AI Safety advocates are *not* really taking seriously the possibility of AI super intelligence. If you do really think that this is a danger, their main policy recommendations may make us much *less* safe. I explain why in this discussion:

→ View original post on X — @jeremyphoward