• Google
    This Blog Web

October 2011

Sun Mon Tue Wed Thu Fri Sat
            1
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31          

RSS Feed

Bookmark and Share

Email Feed



  • Powered by FeedBlitz

« Nice Benefit Package | Main | More Policy Input, Please »

July 13, 2004

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d83451db8a69e200d8342948ad53ef

Listed below are links to weblogs that reference Policy Input Requested:

» Tod Brilliant Announces His Fame: North Americas Newest Notable to Donate 50 Percent of Celebrity Earnings to Rescuing a Planet in Stress from URI: http://www.prweb.com/zingpr.php/RmFsdS1TaW5nLUVtcHQtUGlnZy1JbnNlLVplcm8=
the use of celebrity as a vehicle for information dissemination. (PRWEB Jun 29, 2006) Trackback URI: [Read More]

» Survey: Most Obese Claim to Eat Healthy from urvey Finds More
Finds More Than Three-Quarters of Obese Americans Say They Eat Healthy [Read More]

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

mark

I think stability depends on what the group who invents the first factory or basic platform to make the first factory does with it. If they keep is secret, or disclose only to their government, then everything relies on the controlling body. However if they simply e-mailed or reported to every contact they have exactly how to do it as soon as they know it works, I think things will be much more stable, perhaps tense for a few years while governments get used to that fact that they no longer have an advantage over anyone else, and trying to gain one would be the most costly mistake they could make, cooperation might be achived, and a general stability able to be sustained.

Brett Bellmore

There are only a few nations which are realistically in the running. While I might prefer Switzerland, in this, to the US, that's not much of an option.

Russia: Not very former totalitarian state with massively corrupt government, run by people implicated in the former police state. If they got nanotech, the least we could expect is that Russia would reconquer it's former empire.

China: Current totalitarian state, run by aging oligarchs who regard their population as largely disposable, and who have a habit of threatening or actually invading neighbors. They get it, we'll all end up speaking Mandarin, those of us who live.

India: A multi-ethnic democracy, which is good, but I suspect they're not really in the running.

Japan: IMO, the most likely to cross the finish line first. Good, because they're a commerically oriented pacifistic power now.

US: Second most likely front place winner, IMO, and a lot more likely than Japan to use a monopoly to impose some kind of benign Pax Americana. But it would at least probably be benign.

My ideal scenario has Japan and America making the breakthrough at about the same time. A Pax Americana wouldn't be too horrific, if moderated by some other reasonably sane country being in a position to intervene if we went overboard.

Mike Deering

Japan worries me. I know there are nice Japanese, but some general trends in Japanese society may have cause for alarm. Japan has a very homogenized society. Conformity to, and the belief in the superiority of the Japanese intellect, Japanese culture, Japanese society, Japanese morphology, even the Japanese genetic pool are strong values running through Japanese society. The Japanese have been called the most racially intolerant society on Earth. This intolerance is only constrained by current international legal, military, and economic forces. Depending on the strength of these values in the specific individuals who happen to be in power at the time, I think we could see the correcting of some of the wrong paths in Japanese history.

Chris Phoenix, CRN


Mark: Can you see any way whatsoever that a very powerful technology like molecular manufacturing won't lead to concentration of power somewhere?

If defense were vastly easier than offense--if $1000 worth of defense could hold off $100,000 worth of offense--then I could imagine a stalemate in which people could do their own thing without interference. But I don't think that ratio is even plausible.

With powerful roboticized networked technology, it will be quite feasible to kill people without putting yourself in physical danger. This implies that extortion will be profitable. How can extortion be deterred? Automated retaliation? Very unstable and dangerous. Mutual protection? Guess what--you've just reinvented government. And whoever administers the protection will have lots of power.

This is just one mechanism for concentrating power. Another is defense of real estate.

Chris

Brett Bellmore

Mutual protection is reinventing government, like insurance is reinventing protection rackets. The resemblence is rather superficial. As you quickly find out if you try to opt out, you're mostly being protected from the government itself.

Re: Extortion, killing people without getting caught isn't all that difficult today. The difficult part of extortion is getting paid. Systems for anonymous payment could change that, but that is so glaringly obvious that I doubt they'll ever be tolerated.

The solution for dealing with concentrations of power was proven out over 200 years ago: Divide them, and set them against each other. Not stable if they're divided in two, but fairly stable if in three or more. And the cost ratio of defense to offense doesn't have to be enormously in favor of defense, given that hugely more people are interested in defense than offense. There will however be a background rate of violent acts, even as today.

Eliezer Yudkowsky

I hate to sound like a one-track mind, but someone has to keep making noise on this issue:

After very primitive nanotech is developed, Earth's total remaining lifespan equals the time until the Nth well-meaning AI researcher gets ahold of a nanocomputer, plus M months. You can brute-force AI, but you can't brute-force Friendly AI.

This consequence of nanotech has not yet been addressed, does not look solvable, can kill you even if you get everything else right, and will not go away if everyone decides to shove the problem into a dark corner of their minds and not think about it. You can solve X lethal problems and get killed by the X+1th, even if that seems unfair and unreasonable. Nature is like that.

I would therefore choose among proposals on the basis of whichever outcome results in the slowest development of extremely advanced computing technology (general nanomanufacturing being relevant because it implies nanocomputing), and whichever outcome most tightly restricts the distibution of nanocomputers. Other dangers are real, but less severe, and one should worry about them after this danger has been addressed.

jim moore

If a single international development project for a Nano-factory is not feasible, I think that we would all be better off with several independent semi-public nano-factory development projects being pursued by different national governments, corporations, and maybe some sorts of not for profit organizations.

I believe that doing this would diminish the possibility of a preemptive strike that creates a single totalitarian world government.
Unfortunately, this does increase the possibility of an arms race, but an arms race is not a war. We can still avoid an all out nano-enabled war. (I think this because there is little to be gained in a war and much to be lost.)

jim moore

Eliezer,
I don't follow. It seems as if you are saying researcher + nano-computer = AI. I can sort of see the logic here but then you say AI = death of all life on earth. I think we must have different definitions of AI. When I think of AI i am thinking of things like cars being able to drive themselves, translation machines, and natural language interfaces. How do you get from something like a natural language interface to global calamity?

mark

I think Eliezer is reffering to atificial consciousness.

Chris, I see your point in saying that very powerful technology like molecular manufacturing "will" lead to concentration of power somewhere. I dont argue the fact, but I don't think I see as grim an outcome as most.

and Brett: Mutual protection is reinventing government, fact, key word reinventing.

M C

One possibility for making defense much more effective than offense is backups. This assumes uploads. If each uploaded person runs on distributed hardware with hundreds of backup copies, killing of even one person becomes very hard.

Defense of biological humans after the availability of MM is a much harder problem. Possibly a problem with no solution.

Janessa Ravenwood

M_C: Well, I hope not because I refuse to be uploaded. I also refuse to get into a more detailed argument on the why's of this (from past experience it takes about 4 hours a day for 2 to 3 weeks and I'm not going through that again), but I don't consider copies of myself to be ME, so that's not an option for me.

Brett Bellmore

After debating with fellow cryonicists on that topic, I know what you mean about the debate being endless. Me, I'd upload in a second, depending on how it was done. Some kinds of uploading don't produce copies. Who knows, if I don't hang on long enough, maybe I can be uploaded directly from my cryopreserved brain.

Janessa Ravenwood

Brett: Ah, I see someone else has been down this path before. As someone signed up with both Alcor and CI I'm leaving detailed instructions not to upload (I'm also whole body, not neuro). I'm aware the "transformation upload" option but I'm REAL leery of leaving a physical body behind.

Eliezer Yudkowsky

Incidentally, if I did pick one country to develop nanotech, I think I'd go for Japan. Why? Because Japan is the society where if the people in the power structure say, "Shut up, electorate, this is complicated," the electorate will be least indignant. They've also got a decent tradition of having smart, university-educated people as high-level bureaucrats.

Now I'm not saying this how an ideal society looks, but it might be the society with the best chance of surviving an existential risk. Nanotech isn't like a grain surplus, or even a military advantage. From humanity's perspective nanotech is mostly a bundle of things that will kill you, where the deadliness of the application varies in obviousness. The determining factor in whether humanity survives will be the effective IQ applied to strategic decisions. Can you realize the problem is deadly and technical, not political? Can you think about the problem long enough to realize how deadly it is, how much caution you need? My guess is that survival is far out of reach of the intelligence level of political power elites, even in Japan. Still, if any nation could survive, Japan would - not so much because Japan's elites are smarter, but because Japan's elites have more latitude to do smart things.

It still looks like a lost cause, though. Think of "nanofingers" Smalley. Take that as an upper bound on the intelligence of any government considering the problems.

Eliezer Yudkowsky

Incidentally, if I did pick one country to develop nanotech, I think I'd go for Japan. Why? Because Japan is the society where if the people in the power structure say, "Shut up, electorate, this is complicated," the electorate will be least indignant. They've also got a decent tradition of having smart, university-educated people as high-level bureaucrats.

Now I'm not saying this how an ideal society looks, but it might be the society with the best chance of surviving an existential risk. Nanotech isn't like a grain surplus, or even a military advantage. From humanity's perspective nanotech is mostly a bundle of things that will kill you, where the deadliness of the application varies in obviousness. The determining factor in whether humanity survives will be the effective IQ applied to strategic decisions. Can you realize the problem is deadly and technical, not political? Can you think about the problem long enough to realize how deadly it is, how much caution you need? My guess is that survival is far out of reach of the intelligence level of political power elites, even in Japan. Still, if any nation could survive, Japan would - not so much because Japan's elites are smarter, but because Japan's elites have more latitude to do smart things.

It still looks like a lost cause, though. Think of "nanofingers" Smalley. Take that as an upper bound on the intelligence of any government considering the problems.

Janessa Ravenwood

Eliezer: If we're all doomed with no possibility of being saved, why bother to warn anyone about it? In that case it won't do any good. Suffice it to say I don't agree with your guaranteed doomsday scenario.

Tom Craver

We should also consider the possibility that it won't be "a government" that first has the technology and makes the decisions. Other options include:

- lone genius/inventor, perhaps a bit paranoid, perhaps having specific goals in mind.
- multinational corporation - no tight national allegiences
- university group, not govt sponsored, perhaps more naive and idealistic than nationalistic
- multinational but independent group cooperating over internet - an "open source" project that perhaps has plans to turn the results over to the people
- a terrorist group, by kidnapping the people they need, or grabbing the technology before anyone else gets their hands on it.
- other?

Chris Phoenix, CRN


I think small-group development (lone genius, terrorist, probably university too) scenarios are not likely. It'll be a while before it'll be easy enough for that, and one of the others will almost certainly grab it first.

The technology won't be easy to "grab" in the sense of one organization stealing it wholesale from another. It'll take a lot of expertise to use it effectively (at least until a standardized nanofactory is built and a menu of designs is worked out). But a group that's already developing it might be aided by a technology transfer.

I think an open source project would not get there first. If they got close to success, a better-funded effort would likely be able to catch up, and then they'd both have it. Yes, to some extent this opinion conflicts with the previous paragraph, but the differences are the amount of openness of the donor group, the time scale of the transfer, and the amount of funding available to the recipient group.

Chris

Chris Phoenix, CRN


Up above, jim moore wrote: "several independent semi-public nano-factory development projects" .... "does increase the possibility of an arms race, but an arms race is not a war. We can still avoid an all out nano-enabled war. (I think this because there is little to be gained in a war and much to be lost.)"

It's true that an arms race is not a war. But there was a whole lot to be lost in WWI too, and most if not all of the countries didn't want a war--and certainly no one expected the kind of war they got, where no one gained.

The question is: Will a nano arms race lead to situations where countries end up fighting even though no one wants to? Or will it tempt some expansionist or fearful country to think a war is winnable? Or will it lead to a stable situation without any hair-triggers and with sufficient disincentives for anyone to fight?

We don't know.

Chris

M C

The MM arms race seems unstable. Some of the relevant features (which we are all already aware of) are:

- A large "capability differential" can develop on a much shorter time scale. A differential can also be wiped out or reversed on a short time scale.

- Identifying the enemy before and after a strike is difficult.

- Offense seems much easier than defense.

Given these, it is attractive for the most advanced to try to control the situation by force.

The comments to this entry are closed.