Ginkgo's CEO explains in some interview [1] how they seek to become co-owners in the companies they involve with since apparently they are well funded for now.
"You’re a big company, and you have an idea for what you’d like a cell to do. You come to us, and we agree on the spec. We program the cell to do it. We give you that cell app. That’s our product. It’s essentially an intellectual-property license. We’ve written a piece of custom software for you, except it’s a custom genome inside a cell. You then take that back and build a business around it. The key is: I need to get a piece of that. It could be a royalty or - we just announced a partnership with Y Combinator - we can take equity in companies that are building on top of our platform. We want to make it so that everybody uses our platform to program cells. It’s much cheaper than it’s been historically. You want to start a biotech company? You’ve got to build a lab. You’ve got to get equipment. You’ve got to get all the lab materials. All these things are an enormous upfront expense, and it really limits innovation. So what we’re saying to the Y Combinator companies is: Just use our platform. We’ve already built all that. We have that huge fixed cost, and you get a low marginal cost. And I will program that cell for you in exchange for a chunk of your company. That’s the deal."
In the early 2000's I was seeking to license some ML software that performed 3D reconstruction, similar to photogrammetry, but specific to being accurate for 3D versions of people. In the forensics and medical fields there are companies creating these types of trained algorithms for their industry's needs, and the quality of what they produce is acceptable as a starting point to create a 3D model of an person that looks like the real person.
However, every single company I located providing any useful piece of software wanted a piece of my company, or they wanted a revenue share of anything my company sold. I found outside of "traditional software" there is another software industry where the developers are integrating themselves into industries as key suppliers, never selling their (surprisingly simple and basic) software, but licensing to companies that pay them for every single use - as if the software were a person being paid for every use.
They are one of the first YC
biotech companies with plenty of funding so it is just as likely to be long-term experiment/public relations exercise. Actual profitability may not be relevant.
"You’re a big company, and you have an idea for what you’d like a cell to do. You come to us, and we agree on the spec. We program the cell to do it. We give you that cell app. That’s our product. It’s essentially an intellectual-property license. We’ve written a piece of custom software for you, except it’s a custom genome inside a cell. You then take that back and build a business around it. The key is: I need to get a piece of that. It could be a royalty or - we just announced a partnership with Y Combinator - we can take equity in companies that are building on top of our platform. We want to make it so that everybody uses our platform to program cells. It’s much cheaper than it’s been historically. You want to start a biotech company? You’ve got to build a lab. You’ve got to get equipment. You’ve got to get all the lab materials. All these things are an enormous upfront expense, and it really limits innovation. So what we’re saying to the Y Combinator companies is: Just use our platform. We’ve already built all that. We have that huge fixed cost, and you get a low marginal cost. And I will program that cell for you in exchange for a chunk of your company. That’s the deal."
[1] https://www.bloomberg.com/news/features/2019-11-06/ginkgo-bi...