[saint_abroad]

> Where do you draw the line?

My simplistic view is that the following is legally equivalent:

input -> ai network -> output

input -> huffman coding -> output

So, whilst:

* compressing and decompressing a copyright work is permissible;

* output and weights are deterministic transformations of the inputs;

thus:

* not eligible for copyright (lacking creativity); and

* are derivative works of the inputs;

[kevincox]

> output and weights are deterministic transformations of the inputs;

That may be true but I fail to see how any process that produces the same content that was input into it somehow strips the license. If the generated code is novel, then there is no copyright and it is just the output of the tool. If the code is a copy, but non-creative (example a trivial function) then it isn't covered by copyright in the source anyways, so the output is not protected by copyright either. However if the output is a copy and creative I don't think it matters how complicated your copying process was. What matters is that the code was copied and you need to obey copyright.

Again, I don't think that novel code generated from being trained on copyrighted code is the problem. I think it is just the verbatim (or minimally transformed) copying that is the issue.

[ghoward]

But at the same time, a compiler does a deterministic transformation of its inputs, and we still count its output as under copyright and license.

copyrighted input -> compiler -> copyrighted output

[saint_abroad]

Perhaps I wasn't clear enough on this point: copyright of a derivative work is distinct (but not inseparable) to the copyright of the original work.

So portions of a derivative work are covered by the original copyright, and other portions may be under a distinct copyright as a derivative work, and several copyrights may apply to a work as a whole.

In the case of a Huffman transform, the transformed work does not meet the "creativity" requirements to be eligible for copyright, over that of the original works.