Comment to Elon Musk – My Final Warning
[ 21/April/21 ]
Maximizing freedom is self terminating.
What is needed is to maximize freedom with minimal risk to life.
Every level of structure has necessary constraints required to support that level of structure.
Any “free” action that destroys a necessary constraint leads to self termination.
So simply maximizing freedom, without minimizing risk to life, self terminates.
Life has to be first priority, liberty second to it. Cannot sustain life long term with anything less.
Ted Howard NZ's Blog 
I also noticed that. I thought about maximize freedom, but then realized an AI built with that as the sole basis for its utility function could quite easily reach the conclusion of creating anarchy that, without considering life, leads to everyone dying explosively (and even anarchists generally have some boundaries, which would be necessarily be adults with freedom at least in the short term).
I think I heard or read somewhere (though, if I may say so, I’m fond of how I’m paraphrasing it here) that neither perfect security nor perfect freedom can exist without the annihilation of the other. Since both concepts have their place in the balancing act that is life… they both need to be accounted for in any good utility function.
LikeLiked by 1 person