Okay, so they're showing off the ability for a system to analyze a voice sample and mimic the voice it 'heard' with a reasonable level of fidelity. Intellectual

Okay, so they're showing off the ability for a system to analyze a voice sample and mimic the voice it 'heard' with a reasonable level of fidelity. Intellectually interesting, has some potentially good use cases (IVR call trees, more distinct virtual assistant 'personalities', voice interfaces that seem 'familiar' to dementia patients and the like...) But they choose to present it in the context of perhaps the most ghoulish use case possible? And not only a use case that's entangled with taboos surrounding death and respect for the dead, but one that's been openly lampooned in popular media franchises like Black Mirror? I'm failing to comprehend why this was the chosen approach. Perhaps because it's controversial enough to distract from the implications for more actively malevolent uses of deepfakery?