If you're interested in this concept, it's not new and the alarm has been sounded since the android Facebook app required motion sensor permissions in android 4.

https://par.nsf.gov/servlets/purl/10028982

https://arxiv.org/pdf/2109.13834.pdf

Something to note here that annoys me about the title is that the LLMs aren't taking in the raw data (LLM's are for text, after all). The raw data is fed through audio and motion models that then produce natural language descriptions, that are then fed to the LLM.

Unrelated: yeah, this article is a little creepy, but damn is it interesting technically.

AI will finally allow us to bring 1984's Telescreens into existence, at scale.