4 ms·
I am not a lawyer, but I'm wondering if the same argument can be made about derived works such as text - the AI is learning from open-source code, and outputtin
by samspenc 3y ago
I am not a lawyer, but I'm wondering if the same argument can be made about derived works such as text - the AI is learning from open-source code, and outputting not the exact same code, but just 'derived' code based on patterns in the original code.
In other words, somewhat similar to what a programmer studying existing and open-source code would do: read the code, understand it, and try to reimplement by writing their own code.
- gwnywg 3y agoIn my uni (and I guess in every other uni) they have a system which detects plagiarised parts of mastery thesis (and other too). My first thought is, would text produced by AI pass that test. And also I guess the same could be done with code (lets say plagiarism test done by independent escrow company) I remember people claiming things like github copilot were reproducing original code verbatim and this was the main source of concern. I would see this as a problem if my colleagues were copying code as if it was their own.
- trillic 3y agoIt’s probably Stanford’s MOSS https://theory.stanford.edu/~aiken/moss/ https://theory.stanford.edu/~aiken/moss/