It's unethical (and in many cases illegal) to record and analyze other people's conversations without consent. The intended usage is to only record your own speech but it inevitably leads to compromising others as well.
If I had control over the recordings, I could use it for "good" (useful to me) causes. But I can't make any guarantees as to what other people will do with conversations recorded in public.
1984-esque start-up idea: pay people pennies to forward your company the recordings of their surroundings. Make profiles and sell "anonimized" datasets of the interactions.
Devices such as these and "smart speakers" are sold at a loss because their lifetime value to the vendor is in gathering and analyzing data. They are attacked more emphatically than smart phones by privacy enthusiasts because they are so grossly unbalanced when comparing end-user value versus their cynical treatment of users' data interests. A smart phone can have its "smart agent" features disabled and remain a highly valuable device to its user. But if you take away the smart agent feature from these more recent "smart" devices and you're left with an inert brick.
The only company I'd even think about maybe-sorta-kinda trusting with this kind of device would be Apple, and that's only because they've put a clear priority on making more and more stuff happen on-device and on not selling info on what happens off-device.
Speech as a means of controlling a device also seems down right stupid. Personally I don't see it being superior to a keyboard in any form. Certainly there are scenarios where voice commands are the better/safer choice, but again, those cases are extremely few and not something I encounter.
Honestly what's use case for Echo Frame, who seriously needs this?
While I might be willing to try working on support for a new device provided its interfaces are documented, I have little desire to buy an expensive new device knowing I'd have to reverse-engineer the thing before being able to even use it.