On this, I'd strongly recommend picking which part of game development you're interested in as it can be absolutely massive. Once you have that, then you can decide what are the best steps to take to achieve what you want.
For me personally, I started by making small games on machines back in the 1990s, by trying to replicate games that I loved playing myself. Things are way bigger than they were back then, so it's not as easy as to make a game single handedly (at least not of any scope).
Being me, I decided to develop the tooling for the game first as that's what interested me. It's meant that I can scale up my game development if my games are ever successful, but it wasn't necessary the best idea if I was looking to be the most commercially minded. Once I had that (and it's really a lovely level designer :-) ), I then started work on the actual games…
For my recently released commercial games, as I'm a programmer by training / profession, I started off with placeholder graphics (I'm not an artist of any ability) and then after the game was largely code complete, I then went to freelancer websites to get someone to draw graphics for me (the difference between projects which were fun to play but looked awful and ones which looked nice and fun to play was a few hundred dollars - obviously depending on the requirements for your games)
If I were starting again, look at things like unity as you can get things working very quickly and that's enormously satisfying! Not to mention, they're in-demand skills and you can use anything that you create as a portfolio to get involved in bigger projects.
If you're looking for ideas to start with, replicate the classics, look at itch.io for asset packs etc. and then see what motivates you!