By incorporating the Tkinter module, the Speech_Recognition module, and subprocess module; anyone can make their own miniture Siri or Cortana. Of course you might not want to replicate them, you know, for copyright issues.
The AI is a fairly simple process, and as long as it follows this template it'll work (it will need to be coded and debugged):
1: Get user input
2: compare input to reference sheet
3: analyze reference sheet for corresponding output
4: output result