I think on some level putting work into the public sphere requires letting go of how you hope it will be used. So long as there are incentives to profit off of

I think on some level putting work into the public sphere requires letting go of how you hope it will be used. So long as there are incentives to profit off of "unethical" industries, there will be people looking to exploit those incentives. This is especially true when the published work is general-purpose enough to be reused and the barriers to use are low. Look at Amazon's facial recognition, for instance. OpenCV and the like have existed for years but by tightly integrating with AWS, we've now unlocked brand new use cases and customer bases -- say, startups and researchers on one hand and police departments and authoritarian governments on the other. The pool of qualified developers, and by definition number of developers who might be willing to implement a facial recognition database for dissidents, goes up significantly as the technology becomes easier to use. For a map data format, you might make life easier for people drilling wells in remote rural villages, but you might also be designing a data format for bomb strikes.