I'm a largely self taught PCB design/escape game/controls/etc designer.
Arduino is a wonderful place to start, better is the ESP32 based wifi stuff.
You will want to learn some analog too, but I would stay away from multivibrators and 555s and other low level stuff unless you're really serious about deeply practicing fundamentals. They're easy and kinda fun and cheap enough to do if you see something interesting, but I don't think you need to work your way through a book of them.
Software is free. It can be changed. It doesn't usually wear out. I don't do usually do anything in hardware if I could do it in software.
The biggest challenge on almost any project will usually be power management. You will probably need to learn about inductive and capacitive loads, backfeeding regulators, switching regulators, battery protection, inrush transients, smart driver ICs, relays, etc.
You'll also want to learn all about op amps and RC filters.
I would strongly advise being careful with ANY "big ideas". Not big projects. Those are fine.
I mean big ideas like "I'm gonna do all my future stuff with this connector, better but a ton of them!". It. Is. Very. Very. Hard.
Instead, use standards. Your idea is probably 100% better than USB for this application. Your next idea will be even better. If you want to do anything but fuss with low level stuff... Just power your thing with the USB C trigger and be done with it.
Don't invent a new serial protocol if you don't have to, etc. The people who would be most interested are too busy with their next big thing. Focus on the specific things you most want to innovate on and don't try to reinvent all of embedded tech.
Do NOT buy any really specific parts that you don't have a full design and plan for unless you want to accumulate crap.
The only exception I've found is premade modules. They are so easy and trivial you find yourself using them nkwhile your fancy design gathers dust.
You do NOT need 500 kinds of resistor and capacitor especially if you are going the digital route.
You really have to figure out your goals. If you want to do art installations and home automation or something and are just using tech as a tool, modules are probably going to be your friend.
The best absolute immediate way to get started instantly is going over to falstad's simulator online and playing around there. I would highly suggest it maybe even more than any physical educational tool.
At the basic level, the sim is accurate enough, but much easier and faster, and you can't burn out a part that doesn't exist.
After that, look into things like the Wemos D1 Mini and the absolute hoards of USB-C
Keep things simple and modular and aim to use as many off the shelf unmodified parts as possible.
Want to run on 12v? Maybe don't put a 7805 on your board.
Find a nice 12v to USB converter, base your thing on a USB-C ESP32 type board, and now it's all modular and swappable and the parts are reusable.
Want to build something from scratch for fun? Cool!
Just make sure you know what parts you will enjoy doing from scratch, and that there's a real risk of getting tired of any long term project, thinking of a better way, etc. Staying with off the shelf methods for the big stuff gives you a framework, a starting point for research, keeps you from getting bogged down in arbitrary choices, etc.
From there, you are free to do each individual thing from scratch as much or as little as you want.
Oh, and check out the Pinecil, a clone T12, or a similar iron. Don't let anyone try to make you spend $100 on an some random analog iron with expensive tips just because it sounds like a pro brand!!
Also, solid state and low power is your friend. If it moves or handles power it might wear out. If it moves and it's cheap it probably will wear out. If you want it to move, but high quality even if you have to make the digital stuff lower quality to stay on budget.
This stuff is really... not actually as hard as various books and how tos make it sound, at least at the hobby level.