I think the reason Stephen Wolfram is using "human-aligned" is because that is the only feedback loop that is training the models. If you haven't read "What is

I think the reason Stephen Wolfram is using "human-aligned" is because that is the only feedback loop that is training the models. If you haven't read "What is it like to be a bat?" by Thomas Nagel, I highly recommend it.