Post #1688979
2026-04-25 13:05 UTC
@camertron A number of commenters do not seem to understand the difference between deterministic transformation of an input text and probabilistic text generation given a PRNG seed. I'm not sure if it's genuine or willful ignorance but it's still troubling how many dedicated computer touchers don't understand what a compiler or assembler does and how it does it.
Ir helps to remember that the vast majority of the training set the code assistants were built from was not provided with explicit consent from the original authors. Unlike using a library or modifying open source code, there's no trace left of the original authors, no attribution, no credit given for design and implementation. It's perfectly fine to say one has made something using some set of preexisting parts or using the work of others provided you clearly indicate what is your original contribution and what is the work of others. If you're a degreed engineer with the potential of licensure, part of the professional engineering code of ethics is to give credit where it is due and not take credit for work you did not personally perform. Attribution laundering is an ethical issue in engineering; it seems less important to general software developers who may not be able to point to a clear set of ethical standards they commit to uphold. No amount of pedantry about tool use can weasel around using other people's work without consent and taking credit for it as if it's your own.
Replies (1)
-
@camertron@ruby.social 2026-04-25 18:29
@arclight Yeah, that's a fantastic point. In my original post I was trying not to touch on the ethics of using LLMs since it's such a massive topic, but you're absolutely right. There is no ethical way to use LLMs for the exact reason you mentioned - they are large-scale IP theft machines.