Really nice to see improcements being made to local voice control.
I personally haven't opted for it yet due to the computing reqirements - instead been looking into dedicated offline voice recognition modules, since they use much less power. Downside is that they recognise a mostly predefined set of commands, and are just looking for patterns in speech instead of actually performing voice recognition
Would be interesting to compare the two voice control approaches side by side at some point
I've been eagerly looking forward to the time when I can replay my Echo Dots with a self-hosted solution, but so far I haven't found hardware that I really liked the look and style of.
That is really a missing part of this whole thing. I get that I could build something myself, but I don't want to have the hassle of doing it and keeping it working all the time (because I am able to build it, but not on a "works perfectly all the time" level). I really hope that sometime in the future there is a standard for smart speakers (and screens maybe) that allows me to add them to my cloud service of choice.
