Sounds fairly close to what I think Jablonski is saying, yes.
Preference isn’t the best word choice. Ultimately it comes down to realizing that I want different things at different times, but in English future wanting is sometimes hard to distinguish from present wanting, which can easily result in a subtle equivocation. This semantic slippage is injecting confusion into the discussion.
Perhaps we have all had the experience of thinking something like, “When 11pm rolls around, I want to want to go to sleep.” And it makes sense to ask, “How can I make it so that I want to go to sleep when 11pm rolls around?” Sure, I presently want to go to sleep early tonight, but will I want to then? How can I make sure I will want to? Such questions of pure personal long-term utility seem to exemplify Jablonksi’s definition of morality.
Amanojack has, I think, explained my meaning well. It may be useful to reduce down to physical brains and talk about actual computational facts (i.e. utility function) that lead to behavior rather than use the slippery words “want” or “preference”.
Sounds fairly close to what I think Jablonski is saying, yes.
Preference isn’t the best word choice. Ultimately it comes down to realizing that I want different things at different times, but in English future wanting is sometimes hard to distinguish from present wanting, which can easily result in a subtle equivocation. This semantic slippage is injecting confusion into the discussion.
Perhaps we have all had the experience of thinking something like, “When 11pm rolls around, I want to want to go to sleep.” And it makes sense to ask, “How can I make it so that I want to go to sleep when 11pm rolls around?” Sure, I presently want to go to sleep early tonight, but will I want to then? How can I make sure I will want to? Such questions of pure personal long-term utility seem to exemplify Jablonksi’s definition of morality.
ok cool, replying to the original post then.
Oops, I totally missed this subthread.
Amanojack has, I think, explained my meaning well. It may be useful to reduce down to physical brains and talk about actual computational facts (i.e. utility function) that lead to behavior rather than use the slippery words “want” or “preference”.
Good idea. Like, “My present utility function calls for my future utility function to be such and such”?
I replied to Marius higher up in the thread with my efforts at preference-taboo.