Posts
Comments
Wearing the Patronus isn't any more dubious than casting it inside of Hermione to revive her. You're right about stunners instead of AKs of course, but that can be blocked by a thin invisible tranfigured shield (air into glass, since apparently he can transfigure arbitrary atomic structures). Transfiguration is wordless and he has a wand. I mean, this isn't anywhere near as deus-ex-machina as half of the Azkaban escape anyway.
"Expecto Patronum", at which point Death-Eaters will fire an utterly futile barrage of AKs. Voldy still can't fire directly at Harry due to resonance. Gun is not as much concern if you move fast enough and considering Voldy is some distance away. Gives Harry enough elbow room to get to his nearby (?) Pouch, Cloak and Time-Turner with 1 more hour on it. At this point we're sorta free of any serious constraints.
And the entire HPMOR fanbase has just now googled the concept. Promotion of ideas is what HPMOR's purpose is, after all.
Perfect mathematical reflection, free of Gödel's incompleteness theorem.
checkable partial execution traces with cryptographically strong bounds on honest vs fraudulent work
You may want to talk to these guys: http://www.scipr-lab.org/pub (SNARKs for C)
And to me as well, in due time.
I quit my job for one that I'm much less comfortable with, but with more room for long-term improvement.
In your face, risk aversion!
Assuming somebody would want to take over a bankrupt company with liabilities as nasty as not-quite-dead humans. The liabilities of a bankrupt cryo company would vastly exceed the assets. Also, you can't get rid of those liabilities, not even part of them, in any way which isn't a PR disaster.
Frozen people are liabilities, not assets.
Interesting. I'll look into that when/if I have some free time. In the meantime, may I suggest gamifying this at some point? Let MIRI organize a programming competition in Botworld, preferably with prizes. If this plays well, you'll get a lot of attention from some highly skilled hackers and maybe some publicity.
A simple loop can alter a sizable fraction of the 'world' within short time. Thus no complex analysis of opponents never pays off (except for tests like 'is some opponent at this address').
It's not because a simple loop can alter a lot of space. It's because Core Wars world is crowded both in space and time. Make agents start a lightyear away from each other and/or make a communication/energy/matter bottleneck and all of a sudden it pays off to do a complex code analysis of your opponent!
I was going to start my comment by pointing out the silliness of your real estate system. Namely, that if people/society were really that fast to adapt, they would almost all work remotely by now. But on second thought, I think this part was actually intended as April Fools material.
Then I was going to say that the BDSM reference is somewhat distasteful, at least to some people. Not to mention it smells of internet exhibitionism, even if untrue. You could easily pick something better overall.
Then I'd say that while your education and medical care ideas are good, they fail to account for the world-as-it-is. It's easy to imagine a world-as-it-should-be, especially if you assume that people/society are different. It's not enough. You have to design a smooth transition from a world-as-it-is to the world-as-it-should-be.
I'd finish by saying that I don't agree with you that the April Fools excuse will let you get away with as much as you think, especially not in the minds of people outside of LW. This post lowers my estimate of your sanity waterline and I'm accustomed to your writing. Linking here from bloody everywhere was a mistake.
Instead of elaborating on the above, I'll say this: Eliezer, you really need to get your ego below your IQ.
They are somewhat high on the list of top black swan events, however.
Thank you. I asked because I don't understand other people's attraction to personal details like this. Nothing specific to Satoshi or bitcoin.
(This has, incidentally, been a good reminder of why I don't post my Satoshi research publicly, and generally limit my comments to debunking proposals.)
If you don't mind sharing, what was the reason for research, besides curiosity?
demirationalist - on one hand, something already above average, like in demigod. On the other, leaves the "not quite there" feeling. My second best was epirationalist
Didn't find anything better in my opinion, but in case you want to give it a (somewhat cheap) shot yourself... I just looped over this
I think I disagree. System 1 doesn't care about any particular identity nor any particular action. It cares about general, somewhat vague emotions. To use one of you examples: system 1 will not care about watching TV all day. It will care about relax and entertainment. It will be equally happy if you play relaxing and entertaining computer games instead ... and you could pick those compatible with the "growth mindset"
You guys left one possible loophole: Does "tanks will be involved in combat" mean actually firing shells?
Ture, not yet, at least. Would you agree though, that this could easily escalate out of proportion?
It seems to me that the odds look so grave to you because you gloss over several steps during this potential escalation.
Have a look at this: http://www.zerohedge.com/news/2014-02-28/ukraine-acting-president-says-russia-starts-aggression-against-country-russian-plane Original source: http://www.pravda.com.ua/rus/news/2014/02/28/7016674/
Posturing or not, if this info checks out then by existing treaties USA and UK are obliged to help Ukraine against Russia. You see, Ukraine gave up nukes in exchange for safety guarantees from Russia, UK and USA.
http://en.wikipedia.org/wiki/Nuclear_weapons_and_Ukraine
Before voting on accession, Ukraine demanded from Russia, the USA, France and the United Kingdom a written statement that these powers undertook to extend the security guarantees to Ukraine. Instead security assurances to Ukraine (Ukraine published the documents as guarantees given to Ukraine[5]) were given on 5 December 1994 at a formal ceremony in Budapest (known as the Budapest Memorandum on Security Assurances[6]), may be summarized as follows: Russia, the UK and the USA undertake to respect Ukraine's borders in accordance with the principles of the 1975 CSCE Final Act, to abstain from the use or threat of force against Ukraine, to support Ukraine where an attempt is made to place pressure on it by economic coercion, and to bring any incident of aggression by a nuclear power before the UN Security Council.
Of course more likely then not we'll find out once again that treaties and words aren't worth anything unless you have the upper hand... but this looks scary enough to me.
A skeptic doubts the senses give actual evidence, they doubt math ...
How many skeptics walk off the cliff expecting to continue walking? If you're skepticism is of the purely theoretical kind "sure, I doubt everything, but God (heh) forbid me act on these doubts" then I cannot help you either.
Besides, that's cherry-picking circularities. Let's go meta: don't you doubt your doubts? If you claim you can't calculate or measure the level of anything real because "that's axioms", what makes doubt in math/physics weaker than doubt in doubt in math/physics? And if none is weaker then the other, why don't walk off the cliff?
how are said axioms to be justified?
This is how I'd answer a sceptic:
If I put two apples into a bag that previously had two apples, I can take four apples out of the bag. Thus, I believe that axioms on which basic arithmetic is based are "justified". By the same token I believe axioms of probability and I'm pretty sure you see a close approximation of a "fair coin" on a daily basis, not to mention more complex behaviors which probability theory predicts very well. If after that you're still skeptic of the correlation, I expect you to have strong evidence against the correlation. I predict you'd say that this reasoning is circular because the whole notion of "evidence" is sort of dependent on the axioms (Bayes etc.), in which case I can't help you any more than say that the given axioms are what they are precisely because of empirical observations.
In a huge oversimplification that's how math theories are constructed - you add or remove axioms until the stuff it predicts corresponds to stuff we observe. The correlation is the goal.
Bahahah. Your current neurochemical high will wear off in 2 days.
Current? The event happened months ago, I only wrote it down now, I hope it's obvious you can't keep this up 24/7. A silly example: I learned to put on and take off my jacket while walking the stairs. Saves me few seconds every time I go out. It's a habit now and, most importantly, it was fun to pick that habit up. It's boring to cleanup a desk. It's fun to try to cleanup a desk with only one hand within 20 seconds.
The story is just a feed for thought, it's up to the reader to figure out what works for him.
How about an architect walking his clients though their soon-to-be house?
The Severing Charm wouldn't bring down a tree, so he'd started partially Transfiguring cross-sections through the wood.
Quirell saw that. Partial transfiguration is not the power the dark lord knows not.
I just used charity as an example; the same argument applies to taxes as well.
That does not follow. Why would I care about money I have no control over? Why would a politician care about efficiency over publicity? Why wouldn't the recipient try to take more than he needs? There's no incentive for anyone to do anything right.
Is that right?
Yes, that's correct. I'm arguing that redistribution in any form of giving "stuff" for free makes it worse by providing strong incentives to maintain status quo.
I can't shake off the impression that you're implicitly assuming the thesis. I'll try to answer what I can.
In this account, how did I get so much more healthy, wealthy, and educated than you are?
Why the implicit assumption that redistribution has a net positive impact here? Another implicit assumption here is that we're all born equal. Genetics aside, don't the children inherit the mindset of their close ones to a large degree? Aren't societies semi-stable, self-reinforcing, whatever their current wellbeing is? Africa is still poor, despite years of foreign aid. Middle east is still fighting, despite years of foreign interventions... What you need to do is to create incentives to break out of the current state of things and survival instinct is an excellent incentive which can be applied to poor people. What redistribution does is removing this powerful incentive and creating opposite ones. Basically, you're rewarding people for making poor economic decisions.
Is there any way to do whatever that was to you, as well
Assuming it was redistribution that made you wealthy and educated, does it work reliably on the majority of poor people? Without forcing them somehow out of their current environment? That's along the same lines of thought as that giving someone a million dollars reliably makes them a millionaire.
Would it have similarly been better to let me die (or kill me) before doing whatever it was that made me so much more healthy, wealthy, and educated?
This argument only makes sense if you already believe that redistribution reliably works, or anything reliably works, for that matter. People are notoriously difficult to change.
Of course I won't argue against effective altruism or charity and I suppose charity is technically a kind of "wealth redistribution". However, it's different than taxes in one very important way: it's redistributing excess wealth after my own goals have been achieved, not before.
redistribution of wealth and other oppression-proofing liberal policies
You're a healthy, wealthy, educated person. Being educated, you know you shouldn't have more than, say 2 children, to be able to afford their education and ensure their good standard of living. You'll have first child at age 25+.
I'm poor and uneducated third-world citizen. Being uneducated I don't know how many children I can afford. Or I just don't care, don't think about it. I'll have my first child at age of 18.
Now you give me half of your wealth and now you can only afford one child.
25 years from now you and your one educated child have to support me, 8 of my uneducated children... and 40 of my uneducated grandchildren. Your child can't afford having children at all.
This is what redistribution does, exaggerated. You're assuming the person you'll give wealth to will use it in a sane way from your point of view. They won't. You don't want to admit the possibility that it may be long term better to let them die to stop this. Of course this is not ideal, not even good. Ideally you'd teach them, but will they listen?
Not necessarily. The brain pattern-matches continuous sensory experiences to something already known, which is discrete. I tend to think about it as rounding-to-nearest. Blood gradually transforming into water before your eyes doesn't make sense to the mind.
Yours is more spooky, but I had a similar experience.
In high school I had to get up early to be on time. In winter it meant waking up when there was still pitch black outside. Also, one teacher was exceptionally strict and would be angry at you for months if you were late or missed his class. So, when one day I woke up when there was fully bright outside I freaked out and jumped out of bed with a loud "F...F...F...". And then I woke up and it was pitch black outside. And then the alarm clock rang. I laid there for a while trying to figure out how the hell am I supposed to figure out whether I am awake or still dreaming. I didn't so I carried on ... maybe I'm still asleep, but at least I wasn't late.
This has never occurred to me! Yes, this would be quite likely. On the same note: Shouting at hard drives
I think you misinterpreted the article. The virus can't infect a healthy machine "through the air" (microphone). It can bridge air gaps in the sense that two already infected machines can setup network over microphone, which is orders of magnitude more likely than the former. BIOS infections have been done before, so ...
No, I was never able to repeat that experiment. However:
Zl fgebatrfg ulcbgurfvf vf gung gur pnoyr unq n gval culfvpny qrsrpg, juvpu pnhfrq ovg reebef zber bsgra jura orag "whfg gur evtug jnl". Vg pbhyq or grfgrq ol ybbxvat ng gur engr bs GPC/VC ergenafzvffvbaf pnhfrq ol vainyvq cnpxrg purpxfhzf, juvpu jr qvqa'g gubhtug bs ng gur gvzr. Guvf vf fgvyy ener rabhtu gung V qba'g rkcrpg gb frr vg ntnva, rira vs V unq gur rknpg fnzr pnoyr gb rkcrevzrag jvgu. Guhf gur gehr pnhfr erznvaf haxabja.
Edit: phase of the moon bug
I had a large file to copy from my laptop to a friend. We were on a break between lectures - no external drive, wifi way too slow - so we used cable. The copy was taking a while so eventually we started making jokes about data flowing down the cable faster because of, you know, gravity. It didn't take long before we placed the receiving machine on the floor and... the speed increased. Not radically, but definitely more than what you'd expect from random fluctuations. We replicated the experiment three or four times, in both directions. Every time when the receiving machine was physically below the source the transfer was better. To this day I can't explain it and to me this is a proof of how a relatively simple and well understood system can be unpredictable.
The race to win the Singularity is over, and Google has won
Counter-argument: NSA has a track record of having Math advanced decades beyond public. Also, quantum computing should be within reach in 30 years... might be NSA as well, since it'd make a perfect crypto cracker.
OK, I get it, this is supposed to be one of those self-help thoughts that are supposed to make you better off if you think them (suggestions for a name for such a thing, anyone?), regardless of whether they're actually true. Well... it doesn't work. My thoughts, roughly in order:
WARNING, manipulation / black arts
WARNING, causation != correlation
WARNING, opinion != fact
What about, say, good leaders? 5 people closest to one can't possibly be good leaders themselves because who'd they lead then?
WARNING, thesis likely literally false, seek metaphorical sense?
Silly and bogus example. You don't transfer attributes by osmosis. If anything, to become e.g. creative, you'd be better off by meeting people who are currently becoming creative not who are already creative.
This is an interesting way to setup a lottery while promoting one's ideas.
yes. I tend to reevaluate goals frequently (at least weekly)
yes to the extent of future uncertainty. I am free to change my goals at any point, after all.
yes... (obviously?) I'm having a hard time understanding how anyone could not do that.
if you would call it "a system", I try to restate nearest action that brings me closer to achieving my goal every day. More often than not the action is "wait" (until you acquire enough money/power/influence to make something happen, until you have enough information to make a decision, etc...)
Thus the design (i.e. "The Math") vs implementation (i.e. "The Code") division. I believe design verification suffers from same problems as implementation verification, albeit maybe less severely (though I never worked with really complex, novel, abstract math... it would be interesting to see how many of those, on average, are "proved" correct and then blow up).
Still, I would argue that the problem is not that black-box testing is insufficient - it is where we are currently able to apply it - but rather that we have no idea how to properly black-box-test an abstract, novel, complex system!
PS. Your trivial example is also unfair and trivializes the technique. Black-box testing in no way implies randomizing all tests and I would expect the QuickSort to blow up very very soon in serious testing.
I for one can't agree with the point that transparency does any good in security assessment if we consider implementation of a complex system (design has its own rules though). I believe you underestimate how broken a human mind really is.
Transparency == Priming
The team which does security review of the system will utterly fail the moment they get their hands on the source code, due to suggestion/priming effects.
- comments in source - I won't even argue
- variable and function names will suggest what this part of code "is for". E.g code could say "it was (initially) written to compute X for which it is correct" except that later on it was also made to compute Y and Z where it catches fire 0.01% of the time.
- whitespace. E.g. missing braces because indentation suggested otherwise (yes, really, seen this one)
If you truly consider removing all metadata from the code - the code looses half of its transparency already, so "transparency" doesn't quite apply. Any of that metadata will cause the security team to drop some testing effort. Those tests won't even be designed/written, not to mention "caring enough to make an actual effort". Otoh, if a program passes serious (more below) black-box testing, it doesn't need to pass anything else. Transparency is simply unnecessary.
Hardcore solution to security-critical systems:
- Have a design of the system (this has issues on its own, not covered here)
- Get two teams of programmers
- Have both teams implement and debug the system, separately, without any communication
- Get both teams to review the other teams binary (black-box). If bugs found goto 3
- Obfuscate both binaries (tools easily available) and have both teams review their own (obfuscated) binary while believing it's the other team's binary. If bugs found, goto 3
- At this point you could open up sources and have a final review ("transparency") but honestly... what's the point?
Yes, it's paranoid. Systems can be split into smaller parts though - built and tested separately - so it's not as monumental an effort as it seems.
Is true precommitment possible at all?
Human-wise this is an easy question, human will isn't perfect, but what about an AI? It seems to me that "true precommitment" would require the AI to come up with a probability 100% when it arrives at the decision to precommit, which means at least one prior was 100% and that in turn means no update is possible for this prior.
Your writing feels like it needs a special dictionary.
I would like to offer a distinction between two different kinds of accepting. One is the opposite of denial (which is being called "fighting" in this case). The other is the opposite of changing.
"accepting" != "acceptance". Why relabel "denial"? You attach a second meaning to "accept"... didn't "acknowledge","admit", "avow" or "recognize" work for you?
Ah, I see there was a race condition. I'll retract my comment.
I find "How do I proceed to find out more about X" to give best results. Note: it's important to phrase it so that they understand you are asking for an efficient algorithm to find out about X, not for them to tell you about X!
It works even if you're completely green and talking to a prodigy in the field (which I find to be particularly hard). Otherwise you'll get "RTFM"/"JFGI" at best or they will avoid you entirely at worst.
If you instead went due north and then due east, that would be two left turns (driveways), one right turn, and three straights. Isn't that a strictly better route?
I assumed turning left into/out of a driveway (i.e. "crossing the street when not on a crossroad") is impossible or at least hard (slow). This is often the case in a dense city. If we're not in a dense city then Taxicab assumption is an error as well.
data point: I didn't parse it as condescending at all.
It seems to me that, unless one is already a powerful person, the best thing one can do to gain optimization power is building relationships with people more powerful than oneself.
Depends on how powerful you want to become. Those relationships will be a burden the moment you'll "surpass the masters" so to speak. You may want to avoid building too many.
Before experiments pretty close to 8 on average, maybe slightly less.
No, I didn't know it exists, cool. Looks like one more reason to get myself a smartphone already.