Just a followup I think you (4death4) are missing the original comment that I am replying to. In that the Poster asserts that "I think there is a good amount of

Just a followup I think you (4death4) are missing the original comment that I am replying to. In that the Poster asserts that "I think there is a good amount of content that has been consumed that was neither in the public domain or properly licensed". I am saying this is probably the same for humans using the Internet. They don't have an explicit license to read or learn from much of what they read on the Internet. You are asserting that most of the edge cases are fair use or similar. The problem is that OP is asserting that an explicit license is required for an AI to learn from content. My main point is that it is a contradiction to believe that it is legal for humans to read and learn from anything and illegal for machines to do they same. Either it is legal or illegal for both.