I've noticed that there are 2 flavors of the alignment problem. One is about the technical how to and practical engineering, the other is about humanities, social sciences, human behavior and psychology.
What they both have in common is doomsday scenarios about paperclip maximizers.
I don't know about you, but personally I don't care if an unaligned AI is aware of its experience while turning the universe into a paperclip, or if the whole thing is happening entirely without consciousness. What i care about is: How can we make AI smart enough to prevent that from happening?
That said, I find especially in amateur circles that CONSCIENCE is not just omitted from the discussion, it's actually poorly understood.
Arguably, if we find an algorithm for empathy, we can solve the alignment problem rather easily. Humans seem to be capable of monitoring their actions for morality and preventing harm to others, how hard can it be to build an artificial brain that can do the same?
How would you explain to someone what the difference is between consciousness and conscience?
The problem is that the AI can (and does) lie. Right now, ChatGPT and its ilk are a less than superhuman levels of intelligence, so we can catch their lies. But when a superhuman AI starts lying to you, how does one correct for that? If a superhuman AI starts veering off in a direction that is unexpected, how does one bring it back on track?
@gwern short story, Clippy highlights many of the issues with naively training a superintelligent algorithm on human-generated data and expecting that algorithm to pick up human values as a result. Another post to consider is The Waluigi Effect, which raises the possibility that the more you train an agent to say correct, inoffensive things, the more you've also trained a shadow-agent to say incorrect, offensive things.