Obviously the world was much different in Pythagoras’s day. I am not clear on the distinction between a patentable machine and abstract idea. Patentable machines often have key mathematical and scientific theory. Can a new machine be built that uses some of the theory in another patented machine? The theory may be so general that it can be used in many machines totally unrelated to the patented machine. I do not have answers. I am just posing controversy.
The most valuable patents are the most general. TIVO has patented key elements of the DVR. Can the TIVO patents force anyone who wants to develop a DVR to pay royalties? I do not know the TIVO patents but I do not agree with the claimed generality of its patents. Perhaps my argument is based on prior art such as buffering concepts and prior digital video devices. A British inventor received a patent on hyperlinks, effectively putting a toll on web based information exchange. Even without the existence of prior art, the idea of a hyperlink seems obvious.
Even if patent controversy is resolved, patent examiners are a problem. To properly evaluate patents, you need to be a specialist in the particular field with an advanced degree. It seems that patent examiners lack expertise to evaluate submitted inventions.
There are very, very few developments which do not appear to be obvious in hindsight, yet original thought remains one of the rarest occurrences. Without patents to protect even "simple" thought, the only way to make money would be to have enormous capital and secrecy so that you could instantly flood the market place. But even then, large R&D couldn't be paid for.
Patents turn development into a race. I'm certain that increases vice decreases technological advancement. It also forces people to think in different directions. Look at big screen TV's.