AI, Copyright, and Open Licensing

Engineering notes

Engineering notes will be added by the aipentium editorial team.

Chinese explanation / 中文解读

中文解读待补充：本站会优先为大语言模型、生成式AI、ChatGPT相关技术、计算机视觉、深度学习等高价值论文补充中文说明。

Original abstract

For many years, proponents of open culture suggested that more liberal licenses were in line with their ethos. More restrictive licenses, it was felt, would restrict possible future reuses that could not be anticipated. For the past few years, one such "unanticipated" use has come to the fore: the training of large language models. This chapter theorises the unintended consequences of AI training on open-access cultures and views. While arguing that AI training is actually only a logical extension of the text and data mining that OA advocates promised, the danger foreseen here is that those who might otherwise make their work open will now retreat behind paywalls despite germinative “transformative use” court rulings in the US.

Links and sources

Comments