This is one of those internet “memes” that keeps going on and on no matter what. It’s been fueled lately by statements from Stephen Hawking and Elon Musk that they fear computers would become intelligent and take over the world, making people their slaves. However folks who really know computers like Apple co-founder Steve Wozniak sees it for the ridiculous misconception it is.
Who am I to talk? Well for over 10 years I was one of a very few people in the US who were, on a daily basis, given the task of figuring out why the newest, state-of-the art servers were not operating correctly. I worked for a Fortune 500 company that serviced servers for the largest computer manufacturers in the world (and made our Fortune 500 customers think it was serviced by the manufacturer – but that’s stuff for another article). I remember having a server brought to me with 12 8-core processors and 64 128GB memory sticks. It was not behaving properly, according to the company who had just purchased it. Never for an instant did I assume it had become self-aware and decided on its own not to comply with human commands just as I never assume my can-opener is not working correctly because it had developed intelligence and refused to cooperate.
Add to this that my hobby along the same time period was creating artificial intelligences. So I also know a bit about that subject, too, and would test my artificial intelligences by using them as “bots” in online games. In other words, my characters in MMOGs (massively multiplayer online games) were controlled by my AIs who interacted with other online characters who were controlled by real people. My purpose was not to make fools of other people, but to test my AIs and find areas that needed improvement.
So here’s the technical explanation. Computers, on it’s most basic level, are machines that change 1’s to 0’s and 0’s to 1’s. The CPU (central processing unit) has several different mathematical equations for changing 1’s to 0’s, and depending on which part of the CPU the numbers are sent to, decides how the 1’s and 0’s are changed. With the original 8-bit processing, the CPU could transform a binary number with 8 places. (An example is 10011101). With the new, amazing 64 bit processors, CPUs can transform a number that has 64 places. There is no area in a CPU that processes anything other than numbers which means all data, whether it is images, sounds, videos, or anything else has to be broken down into 1’s and 0’s before it can be sent to the CPU, and when it comes out of the CPU the other components in the computer has to change them back from the 1’s and 0’s back into words, images, video, and sounds. In other words, the CPU can only handle 1’s and 0’s and can’t do concepts, ideas, or anything else whatsoever.
Intelligent beings, on the other hand, think analogically, which means our thoughts are in waves that snake up and down and can have great lengths which means we are not limited to 1’s and 0’s and our brainwaves can be so complex we can think of an entire idea or concept in a single wave.
For example, picture your dog. You can imagine an image of your dog in a single wave. A computer must have a file image of a dog, then the file has to be read like a recipe and the corresponding 1’s and 0’s has to be sent to memory chips. These memory chips then send them to the CPU (or in certain cases, bypass the CPU and send them directly to the graphics chip). The graphics chip then has to paint a pixel to the monitor by sending a signal on the wire to the monitor. It has to repeat this process again for the second pixel, the third, etc, etc until the screen is filled with an image of a dog. It does this so fast that it seems instantaneous to us, but it actually requires thousands of steps. In many cases where the CPU is bypassed and the information goes directly to the graphics chip, the CPU is totally “unaware” of what the rest of the computer is doing.
Well what about Artificial Intelligence? Isn’t this a type of thinking?
Actually no, and it doesn’t even take place in the CPU.
If you ask a computer running an artificial intelligence program “Hey what do you think of Angelina Jolie?” it has to search through files already in the computer to find a match. If it does not find an exact match, it has to go through a subroutine which tells it what to do in the case of an inexact match or no match. It goes through pre-programmed processes which have pre-decided what to do in the case of an inexact match. If you asked one of my bots this question, it would probably reply with “I don’t know who Angelina Jolie is. Tell me about Angelina Jolie.”
It will “learn” what you say about Angelina Jolie and in future it will know who she is and reply with an opinion – because it has created a new file with your information, but this is not actually learning. It just created a file.
This is in no way intelligence as we generally think of it, but it is an emulation designed to fool humans who, because of our unique intelligence, can be pretty easily fooled. Magicians and con men make a living because of our own fallibility.
No matter how sophisticated and “intelligent” your computer is, you can’t play chess with your computer until you load and run a chess program. It can’t show you a picture of a dog unless it has a picture in it’s files. It can’t “take a guess” unless it has a matching program designed to guess, and then it’s guesses are limited to the pre-loaded choices in it’s program. It’s no “smarter” than a can opener, it’s simply more sophisticated.
A computer is unable to have any kind of intelligence because it is purely a mathematical device. It can’t even come up with a truly random number because there is no mathematical formula that has a different result each time the equation is solved.
Random numbers have always been a problem for computer programmers. People who use Excel and have used the random number function have noticed it always produces the same numbers every time the program is run. That’s because Excel tried to solve the problem by using a random number table. Excel uses the first number in the table the first time it is run, the second number the second time, etc etc. Programmers have tried to get around this by using a subroutine that uses things like the time of day, countdown timers, and so on.
Computers are imprisoned by mathematics. On the component level, a computer cannot do anything that is not mathematical and cannot come to a conclusion that does not have a predetermined mathematical equation. No matter how sophisticated or complex computers become, they will always have no intelligence whatsoever because they are bound by mathematics and cannot do anything unless it has a matching mathematical equation. There is no originality at all in a mathematical equation – they always solve to the very same number every single time. An earthworm will always be more intelligent because it is not bound by mathematics and can come up with a different answer each time a problem is faced. So continue to enjoy the movies and novels about computers becoming intelligent and taking over the world, but don’t worry about them ever actually doing anything like that.
Go forward and think free.