AI has a much wider range of applications than most people realise. For instance, IBM created the first fully synthetic singing voice using voice synthesis in 1961. In 2001: A Space Odyssey was created with the same technology.

There are numerous other types of sound synthesis that have been around for a long time. Using machine learning techniques, the Kurzweil synthesiser was the first to successfully reproduce piano sounds in the early 1980s.

Thanks to developments in machine learning, there is now a wide variety of much more sophisticated audio and visual special effects available. It's not as bizarre of a concept as you think it is. It's already prevalent, and it's just adding to the industry.