How would one person bootstrap a venture in medical ML without initial funding or datasets?
How about starting with a problem to solve? You don't need anything but a pen and paper. Gotta start somewhere.
In medical AI, you really need datasets. Of the most expensive kind, and lots of it.