IMHO . . . this is a bunch of crap !
First and foremost is the question of who gets to define what AI is for bots . . . next is how we define "sentient" !
A robot will ALWAYS and only have the "knowledge" that a human programs it with !
If it is programmed to kill . . . it will kill !
And will they make them waterproof . . . so they can rescue wetbacks trying to cross the Rio Grande ?
And light enough not to sink while at it ?