Computers as Logical Extreme of Autism

Post Reply New Topic

I had never heard that one, hilarous thank you !



Couldn't agree more, as for your previous post. But that does lead to interesting questions.

A computer is, as you say, a dumb machine. It obeys orders, we know why, we know how, its behaviour is completely determined by logic rules we completely know and understand. It just performs a lot of logic and symbolic operations much, much faster that we can. No human person, autistic or not, behaves in this way, at least according to our limited cognitive capabilities.

Being a computer scientist with a background in AI, I have thought a lot about this. However you phrase it, you end up with the philosophy of mind problem known as the symbol grounding problem. The idea is that computers can only perform logic operations on arbitrary and meaningless symbols, and the only thing that can give these symbols a "meaning" is ourselves. For example if I have a rocket and I want to go to the moon, I can program a computer to calculate the optimal launch date, engine ignition times, etc. In order to to this, I will provide it with data about earth and moon orbits, physical laws, the characteristics and behaviour of my rocket, etc. I will put the problem in the form of a program and data structures my computer can understand, which means translating the world into a shape that both makes sense for me and is workable for the computer. Then, my computer will do its work, which means that some patterns of electrical currents and potentials will evolve in a way that depends on how I programmed the computer. Then those patterns of electrical activity will be translated and displayed in a user-friendly way (printer, screen, etc.) and since that data represent a solution to my problem, I can use it to setup my rocket. But the point is that the input and output steps (translating the world into computer states and translating back computer states into something I can use in the world) was done thanks to my own abilities ; all the computer did was apply arbitrary rules I specified on meaningless symbols I provided, thanks to semiconductor technology. If I wasn't here with my rocket, my computer and my desire to go to the moon, there would be no semantic grounding to the computer's symbols and syntactic rules, and the computer wouldn't do anything interesting, or at least nothing relevant to what I perceive as the real world.

In order to have a strong AI, which is an artificial system fully "self-aware" in a way at least comparable to us, it would need to have an ability to relate to the world and take "meaningful" decisions in a way that does not depend on our own symbol grounding abilities. Right now I have no idea know if this is possible, or if we could recognize such system as self-aware. Everyone is so engrossed in his own reality and sensorimotor world. But what I do think is that any artificial system simulating and imitating life or counsciousness as we understand it is just a tool we use to trick our own perceptions and not a true AI. As John Searle wrote, a simulation of a human brain has no more chance of being self-aware than a simulation of a wildfire has of putting the lab on fire.

No. It could not. Your computer , however fast and slick is a dumb machine. Autistics, whatever level they function at are people.

ruveyn

Last edited by ruveyn on 03 Jan 2014, 4:10 pm, edited 1 time in total.

The difference between women and computers:

A computer doesn't complain when you stick a floppy into it. Women prefer hard drives.

This is something I've thought about a lot. My computer program is made from tiny logical instructions. My brain is made out of tiny logical cells. There would appear to be no way to reach consciousness simply by piling more and more tiny mindless blocks on top of each other. This is one of the reasons I believe in God.

Exactly. We may know quite a bit about what regions and receptors in the brain are responsible for what, but we know almost nothing about how this leads to consciousness! Ever since I was 5, I've wondered why I'm consciously experiencing life as me, instead of my father, a neighbor across the street, someone living in a different country, etc. And like you, I like to believe that there's some kind of higher power who is at least partially responsible for consciousness. Life requires a certain faith in the unknown (regardless of whether or not it's logically possible to know it!)

There is a functionalist theory that says that consciousness is the result of the collective activity of our cognitive apparatus. Without these apparati there would be no consciousness. In other words, consciousness is greater than merely the sum of its constituent parts. All our brain functions work together to produce this greater result.

I read a book several years ago on this topic; it was called The Emperor's New Mind by Roger Penrose. I highly recommend it! Penrose believes that our neurology and psychology are so complex that it is very unlikely we would ever create equally powerful robots using the binary logic that we use in engineering our computers, because our consciousness is a non-algorithmic process. He explains this by using examples from theoretical physics, logic, and mathematics, including Goedel's incompleteness theorems(!). He believes Quantum physics is important for understanding consciousness (in particular, the collapse of the wave function). There is also Penrose's subsequent book on consciousness called Shadows of the Mind, which I also recommend but haven't read yet.

This is something I've been pondering as well. Conciousness and whether it would be possible to make a computer that is conscious. I think a robot could be made to simulate human emotions and their physiological responses. But that's it. Only a simulation. Consciousness is free will. To go against your nature. A program with intricate algorithms could simulate the human brain, but you would always define its parameters.

Could you define free will there? If the robot was allowed to self modify in response to sensory input, it would appear to have free will, though in reality it wouldn’t. Is that any more than what we do?

I don’t see how free will is essential for consciousness. I can be conscious of events that I am not free to influence.

Penrose would probably say that randomness is built into nature at a quantum level. Brains are influenced by quantum events, computers are not.

When I was at university, one of the lecturers hooked a programmable gate array, essentially a physical chip that can be reconfigured programatically, up to an evolutionary algorithm and tried to make a radio.

The chip learned to be a pretty good radio, but the circuit made no sense. It had random loops in it which didn’t connect to anything. Removing the loops broke the radio, and programming another array with the same circuit also broke the radio.

It turned out that evolution was using the quantum properties of the chip to optimise the circuit. A different physical chip had different quantum properties.