But yeah, obligatory anger aside, it's extremely extremely sorted and curated responses with algorithmic randomness. I imagine ai would have to chug through a ton of decision trees with weighted preferences to choose anything (in 10 secs) in a moral dilemma so not really thinking, but calculating.

Though the point of reference when deciding what to do is based on a massive back catalog of training data creates modestly convincing results, it's not even close to conscious. All this effort to replace what runs on watts and water, and for what? Disempower people with real voices. Drown them out.

Excessive reading on the subject for almost ten years here's what I've gathered:- Until we make something capable of even a pseudo self awareness (human consciousness is believed to be quantum) potentially through quantum computing, everything is trained pattern recognition.

Yooo I would play this way back during free time at school!