I'm happy you appreciate this, Joris. The issue of alignment is one dear to Prof. Stuart Russel, esp. in his book "Human Compatible." What I'm trying to get across plainly is that AI is a tool for thinking. It is capable of amplification of all thoughts: good or bad... and as many wiser than me have pointed out, the line dividing those two runs across every human heart.

Bad thinking is detrimental to individuals and groups... This is why AI, a powerful amplifier of good and bad, is both exciting and terrifying. And also, as numerous thinkers have already stated throughout history of human culture, thinking badly is one of the most dangerous application of the power of thought... And here is the catch: you cannot fight bad thinking by not thinking at all... It's like throwing ignorance to tackle logical fallacy and expecting justice will prevail!

BTW, I appreciate you comment and it gives me hope that maybe reason still has a chance to save our species without killing our curiosity in the process by insisting we are only allowed to be pragmatic! I think Socrates would have much to say about that...