They are able to make someone believe they have some kind of mental illness. They dub every time a sound is made by anything, or anything changes position, it will dub every time. They can even be scanning in 3D through walls at the person and watching them, a computer helps them dub the objects falling and making noises, or engines. They can produce voices from the sounds, by matching the time they happen using new technology only not, that new, or they can see in plain view, or on camera, security cameras, which is all combined in some AI program to see through all walls and see every thing that happens and DUB over it, then they can make a radio soundwave, which goes through walls too. It is real and terrible. I am pretty sure that if you whisper to yourself for one, you might be repeating something the radio is saying to you. Also the tiny jaw and tongue positioning can be scanned in 3D and some AI can read what you are saying and things like that. They spy and collect all sorts of data and then change recommended videos, and hack websites to show numbers and things they change to produce fear. Even dubbing humming, music notes, to say words, to make it sound like you are saying them while humming, and only using "La la la la la" will defeat that one technique.