5 ms·
Making your code open source was consenting to have people (and AIs) read it and learn from it. As an open source developer myself, I always understood that.
by jujube3 1mo ago
Making your code open source was consenting to have people (and AIs) read it and learn from it. As an open source developer myself, I always understood that.
- _luci_ 1mo agoI consented to PEOPLE reading it. How could I have predicted that AI would take off so rapidly and get trained on my repos? How could I have predicted that AI would be trained on code without a license attached, or on code with strict licenses attached that do not allow for such use? A silly response if you ask me
- frozenseven 1mo agoThe field of ML is probably older than your parents and IP law doesn't grant people immunity from it (or any other form of data analysis for that matter). If a person can read it, so can AI.
- jujube3 1mo agoSaying that you can't copy something is different from saying that it is illegal to read it. For example, it is legal for me to read Ray Bradbury's book, "Fahrenheit 451" when I check it out from the library. It is not legal for me to copy the book. This is why it's called "copyright" and not "readright". So then the question becomes what "copying" means. There is an extensive body of legal precedent on this matter, and it mostly all points to "transformative uses" being exempt from copyright. For example, if I count the number of words in Fahrenheit 451, that number is not copyrighted. If I reproduce just a sentence, or maybe even a paragraph, that also doesn't count. But at a certain threshold, once I've reproduced enough of the book verbatim, I have violated copyright. Training an LLM is also a transformative use which means that it is simply not covered by copyright.