I think that people will still develop the theory. It still needs a bunch of work but it seems popular enough to stay around.
One of Voevodsky's views was that all the foundation for a dependently typed formalization of maths is already there, and that it mainly needs to be organized in a better way (cf his unimath project).