The problem really, is that people want to believe. So it doesn't matter how many PhDs pile in to explain that "no, this isn't sentience, it's just statistical

The problem really, is that people want to believe. So it doesn't matter how many PhDs pile in to explain that "no, this isn't sentience, it's just statistical modelling", a whole class of people (including very smart non-domain-experts) go "yes but... <insert opinion/hope/imagination>", and the whole thing hangs up on semantics. Google firing this guy simply stokes the flames.