I don't just want to work on one idea at a time, because then I'm learning slowly and what I learn may be confined to that specific context. If I am able to implement many ideas at once and draw connections across them, then I'll build better mental models of how the world works.
Likewise, not every idea is a winner. That being said, if I shoot enough shots, I'll find the winners and learn how to more reliably produce them. In general, this is how I approach creativity as well. I generate more ideas than anybody could possibly use, and then I narrow down.
Portfolio effects are a big reason why I’m consulting now. By implementing my ideas with of multiple products experiencing varied problems with unique user groups and contexts, I am able to see which ideas are robust, which have value, and which ideas deserve to be scaled. This allows me to usefully Learn by going up and down the ladder of abstraction.
Long term, I see myself running a startup studio / consultancy hybrid firm.