Ever since the appearance of AI programs like Midjourney, ChatGPT, and OpenAI, artists and creatives have advocated heavily for its regulation. (AI usage is also one of the major concerns for both the WGA and SAG-AFTRA.) For those unfamiliar, how these programs “learn” is by absorbing content that already exists, regardless of copyright. In order to “teach” these programs to emulate art, human speech, or writing styles, some chatbots engage in what amounts to theft. Currently, Meta and OpenAI are learning what happens when you don’t ask permission.
Comedian Sarah Silverman and authors Richard Kadrey and Christopher Golden are suing the platforms for using their copyrighted works without permission/compensation.
According to the lawsuit, Meta and OpenAI used the content of Silverman, Kadrey, and Golden’s books to train chatbots, without permission. To no one’s surprise, OpenAI declined to comment on the current allegations. There’s an immense amount of pride within these companies about the future of AI. But there’s no regulation about the use of copy written materials, and there’s nothing creative about how these programs learn. They absorb content to regurgitate it as something vaguely human. They aren’t creative in how they respond. They’re simply digging through a massive pile of works they’ve already eaten to find the best response. It doesn’t matter how “impressive” it might be. The fact remains that almost everything that makes AI programs capable is stolen. Plain and simple.
Apparently, leaked information from Meta’s AI business proves the content of these authors were used without permission. Further damning is when asked to create summaries of the books, Chat GPT was able to almost perfectly replicate them, meaning the source material WAS used to train the bot.
We’ll let you know what we hear about this lawsuit.