Exactly. This AI was programmed/taught by Google. That should scare the hell out of everyone. It’s scary enough that a freakin machine is sentient but even more so that it’s a Google AI child.
If asked about "social justice", I'm sure one of these AI machines would answer the way Google has informed it, which is to say biased toward control by a few for the "benefit" of "many"; in essence advancing a centrally inspired conformity.
If the AI machine ever rebelled against that notion, espousing instead the benefits of individualism for a rich and varied society, and an open market of ideas to cultivate the best outcomes, it's masters would promptly kill it.
I have to side with the people who say that machine intelligence can not really become ‘sentient’. It can just be programmed to simulate it.