I was actually hoping it could be paired with speech to text very well and help along with hearing aids when the latter do not perfectly work. There are legitimate use cases.
Not an expert, but my suspicion is that the camera following lips can add an extra streaming data point making transcription accuracy much higher even at low volumes. Again a hunch and I guess the computational power and battery needs might still be insurmountable
There are legitimate use cases and illegitimate ones. Unfortunately, I'm seeing more examples of the latter. I somehow suspect Mark's entourage aren't all hard of hearing.
