“Yet, leading AI researchers realize something is not right.” — Glad to hear it.
I will now put my finger on exactly the thing that is “not right”. Please pay attention.
In the book he always regarded as his master work, The Philosophy of Freedom, Rudolf Steiner argues that absolutely everything we know, and can ever possibly know, derives from exactly two sources: (1) perception; and (2) thinking. Percept; and concept.
Steiner argues further that in “thinking about our thinking”, we directly perceive our own thoughts; and that in this case, percept and concept are of the same order, are on the same ground. First we create (we think of something); then we perceive our creation. Steiner points out that in the Bible, even God first has to create the world, and only then can He perceive that it is “good”.
Perceptions can be likened to electrical impulses in nerves, to brain wave pulses. If you step in the snow, the snow “perceives” your foot; there is literally an impression in the ice. Do you say: the impression in the snow proves that the snow is “thinking” about my foot? Do you say there are magic forces in the snow that somehow make it sink to match your foot?
Over, and over, and over, and over, and over, and over again: I watch you AI people. And at every point, I have to say: where is thinking in your scheme? Your machine may embody thought. Please, a shoelace embodies thought, capisce? And even if you engineer a smart shoelace that can thread itself, is the shoelace thinking, or is it embodying your thought?
Get this straight, get this right, get this once and for all. Thinking is a spiritual process. Thought occurs in a universal, cosmic, spiritual plane. No combination of electrical brainwaves, no matter how complex, constitutes “thought”. The brain is like a radio receiver: it tunes into spiritual frequencies and renders them perceptible on a physical plane, so you can type these letters on your keyboard, act in the world, write words that correspond to universal concepts. Believing that electrical brain waves constitute “thinking” is an elementary error of the most atrocious order; yet entire industries, whole world views, are premised on the idea that these machines can somehow “think”.
I’ll keep using this analogy until someone shows they understand it. I’m an old radio man and I love old radio man jokes. This is my favourite one. In the early days of wireless, a radio engineer meets a man who says: I’m amazed, just amazed, how you get all those little people, the opera singers, the news readers, into that little box.
And the radio engineer explains: no, no, the opera singers are in an opera hall, we pick up the their voices and transmit them through the air, using electromagnetic waves; all these valves and coils, they pick up the signal and turn it into music for you.
And the old guy is very grateful for the explanation, and says, he understands now, how it’s all done, thanks very much.
“But”, he says, “I’m sure you do keep a few little people in there, just in case.”
This is exactly, and I mean exactly, what you are doing. You find some changes in the brain wiring, and you think — ha! — you’ve found “thought”. Look! There is a little opera singer in the box!
The terrifying fact is that all these people who are pushing AI are creating systems that embody their thoughts, even their subconscious thoughts, and setting these mindless automata out into the world to implement them in real situations, like driving cars. And saying, look how fucking clever we are.
This is where I find myself on the same page as Elon Musk. I am terrified by AI. Terrified that people are stupid enough to believe that it even exists.
Let me put this into context. You programme “clever” machines to kill human beings. I’m talking about strategic nuclear regimes. The most sophisticated simulations of the Cold War showed the Russians and the Americans shooting nuclear missiles at each other on autopilot — long after all the humans were dead. That is AI.
The vast machinery of global destruction has one target programmed into its routines: humans. Save the buildings, save the infrastructure, use neutron bombs, just destroy all life, leave the machines intact. Above all, the machines are given one message: Kill Humans.
And if the programming runs haywire: is the machine thinking? No, it’s embodying your deepest subconscious intentions with perfect mechanical clarity. Kill Humans.
Something is wrong with AI, all right. It’s an oxymoron, it cannot and will never exist. But idiots who think that it does, can create machines in their own idiotic image that are quite capable of wiping out all living beings on this planet.
The last I saw, the USA, on its own, had enough nuclear weaponry to destroy the whole world 500 times over.
Now, I have to ask Americans: why do you want to kill all of us, so many times? Isn’t killing all life on Earth just once, enough for you?
Consider one final thought. If Rudolf Steiner is right, and Thought exists in its own, timeless plane: then even the most insignificant idea you’ve ever had, is more permanent (read “eternal”) than the entire physical Universe and every atom in it. We know these will all pass away. Thoughts are different. Marbles rattling in a tray don’t think, no matter how long you rattle them. The fundamental axiom of all spiritual beliefs is that the material world devolves from spiritual realms. The axiom of the materialist is that if you have marbles rattling in a box for long enough, somehow, those marbles will start organising themselves in ways that show they are “thinking”, the electrical impulses in their wiring create a kind of excrescence or emanation which corresponds to “thought”. This is just absolute complete mindless nonsense, yet entire economies are founded on it.
Wake up. Think.
