A good example is astrophysics. There are so many nice theories around, but they are all wrong. If you want to theorycraft the universe now, you need to deal with an incredible amount of very precise data, advanced maths and billion dollar machines. If you don't, the best you get is to have you theory proven wrong by an actual specialist, most of the times it is "not even wrong".
Researchers are now working on a subdomain of a subdomain, and they spend decades studying just to reach the level where they can start having good ideas.
Most of the easy good ideas are already taken. What remain are the hard ones, and those that require a lot of luck and/or hard work to materialize.
I think this is slightly specific to astrophysics and cosmology. In Neuroscience, it definitely feels like there is plenty of headway to be made by individuals with good ideas, although I admit that it feels as though ideas are harder to come by as more is published and the bar is continually raised.
A good example is astrophysics. There are so many nice theories around, but they are all wrong. If you want to theorycraft the universe now, you need to deal with an incredible amount of very precise data, advanced maths and billion dollar machines. If you don't, the best you get is to have you theory proven wrong by an actual specialist, most of the times it is "not even wrong".
Researchers are now working on a subdomain of a subdomain, and they spend decades studying just to reach the level where they can start having good ideas.
Most of the easy good ideas are already taken. What remain are the hard ones, and those that require a lot of luck and/or hard work to materialize.