Do you think TDT/FDT imply magic is real?

May 10, 2021

Leaving "magic" as a pretheoretic pointer doesn't get you unbiased results, it makes your question incoherent. You have to tell me what you mean by "magic" before I can attempt an answer.

If you leave it to me, then I will define magic as "Humans doing things that violate the laws of physics as we know them", in which case the answer is trivially "No, TDT/FDT do not imply that magic is real."

gjm

May 10, 2021

TDT and FDT don't state any propositions, they prescribe behaviours. So they could only "imply magic is real" in three ways that I can see.

They could require you to believe some proposition close to "magic is real".
They could require you to do things that, if done by someone using a more "normal" decision theory, would be a clear indication that they think magic is real.
They could presuppose something close to "magic is real".

I am not an expert on either TDT or FDT, and my understanding is that neither is actually well enough defined to be very sure about exactly what they require or presuppose, but I am fairly sure that neither 1 nor 3 is the case.

2 might be; e.g., allegedly TDT/FDT one-boxes on Newcomb, and you might say "one-boxing on Newcomb implies believing in causation that reaches backwards through time, and that would be magic". (Maybe this is a bad example; to my mind, the thing in Newcomb that kinda-implies backward causation is the original setup, not any particular decision. Perhaps something involving "acausal trade" or basilisks is a better example, but I don't understand those well enough to be sure.) But I don't think this sort of thing justifies TDT/FDT practitioners if they say that magic is real, nor justifies other people if they say TDT/FDT practitioners think magic is real. If you follow FDT, and reckon it tells you to one-box on Newcomb, and then do so when actually confronted with a Newcomb-like situation, at no point does your thought process involve contemplating anything I would regard as "magic".

kithpendragon

May 10, 2021

"Magic" is the part of a system that you don't have a gears level understanding of.

https://www.lesswrong.com/posts/kpRSCH7ALLcb6ucWM/say-not-complexity https://www.lesswrong.com/posts/B7P97C27rvHPz3s9B/gears-in-understanding

What do your acronyms mean?

[-]emmab5y30

https://www.lesswrong.com/tag/timeless-decision-theory
https://www.lesswrong.com/tag/functional-decision-theory

1kithpendragon5y

Thanks! 3-letter acronyms are really hard to Google for clarification.

5ChristianKl5y

LessWrong's search helps with searching for LessWrong jargon.

1kithpendragon5y

Haha, so it does! Good call!

[-]emmab5y10

So what does "is magic real?", as a child might ask the question, correspond to?

3kithpendragon5y

Depends on the context. If the child is asking... * ... about mythological stories being true, the answer is that they are not factual, but rather correspond to important aspects of the culture. * ... about the ability to affect things without apparently interacting with them, the answer is to point at the TV remote. [My kid still doesn't believe me that talking and writing are psychic powers, but that's the answer I'm giving on that front.] * ... if things happen in the world that are not trivial to understand or control, the answer is simply "Yes". [If my kid asked "Is lightning magic," I would have to answer "Yes" because I cannot make lightning do what I want: as far as I'm concerned it "just happens" when conditions are exactly right. As far as I've been able to find out, nobody fully understands it.] * ... if actions can have strongly unexpected results, of course they can! * ... if waving my hand was the proximate cause for X happening, I'd have to admit most of the time that it was just good timing on my part. But I can wave my hand at my cell phone to have it snap a selfie, and if I have an NFC chip I can sometimes wave my hand to tell a computer what to do. In general, I think "magic" is best used to point at a state of non-understanding. I don't think the "is it real" query is really thinking about it in the right category. It's a lot like asking, "Is a binary search tree real?" or "Is the color red real?" All three are data processing questions of one sort or another; but the color red, binary search trees, and magic all don't directly talk about anything in the territory. Rather: "red" is a perception, "binary search tree" is a data structure, and "magic" indicates non-understanding. In terms of decision theory, either the "magic" process gives known or unknown outputs for specific inputs. If the outputs are known, as in "When I twist the fluff in this particular way I get thread", the gears might not matter for your purposes. If unknown, as in

LESSWRONG
LW

LESSWRONG
LW

-15

[ Question ]

Do you think TDT/FDT imply magic is real?

-15

-15

3 Answers sorted by
top scoring

May 10, 2021

May 10, 2021

May 10, 2021

-15

[ Question ]

Do you think TDT/FDT imply magic is real?

-15

-15

3 Answers sorted by top scoring

May 10, 2021

May 10, 2021

May 10, 2021

3 Answers sorted by
top scoring