I generally loathe voice input on computers & phones. I have very specific use cases for the two assistants I've created. "Open the pod bay doors Hal" = open th

I generally loathe voice input on computers & phones. I have very specific use cases for the two assistants I've created. "Open the pod bay doors Hal" = open the garage door (but only because it is funny). "Start the dust extractor Hal" = start the workshop dust extractor, but I no longer use this since upgrading to remote start/power line detection on my dust extractor. "Hey producer, <various commands to control video production software and PTZ cameras>" = start recording/cut cameras/re read the prompt/extreme close-up on keyboard/pan the camera to me/etc for video production in a one-man multi-camera recording setup. Everything else, when it comes to NLP, chat bots and voice recognition is "too much trouble" and I find hitting a dedicated button or punching in to a few numbers on a physical keypad to be easier and more reliable than any voice interface.