next page (older posts) →
Eliezer Yudkowsky (Eliezer_Yudkowsky) · 2007-02-25T00:39:33.000Z · score: 71 (66 votes) · comments (77)
Eliezer Yudkowsky (Eliezer_Yudkowsky) · 2007-02-18T21:23:00.000Z · score: 119 (98 votes) · comments (231)
I was trying to convey the same problem, although the underlying issue has much broader implications. Apparently johnswentworth [LW · GW] is trying to solve a related problem but I'm currently not up to date with his posts so I can't vouch for the quality. Being able to quantify empirical differences would solve a lot of different philosophical problems in one fell swoop, so that might be something I should look into for my masters degree.self-embedded-agent on [Reference request] Can Love be Explained?
I mean romantic attraction.mathieuroy on When a status symbol loses its plausible deniability, how much power does it lose?
Maybe. But I'm a bit skeptical. I have the impression there are more time-efficient ways to meet high-status people.jameson-quinn on Has anyone written up a consideration of Downs's "Paradox of Voting" from the perspective of MIRI-ish decision theories (UDT, FDT, or even just EDT)?
That book is from 2006. I understand that it deals with the Paradox of Voting, but does it have anything that would be directly relevant to considering it in light of "acausal decision theories"? As far as I know, such theories pretty much didn't exist back then.klen-salubri on Harry Potter and methods of rationality alternative ending.
I c. Thank youklen-salubri on Harry Potter and methods of rationality alternative ending.
Yeah, on second thought I think you're right. He already had a hard time interacting with them before the ending; now I imagine the gulf will be too wide to overcome. But I do support him telling Draco -- the choice was either tell draco + obliviation or keep it a secret. Harry also went to extraordinary lengths to bring Hermione back, when everyone else, even Quirrel, told him he should move on. That's the kind of friend I want in my life.
Obliviation, right. Does anyone ask Draco where he was and what he was doing? I mean, don't think that this spell can actually work, or, if it does, should cause unexpected side effects. Probably, appearing of fake memories which will dwell like ghosts in subconsciousness unable to integrate into a chain of causes and consequences until the moment victim would figure out what have happen or decide to overcome fear and gaze into the eyes of truth. Imagine you had zero apples, then one apple fell on your head, which led to a loss of memory. You wake up and find yourself with an apple in your arms. Will you start investigation or simply eat an apple? Or you find an apple tree, then you decide to erase your memories so no one will find an apple tree until the moment people learn how to preserve and cultivate trees. Yet you take one apple with you. I guess, in both cases it is just a matter of time when your memories will be logically reconstructed. "rien ne se perd, rien ne se cree, tout se transforme".
I'm not sure that i'm the right kind of a person you need, nor can't guarantee that i always be there when you need me (life is full of unexpected stuff and i live in pretty dangerous and unstable place), yet i could try to be your friend.
But he abandoned that plan after killing Hermione, because he heard a prophecy (narrated to us in the Defense Professor interlude)
Troll did. Interesting that Dumbledore offered Hermione to try to be a hero herself, yet forgot to mention that Harry have several powerful "white" mages to cover his back and most powerful "black" mage to guide him, teach how to protect himself, how to surrender saving dignity, retreat when it is necessary and keep emotions under control. Also, i can't remember clearly, have Voldemort killed anyone during the book?
He freaked out because the last prophecy saw him end up in horcruxes for ten years. He decided he had to stop Harry, because otherwise he would "end" the world. The most likely interpretation is a transhumanist one, but Voldemort doesn't realize that and thinks Harry will just kill everyone and destroy everything.
By "transhumanist" you mean something like denying death as essential part of birth-to-death cycle? Agree, most probably it will ruin natural cycle of life, not sure about "end of the world". Also it could mean like getting everything under full control to exclude possibility of unexpected events, or “reversal of time” to fix previous errors, or going with his nano researches to deep to the places where flux of information will be too powerful for his brains to handle, or any other drastic change to the order of things his dark ego will decide to make without complete awareness and evaluation of consequences for closed ecosystem and universe he lives in. For similar reason getting rid of Dementors could be not that good idea if we imagine that they are projection of fear of death and oblivion, since this fear could be one of the moving forces for creativity (Parkinson's law, you know. The more time you have to do the job, the more time you will spend to do it).
Voldemort's evil, but he doesn't want that.
People often call "evil" something they can't control or understand. Precisely, they call something "evil" if it takes something valuable away and call it "miracle" if it adds something valuable.
When your body dies, your mind dies with it
Yep. But strictly in terms of theorizing, can we consider the body and brain as a capacitor which stores some type of energy with unique properties? Can combination of this proberties store some information? What will happen when capacitor is destroyed, will this cause some effect to surrounding field?slider on Dynamic inconsistency of the stepwise inaction baseline
I got confused on what is the reward scheme [quote] It gets a reward of r>0 for standing on the blue button for the first time. [/quote] would read to me that blue button gives a reward one time when stepped on but red button does nothing. The story of the post seems to intend that first stepping on the red button will prevent the blue button from giving out any rewards. "blue button is the first button pressed" vs "what happens when blue button is entered for the first time"jameson-quinn on Has anyone written up a consideration of Downs's "Paradox of Voting" from the perspective of MIRI-ish decision theories (UDT, FDT, or even just EDT)?
"You play the game with many other CDT agents" — this seems demonstrably false, at least, if we accept the Paradox of Voting as being a thing, in which case, CDT agents have by assumption removed themselves from the game. (I understand your response that voting may be altruistically-CDT-rational; as you know, it's been discussed before, and very rightly so. But I also think it's still worth considering the boundedly-altruistic/diagonally-dominant case.)
It seems to me that the only way you can claim there's "many other CDT agents" is if "CDT" is being used as a catch-all for "not explicitly FDT/UDT", and I'd strongly dispute that usage. I think that memetically/genetically evolved heuristics are likely to differ systematically from CDT. It may be best to create an entirely separate model for people operating under such heuristics, but if you want to force them into a pure CDT-vs-UDT-vs-random-noise (ie, mixture distribution) paradigm, I'd say they would be substantially more than 0% UDT.
ETA: I guess I can parse "other voters are CDT" as a sensible assumption if you're explicitly doing repeated-game analysis, but such an analysis would pretty much dissolve both the Paradox of Voting and the CDT vs. acausal-DTs distinction.turntrout on Dynamic inconsistency of the stepwise inaction baseline
Nice post! I think this notion of time-inconsistency points to a key problem in impact measurement, and if we could solve it (without backtracking on other problems, like interference/offsetting), we would be a lot closer to dealing with subagent issues.
I think the other baselines can also induce time-inconsistent behavior, for the same reason: if reaching the main goal has a side effect of allowing the agent to better achieve the auxiliary goal (compared to starting state / inaction / stepwise inaction), the agent is willing to pay a small amount to restrict its later capabilities. Sometimes this is even a good thing - the agent might "pay" by increasing its power in a very specialized and narrow manner, instead of gaining power in general, and we want that.
Here are some technical quibbles which don't affect the conclusion (yay).
If using an inaction rollout of length l, just multiply that penalty by γl
I don't think so - the inaction rollout formulation (as I think of it) compares the optimal value after taking action a and waiting for N−1 steps, with the optimal value after N steps of waiting. There's no additional discount there.
Fortunately, when summing up the penalties, you sum terms like …p|γn−1−γn|+p|γn−γn+1|…, so a lot of the terms cancel.
Why do the absolute values cancel?tag on Situating LessWrong in contemporary philosophy: An interview with Jon Livengood
But past mathematicians already just taught what they thought was true then.
But we don't know that we are living in the optimal timeline. Maybe relativity would have arrived sooner with fewer people in the past insisting that space is necessarily Euclidean.
I’m asking what relevance you think it has for current math education.
The topic is philosophy education. Science can test its theories empirically. Philosophy can't. Mathematics can take its axioms for granted. Philosophy can't.
As it is said, keep an open mind, but not so open your brain falls out. Teaching a specific thing impedes progress when that thing is wrong or useless, but it aids progress when that thing is a foundation for later good things.
The difficulty is that we don't have certain knowledge of what is in fact right or wrong: we have to use something like popularity or consensus as a substitute for "right".
It may well be the case that one can go too far in teaching unpopular ideas, but it doesn't follow that the optimal approach is to teach only "right" ideas, because that means teaching only the current consensus, and the consensus sometimes needs to be overthrown.
The optimal point is usually not an extreme, or otherwise easy to find.