The way I frame this for myself is - just get the stupid solutions out of the way first. Do them so you can see how dumb they are, and move on to the better ones.

And that comes with the proviso that sometimes, the stupid solution is actually just fine, and you didn’t even need to make it so complicated.