"A rat isn't exactly seeking an optimum level of food, it's seeking an optimum ratio of ventromedial to ventrolateral hypothalamic stimulation, or, in rat terms, a nice, well-fed feeling."

So if I move my hand away from a hot pan, am I actually seeking to: "move my hand away from a hot pan" or

"avoid touching the pan" or

"avoid being burnt" or

"avoid pain receptors in my hand being activated" or

"avoid neural signals in my brain that correspond to pain" or

"avoid the feeling of pain"?

Someone needs to do some buck-stopping or else the master-slave model will turn into a master-slave1-slave2-slave3... model. Although come to think of it, that might me more correct. (EDIT: Note to self, line spacing is weird, I'm off to look in the wiki)

If someone starts a post with "Rationality is wrong..." or similar, I'm much more likely to downvote it than if they start it with "I've got these scenarios where standard rationality techniques seem not to work..." To this extent at least the presentation of the ideas matters as much as the content. So I hope that these rules will cause people to present their ideas more cautiously, while still posting experimentally. If you are thinking of posting something controversial, it might be worth seeking advice from the other users who you think will be interested.

Basically, I think downvotes should work as an "utter stupidity" filter.

Being a science and maths geek, I've tended to dismiss a lot of philosophy as bullshit, and have only recently begun to realise that (some of) what I've dismissed is actually valid and interesting. Of course, one place where this effect is incredibly strong is when a parent is arguing with a child, i.e. the "teenagers think they know everything" syndrome.

Surely that only works if the probability of winning a case depends only on the skill of the lawyers, and not on the actual facts of the cases. I imagine a lawyer with no training at all could unravel your plan and make it clear that your hobos had nothing to back up their case.

Also, being English myself, it hadn't dawned on me that the losers-pay rule doesn't apply everywhere. Having no such system at all seems really stupid.

It also occurs to me that hiring expensive lawyers under losers-pay is like trying to fix a futarchy: you don't lose anything if you succeeded, but you stand to lose a lot if you fail.

I imagine by declaring that they would power up the LHC iff the Green party won, thus forcing everyone who would vote Blue to come down with a fever on election day.

Scott Aaronson has a nice post about the differences between gravity and electromagnetism. It seems his thoughts were running along the same lines as yours when he wrote it; he asks almost all the same questions. http://www.scottaaronson.com/blog/?p=244

Is it equivalent to state that leaning about the origins of your priors "screens off" the priors themselves?

In the same way, it's hopeless to try to assign probabilities to events and do a Bayesian update on everything. But you can still take advice from theorems like "Conservation of expected evidence" and the like. Formalisations might not be good for specifics, but they're good for telling you if you're going wrong in some more general manner.