How many propriatary solutions do you think will hand over the code for them to train, thats the point that i'm trying to make.
Opensource has provided training data, closed source code not so much. They will have a bad time learning.
Opensource has provided training data, closed source code not so much. They will have a bad time learning.