Yesterday, six authors filed new individual copyright infringement actions against Anthropic, OpenAI, Google, Meta, xAI, and Perplexity AI. The suits, which were filed in the Northern District of California, states the companies copied authors’ books from well-known pirate libraries—including LibGen, Z-Library, and OceanofPDF—to train their large language models without permission, licensing, or compensation.
I will admit to experiencing a great deal of dread about AI in general, but especially for the outright, and what appear to be quite cavalier infringements and violations of copyright. I won’t ask how we got to this dangerous point in time and practice, but would genuinely welcome a nugget of hope if anyone can toss one out. 
I wish I had something positive to say about it, @Sunny, but a professor recently did an experiment whereby he fed his AI a bunch of high-quality literature, then gave the AI a scenario. He asked it to write in a specific author’s style. It ended up that his class preferred the AI’s version in a blind test.
I have to say I’m not surprised. Having worked in public libraries for 35+ years, I had a daily, front row seat for witnessing the slow, but inexorable shrinking of the ability of young people (students) to concentrate, to focus, to truly and deeply appreciate the beauty of lyrical writing, the magic of well-developed characters, the intensity and allure of magnetic settings, and the breathless desire to discuss what they had read! Of course there are exceptions, thankfully so. And perhaps it is there–in the existence of the exceptions–where I shall find refuge.
1 Like