Coffee with JP: What's going on with copyright and the internet?
Coffee with JP returns with JP discussing current and historic events in light of the Hachette vs. Internet Archive decision. In it JP covers:
1) Singular XQ and the need for disinterested (read not financially vested) research.
2) Aaron Swartz and Open Access
3) Fair Use and copyright infringement
4) Large Language Models and plagiarism
5) Gaslighting of LLM critics
6) The lack of transparency around training corpora and the tech in commercial LLMs
7)The unclear level at which deterministic and supervisory learning are being used as "guardrails," "customization," and "fine-tuning,"
8) The philosophical complexity around creativity and intellectual property
9) Sharks, fish tank ecosystems, and more balanced approaches to the tech innovation ecosystem
In the piece she notes authors Jooyoung Lee, Thai Le, Jinghui Chen, and Dongwon Lee and their research at https://pike.psu.edu/publications/www23.pdf who determined that LLMs measurably exhibit plagiaristic behaviors in three different species.