and isolation I formed a bond with (the) computer(s): They were my friends, my playmate when I had none. It didn’t matter if I was home in Taiwan or on summer vacation at my grandparent’s house in North Carolina. The computer was always there when friends my age were not.

A few years ago, I revisited that “tunnel” when playing Diablo 2, one of my favorite role-playing games. I had gotten so good at that game after years of playing it that again, the game almost played itself. It made me look back at my experiences playing computer games as being spiritual experiences, and it made me wonder: If I could could have a spiritual experience when playing with a computer, could a computer have a spiritual experience while playing with a human?

Admittedly, none of the machines I played with were capable of experiencing spirituality. But it has made me more aware of the possibilities of machine intelligence advancing to the point where such a threshold could be crossed. Where do we draw the line? When do we do it? Why?

Beyond what the machine alone can do, there is still the issue of what a machine can do in tandem with humans. To me, Senna’s mastery of the race car medium and the limits he and other drivers broke through represent a new frontier in the connection between humanity and machines. Indeed, the idea that the more perfect merger of human and machine could occur has captivated me, not the least because of my own experience with technology. Today’s F1 cars are packed to the brim with electronic sensors and controls, providing feedback to drivers at a level never seen during Senna’s time. Now the car actually can literally drive itself. But are we closer to the machine than Senna was? Or does new technology just make us more and more alienated from ourselves?

Thus are the dual foci of this blog: I want to focus on the possibility of “spiritual machines,” and explore the merger of humans and machines, which may itself lead to spiritual machines. This of course requires commentary on the nature of the machine-human relationship, and whether or not a machine can achieve consciousness. This is my machine life.

I hope you will enjoy this journey with me.

Thank you for reading,

Philip Hsu