thats what start up capital is for. thats what the venture capitalist job description is.
typically good IP to start a company based on is based on white paper research which a university will patent. then they will sell that patent to someone who thinks it can actually work, and there is your start up. or sometimes the researcher who wrote the papers, and the patents, will buy the IP from his own university, and start a company around it.
another route is research at a government funded lab. like lawrence livermore, or oak ridge or something. but a similar path of white paper, to patent, to company.
and then you take that company, and tell everyone everywhere what you are doing in order to sell yourself to someone big.
this is the eco-system of technology development.
research costs big money. we are not talking about "oh i thought a new clever way to make a paintball gun valve close slightly quicker" technology development. we are talking "i have this idea for a new display technology and i've demonstrated that it does work" or "here is a new medical process that no one has ever done before and i can invision building a machine to do it" type research.
this type of research is funded by universities, NSF, NIST, government labs, etc, and protected by them. and then sold to venture capitalists or directly to larger companies. sometimes a university or group will license, but its less common.
sorry, i happen to be way too familiar with this and have been spending a lot of time researching the IP ecosystem of nano-imprint lithography.