AI has no body

New Year in The Netherlands has always had a problematic association for me because of an experience I had as a small child. My parents and I stayed at my grandmother’s for the occasion, and I was put to bed on New Year’s Eve being to young to be able to stay up.

At midnight hell broke loose. The room in the apartment of my grandmother where I slept looked out over the sports grounds where the main fireworks event took place each year.

It was the most terrifying thing I had experienced in my short stay on this planet. Things were exploding, burning, and everybody seemed to be screaming, and I was choked by a pungent smoke.

Maybe my parents had tried to explain to me what would be happening. I cannot remember. I do remember that I was alone and terrified for a few moments that felt to me like a very long time before my mother came in and attempted to comfort me.

That night was the first time I realised that life was finite, that there was something that could never be avoided: I would eventually die. The absoluteness of it felt like it was crushing me, and all joy was extinguished for several days.

To deal with the weight of this awareness, I, as many children do in some way or another, started looking for a way out, and entertained various fantasies on the problem. They converged in one that persisted for many years: I would upload my consciousness into some kind of machine that I would launch into orbit. That way I would be completely liberated from this world. I would live as pure consciousness, without a body that could become sick, age and die, looking out over the world and using all kinds of sensors to see what was happening. But never, ever, partaking.

This fantasy, I later realised, fitted in quite nicely with the Zeitgeist, the spirit of my time. We were starting to build machines that were called “intelligent”: computers that could do seemingly very smart things. And these computers would be doubling their power each year! The possibilities were staggering! Intelligence was felt to be a set of algorithms, a disembodied processing power, a kind of hypercharged clockwork or engine but now using this new and mesmerising power: electricity, power that could move with the speed of light. 

The great Alan Turing did a nice thought experiment around the year 1936, called the Turing Machine. This concept was one of the fundamental ideas that led to the modern computer (in fact I noted in several articles that maybe is was not so much that idea as the idea of the Meme Machine put forward by Vannevar Bush in 1945). It was an idealised, “bodyless” thought construct. It did not live (or was thought to live) in the physical world, but instead was more like a mathematical thought experiment, much as the thought experiment that led to the theory of relativity, done by Alfred Einstein in 1905.

The Turing Machine was a machine with “infinite” memory (which Alan Turing labelled a “tape”). One could write, read, and erase strings from this tape. However, as conceptualised by Alan Turing, this writing and so on was frictionless and immediate. The machine, so to say, did not have a body. The infinite operations it was able to do generated no heat.

In reality, all functioning systems generate heat. This is a consequence of the second law of thermodynamics which states that in any closed system entropy can never decrease. So as soon as you embody this mathematical concept of a Turing Machine (or any machine for that matter) in the physical world, these limits arise. This has led to the formulation of the second law of computing, the Landauer’s Principle, which relates computing (let us say the manipulation of bits) to entropy, and the real, physical world limitations that arise from that.

The concept of mind, consciousness or intelligence as something that is disembodied, is actually one that is intimately related to Western culture. Maybe you find the concept obvious. That shows how it is part of the Zeitgeist. We talk about “pure consciousness” (as if any relationship with the physical somehow dirties it, makes it “impure”!). 

Western Culture, which might be seen as originating in the Middle East (Sumeria), was intimately associated with a revolutionary human invention: writing. Communication based on complex symbolic language has always been a distinguishing characteristic of being human. In fact some argue that being human means using language.

Language enables us to create purely fictional narratives, of which “fake news” is just another label. In a way “fake news” is a human trait. In fact humans have devised elaborate cosmologies in their religions that are difficult to relate to anything tangible, visible or measurable.

Computers run “software”: an almost purely virtual power, consisting of instructions for processing units (processors). The instructions and the processing hardware are in no way aware of each other. The instructions are at the most hampered by the set of instructions that the hardware allows, but apart from that it knows nothing whatsoever about the processing unit: nothing about its layout, nothing about how it is connected through a data bus, and certainly nothing at all about their environment. There is no body awareness in computers. There is also no awareness, and certainly no feedback from that awareness, of the increased entropy that is the inevitable result of all the computing being done.

Living beings, from the smallest units such as bacteria to complex multicellular life-forms as ourselves, all seem to have this one thing in common: they have awareness of their own body in its environment: proprioception. Also plants seem to have this innate ability.

What if intelligence, consciousness if you want, is intimately related to this awareness? What if the entire concept of “disembodied” intelligence is fundamentally flawed and can only lead to madness? Is it even possible to create “intelligence” that has no body?

The concept of intelligence, and certainly of consciousness, is by no means explained or understood. It eludes us to find a unified theory of consciousness. But whatever that theory will contain if it will ever come to be, I believe that an essential part of it is the fact that to be conscious is to have a body, to be aware of that body, and for that body to be aware of itself and its environment. We urgently need to step away from the idea that “intelligence” is disembodied and lives in a purely idealised state or in the form of mathematical equations.

reflektis Logo

Copyright © 2021, reflektis & Rob Vens

%d bloggers liken dit: