Universe A
“Quit my job?”
YesDo you have a hard choice to make?
Want to make sure even if you pick wrong, some other version of you picks right?
It might sound crazy, but below, we call IBM's free online quantum computer to split the universe and tell you which one you're in. Choose based on that, and know there's another you out there who made the other choice.
“Quit my job?”
Yes“Quit my job?”
NoYes, we use IBM's real quantum computer every time you submit a question, and it really does fork the universe if you believe in many-worlds.
No. Imagine if you picked 'Yes' now. Then no version of you would be willing to do 'No'! How would you expect a version of you to do 'No' if you're never willing to do it?
Coins are mostly governed by classical mechanics because they're macroscopic. In classical mechanics everything is deterministic, meaning there is no way for two different outcomes to happen from the same starting point. Nondeterminism (and thus branching) comes from quantum mechanics.
We run your query through the quantum gate below. It simply prepares a qubit in 50/50 superposition of 0 and 1. We then measure that qubit. You can't have half a qubit, so the universe is forced to decide whether it's really 0 or 1 and thus split.
We interpret 0 to mean Yes and 1 to mean No.

It's not passed, it's just there for clarity so you can tell what to do based on the universe you're in.
In order for the universe to fork, IBM's quantum computer needs to work. If it doesn't work properly, it's just as good as just doing a normal coin toss. IBM's quantum computer has very low error rates (0.56% measurement and 0.08% decoherence loss), so there's very high probability that it works on this trivial single-gate circuit.
If many-worlds is true, then the universe is just branching the way we show in the picture.
If many-worlds is false, the outcome provably does not exist until you measure it*, which means the universe is making a decision for you on demand. There is room here for higher order intent unlike in classical physics.
* The GHZ experiment is classic proof, but we prefer the delayed double slit experiment.