Earlier quoted context omitted.
Right, the fidelity is not that good, when just sampling arbitrary vibrating objects. For example a simple accelerometer attached to the center of a lamp shade might create an intelligible signal, but one attached to a heavy table might not. Remember, AI can build speech from a very noisy signal, however. To pick up a better signal you need either a large flat or concave object (like even a computer case), or else so…
Why not? You can't hide microcode or MEMS IMU elements or analog circuitry in a way that decapping and examination by experts cannot uncover. This very article shows acid etching the epoxy packages to inspect the die. There are experts like Ken Shirriff (kens on HN) who reverse engineer silicon devices at the transistor level. How do you arrive at the conclusion that microcode or software cannot be detected even by e…
I was just saying it's a practical impossibility for even an electronics expert and even after disassembling something, to detect if a device had been recording or not, just by examining the device physically.
Once you start reverse engineering the code itself you're likely to find any hidden surveillance, but even then not guaranteed, because sufficiently obfuscated code can camouflage what it's true purpose was.