Hollywood dialogue is being used to train AI systems, according to research from The Atlantic. Companies like Apple and Anthropic have used more than 53,000 films and 85,000 TV episodes, including dialogue from well-known works like The Godfather and Alf. This information comes from the website OpenSubtitles.org, where users upload subtitles. These subtitles are valuable for training AI because they represent spoken conversation. Despite the use of this data for AI development, there is little transparency and the legality of using copyrighted work remains an open question.

read more (theatlantic) »