I didn't say technology would solve those problems. I said technology can be used to explore potential solutions.
I understand the point; you've restated it exactly as I understand it. That's why I brought up unsolved problems. The state of the art here is very often new and novel approaches that have not been explored in enough depth.
Take color theory, for example. I want a function that maps points outside visible Lab to the nearest visible point in Lab. Sure, that's a solved problem, but if you don't already know how to solve it, you've got something to learn. More importantly, can it be done quickly? Can it be generalized to other color spaces? What is the best data structure for encoding only visible points in Lab? Then there's a problem of generating maximally orthogonal gradient curves covering an image.
There are purely technical problems. Can you solve them analytically? Will it help you to learn Topology or Tensor calculus? You won't know until you try. And there are new results in any of these fields coming out every day. Pick up a textbook and think about how to write software that will be capable of solving all the problems in that book. Does it already exist? Is it optimal? Easy to use?
Everything looks the same when you're saying "I'm just applying some abstraction to a problem and encoding it into a machine using several layers of translation." But that shouldn't make you cynical, as that leaves plenty of room for innovation and learning. But if you're thinking "this new framework/language is just doing more of the same," then maybe stop focussing on frameworks and languages and instead focus on the problems they should be solving rather than the ones they actually do. And even there, you should have no problem seeing syntactic and efficiency hiccups that might be ameliorated through an intense exploratory creative design session.
I understand the point; you've restated it exactly as I understand it. That's why I brought up unsolved problems. The state of the art here is very often new and novel approaches that have not been explored in enough depth.
Take color theory, for example. I want a function that maps points outside visible Lab to the nearest visible point in Lab. Sure, that's a solved problem, but if you don't already know how to solve it, you've got something to learn. More importantly, can it be done quickly? Can it be generalized to other color spaces? What is the best data structure for encoding only visible points in Lab? Then there's a problem of generating maximally orthogonal gradient curves covering an image.
There are purely technical problems. Can you solve them analytically? Will it help you to learn Topology or Tensor calculus? You won't know until you try. And there are new results in any of these fields coming out every day. Pick up a textbook and think about how to write software that will be capable of solving all the problems in that book. Does it already exist? Is it optimal? Easy to use?
Everything looks the same when you're saying "I'm just applying some abstraction to a problem and encoding it into a machine using several layers of translation." But that shouldn't make you cynical, as that leaves plenty of room for innovation and learning. But if you're thinking "this new framework/language is just doing more of the same," then maybe stop focussing on frameworks and languages and instead focus on the problems they should be solving rather than the ones they actually do. And even there, you should have no problem seeing syntactic and efficiency hiccups that might be ameliorated through an intense exploratory creative design session.