Kevin Bacon, Kate McKinnon, and Other Creatives Warn About the 'Unfair' Threat of Artificial Intelligence.
Unauthorized training of artificial intelligence puts livelihoods at risk, they claim.
A large number of creatives, including well-known actors like Kevin Bacon and Kate McKinnon, as well as other actors, authors, and musicians, have signed a manifesto warning about the dangers of unauthorized use of copyright-protected materials to train artificial intelligence models. To date, the list of signatories includes 11,500 names.
The message expressed in a brief statement is clear: “The unlicensed use of creative works for training generative artificial intelligence represents an unfair and significant threat to the livelihoods of those behind these works, and it should not be allowed.” This statement was published by Fairly Trained, a group advocating for equitable use of training data by artificial intelligence companies.
Ed Newton-Rex, CEO of Fairly Trained, commented that generative AI companies require “people, computing, and data” to build their models. Although they invest “huge sums” in the first two, they seem to expect to obtain the third, i.e., the training data, for free. Newton-Rex founded Fairly Trained after leaving Stability AI, where he accused generative artificial intelligence of “exploiting creators.”
Various professionals and organizations, such as News Corp and the Recording Industry Association of America (RIAA), have sued AI companies for using protected works while training their models. The RIAA has joined the statement from Fairly Trained and has made its stance public, as has the News/Media Alliance. However, there are some notable figures who are not among the signatories; for example, Scarlett Johansson, who had a media conflict with OpenAI after accusations that GPT-4 was modeled to replicate her voice. Actors like Dame Judi Dench and John Cena also do not appear among the signatories, despite having signed for Meta AI's voice chat system to replicate them.
