1. Join a company (usually big) that has a culture of open sourcing softwares.
2. Start an open source project you are keen on, and make use of the resource to grow the project/community/contributors.
3. Quit with other core maintainers and start your own company that commercializes the open source project.
Remember also, that large companies have experimental expansion phases which they usually abandon once economic crisis dawns. If the thing you build is such a "experimental fort" - your parent company might be open to a "buy out", were you promise them favorable conditions by your future company with whom they share the IP.
If you are truly developing tools innovative enough to battle against the existential crises of human civilization (as first enumerated by the Club of Rome in the early 70's and currently unfolding today), then all problems of law/licensing and finance/pricing fall by the wayside as your on-the-ground presence starts to ascend above your local competitors in the battle for the hearts and minds of the younger generations in your midst who are beginning to sense how infinitely screwed their future prospects are. What I'm trying to say is that it might make more sense to submit one's innovations through direct community engagement rather than writing up long articles to be posted on the web meant for consumption by perfect strangers.
What is “an open-source business”? I.e. what do you want your business model to be? To be precise, when your customers pay you, what do they expect, and receive, in return? This could be consulting, bespoke development, etc. That, whatever that is, is your company, not “open source”. The fact that you publish some code as open source is mostly irrelevant. No company is “an open source company”, they are instead companies with business models which allow them to publish source code under open and free licenses. A woodworking shop could be such a company if they also publish whatever software they write (during the course of their normal work) under open and free licenses.
In other words, open source is not a business model¹.
Well it's either that or write software that's useful enough to business that they want to pay money for the features it provides, but arcane and poorly documented enough that they need to pay your consultancy fees to use it effectively...
