3 ms·
It's not that clear cut though. I've been thinking about this question since I was about 9 and the conclusion I came to as I stepped off that circle is that the
by Dn_Ab 16y ago
It's not that clear cut though. I've been thinking about this question since I was about 9 and the conclusion I came to as I stepped off that circle is that the distinction is meaningless because they are dual to one another and hence not separable. But if I had to choose I would say everything is a discovery of how to put or see a series of things in a certain way.
Consider the light bulb. It could be described that someone discovered that if you run so much current through this material you get light. And we can ask the same questions about math. Is the matrix a discovery or an invention? What about zero or logarithms or Aleph Null or electricity? The invention part is the series of steps that take you to the discovery, if these steps arrive at a construct then this construct is the invention. Invention is the centering of discovery to the human condition but everything is already there so to speak. Things like zero or a matrix or complex numbers or quaternions or taylor polynomials or the combustion engine are inventions or tools to allow us to see, do and think about certain things.
I am not for software patents because they are used in the opposite way than they were intended. Patents were an incentive to get people to share but now they are used as a tool to stop people from sharing and creating. People associate inventions with physical things and material costs but really every reproducible invention implicitly contains an algorithm and that is what the patent is on. The thing with software is that there are no material costs and so the constructs can be built and arrived at with pure thought like in maths (which makes sense since algorithms are about math in a constructive universe). Here the algorithm and the result are one and the same where as in the old world the algorithm yielded the result.
This means that a lot of its constructs can be arrived at with trivial costs and then hoarded and used as ammunition by better captilized entities. Counterproductive. In areas where costs are non trivial and barriers are high then patents make sense as a motive to share and incentivize. But information moves so well now that the old time scales no longer make sense. And in particular, because like maths nearly all software are constructs of thought progress is made very rapidly via composition of pure concepts and so the time scale must be zero or the opposite effect - one of anti progress - is had when patents are applied.