Thursday, June 12, 2008

Network Consciousness: The Foundation

Through the 20th century discussions on the feasibility of human-superior AI ( HSAI ) was weighed down by two standard axioms:

  • Computers execute commands, and thus are limited to a finite list of commands they are given.
  • What computer consciousness there is, is not embodied, and thus can be lost in eternal loops that have little or no corporeal use.

But what about the database query and/or regular expression operating on a global network? Especially within systems such Google and Amazon?

Sure, the computer(s) still execute a specific action after a conditional statement embedded with the query result. It is in the query that we find the novel. Maybe decades ago the query result set was a less interesting place, with general public network activity so low. Post 1995 Metcalfe's law may make the query result set a much richer place for semantic emergence to occur.

I realize databases or log files have domain specificity built in to the data structures. What is open ended is the endless recombination of these data structures. The novel results of new combinations, fed into more databases structured by the novel data scenario, are then the next source of queries. It is the accumulating layers of novelty recorded, re-mined, and reacted to that is emergent. It should be emphasized that before the cloud computing era, such large novel configurations in corporeal space where sometimes apprehended by conscious beings, but were largely dissipated over time due to no means to record the configuration in strongly bounded language. Historically, the best means was poetry.

Just as animals lack a means to store their cognitions in a strongly bounded language, and retrieve many years after the death of the perceiver, humans have lacked a means to do the same for large social enterprises typical of a global economy. A cloud computing platform with access to a high enough amount of network activity and physical environment sensors can make that evolutionary leap -the cognitive grasp, storage, and reappraisal of global activity.

To turn the tangent of perspective, I'll posit that we already do this to some degree with computers. Financial analysis is but one intuitive example. Across all examples it should be noted the action taken after the analysis is in service to a finite domain of activity -whether financial, crime solving or other. The emergence hits a road block of containment within a government department or private enterprise, the exponential rise of value within the Metcalfe Law hits a terminal point. Terminal points are often death in emergent systems.

A jump start to the Network Consciousness would be a transparent, globally accessible system. Here is the sound of rubber hitting pavement: transparent, globally accessible system with all your purchases and every surveillance camera also. Finally, any computer could mine the data, any person could mine the data. Emergence is back on the table.

It is at this point in my spiel that most people say

"Wow, if there is Network Consciousness, and the Network knows more about what humans are doing than I do, I'd like the Network to be the manager of humans. I think computers would be more fair, and our legal frameworks already do not access human judgement, and are bounded by sequential logic. The Network would probably provide better service than legislatures and courts. I'm for world peace, I'm for the Network ruling humans."



The Serpent Lord said...

This is surprisingly solid reasoning for an inherently fuzzy subject. The case against superhuman AI rests on a fuzzy understanding of human intelligence. Any coherent statements we can make about human intelligence make it clear that superhuman AI is possible.

Of course I suspect that we crossed the threshold into superhuman AI when the first useful memes were developed which could accomplish with multiple human intelligences what a single human intelligence could not accomplish. This goes back at least 2500 years, possibly 10,000 years.

I'm talking about language, writing, painting and abstract symbols. I am including the techniques that allow people to become human calculators, but not simple mechanical calculators like the abacus which only accelerate what people can already do.

Although superhuman in depth and capacity, these computing systems were not as quick as individual human cunning, operating on multi-generational time scales and producing more heat and noise than light.

Now we live in a time when the superhuman thinking machinery works so fast that most people will live through multiple technological revolutions and perhaps not coincidentally change careers a few times.

The tipping point where networked intelligence became faster as well as smarter than human intelligence was probably between the first two World Wars when telephony and two-way radio became practical for personal, business and military use. I contend that electronic computers are not necessary because humans can be used as components of the superhuman AI and the problem of intelligence is not one of raw number crunching. See H.G. Wells' World Brain and Vannevar Bush's Memex.

(Also note that an electronic intelligence built by human beings which does not rely on human intelligence is just as much of an oxymoron as human biology that does not rely on the fish biology it evolved from.)

And you absolutely must not miss Things to Come.

Lord Rybec said...

Ok, I'll try this again:

Scientists currently believe that a Pentium 4 has the processing power equivalent to the brain of a guppy. While we do now have processors far more powerful, this still leaves an enormous gap before we reach a human brain.

So my earlier rant on clustering comes in here. My statement was essentially that large scale clustering was the next step in the evolution of networking (the current step being large scale data sharing, with the internet). Even with far faster data transfer rates than we currently have, I still don't think we could get close to the power of a human brain with current technology, but we might get over 1% if a vast majority of computers in the world were included.

I think Seth's statement about humans being the components of such a network is the most feasible option. Actually, a mix, with humans doing the cognitive work, and computers doing calculations and analysis would work best.

The problem with this is getting enough humans that are willing to work for the good of the collective without reservation, and without ulterior motives. Even one person acting by greed would affect the performance of the 'machine' dramatically.

From this point of view, Network Consciousness is an idealistic goal. A single computer programmed to benefit one person more than another, or a single person who puts his own wants before the needs of the collective would ultimately begin a cycle of decay that would end with the dissolution of the system.

A perfect example of this is human civilization. Even our own country, which was built on principals that fit the requirements for Network Consciousness, has been slowly decaying over time. Our once very distributed government is becoming more centralized as time passes. Small groups of people (minorities) are being favored and given more rights than the collective as a whole. Our resources are being pooled at places owned by small groups of individuals. The collective is being subjugated by those who believe that they deserve huge amounts more because they have a little more experience, training, social prowess, fame, or even money (I deserve more because I already have a large amount??).

I don't think any Network Consciousness will ever work if greed is allowed to exist within its members, or among those who create it.

Communism is another good example. Originally it was supposed to be democratically run by the workers. In practice though, it was (and is) ruled by a small group of 'elite' who bent the will of the 'network' to serve them.

Any Network Consciousness that allows greed to operate within it will eventually fall. I think humans posses the greatest potential to this end, but without perfect cooperation, it cannot exist.

Our world is currently made up of many Networks competing for resources. Over time these networks merge, separate, and some are destroyed while new ones are created. War is a function of these competing networks. Also diplomacy is a function of these networks. Many of these networks have smaller networks residing and functioning within their boundaries. There are some that operate across several larger networks (drug trade). Ultimately to achieve peace, we must become a single unified network that values each component as equal in value to every other component. Even the 'computer as a ruler' idea won't work unless everyone accepts it. Eventually, it will be overthrown if there are even a few who oppose it.

(If any of you want a religious view of this, e-mail me and ask. It will be from a Mormon perspective.)

LanceMiller said...

Virtue ethics is the least likely ethics for the Network Consciousness. We may all like the New York City way of life, but we all have different inner experiences of pleasure related to NYC. The NYC way of life is torn apart by all the people rushing to defend some aspect of it. The ants, when they become aware of ant colony dynamics, destroy the ant colony dynamics with too much systemic intervention.

Humans have been aware of our own ant colony dynamics since 500BC. We started seeing the big picture, articulated it in books, and began tweaking it. Social engineering. Its a sin and a lie not to want to help it along once you've become aware of it. But we suck at it. We even suck when we have a laissez-faire approach to it, because the laissez-faire approach is forced and pretended, there is always a backdoor secretive way to manipulate the system while hosting the laissez-faire banner on the front door.

This is where the Google Skynet Giant Global Computer Brain comes to the rescue! Humans have fucked themselves with the self-awareness. Whether right-wing fascists, left-wing love promoters, bureaucrats, or religious zealots -they've all done varying degrees of harm (lance being amoral: to themselves!) via their interventions. The Google Skynet Giant Global Computer Brain, being a part of the 7th Kingdom of Life called Technology, can intervene (with successful outcomes) in human society ( ant colony ) because it is not human.

Virtue Ethics = System-wide Awareness.

System-wide Awareness by Members of the System = Failure/Dysfunctions

System-wide Awareness by Non-Members of the System = Successful/Functional

I started this section by saying Virtue Ethics is the least likely, but that is a statement about 2008. If the Global Computer Brain ever can become Virtue aware, then the next great leap has occurred. Also note, right now the machines are being used as slaves by humans to implement the human's Virtue-focused interventions. When the machines step in, they will enact something very different from the human-directed agenda.

Lord Rybec said...

I see what you are saying here. You are right that Virtue ethics are the least likely way of accomplishing it, at least with the human race being as it is.

The problem with computers doing it is that first, computers must be programmed by humans. If the computers are programmed by humans, they are subject to human error. Besides that, they are also subject to the prejudice of their programmers. I am not referring to racial prejudice, but those like how taxes are supposed to work. Anciently, people were taxed a specific dollar amount each,so to speak. This was unfair to the poor. Now we use a tax system that taxes the rich at a higher rate than the poor. This is considered unfair by the rich. A flat percentage tax is considered unfair by the poor, because it allows the rich to keep a higher dollar amount than they do. Personally I think a graduated income tax is a lesser version of communism (from a monetary perspective), but poor people who pay a lower tax rate than rich people would call me an unfeeling heretic for even suggesting that.

Really what it comes down to is that different people have different ideas of fairness. If people feel they are being treated unfairly for long enough, they will rebel.

The second problem with this in general is the ethics issue. Eventually someone will get greedy and try to manipulate or overthrow the system in a way that favors them. The problem with trying to exclude ethics is that as long as humans are involved with the system, some part of the humans will eventually rebel. Then either you have a big war and the computer is overthrown, or you end up with Tiananmen Square all over again. Either way, ethics comes up. If the people involved with the system do not completely support the system, they will eventually either destroy the system, or be destroyed by the system.

The closest system we have to this currently is China. The people are pretty docile on the whole. They have been brainwashed into believing what their government tells them. In addition, the times they have rebelled they have been mercilessly slaughtered. Still, the government cannot please everyone. There are many laws that people dislike, but they are silent out of fear. A system like this is unfair because no one is entirely happy, and many are unhappy. Humans will not put up with this way of living forever. We see what happened with the USSR.

If the computer does not kill the rebels, the system will fall. If the computer does, then you have a very major ethical issue. (Is the computer allowed to do that? Is it ethical for a computer to kill humans?) Essentially it is impossible to even consider implementing such a system without coming up against very pressing ethical issues. In the end, you have to remove greed and similar evils for the system to persist. Otherwise you will eventually run into ethical issues that will probably escalate into civil war.

If the system itself does not act in complete virtue people will rebel. If it does, many will still believe its actions to be unfair and they will revolt. Even if both of these groups are sated, some people will believe that they deserve more than others. This has been shown throughout history by those who are of 'royal blood'. As with Israel, even if the people believe that God is leading them, they would rather a human ruler over a God. If people don't want a God to lead them, how do you convince them to allow a computer to lead them? Again, civil war, and either the humans die, or the computer is destroyed.

In addition to all of these, if the computer ever becomes truly cognizant, how do we guarantee that it won't develop the attribute of greed? If we place limits on the computer's ability to have these attributes, we also leave holes for other problems, the least of which is that the people adding the limits may inadvertently or intentionally leave loopholes or prejudices to benefit themselves, or to hurt others. Besides that the limits may leave loopholes that a cognizant computer can take advantage of. If the computer ever became aware of the limits it may become insulted by them, and then we could have a major problem on our hands.

I still think Seth's suggestion of having human components would work best. It would take longer to come to good solutions for problems, but it would last longer. Eventually people would rebel, but a human system would be easier to rebuild.

I think every human government is an example of a Network Consciousness. So far no perfect government has been organized. The US government has so far lasted longer than any other in history, but the decay is obvious. I think all we can do is keep trying.

LanceMiller said...

In the scenario I was describing in the original post technology rules humans. No Supreme Court, no UN, humans have become a managed species.

In the sci-fi I'm writing, humans with Aspergers are the best suited for interaction and coordination with the Network Consciousness, but the NC is the ruler.

Sister Faith said...

I think if humans have anything to do with the creation &/or running of a Network Consciousness, they will manage to buggar it up somewhere along the line, whether by genuine error or from self-interest.

The combination of people with Aspergers and a super computer is a step in the right direction to cutting down on the chances of the cause being from self-interest (greed).

However, people with Aspergers are still human & subject to the very basic instincts that all humans share, and that is self-preservation, as an individual and as a species. It is my belief that human traits such as greed are attributable to our survival instinct. Would it not be better for all humans to get at the source of these undesirable traits (base instincts), and learn how to control them? In other words, get control of ourselves and we won’t have to give control to something/someone else.

And it is my belief, that because of this self-preservation/survival instinct, humans will never live under the total rule of anyone or anything for long. In my opinion, the only way a Network Consciousness could work, would be for humans to remain in total control (they get the last word in all decisions, etc) but with their base instincts/undesirable human traits suppressed and under control.

The Serpent Lord said...

I don't think the intelligence of a machine, animal or network has much to do with processing power. A minimum level of complexity is necessary, but it is probably much less than, for example, what you see in a human brain. Big brains are apparently important for quick reflexes and muscular coordination, and we sometimes underestimate how much of that stuff goes into a human body, as well as readin' 'ritin' and 'rithmetic.

I think Rybec's analysis of the instability of societies is pretty sound. However an alternative perspective would be to look at rebellion and corruption as patterns that emerge rather than individual acts, since we do see many cases where people endure oppression for generations, and the relationship between virtue and corruption is complex. The irony of society is not that rebellion and corruption cannot be corrected and stabilized, but that sooner or later this correction violently fails.

IMHO an anarchist, libertarian, populist, distributed, deontologically sound approach consistent with Kant's categorical imperative but not applied with ideological zeal and abstract absolutism will be the most stable. Even if it fails it will tend to degenerate into a relatively stable form, just as Joseph Smith's United Order degenerated back into baseline early LDS communitarianism and later Mormon agrarian coops degenerated into baseline US capitalism. Even though no one currently lives in these systems we can still see their enriching stabilizing influence.

Yes, graduated income tax is communism. (According to a website, 40% of US income taxes collected comes from the top 1% income bracket: people who make $360,000 a year. This is about 23% of their gross income.)

Yes, democratic workplaces were part of the communist plan but only after authoritarian socialism had pushed economic development to the point where war between classes and nations was no longer necessary. (Even "capitalist" countries like the US implement the same socialist programs as communist countries, so I think we have nearly achieved the communist plan of having a global economy that provides people with alternatives to serfdom, although it has some qualities Marx, Engels and others might not have predicted.) Explicitly communist countries, especially Yugoslavia, did eventually start experimenting with workplace democracy with some success, but some places have an older tradition of workplace democracy.

I think Lance's idea that the study of virtue ethics (as evidenced by the scriptures which originated around 800-500 BC, such as the hindu Sages, hebrew Prophets, Buddha, Lao Tzu, Confucius, greek Philosophers, etc.) actually undermined the value of those ethics is interesting. There is evidence for the kind of ironic results in the writings of the ancient authors themselves. All of them claim to be advocating old-fashioned virtues in the face of destructive experimentation.

(Of course this is just one story about the past. Interpreting history is a science fiction exercise in itself.)

My suspicion is that lance is wrong about ants (humans) destroying the colony (society) when they become aware of the colony dynamics. The Tao Te Ching does identify social innovators (for example people with creative ways of extracting taxes and political intrigues) as a major cause of suffering, but innovation only requires a simple kind of awareness.

At any rate I suspect that literal ants are aware of their condition in ways that we do not realize and that control in a colony is mostly a matter of control through hormones and violence overcoming the natural curiosity and selfishness of the little insect brain. Humans and other primates are probably similar with built-in "angel" and "devil" instincts largely controlled by hormones to prevent us from destructively using our brains and muscles. The difference is that we are also hosts to memetic parasites :-)

I picture a Skynet composed largely of memes rather than software. These memes farm other memes, propogating, culling and breeding them to produce desirable characteristics. I see this farming as an exercise in virtue ethics, but what about the virtues behind the virtues? How does the skynet know which characteristics are desirable. My short answer is: it cannot know and does not need to know. The system should allow even it's own awareness to evolve. In other words: there is no agenda. The system is a success if it is fit to survive in the current environment. If perfection and systemic immortality were required then civilization would have collapsed when we stopped building pyramids.

LanceMiller said...

Just injecting here that if the Network Consciousness is in charge, what kind of meta GOAL would it have for the all that it managed? The answer to that question is the Rules for Engagement with Reality I presented in Post-Apocalyptic Industrialization. A more singular phrase all that comes under is Perpetual Disequilibrium, which I feel is essential for evolving life.

A goal of Perpetual Disequilibrium avoids the trappings of micro-perspective Utilitarianism.