0. Pick an area in which you have domain expertise. If you're building something you actually want to use yourself, that's even better. As a solo founder, it's even more important that you have a strong intuition for your users and for your problem space.
1. Make something people want, get people to use it, and talk to them. This comes first, and almost everything else can and should be ignored until you've done this. This is the best way to attract good people. This is also the best way to attract funding. This should be your north star. If you're struggling to find people and/or funding, goto #1.
2. You have advantages, too: as a solo founder, you should be able to move faster than anyone else - even companies with cofounders or entire teams - in the early stages. When you talk to people and they make suggestions, you can and should literally immediately make them, deploy them, and tell them that you did so. This may feel strange once it starts to feel like you're a glorified CS rep, but this is actually the iteration loop that you want to be in. You can do this faster than anyone else because you have zero communication overhead and zero friction between the point at which feedback is received and the point at which improvements are shipped. When you find yourself in this state, bask in it for as long as possible.
3. Give your early people a lot of equity and make sure they feel like they own the company too.
The good thing is that it's been more straightforward than I expected (though certainly not easier than I expected).