I follow, but what exactly is the real privacy threat here? If we look at an echo for instance: * The device is hardwired to light up whenever it is listening

I follow, but what exactly is the real privacy threat here? If we look at an echo for instance: * The device is hardwired to light up whenever it is listening so you can't be eavesdropped on without some indicator. * The device only transmits when it is in listening mode. This has been validated independently by multiple people. * The device has a mute function (granted, software). * The device can be registered under a bogus account. * You can turn off all phone call capabilities / external listening capabilities. * You can pihole it so it won't reach back to Amazon's ad services. * You can delete any conversation you don't want stored manually, including one click deleting your entire history. * You can have all content automatically deleted after X days. So what exactly are people concerned about? Out of all the convos someone could have, a negative one just happens to probabilistically be recorded, you don't notice, and within 90 days someone in law enforcement or entity X figures it out and uses it against you? Someone at Amazon decides to look through your stored conversations and sees you turn on the lights at 5 PM each day? I'm not saying there are 0 privacy implications with this device, as any device that has a mic and can record remotely has privacy implications. I just feel like the level of angst generated by smart speakers isn't warranted in the grander scheme of things.