- cross-posted to:
- news
- foss@beehaw.org
- cross-posted to:
- news
- foss@beehaw.org
Malus, which is a piece of “satire” but also fully functional, performs a “clean room” clone of open source software, meaning users could then sell, redistribute, etc. the software without crediting the original developers. But I have a hard time with the “clean room” argument since the LLM doing the behind-the-scenes work has already ingested the entire corpus of open source software – and somehow the output of the LLMs isn’t considered a derivative work.


How isethat a “clean room”? The program is scanning the actual software and making a version based on what it learned from the scan.
I think for this to be legit, you’d have to give malus a spec (no source code) of the program and then have it generate new code from that.
Malus uses two AIs for that. One creates a spec and the other implements that spec.
But that doesn’t even work, because you would have to prove that the original software was not part of the training set. And with it being an LLM from a big corporation, that chance is close to zero.