I agree that the decrease in funding for big and vague projects feels like a big societal mistake. So many hugely influential inventions like ARPAnet, the networked office (PARC), and a whole book full of stuff from Bell Labs, came from a place where there were lots of smart people working on big long-term problems in the same context. PARC made WYSIWYG editors for the bitmap displays they made for the workstations they made to connect up with the network they made — none of those would have taken off the same way individually. And VC-funded startups that focus on one small problem for a few years are never going to fill that gap.