"I'm continually amazed that while MS could apparently get it right on a sclerotic ARM core with sod-all memory,"
My Motorola Defy from years back had voice control. Only worked with Bluetooth but it could do a number of things and was quite accurate. All within the phone, all about as powerful as the original Pi.
My Samsung has loads more memory, loads more cores, loads more megahertz, and a voice assistant that can work any time. But I refuse to use it as it is utterly incapable of doing anything for itself. The first use requires you to agree to a bunch of T&C for having your voice data processed by Nuance. Why? Oh, I can understand if you ask "when is the next train from Paddington with first class carriages?" then it might need to do some work; but why can't it so "what's the time" or "call [name]" for itself?
Seems like we're going backwards - perhaps because data grabbing and profiling is the more important thing these days?
Biting the hand that feeds IT
