I misunderstood you as saying that it was a true account of reality (and not merely useful in practice). I guess I am missing your point if that is not the case?
-
-
Replying to @Meaningness @juliagalef
I'm missing your point because I don't understand what it means for decision theory to be a "true account of reality", and hence I can neither confirm nor deny that I believe this to be the case.
2 replies 0 retweets 0 likes -
Replying to @ESYudkowsky @juliagalef
Oh well. I guess we’re both talking past each other. It’s a bit puzzling. I can usually eventually understand other people’s worldviews, but I find yours unusually resistant :)
2 replies 0 retweets 1 like -
(?) It sounds like 'true theory of reality' is something that could be occupied only by a final theory of physics. Decision theory just tells you the objectively best way of acting to satisfy a set of preferences.
1 reply 0 retweets 0 likes -
“Objectively best” assuming a set of axioms is satisfied. There can sometimes be a bait-and-switch or motte-and-bailey here when you try to apply this to a concrete real-world situation.
1 reply 0 retweets 2 likes -
Do you have a specific example in mind?
1 reply 0 retweets 0 likes -
Well, to get decision theory to apply, you have to characterize the situation in terms of a set of well-defined actions, well-defined outcomes, well-defined goodnesses, and you need some meaningful way of estimating probabilities. None of those are objectively given.
2 replies 0 retweets 0 likes -
Replying to @Meaningness @ArtirKel and
Writing this tweet, I have an unbounded number of possible things to say; an inconceivable set of possible outcomes; no clearly-defined goals; and any numerical probabilites would be meaningless. Do you know about aardvark cucumbers?
2 replies 0 retweets 0 likes -
That is, an epistemic, not ontological problem. One can hold that 1)DT is the best way of deciding, objectively 2)DT cannot be applied in its textbook form, just approximated As GA Cohen said, just because one cannot reach some tasty grapes doesn't mean they are less tasty
3 replies 0 retweets 0 likes -
If you define “best way of deciding” simply as “deciding according to what DT says” then that is the best way of deciding. But if no physically realizable agent can do that, this is meaningless.
1 reply 0 retweets 0 likes
It becomes meaningful only if you can argue that approximating DT is both (1) practical and (2) approximately optimal according to some criterion other than “approximates DT.” That is true in some situations, and not in others.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.