“Some of it, like the genetic and fingerprint data and some of the eye stuff. Most of it is for calibrating the controls.”

Especially the voiceprint, since un-augmented humans typically send messages in the feed by subvocalizing.