Not sure if I’m just impatient or if it conceptually won’t work as a universal jamming solution. Basically I wanted to have a jammer so that when you are in a call you can be sort of sure that the other party is not transcribing without your consent. More accurately ensure that their transcribing results in very poor results. The following code base has a failed AI based experiment for this:
https://github.com/DiabloHorn/jamscriber
It does *sort of work*, however I don’t have the patience to tune it properly since that is a lot of manual work. Normally failed projects are not published that much, I thought let me do exactly that. Who knows, someone might benefit from it.
