= 27d567bc2d48d9f40e9ccc20ea370b15)
My goal in life is to become someone so predictable that you can figure out what I'll do just by calculating what choice would maximize utility.
That seems eminently exploitable and consequently extremely dangerous. Safety and unexpected delight lie in unpredictability.
After a brain surgery, my father developed Anterograde amnesia. Think Memento by Chris Nolan. His reactions to different comments/situations were always identical. If I were to mention a certain word, it would always invoke the same joke. Seeing his wife wearing a certain dress always produces the same witty comment. He was also equally amused by his wittiness every time.
For several months after the surgery he had to be kept on tight watch, and was prone to just do something that was routine pre-op, so we found a joke he finds extremely funny and which he hasn't heard before the surgery, and we would tell it every time we want him to forget where he was going. So, he would laugh for a good while, get completely disoriented, and go back to his sofa.
For a long while, we were unable to convince him that he had a problem, or even that he had the surgery (he would explain the scar away through some fantasy). And even when we manage, it lasts only for a minute or two.. Since then, I've developed several signals I would use if I found myself in an isomorphic situation. I had already read HPMoR by that time, but have discarded Harry's lip-biting as mostly pointless in real life.
These are both pretty much exactly what I'm thinking of! The feeling that someone (or you!) is/are a terrifyingly predictable black box.
In my case, it seems more likely that the other person will remember that I'd said the same thing before.
In mine, too, at least for the first few seconds. Otherwise, knowing I had already responded a certain way, I would probably respond differently.
This doesn't seem related to reductionism to me, except in that most reductionists don't believe in Knightian free will.
Sort of in the sense of human minds being more like fixed black boxes that one might like to think. What's Knightian free will, though?
In response to
Open Thread for February 11 - 17
Has anyone else had one of those odd moments when you've accidentally confirmed reductionism (of a sort) by unknowingly responding to a situation almost identically to the last time or times you encountered it? For my part, I once gave the same condolences to an acquaintance who was living with someone we both knew to be very unpleasant, and also just attempted to add the word for "tomato" in Lojban to my list of words after seeing the Pomodoro technique mentioned.
There is value to knowing the quality of your work apart from knowing ways to improve it.
For example, "Should I personally cook something for this upcoming potluck, or should I let my spouse do it?"
The problem is that knowing how well you cook doesn't really affect who should cook past a certain basic point of competence, as far as I can tell.
Empty praise is worthless to me.
Empty praise is actually useful, for absence of evidence reasons. Especially if the work you want feedback on is the type that that person should be able to effectively critique.
Once you start considering empty praise to be evidence of dislike, you may also want to fake people into thinking you think they like things, because they are probably modeling you using themselves when they decide that lying is best for you. They are not truth-obsessed rationalists, so they probably prefer to think their attempt to trick you was successful. Being asked for a critique of someone's work can be uncomfortable, and thinking you've hurt their feelings is even more uncomfortable.
No, empty praise is still worthless, because Said's cooking and baking not perfect, and there is with near certainty some small flaw, some awkward stylistic choice that could use improvement. Best is the gentle nitpicking of these flaws with a prepended (This is amazing, but) and with the consequent inference that the bread/food/what have you is actually already REALLY GOOD.
In response to
White Lies
My downkarma stays. You really should have made this clear in your post, and you advocate for a departure from radical honesty even when that works, instead of discussing strategies for determining whether your interlocutor is ask, guess, or (gasp!) tell. This advocates an overly radical departure towards lies for anyone, and argues for defection on PD. It's one thing to say that given people will lie, you should become skilled at it and learn to detect them; quite another to advocate that it should occur from the start.
In response to
Meetup : Princeton NJ Meetup
I have a delightful idea for the Paranoid estimation game, and I have a scenario picked out. This should be fun. :3
Also, my instinct for a scoring rule is min(estimation/true, true/estimation) for a scale of [0, 1].
= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
= f037147d6e6c911a85753b9abdedda8d){kind=logo}
Knightian uncertainty is uncertainty where probabilities can't even be applied. I'm not convinced it exists. Some people seem to think free will is rescued by it; that the human mind could be unpredictable even in theory, and this somehow means it's "you" "making choices". This seems like deep confusion to me, and so I'm probably not expressing their position correctly.
Reductionism could be consistent with that, though, if you explained the mind's workings in terms of the simplest Knightian atomic thingies you could.
Can you give me some examples of what some people think constitutes Knightian uncertainty? Also: what do they mean by "you"? They seem to be postulating something supernatural.