Assumption: you are intending to build a software product
1) learn about how to do research for opportunities, specifically identifying valuable problems. Valuable problems usually impact lots of people and/or cost people lots of money.
2) learn how to write a problem definition (hypothesis) and then talk with people (surveys) about the defined problem. This helps to refine the problem definition and/or increase the potential value. Try to do this with lots of people, preferably those who might be actual users/customers.
3) learn how to do rapid prototyping. This could be anywhere from interactive presentations to low-code initial development. The goal here is to get something demonstrable that can be shown to potential users/customers so that you can get feedback and measure interest. The prototype might be throw-away. That’s ok, because you are not going to spend much time on these initial prototypes.
This whole concept is typically the opposite of how an engineer/programmer thinks. In other words, we typically think our knowledge and expertise will allow us to build something that people should like. But that’s usually the wrong approach. Let the people tell you what to build.
If you can master these initial steps, then you will have a strong start in your project and increase the chances of success. Good luck.
A web search on “problem definition business” will give you a list of over 100 books on this topic, alone.