Real question here: As I understand, voice (sound) recognition technology mostly uses hidden Markov models, instead of LLM/AI. Do I misunderstand? I think thi

Real question here: As I understand, voice (sound) recognition technology mostly uses hidden Markov models, instead of LLM/AI. Do I misunderstand? I think this is how Dragon NaturallySpeaking and Shazam works. To be clear: I do not agree with any of Samsung's and LG's efforts to record sounds/words from user homes!