How do I build a tech startup if the products/services are too complicated for an initial proof of concept?

I'm currently a graduate student and let's say I want to develop a startup that focuses of developing small nuclear reactors for data centers. As a nuclear engineering student, I might have a lot of knowledge about the feasibility of the project, what materials we might need and maybe I can run engineering simulations to demonstrate my idea. But there is no way I can develop a proof of concept, definitely not with the budget and support I get from my university, not to mention the safety hazard working on a nuclear system. Moreover, even a simple proof of concept device would require significant expertise from a variety of fields: electrical, mechanical, physics, chemistry, etc. While I might have a general idea, its nearly impossible to have all the knowledge to build it from scratch. Even if I work with a group of students, it would still be very hard with lack of tools, skills, and knowledge. In this case, what's the best way to go about initiating a startup? I would require multiple licenses, personnel, ample funding to even get to a proof of concept phase, not to mention building an actual reactor that is connected to the grid. This is just an example, but there are a lot of projects like this that are large in nature and might not follow the traditional pathway: research, develop proof of concept, pitch concept to investors, develop and refine, and then market. Many do, for instance, you can develop a software all by yourself and pitch it to investors, even small robots you can make a small scale prototype in a lab. But what about large projects like this? 3D printed rocket launch company, EV manufacturing company. Where do you even begin? Without a proof of concept, would investors give hundreds of thousands of dollars to a college student to start a capital intensive company?