John Koza has made a serious breakthrough when it comes to the computing power of today and our hopes for it in the future. Studying computers since his high school tutelage in the 1950s, Koza was a man who yearned to make advancements in the world around him. As a clear example of his ingenuity, he was unable to afford a computer while in high school, since the mainframes of the time took up entire rooms and cost far more than the modern day computer. To that end, John assembled bits and pieces from the gadgets around him, and with parts from old pinball machines and jukeboxes, he found himself with a computer that could accurately compute what day of the week a certain date would fall on.
John's latest invention far surpasses his date-decoding computer of the fifties. Utilizing 1,000 networked computers operating in tandem, his new device uses genetic programming to solve problems. Essentially, what the machine does is follow Darwin's laws of survival of the fittest - it compiles code, tests it against other code, and thereby determines the most accurate usage of code. And the machine wasn't built with one purpose in mind; instead, it utilizes it's ‘thinking' process to solve complex engineering problems of manufacturing and production of whatever you can think of. And it does it arguably more efficiently than a human. The machine is well within operating conditions, as Koza has already made an engineering breakthrough with it. Most recently, the machine has come up with a type of lens system that can be used in telescopes and binoculars that offers a better field of vision than any type of lens created before then, and it does not infringe on any patent. To perform this feat, the program generated 75,000 random types of lenses. From there, it ‘mated' the designs, taking the features that were helpful in one and placing them with features that were necessary in another. From there, generation after generation is created until every part of the design specifications is met. For the lens, the machine processed 295 generations before reaching a final solution. The machine has been proved useful in other inventions, as well. This heralds a scientific breakthrough in which computers may soon be doing much of our thinking for us. The machine has already received the first patent given to a non-human entity, and the sky is the limit.