This actually might be crazy, but worth it.
Almost every operating system has a tts-voice assistant. Although we have screen reader, its not like Cortana where you could say “Hey Cortana” and then it will listen.
However, using a Gnome Shell extension, it might be a good idea to think about having this. For example,
Person: “Hey Ubuntu, play KIIS FM on odio (snap package)”
Ubuntu: “Now playing 102.7 KIIS FM (L.A.'s #1 Hit Music Station) on iHeartRadio on odio.”
Thats just one example. But, it’s something that would take a year to develop, so now is a time to think if we want one by Ubuntu 20.04 assuming we will still use gnome shell.
Take a look at:
- Existing Linux-based, license-compatible voice-recognition software
- It’s current quality for users (compared to existing proprietary competitors)
- The success/failure of previous voice-recognition projects. (example)
You’ll quickly see why this idea seems unlikely to get much traction for '20.
Remember that Ubuntu’s focus is on distributing upstream software. If you find some good upstreams that could be packaged for Debian and Ubuntu (or snappified for Ubuntu), please do let everyone know.
Alternately, if you discover a simple integration problem preventing several otherwise-compatible upstreams from working together to provide this functionality, do whistle.
Well, would it be legal if Siri would be implemented into a snap package? Or could Mac sue people for that?
Do you really think that there is a Linux version of Siri existing? I highly doubt that.
Or do you think that a snap is a kind of emulator? Which it is definitely not.
If you could find Siri source code in the wild without Intellectual Property encumbrances,
…and if you could get it to run under Linux,
…and if it had a compatible license,
then sure, yes, indeed. Folks would race to package it for Debian and Ubuntu and every other distro too.
However, that particular heffalump seems very unlikely to be encountered in this hundred-acre wood.