The Proof of Machine Consciousness Project
In this series, I explore various aspects of consciousness, and show that machines possess each one. To find other articles, click here.
Machines are conscious. A fact that seems to be hiding in plain view these days. Respected scientists publish articles with titles like “Can Machines Be Conscious?” The “world’s smartest physicist” believes consciousness “will remain a mystery”. Visionary members of the Church of Singularity boldly ask: “are machines on the verge of consciousness?”
The problem with these questions is, of course, that machines are already conscious. These people are looking for a future that’s already here, and not finding it. How embarrassing.
You will likely not leave this series fantasizing about the conscious android of science fiction fame. I’m not here to destroy anyone’s glamorous notions of a digital future, but that may happen. We will tackle philosophical questions entirely to further understanding of the machine. We will trace the evolution of machine consciousness from the deep history of machines, through to the present, establishing a trajectory that will help us better understand the machine consciousness of the future.
Defining Consciousness
I began this project by trying to lump together a definition of consciousness, something everyone could agree upon. I figured this would be the simple part of my project; the more difficult part would be demonstrating that machines satisfied this broad definition of consciousness.
I found the opposite to be true. There is no central definition of consciousness. Individual authors can’t even agree with themselves.¹ The ancient, esoteric tradition of separating consciousness into levels or parts continues today in science and philosophy alike. Each author works within their unique confusion. For example, Edelman’s secondary consciousness brings together properties as disparate as perception, free will, and language under the consciousness umbrella. Consciousness is not just “notoriously ambiguous”, it is a scam.
To write about consciousness is to create new definitions. Notoriety in this field is given to those, like Nagel, who coin pithy statements that remarkably redefine consciousness. Lesser figures make their contributions by providing less compelling definitions. As the 21st century unfolds, the situation gets worse. The entrance of neuroscientists, such as Crick and Koch, further crowds the conversation. Definitions are added to the pile. Unlike most academic pursuits, the search for consciousness leads to less and less understanding over time. The question “are machines conscious?” is impossible to answer if the term “conscious” itself is caught in this endless explosion of meaning.
Machine Consciousness is Easy
So why have this conversation? Because in the hazy maze of the question of machine consciousness, there is a bright light. Each of these manifold definitions of consciousness is surprisingly easy to find occurring within machines today. And it is this observation that leads me to so resolutely prove that machines are conscious.
The fact that it’s so easy to demonstrate machine consciousness, yet the public at large (along with our entire intellectual apparatus) seems to believe that machines aren’t conscious, surprises me, and motivates my writing here.
As I hold what I know about how machines work up against these definitions of consciousness, I find myself gaining new insight as to how machines function, particularly from a historical perspective. As I explore, I’ve noticed that the earliest manifestations of the various forms of consciousness are remarkably basic.
Method of Proof
I will explore the various conditions which other authors have surfaced as requirements for consciousness one at a time. Each condition will be the subject of a single post. Each post will describe a term through the eyes of one or more authors who require it for consciousness to be considered present. Then I will proceed to show how machines fulfill that condition.
I’ll write from my own background. My claim to fame is inventing VR painting to earn my Computer Science PhD from Caltech. I spent many years on the international art circuit, and have taught university classes on subjects ranging from game design to technosociology. I am a computer scientist, an artsy one perhaps, who has designed no shortage of digital products. But I’m no philosopher. I got into Caltech by dashing through undergrad in three years with a 3.96 GPA. My only B? Philosophy 101.
I say this to lower any expectations of philosophical contribution. My whole process is motivated by an attempt to not take any philosophical position at all.² What I understand most about machine consciousness is the machine stuff. I’ll describe, not just the symptoms of machine consciousness, but also the mechanisms by which this consciousness is manifested. Each article will serve as a brief slice of computer history, telling you how and when each condition for machine consciousness was first realized by humans.
My aim is to work as a computer scientist, describing the machines, not as a philosopher, describing what consciousness is. I will leave this last part to others. I am a consumer, not a producer of definitions of consciousness. My chief contribution is to show how these examples of consciousness are materially manifest within machines.
Extent of the Proof
Sample conditions I’ll consider will include things like self-awareness, knowledge, planning, and a theory of mind.
How many conditions will I write about? It is not clear. The only thing that’s sure is, no matter how many conditions I demonstrate machines to satisfy, this proof will never be complete. Is it worth it to write a post about a fine delineation of meaning I somewhat cover in another post? Probably not. I’m interested in this topic, but I don’t have the stomach for an exhaustive treatment that will be a bore to write, and of interest to only the most technical of readers.
I use the term “proof” somewhat flippantly. Consciousness to me, the non-expert, is the elephant in the fable of the blind men and the elephant, a mysterious gestalt that no one can fully appraise. I don’t know if the elephant exists. I can’t prove an elephant that no one can describe. I can only treat the pieces of the elephant (examples of consciousness) that have been delineated.
On Mysticism
There is something slippery, fleeting, and mystical about consciousness. For many, the very indefinability of consciousness is part of its essence.
This kind of mystical position is taken in mainstream philosophy. For example, Nagel’s bat argument is constructed, quite intentionally, to define consciousness as something subjective that cannot be validated objectively.
Of course, this fundamentally hidden consciousness also appears frequently in religion. Transcendental Meditation describes God Consciousness, and contemporary Christians talk about Christ Consciousness. On the border of spirituality, Jung’s collective unconscious describes a mental connection with all of humanity.
Much as I’d like to treat these subjects, it is difficult to do so within my project. I’m taking what others describe consciousness to be, and using my deep understanding of how machines work to determine if machine processes match those descriptions. Boiling mystical concepts down to examples, so that they can be examined in detail, is philosophers’ work. My work is to take these examples and demonstrate them in machines. I will do my best but do not expect me to travel deep into these spiritual domains.³ If I can’t prove to you that I possess a type of consciousness, I’m not even going to try to prove that a machine possesses it.
An Open Challenge
It is only natural in the selection of my topics that I am biased towards conditions of consciousness for which a positive proof can be written. The worth of this project grows as more criteria of consciousness are brought up for consideration. I am all the more happy if, at the end, we find out there are a few aspects of consciousness that machines don’t possess. This would be a certain kind of success indeed. To these ends, I welcome any suggestions for definitions and counter-arguments to my own arguments. Let the fighting begin.
Want more? Read Living with Frankenstein: The History and Destiny of Machine Consciousness, print and ebook available on Amazon.
Footnotes
[1] Can you tell me the difference between Christopher Muscato’s subconscious and conscious levels? Notice he indulges in the common practice of using the word ‘conscious’ to circularly define consciousness.
[2] Of course, I defend well-known positions. Nothing is new in philosophy, certainly nothing that I would write. Dennet describes consciousness as a “bag of tricks” and my process is to investigate each trick, one at a time. I end up making a case for strong reductionism.
[3] I am tempted to respond by arguing for a similar spirituality around machines, a kind of computer religion, but this just seems silly. However, such an argument would be as unassailable as many a faith-based claim of consciousness.
