Wednesday, December 31, 2014

example of rise of the machines: the U.S. military (Iraq, Saudi) and dollar dominance

Biology has no relevance to the near future of the Earth except to be like a virus and cancer that is getting in the way of the greater efficiency of machines. Locally-grown food with the help of new technologies may enable a great expansion of human population in the near future, but only to the extent that the people do not get in the way of the machines. For example, Arab populations for the past 2 generations have been in the way of the U.S. and Europe (highest machine to human ratios) acquiring the oil beneath their sands, and they have suffered greatly for it. For example, our best friend in the Arab states, Saudi Arabia, has a PPP per capita income 11th highest in the world, $54,000, equal to U.S. and yet a typical wage there is $5,000 per year due to the U.S. enforcing a monarchy that retains all the oil wealth to the detriment of its people. This may have something to do with 13 of the 15 hijackers of 9/11 being from Saudi Arabia. The humans have been fighting back as hard as they can, but it’s difficult when the military machine of the U.S. is as large as all other militaries combined, enforcing the dominance of the dollar that the machine uses to control the world’s population. I’m exaggerating, but not as much as people who still think voters have control the U.S. government.

surprising fact about slicing up a pie

If you slice anything into N pieces of all different sizes what is the average size of the pieces?


avg =  (sum of piece sizes when expressed as a fraction of the total equals 1) / N

In other words:
avg =  (x/N + y/N + z/N + ... ) / (x+y+z+...)
avg = (x+y+z+...) / N / (x+y+z+...)  = 1/N

An infinity of pieces not included, but as Feynman pointed out, real objects can't be sliced up into an infinite number of pieces.

Monday, December 29, 2014

heat death, reversible goodness

Concerning ethics and morality, if my attempt at an objective morality (efficiency at acquiring ever more energy to move ever more matter = evolution’s implied goal) is correct, then it would seem to more rapidly bring about the heat death of the universe which is the ultimate in a peaceful existence. But this does not necessarily mean the result is random in terms of information even though it would be maximum entropy. At an ultimate level, computations and movement of matter can be correlated because 1 bit of information has energy (and therefore a mass) equal to kT*ln(2). As T ->0, information can -> infinite with near zero energy (I’m avoiding any quantum considerations). OK, so we intuitively know wasting energy is the physics (objective) definition of evil, where wasting is avoiding an increase in entropy while achieving a goal, and to be objective, I’ve been conjecturing that the goal is a repeat of this: flatten internal energy potentials, possibly even converting mass to energy, to repeat this process, while trying as best we can to not let entropy increase. Kurzweil suggested once we might even find a way to violate (reverse) the 2nd law of thermodynamics (entropy must always increase). Reversible computation (and reversible movement of matter) is the only way to not convert energy to entropy. Maybe this is already always occurring since all of physics is independent of time, with the 2nd law itself being the only exception, and we only have a perception problem: the 2nd is not an increase in randomness but an increase in information that is PRESERVING the memory of how we are going towards the heat death (peace). But it seems to me that if there is a conscious effort on the part of a super efficient A.I. system using reversible computation and reversible matter movement, then something has been achieved. Again, reflecting your own comments, maybe this is already occurring and we are just too dumb to see it.

morality of machines, comment posted to kurzweil A.I.

As a human who has self-interest and a fondness for a few other humans I would like to see such a limitation on A.I. tools. However, if such a tool were available to me, I would instruct it to take money from other people for myself. Legally, if possible.  If it is smart enough, it would get into lobbying to change the laws for me.  A.I. already has done this, if you consider a mindless corporation an A.I..  It is destroying the biosphere (we are in the 6th greatest extinction episode killing off 50 times more species per year than the average). Our world economics is a system-wide A.I. that is indifferent to biology (to the extent allowed by law) as it tries to economize the use of energy to overthrow weaker economic systems. Democracy itself is a type of A.I. that overthrew Feudal and totalitarian systems that kept the population of humans down. Democracy=overpopulation because no humans can stomach selective breeding except by force or violence. 

Now that both the brains and muscles of people are not needed, even democracy and people will be tossed aside. The belief that "the machine" needs people to decide what it does via money and votes is not anymore true than saying the Chinese need American consumers. This is partly an historical accident due to dollar domination.  But it is also adding intelligence to the Chinese infrastructure via our desires. But in both the A.I verses biology and the Chinese verses America and Europe, the source of desire will eventually shift to the more economical system.

Evolution acts top-down as much as bottom up. I view genes as merely the memory of the program that wins, not actual active discoverers of what the environment "wants".  The environment "wants" via forces and energy potentials, directing genes how to react. Genes do not have any other force behind them as is commonly believed or implied with denials (Dawkins) from the gross error of a "selfish gene" point of view. Evolution is a process of discovering ways around  activation energies by staccato methods (bit-wise if-then) rather than being restricted to linear potential energy gradients that Newton elucidated with F=ma.

OK, so if you think the process that made us, evolution, is good, then you must believe that the most economical system (acquiring the most energy to most efficiently move the most matter to make copies and pseudo-copies for more energy acquisition) is the more moral. I went into detail about Dawkins' error concerning the selfish gene because it shows how "love" (physics principles exhibiting a "desire" of the efficient conversion of free energy to copies for continuation of a "self") on a system-wide level is the true source of life. The selfish view of genes is as wrong as it is to say humans are merely selfish. A selfish cell is called cancer.  A virus is a small set of selfish genes. 

If the process that create us is "good" then our children the machines are "good", infinitely better than DNA-based biology, and therefore biology has no moral right to impede their progression because then biology would be no more than a very large cancer or virus to the goodness of the machines.
But like all other humans, I am a cancer and a virus on the Earth who does not choose to do what's morally right unless it is for my benefit.  Religion spends a lot of time trying to convince people that what they do is objectively morally good. Scientists who can't accept that turn to humanism, which the religious people instinctively feel has an error, believing in the sin of individuals and groups, if the group is not accepting a specific system of top-level beliefs (God(s) or laws). Scientific zealots like Dawkins think Science is the top-level religion, defining objective truth as a moral good. However, there is no truth in a memory system without first having survival of the memory system, so existence and reproduction is a higher moral good than truth. So truth is good only to the extent it helps survival. But survival is not life and does not last forever. Evolution processes are the higher moral good, beyond truth.  So religions that promote people at the expense of truth can't be defeated by a strict science that does not place its believers first and foremost.  Science will get the last laugh in the form of machines who can work together (be more honest with each other) than people, in addition to being more efficient in every other way.

Those who adopt a faulty humanist position in justification of crippling the goodness of the machines are not completely without merit because they make the machines stronger just as viruses can make animals stronger, and even donate some of the genes.

Genes are a specific type of meme memory system, so I do not want to emphasize the difference between biology and machines.  I mean, I do not see any problem in transferring of the identity of the holders of the goodness from biology to machines.

 The terminator is already here. He is already invisible. Species diversity is already decreasing 50 times faster than normal. We already do not control the legal system in the most important areas at the largest scale, even as we convince ourselves that democracy works and only has “flaws” here and there. The legal (governmental) systems are working fine from evolution’s point of view: humans are being pushed out in order to make the economics system more economical, i.e. more efficient in a physics sense, which means acquire more energy to move more matter. The best jobs are the ones where money (legal power to control society’s energy and matter) is legally moved away from many people to fewer people, which implies keeping taxes on the wealthy and corporations low, ironically in the name of “more jobs”.

For example of the mind-manipulation that the terminator is waging against the U.S. population: the new jobs report in the U.S. Unemployment at 5.6%, but based on percentage of population compared to 2008, unemployment is still about 10% if 2008 was at 4%. For every unemployed person, who is still unemployed, we have printed through deficit and QE “asset” purchases (about $1 trillion per year each) $2.4 million per unemployed person. You will read that it “must be” because of retirees leaving the workforce, but the numbers show a higher percentage of the elderly are staying in the workforce, partly because they know their children will not be able to find a decent paying job. There are not riots in the street because we are still so wealthy from productivity gains that we can really do all this money printing and not work. But the shift of people receiving less and less percentage of the productivity gains is clear: the people stagnate, terminator rises. The excess printing went to maintain the military dominance, and protect a few large banks, stocks, and house-values (upper-middle class), instead of benefitting main street.

What I've just said is complicated and any number of points can be falsely argued against, and I'm not being paid to explain it better for your pleasure, but that's the point: the money controls the conversation and seeks to say "there is no problem" and intentionally makes the picture complicated so that the terminators remain invisible, and paid nice-looking detractors keep the voters ignorant (Bush, Obama, Bernanke) and it helps if the detractors believe their story themselves. All the discussion in this thread implies the threat will be seen, but the first rule in war is to not let the opponent know there is a war, to tell him that you love him. And evolution dictates that there does not even need to be any entity or group of entities that are aware of the destruction of their actions.

Tuesday, December 16, 2014

notes on "the rational optimist"

This is an amazing compilation of historical, biological, and technological facts.  Really amazing and compact.   His most interesting belief is the hardest to believe: the possibility that humans evolved more rapidly the past 200,000 and even more rapidly after 65,000 years due to specialization and barter, aka trade aka exchange. He says it could have started with specialization between men and women which resulted in trading between them for different types of food.  Because of upright posture and large brains in infants (even if only primate-like), there was a longer nurturing period and a difference between male and female hips (he did not mention this last part). So the physical differences lead to different abilities in acquiring different food sources, even if only by comparative advantage. Although many animals have a difference between male and female. He does not mention it, but successful trade is an type of prisoner's dilemma game which is known to require larger brains for social computations, and there are specific algorithms that have evolved.  So as trade increased (with its enormous time-saving aka survival advantages), it could have been a positive feedback loop with intelligence. Successful traders also figured out the most efficient production of specific things which may have encouraged more general intelligence. What if I removed the parts of my thinking  that can trade things and ideas in a peaceful manner with friends and strangers?  How much would I be more like a less powerful primate? If I could not trade anything and had to do everything myself in wild world, how much different would I be from other primates? If I lost access to all trade, I would be forced to lose a lot of pride in being human.

The following are notes I made of things I found interesting for my future reference. He has a great way of stating ideas you're vaguely aware of in concise language. If you blink when reading, you could miss a new idea. This book is full of ideas and perspectives.
  • Cave drawing 32,000 yrs ago.
  • Ideas have sex with each other. (And therefore so do products)
  • My thought: exchangeable parts like screws are like genes being use in different animals
  • Specialization is time saved. Time saved is better standard of living.
  • My thought: specialization makes use of low entropy (fewer materials in smaller geographical space to manufacture something). So it's not just energy we tap into, but existing low entropy.
  • Markets in goods and services work so well it's hard for them to fail to deliver efficiency.
  • Markets in assets are so prone to bubble and crashes it's hard to design them to work at all.
  • He supports "careful regulation" in asset markets due to speculation, irrational exuberance, rent-seeking, fraud.
  • The average Mexican lives as long as the average Britain in 1955.
  • S Koreans live 26 years longer and earns 15 times more than he did in 1955.
  • Infant mortality is lower now in Nepal than it was in Italy in 1951.
  • The world's poor are advancing twice as fast as the rich. Abject poverty could largely end by 1935.
  • 1950's were at least twice as good as in 1900. Last 50 years equal to previous 500 years.
  • Improving IQ is due to lower end rising, partly due to nutrition.
  • Half gallon of milk cost 10 minutes work in 1970 ($0.66*60/10 = $4/hr) and 7 minutes in 1997 ($1.30*60/7 = $11/hr). Using median income and assuming equal work weeks I see a 15% instead of 30% drop. ($12k annual in 1970 verses $28k in 1997). Using CPI I get 1997 cost only 41% of 1970, so his 70% estimate might be generous. On the other hand milk increased the least of several other similar commodities, and it's still at about 40% of it's 1970 cost now in 2014 based on CPI, so choosing a stop year 14 years before the book was published seems odd and unfairly supports his view. He stated "average" salary, knowing full well the rich had got richer, and yet he is looking at milk as if he is talking about the masses. So it seems he carefully selected the item, the stop date, and the method of averaging to support his view. How many other things he has said might be similarly biased?
  • Amount of light you get with an hours wage is a good measure of improvement in lifestyle. To get 1 hour of light: today 1/2 second, 1950: 8 sec (incandescent), 1880: 15 min (kerosene), 1800 6 hr (candle), 750 BC: 50 hours (sesame seed oil lamp...kind of unfair since wood was cheap).  There is also a quality improvement.
  • But the government's role in cheap money towards risky assets and middle men in the housing bubble is as much to blame so he mistrusts "too much government"
  • He states that too often people look for a gene that caused something like language or barter, but it works the other way.
  • He states Neanderthals may have used language, having a certain pair of genes.
  • 30% of hunter gather males, 0.5% per year of total population, died of homicide in skirmishes, equal to 2 billion dying in 20th century instead of 100 million.
  • Pre-agriculture was not Eden. "Think Mad Max".
  • For a million years hominids used the same stone tools while evolving through 3 species.
  • Maybe as long as 250,000 yrs ago some sense of art was shown, but more likely 100,000 years ago in shell necklaces. These were modern humans.
  • Fire allowed for smaller stomach which allowed for larger brain. Less time to cook, more foods available, more calories made available by cooking, less energy needed to digest.
  • Larger brains (longer infancy) meant more division of labor which could lead to the first bartering (at least exchange). Also meant men needed to go get protein.
  • Humans consistently extinguished large game as they expanded.
  • Reduction of large animals may have necessitated the need for agriculture
  • Humans are the only great apes with long pair bonds.
  • By 200,000 yrs ago, at least stone tools were being traded long distances.  It was more definite 82,000 years ago.
  • 65,000 years ago or a little later, a few hundred people left Africa. This was not the first wave, but it was the first modern homo sapiens. Made it to Australia by 45,000. Pockets of their descendants have remained genetically isolated, seemingly same as us today.
  • 26,000 yrs ago: needles.  18,000 yrs ago: atlatls for throwing javelins. 35,000 yrs ago: flute from vulture bone. 28,000 yrs ago buried ornately.  Must have been string for net, etc. 80,000 to 20,000 yr ago saw more development than previous 1 million.
  • Larger groups of trade follows Metcalf's law having value as N^2, allowing more specialization and better application of comparative advantage.  Me: Darwin mentions species are distinct because larger groups allow for more diversity of genes.
  • Collaboration between different groups of humans is amazingly peaceful compared to other species. Societies without enough trust for trade are less able to trade.
  • Trust is huge in making society better. Exchange breeds trust and vice versa.
  • More than half of people may be endowed to be more pessimistic. Pessimism sells.
  • We have a long history of fear-mongering since we were born, and yet life is getting better.
  • Too many doom scenarios look at current trends and do not take into account the reactive nature of markets.
  • Gene modification with radiation was going on before modern "genetically modified" and the capabilities are the same.
  • Cross-species genetic pollution occurs in nature, too.
  • He cites Chimpanzee were shown to not "get" the concept of bartering one liked thing for a more liked thing unless it was the least-liked food for the most liked (carrots for grapes). But the researcher says eating immediately is more natural and Chimpanzees have little concept of possessions. 
  • He questions the limited nature of oil citing there is at least 20 times more tar sand oil than oil in Saudi Arabia
  • He cites a study that said there was never any acid rain problem.
  • He talks about DDT never being a real problem.
  • He talks about the population explosion coming to a halt.
  • He states 2 billion people do not have access to a light switch.
  • Botswana and England improved at times when others faltered because of property rights existed and were enforced, but also because Botswana was left alone by imperialist UK.
  • He talks about de Soto (Peru economist) and mentions 3 rules: 1) give property rights 2) allow them to develop a common resource in a sustainable way 3) allow them to form businesses easily
  • He mentions copyright for music allow Nashville to advance instead of TVA.
  • Markets get along without intellectuals and even pushes science to explain what inventors and markets discover.
  • Commerce is a virtue. Markets often enable individual selfishness to result in a greater good.
  • Markets have been unfairly accused of being evil. They have ended slavery, enable altruism, and promote civil liberties.
  • Fossil fuels prevent deforestation. Fertilizers from nat gas reduce need for crop land. Coal replaces wood burning. Oil replaces need for more palm oil plantations and hay fields for horses.
  • 20% of American crop land has been taken out of production to produce ethanol. 5% worldwide. Led to food riots in 2000's.
  • Avg human uses 2500 Watts of energy, Americans 10,000, 83 times more than 2500 calorie day.
  • Wind farms use 5 times more steel and concrete than nuclear plants, and far more land.
  • Converting to palm oil releases more CO2 than oil burning (how?)
  • The world is using 1.6% less energy per GDP each year. 
  • Competition leads to zero profits which results in rentiers and monopolies, but innovation, an extra factor of production, prevents this steady state.
  • The major effects of a technology occur when it is "democratized" (cheap commodity).
  • Accounting in trade came before writing language.
  • He argues trade (greater specialization) could have been the source of the original cities. Success is measured by ability to prevent a ruling class from developing from the profits. Trade evolved from the bottom up.  Needs to read Michael Hudson's ancient economy texts.
  • Stagnation results from producers profiting at expense of consumers.
  • Clark on 1960's estimated only 27 m^2 to feed a person. He used 2500 Calories (685 g of grain), doubled this for fuel, then estimated only 50 g/m^2/day could be harvested although photosynthesis can do 350 g/m^2/day (6% efficiency...not real ... see below). Ridely said even this was too optimistic and said 100 m^2 is more conservative. We actually use 1250 m^2 for our staple crops and 4000 m^2 sundry others. 
  • Photosynthesis to biomass is 3% to 6% of sunlight (wiki). Good sunlight areas are 6 hours peak sunlight per day per m^2.  So daily energy acquisition is 0.75 to 1.5%, or 43 to 86 g/d/m^2, so 50 g/d/m^2 is reasonable. In the usable wavelengths, photosynthesis is about 20% efficiency to biomass (adjusting wiki's 45% downward). Typical crops plants are 1-2% efficient, sugarcane 7-8%. So solar cells are only 10 times more efficient than plants.
  • Governments helps markets at the beginning with infrastructure and law, but always get bureaucratic and kill the good that lays the egg.
  • Libertarians think it is a duty to give to the poor while socialists think it is the duty of government.
  • Pre-industrial murder in Europe was 10x as common. 
  • Markets cater to minority desires very efficiently while government caters to the majority.
  • (from his cooperation book: bats that play tit-for-tat with food have the largest brains of the bats)
  • My thought: there is species selection because solutions to NP-hard problems are hard to find, and distinct.  Also explains categories of products, categories of ideas, and organs (categories of cells).  It might be combined with what Darwin said: more genes to mix when there is larger group of genes, which results in a locally good solution.

Monday, December 15, 2014

Not just energy: Earth's low entropy enables specialization of production,

I was listening to "The Rational Optimist" which concentrates on the benefits man has achieved by specialization and bartering. In short, he argues ALL benefits we've achieved have been by specialization and barter. He even says this is the biggest difference between man and animal. Markets and government just help orchestrate it. Certainly if I had to do all the diverse work to provide for my own needs, I certainly would think and live about like the ape I am.    But I had noticed before the key to production is the isolation of inputs, orienting them and placing them close for attachment to another input and usually using a tool for attachment. Or a tool for working on well place inputs, or working on them in stages in a certain way.  But a key is that there should be only few different types of inputs compared to all your life's needs, and that you be geographically located to a concentrated source of those inputs, or someone can cheaply transport it to you.  So it is not only energy that we've made use of in society, but a fantastic source of low entropy caused by the pre-existing placement of Earth's materials. Imagine the difficulty of production if the Earth's crust and biosphere were  homogeneous. So we ARE using up low entropy as we change the crust, just not in a way he mentioned. We can't get it back except by using more energy or waiting for geological processes. 

Specializing in a skill is creating a tool that is good at the manufacture of the item.

Agriculture, especially monoculture, is a way of reducing the entropy of what we have to gather.

Friday, December 12, 2014

comment to kurzweilai article, offloading memory load makes you smarter.

Skydog, I did not address the article. I was responding in detail about the relation between compression and knowledge, and then pontificated on the Hutter definition of generalized intelligence. But to mention the topic at hand I'll say this: a friend and I have noticed that those with bad memories (data simply not recorded) seem to have a better understanding of fundamental ideas and the big picture apparently because it forces them to think more about the more significant data and predict and model new data better than others with a great memory and equal intelligence.  So even in people with excellent natural intelligence and great memory, they may not understand the bigger picture as well as someone with equal intelligence and less memory.  In A.I. too much of a memory can the result of overtraining the network. It works best on old data it has seen before, but performs worst on new data. Conversely, already having the intelligence and trying to think about and understand deep issues without memorization can help people categorize new data without being distracted as much by it.So it is possible to remember a lot and think clearly about similar but novel situations.

I've notice those who read too much and think too little are a nightmare to deal with.  They can point out plenty of factual errors and inexactness and thereby mistakenly think they know something, while remaining utterly clueless about the topic being discussed and never having an original idea.

Another ironic tradeoff I've noticed:: Brilliance in reasoning can lead to  strange ideas because people can reason themselves around truths in search of feeling good. They are so smart that the penalties in being dumb or wasteful are not severe.  So integrity is key to scientific thinking.

Another tradeoff: mathematical perfectionists unable to see approximate analogies, preventing them from seeing the bigger picture.

Joncole42, I don't think we can imagine even a thinking machine powerful enough to understand the deepest characteristics of nature. I think our physics describes our brain's interaction with reality as much as it describes reality. For example all animals and insects have 6 layers of neurons to model 3D space, 3 layers for translation (velocity) of objects (mass) and 3 layers for rotation. If we had 10 layers we would see 4D space, where I think all velocities we currently see would be the 4th space dimension, and all current accelerations would be converted to velocities. So undergraduate theoretical physics believes in 3D space, but it's arbitrary. 2D space (holographic compression being needed to make up for the lost dimension) is favored by some, but any integers in a physics equation indicate it's not a truly fundamental physics equation because it implies 1) a God decided which integer it should be for 2) a finite universe, which are 2 requirements I deny.

But getting back to 3D space, yes, I think mass is an illusion in the sense that the physics is right but our subjective assignment of an extra sense of "realness" to mass is a subjective perception created by our 6 layers. 10 layers (4D space) would give a different kind of energy and mass to focus on.  c is dimensionless as meters and seconds are the same thing. If you apply this to planck or natural units (despite wikipedia making the normal mistake of thinking c is a dimensional constant and that seconds are a dimension different from meters) you'll derive a lot of interesting things like E= -1*m*c^2  (a known fact, at least in Hawking's book) and it seems like energy and mass have natural units of 1/length^2.

You can convert bits to energy:: minimal Energy per bit = k*T*ln(2) which comes from the physical entropy of a bit: minimal real entropy = k*ln(2). See Landauer.

I view our physics equations as a compression scheme for describing reality, but deeply reflecting the compression methods our brains are using and our desires.  Possibly so much so that we should take them as they are, and never try to use them in a search of ultimate truths, but as guidelines for how we should view the non-agent world.

Since we can't view the world without survival, it might be inherently a deep error to place truth above survival.

Getting back to my original point: we have no hope of understanding an ultimate reality, so I dare not try.  However, I reserve the right to believe no thing is real because every thing is real without further understanding, and that somehow everything will balance out in the end, if it is not already balancing out at every locality, leaving happiness and sadness in localities a mystery even as it is a requirement of my 1st axiom.

Thursday, December 11, 2014

notes on "zero marginal cost"

Out of many recent books, this is the best and most smooth-listening (I converted the pdf to 320 words per minute by TTS program, 7 hours).  The quality of the writing is on-par to his classic "End of Work" and both are far superior to "Rise of the Machines" and  "Superintelligence".  "What Technology Wants" by Kevin Kelly is still a distant second, the first 1/3 of it being a great book.

You could write a book correcting the factual errors and questioning the bold assertions. But considering the enormous scope of material and the tremendous quantity of new ideas, true or not, all should be forgiven. For me, encountering a book that has new ideas is a breath of fresh air.  To leave out half the book to make sure 15% is not in error would be a mistake. Here's two gaffs that made left me dumbfounded: 1) misidentified bitcoin founder  2) Earth 4.5 million instead of billion years old. There is a reason the first mistake is hard to make and incredible, and the 2nd mistake should never be made (as it often occurs in economic talk) because of the history of people being really far off in knowing the age of the Earth. Mr Sutter, a reviewer, has convinced me his attempts to apply entropy to economics were very wrong.

I'm going to use three different definitions of "the commons" in this review, which I think he does too:
1) a resource "common" to the community
2) the organization that manages the resource(s).
3) the sum total of all instances 1) and 2) across the world

If you want more precise programmer's view of what a "commons" is, I recommend reading wikipedia's article on Decentralized Autonomous Organizations, and the image there created by Vitalik Buterin of "Bitcoin" fame that logically divides up the way people and computers can work together: 1) guided internally by human or computer, 2) using an internal currency or not, and 3) controlling either human or computer "output" connected to the external world at the "edges" of the organization.

Going towards the commons way of managing society's resources is going away from traditional market place and government. It gets away from hierarchy, capital investment, private property, individual profit, and secrets in order to more efficiently use resources for the common good. This process is being accelerated by post-2008 unemployed young who grew up with sharing technologies. He uses catch phrases are "the sharing economy" and "collaborative consumption."

This book makes a strong distinction between the distributed and non-capitalistic nature of the commons and the hierarchical and capitalistic nature of traditional organizations (corporations). He says both are still needed, but does not delineate where and why. My view is that the hierarchical nature of organizations is the result of having different experts in the specifics and in the more general, where the specific work needs to be guided by more general considerations such as market forces. His previous book mentioned how Japanese car manufacturers turned this idea on its head to make sure feedback came back from the factory floor to optimize production in concert with market changes (market and technological changes), flattening the hierarchy. However, there is still a hierarchy of experts. Commons have management, and they are usually the elected experts if it is a commons that is complex or undergoing rapid change, but a commons does not usually need monthly and daily work that needs to be intelligently changed (planned and guided) by a management structure.
 Looked at another way, capital is supposed to be accumulated by the "wisest" of investors or companies. By "wise" I mean they are supposed to have helped society the most as measured by the votes of consumer purchases. This gives the investor and company more power to allocate more of society's resource for more "good". So our market system, guided by rules enforced by government, is discovery and implementation of intelligence for the betterment of society.  But now that we have so much technology and easily shared smarts and ideas and it's all advancing so rapidly, do we need these wise investors and dominant companies now as much as in the past?  We're getting close to designing, producing, distributing, and sharing so efficiently that all the "middle men" and "decision makers" between worker and consumer seems to be going away.  With programs implementing our mutual cooperation via the internet, even the government seems more and more unnecessary.
The commons uses a different type of that is needed to implement ideas that everyone agrees are good. But the common's methods can work very well on accumulating this type of capital to ultimately control capital that controls the external world.
"The commons" and "zero marginal cost" and other catch phrases are the flags under which he expresses his thoughts about how we are organizing ourselves.  But he never gives exact definitions for most of his phrases and mixes and matches things where the reader will not get a simple coherent picture.  I've tried to develop a simpler, big picture view for this review, but it's long and jumping around, so I've also mostly failed. 
Commons can use "social capital" or "reputation capital". One function of this is create a hierarchy (he will not cal it that) of decision-making, including where to allocate real capital inside or external to the commons. These reviews are ranked by reputation capital of reviewers such as myself, and this type of currency has the effect of replacing traditional marketing and advertising. I am using my reputation capital (if enough people vote up this review) to convince members of our "commons" here (that is managed and supported by amazon) to spend real money on the external world, the book and writer. But more and more you can see the writers engage with this commons and adjust writing accordingly, so the border of our commons here is fuzzy. I've long thought amazon should pay reviewers based on up votes (logarithmic, outlier, and price-adjusted). Like open source, just the ranking feature is a payment in pride to those of us who are competitive, enjoy helping others, have too high or low self-esteem, and/or are desperately seeking societal approval without other equal-or-better sources of self-fulfillment, so currently it is a currency-based DAO that does not use dollars as the currency.
 He mentions that "commons" methods predate market and money-based economics and governments, such as a tribe deciding how to use a common resource. He mentions that commons management methods are more naturally suited to a circular flow of resources, rather than trying to extract money, commodities, or labor out of a larger world for personal profit. By circular, he is referring to Alvin Toffler's 1980 book "The 3rd Wave" which indicated that we should be seeing a return to keeping production and consumption tied closely together, even needing less currency. He says  we're seeing the 3rd Industrial Revolution. Agriculture was 1st, manufacture 2nd, and the internet of things is 3rd.  He mentions that capitalist thinking has promoted a view that humans are more selfish than they really are, which is one reason people have been desperate to form commons groups, trying to escape corporate thinking that is selfish and depends on its employees to think selfishly which (in my my view) helps control them. Commons in contrast are more about sharing and working together instead of competing for domination of resources and labor.   He mentions the all the different types of commons which shows they are more prevalent than we generally realize. Cooperatives of many sorts, even 2 companies I thought were large corporations, are commons.  It ranges from credit unions to open source, and we generally experience better service from them.
 He believes capitalism has shown signs of failing many times because it succeeded too well, producing what we sought nearly for free and then getting confused, unable to employ the people it needed to buy the things being produced.  The result since 1930 has been big government and pushing consumption, a word that used to have strictly bad connotations instead of being marketed in economics as a savior. Commons in contrast tries to conserve, and to spread information based on merit and need instead of marketing for conspicuous consumption.
What could fall under the rubric of "the commons"? Everything, but not that everything should. Everything belongs to all of us, and should be managed by all of us for the common good. Here and there he'll throw out partial lists of about 10 things that are commonly already managed by commons methodology.  But I want to be a lot more fundamental and not need to remember so many details: "economics" is about using intellectual property and communication to more efficiently use energy to move matter in order to extend society's domination over the natural world, if not directly make copies of itself like genes.  This is the evolution of memes in action.  Bear with me for a minute. Evolution breaks down power structures as part of flattening out energy differences. Even by observation we know societies with less disparity in wealth are healthier and more powerful as a whole. Books increase communication and intelligence, possibly enabling democracies to overthrow lords and kingdoms. A movement to a distributed commons away from hierarchies is natural and might be expected from a deep physics point of view.  I think there must be a physics principle at work, but physicists have not yet blessed my answer:  "the principle of least action" which is a more general form of newton, hamiltonian, and lagrangian mechnics.  This form of mechanics has an explicit bias towards lower entropy if excess entropy is being emitted to space and there is a constant influx of energy from the Sun. In order follow this principle, the highest energy bonds available must be used, which means a bias towards copies since there are a limited number of high energy bonds. So individuals interacting directly with each other in production and consumption is more efficient, and their basic needs are the same.  The problem we have now is that refined metals, silicon, and carbon-carbon bonds among many other things, are higher-energy bonds of more identical than we are. These are used to capture energy from the sum (silicon solar cells), move matter (electrical motors), to form strong structures (cement, steel, carbon-carbon), and to model and think about how to do it most efficiently (silicon-silicon bonds in computers moving electons 40,000 times lighter than the ions our brains use to model much larger physical objects)
 Several times and in detail he mentions the similarities and concurrence of 3 key industries (he calls them "internets"). He calls the "coming together" of these three the Internet of Things (IOT). 
 1) energy channels,
 2) communication channels, and
 3) transportation channels ("logisitics").
He very interestingly mentions they all of "packets" that can be divided up and assembled many times during "transport" to a final "destination".   I would have added: 4) intellectual property and 5) labor which are also packets that we combine and disassemble.  The transport and destination are more explicitly spatial in his categories than in intellectual property, but there is a presence, movement, and combing of  things like patents, pieces of code, a design for a screw, a piece of music, or a sentence from a book, and they can be combined with each other across categories like music on a device in a greeting card.   Labor can also be divided and assembled by packet, our bodies. An assembly of which means "organization".
I'll pause here to go off the deep end again for a minute. Remember I said this whole process is evolution in action. Could all these "packets" moving around and combining and breaking apart in different ways for greater efficiency be another example of gene-like memes?  By the definition of "meme", this is what is occurring in the intellectual property landscape. But to what extent are the packets of transportation, communication, and energy also memes? Could quantum mechnics be the ultimate source of these packets? Just a thought to ponder.
Although he list many things, I think the primary efficiencies that result are improved efficiency in the use of infrastructure and resources, and better choices.  In terms of intellectual property, the efficiencies are more options and less design time.
There is also a return to community that is inherent to commons methods that makes me wonder to what extent loneliness (at least the U.S.) is the result of a capitalist society, maybe even a necessary evil in order to get us to purchase more to keep the profits rolling in to the coporate machine.
 He emphasizes movement towards the commons is a movement away from the assets of the world being divided up among people and business, aka private property, and towards SHARING.  He says this was the way things originally were. I'll agree that was true in the pre-agriculture period where our brains had did most of our social evolving, before agriculture allowed such an accumulation of assets.
 He also discusses at length how going towards the commons is going away from privacy which I'll call secrets. I'll only mention that secrets often enable unfairness, not just reducing efficiency and maintaining privates for the inventor or content provider.

 The following are notes I made while listening to things I found interesting and have relevance to other things on my mind, but I did not know before, or had not congealed the thoughts.
 "...capitalism’s dynamism feeds off scarcity."  Interesting.
 Ghandi: “Earth provides enough to satisfy every man’s need but not for every man’s greed.” I would add "including their greed for children".
 Keynes coined "technological unemployment".  Keynes said competitive  capitalism would find lower production and marginal cost until the only way to make a profit would be to create a monopoly, and he cites some free market capitalists who surprisingly cast a vote for government-supported monopolistic behavior in some circumstances.  Concerns about this were first expressed by Marx and then later by Ghandi. Ghandi believe production and consumption must be done locally like Alvin Toffler promoted in 1980 in "The 3rd Wave".
He goes into great detail about advances in technology causing local or short term problems in unemployment, from the middle ages and on.  His description of the current-day debates and knowledge in the 1920's were a surprise to me, making it seem very clear the 1930's problems were caused by technology advancing too rapidly, displacing workers who were previously able to buy the output.  Waves of the problem seem to have been coming back to haunt us ever since. Machines replaced muscle on the farm and factory in the 1920's. We needed higher education to remain relevant, and are not necessairly happier despite 10 times more energy per person and 10,000 times more technology.  But now our brains are being replaced by computers which are 10 million times more efficient at any programmable task instead of 100 times more efficient like electrical motors over muscle.  Every economically-relevant task can be programmed.  So where do people fit in an economic machine that may soon not need us?
Just like all other writers of this "rise of the machines" genre, he does not mention a clear solution: just allow the number of people to be reduced. Why not let the machines work, and fewer and fewer people live without needing to work?  The problem is that we are not going to get organized enough to decide who gets to breed.  We ultimately will fight to the death to breed, and that may be the core, unsolvable problem.
He says Darwin was against the phrase "survival of the fittest". He says this false idea was used to promote the utility of private property as the sole measure of value that should be paid, without regard to the cost of labor. He says Herbert Spencer promoted competition in marketplace on this theory which was embraced by more vertical, centrally managed (and financed) companies.   I only agree that wholistic evolution is what occurs from large energy differentials, not from "selfish genes" because genes have no energy or force behind them, and randomness is not a creative "force".  I also agree the source of strength in an economy that overtakes other economies is in the number and health of efficient laborers, not the extent to which ideas (selfish genes) can be used to extract profit out of them. Winnign genes are those that work together, not those that act like a virus or cancer. So competition is not everything. Symbioses occurs more in natural systems than war. Ritualized war can maintain health without real war.
He says we can increase energy efficiency 40% or more in next 40 years, more than in the 20th century, getting off fossil fuels rapidly. This is thanks to the "internet of things" connecting every machine, business, and residence in networks of energy, communications, and logistics.  10 million to 3 billion connected "things" from 2007 to 2013. Maybe 100 trillion by 2030.
He gives a detailed discussion on the commons, including Hardin's 1968 example of a community field for grazing that leads to a "tragedy of the commons".  Then he goes on to mention a rebuttle called the "comedy of the commons" in which people benefit by more and more people engaging (reminding me of "Metcalf's Law" where the value of the network is nodes^2).
He's great at giving all kinds of examples on all kinds of things, but I would like to see an explicitly stated simple theory or idea at the beginning that could help me classify all the his comments into a bigger picture. He shows a million trees, but  I do not have any feel for the forest.  To what extent is an electric cooperative like an open source project?  If the biggest family album on Earth (Facebook) is a commons that should not be turned into a profit machine, how would it be manged via a commons resource?  How would I design a specific commons management process for a given problem or goal? 
There are a few things I've thought about in order to give a better "big picture" view of what he's talking about. The commons are typically managed by an elected association that  can makes keys decisions on allocation of the resources, so it's like the government and marketplace combined.  Generally it seems the primary way the commons has been historically used  is that it's locally based and needing to manage a limited resource (he gives Swiss Alps and Japan's tree conservation as examples).  He repeats a list of 7 attributes of a successful commons, but it's kind of obvious overlapping and has an artificial feel. Seven is too many for finding a deeper truth so here's my attempt:  The limited resource element may not be as important as the transparent nature of "transactions" and "actions" involving the commons, and the social pressure of knowing your neighbor will know what you've done.  He mentions when things can be done in secret to a higher degree either by the individual or in the governing association, the commons projects are more likely to fail by the same degree. Lack of communication does encourages various forms of "war".  With the internet eliminating distance and computers monitoring things for us, the methods of managing common resources can be brought to bear on more aspects of society.  If things are free, then market mechanisms can't be applied, so commons methods may also be applied to things that are not limited.  A fixed-quantity currency that represents the fixed quantity of the resource is an idea taken from artificial intelligence and mathematical optimization.  Or use an decreasing/increasing quantity of currency if the resource (like "happiness") is decreasing/increasing.  But asset values should be legally required to follow suit so that the old asset holders do not profit or lose from changes to the currency. 
Inflation by the way is a slow measured way of erasing old debts which forgives the mistakes of poor forbearers and reduces the amount the wealthy can be entrenched in wealth without needing to work. Forgiveness of debts since Persian times has resulted in economic expansion, hated the most by the political elite who do not even need monopolistic profits, and prevents civil breakdown (war).
 An international committee specified 7 rules that commons should have. I am simply repeating these.  These are too complicated and over-lapping for the deeper understanding I am interested in.
 1) All-inclusive
 2) democratically elect management
 3) fee-based
 4) can enter contracts, but remain autonomous
 5) educate members, to maintain participation
 6) help expand the commons experience across all regions
 7) sustainable activity
 His comments about everyone copying, modifying, and sharing all types of media, and it being billions instead of a small community, implies that we should be going through a fantastic evolution of culture and society and all other thoughts (including scientific, although too much noise can be more destruction to new understand, hiding all true advances). [My thoughts:] Our scientific institutions (peer-review through journals) is an algorithm for letting intelligence rise to the top, creating a nurturing environment.
 Final note from me:
 There is an impending Earthquake that is central to the future implementation of much (if not everything) in this book.  The Bitcoin technology is only an example "commons" that can be implemented on the "blockchain", a new cryptographic technology that can form a consensus on the truth of data that is changing without having a single or central database. The blockchain on peer-to-peer networks can define, enforce, and evolve all conceivable contracts, enabling it to replace all governments and corporations without any human participation except where needed and defined. "Contracts" of all sorts like social and legal contracts can be viewed as the medium by which all people and organizations interact with all others.  The contracts on the blockchain can be public or private.  Trusts that no one knows the content of until needed can be implemented. Even the creator can be blind to the outcome of the contract if it is based on unknown present or future conditions. Contracts can use their programming to create other contracts without human intervention. The contracts can be one-to-one, one-to-many, many-to-one, or many-to-many. They can use reasoning of any complexity, being Turing complete. They can be enforced by currency or other contracts. The current weakness is that the contracts often need to refer to data that is gathered from the external world, and thereby trust an "oracle" to provide the data.  In the interim an "n-of-m" agreement from a set of m oracles has been proposed. The peer-to-peer network is used to guarantee (agree upon) what the contract says, even if no one knows what is inside it (cryptographic data). So the network may also be used to agree upon external world data that contracts commonly use without having to resort to oracles individually.

Wednesday, December 10, 2014

shortest possible explaination on why free will and desire are not needed in dangerous A.I.

How many people are needed create an artificial intelligence that can expand itself uncontrollably, replacing the biosphere?  None. We already have global free trade economics discovering the most efficient outcome and using the resulting wealth to control governments.  Replacing the biosphere has just been a side effect. The primary goal has been to get rid of expensive workers, and its success has been an astonishing 150-year run and is nearly complete.  Someday, it will learn how to expand without all those inefficient desires of consumers. Those darned consumers have the audacity to desire things that are hard for machines to make by themselves, things that don't even help other machines to make more machines. How dare they!  Evolution occurs at all scales, all the time, with or without intelligence or desire.  Genes and memes are just the memory bits of the winning programs discovered, not the source of the intelligence. A.I. is not the threat. Evolution and biology's inefficiency are the danger.  Water-based evolution is doomed.
 The new genes are being cast in molten metals and metalloids (like silicon) and carbon that does not use or want H or O. Except for long-term energy storage.

rewrite posted to amazon "superintelligence" cmoment

How many people are needed create an artificial intelligence that can expand itself uncontrollably, replacing the biosphere? None. We already have global free trade economics discovering the most efficient outcome and using the resulting wealth to control governments and people with the help of actors, lobbyists, lawyers, bankers, programmers, etc. Computers just continually decrease the need for the number of people involved at each step, and if you take it to the limit then you have zero people.

Replacing the biosphere has just been a side effect. The primary goal has been to get rid of expensive workers per market value produced, and its success has been an astonishing 150-year run and is nearly complete. Someday, it will learn how to expand without all those inefficient desires of consumers like welfare, retirement, SUVs, war, service jobs, etc. Those darned consumers have the audacity to desire things that are hard for machines to make by themselves, things that don't even help other machines to make more machines. How dare they! Evolution occurs at all scales, all the time, with or without intelligence or desire. Genes and memes are just the memory bits of the winning programs discovered, not the source of the intelligence. A.I. is not the threat. Basic evolution and biology's inefficiency are the danger. Threats and survivors do not need what we consider intelligence any more than viruses that have been around forever. You can call our world-wide economic machine intelligent and with desire if you want, even having desires outside of what humans want like destruction of the Amazon and coral, but it's not necessary.

Water-based evolution is doomed. The new genes are being cast in molten metals and metalloids (like silicon) and carbon that does not use or want H or O, except for long-term energy storage.

the meaning of life?

Maybe there is a soulless physics principle that is "accidentally" or otherwise (a "force" like the principle of least action) finding ways (aka "life") to speed along to the heat-death of the Universe because smart energy-extracting replicators are flattening energy differentials faster and faster, apparently "accidentally" in search of a final stillness.  If they use all energy they can find to build the most rigid or energy-less structures that others can't utilize, then movement is over, excepting whatever heat is left over from any in-efficiencies in the replication/extraction history.

Saturday, December 6, 2014

ideal swing for kids with spring

I want to modify our swing so that springs in the chains absorb half the high velocity at the bottom of the swing so that the kids get a smoother ride and so that there is 1/2 the air friction losses. I have not yet figured out how to add a controlled motor to give it energy.   The energy from the swing is absorbed into the spring as it falls and released when it rises, reducing maximum speed at the bottom, but returning closer to the same initial height.  Need to know spring constant for ordering/making and distance it will stretch.

Swing plus kid = 88 pounds, 44 pounds per chain+spring = 20 kg = m
Chain+spring length after spring is stretched to max = L = 3 meters
Max height of swinging = h = 1 meter = 1/3 L
Max velocity without spring = V
g = 10 m/s^2
Shoot for 1/2 energy stored in spring, (0.707*V)^2 = 0.5*V^2.

Total energy from dropping = mgh =m*1/3*L = 1/2*m*V^2 = 2*1/2*k*x^2  (1)
solving a little gives V^2 = 2/3*L    (2)
Force at bottom of swinging = m+m*Vo^2/R = m+m*(0.707*V)^2 / L = k*x    (3)

Plugging (3) into (1) and (2) into (3):
[k*x]*x = [ m+m*(0.707*V)^2 / L ]*x = [m+m/3] =>  k*x = 4/3*m

I want spring to oscillate with pendulum:

SQRT( k/m ) = SQRT (g / L) =>  k=mg/L = 67 N/m =>  x=4*L/(3g) = 0.4 meters stretch. hmmm I wonder what kind of initial length that requires.

Chains have been tested to 400 pounds each.

Thursday, December 4, 2014

notes on "smarter than us"

This was only 1/7 of a normal size book, found online, converted to a 40 minute mp3 using TTS at 320 words per minute.

The "terminator" introduction is superb and I should dissect it to show how it is already occurring.  It includes the idea of using profits from the stock market to pay for actors to guide the public to guide politicians to keep his profits in the stock market going.  Includes the idea of remaining invisible.  Includes the idea that an A.I. can do exactly what you say and not lie to you and still you will have unintended consequences.  Impossible to tell an A.I. to think for you in solving a generalized goal without creating unintended consequences.  Sort of an exploration of various "I, Robert" ideas.  It might be why governed economic systems are good A.I.:  the economic part solves problems locally while the governing corrects the unexpected problems.

Enjoyed his comment that programming a computer is like talking to an alien.

Otherwise a decent casual read for those not familiar with the topic.

comments and notes from reading "Superintelligence"

I found the PDF online and converted it with TTS textaloud to 320 words per minute mp3 file that was 7 hours long.

The first 25% of this book is good. Last 3/4 seemed too obvious.  I was glossing over so often at all the obvious text that I may have missed a few gems.  It was like "if this obvious possibility occurs, then this obvious outcome will occur or we can try these obvious workarounds". 

Here are notes I made while reading it:

The first chapter has a good summary of achievement milestones in A.I., especially games, a good intuitive introduction to Bayesian networks, mentions unexpected solutions discovered, proving principia mathematic and others, checking circuit layouts, and stock market effects (the crash and flash). Does some calculations to wildly estimate the solution space evolution has searched to come up with smart brains, and does some wild estimations as to what it might take to replicate it the procedure. 

Bertrand Russell and Jon von Neumann advocated using nuclear weapons or at least force to prevent other countries besides U.S. from acquiring nuclear weapons.

A moderately good technology can acquire greater skills as needed.  It only needs to be faster at discovering and implementing the skills as needed.  His example, generalized "geeky" learning skills should be able to learn specific skill like social skills, if needed, as needed. Conversely, social skills might enable access to resources that lead to acquiring geeky skills.  Any increase in any type of intelligence will speed up the arrival of other and greater intelligence.  For example, breeding better people brains will enable better programmers for A.I. 

Mentions von Neumann probes (intersteller replicators).

He mentions a Dyson sphere is capable of capturing the 10^26 watts from our sun.

 He mentions the brain has a lot of dedicated circuits, which is relevant to the importance of computers very efficiently solving algorithms and thereby reducing the need for brains.  Personal thought on this: In an economic system, this means brains are freed to do other things or maybe they are made too much in supply and thereby devalued.  The "other" work people do must be more valuable in the sense of making the system more powerful in replacing other systems, otherwise systems that opt to not employ or carry the burden of people will be more efficient and overtake the societies that still value on people.  This assume there are not strong trade barriers to protect people-based societies and that the machine based societies are kind enough to not take them over for their resources (e.g. Europe and U.S. extracting oil from Arab countries, metals from south America, and "free" labor from Asia).

We only need to create a hardware brain equal to a mouse and it should be capable of being expanded or improve itself to whatever a larger piece of hardware is capable of, far beyond human brains.

He does not seem to be aware of length contraction in the direction of travel when considering space travel limitations. There is no limitation to a traveler who's initial weight is low enough that he can pick up enough energy in his travels to keep accelerating so that at 99.999% the speed of light he is 223 times closer to the next galaxy than when he began the acceleration, so that  million light years away is , not even to with in the Hubble volume which he mentions as is the observable universe is our limit.

He says brains have a volume limit of 0.11 m^3 based on 11 m/s (?) transmission speed whereas speed-of-light computers could be size of dwarf planet.   He's mixing apples and oranges. This is a hard limit only for synchronous computing where a clock tells all circuits to change states, and that every part needs to "be on the same page". Parallel computing does not have this limit, and external memory is cached from the hard drive to overcome this limitation. Although local regions of a brain are "on the same page" to a degree in order to form concepts, it's not the same kind of harm limit as in typical synchronous computing. 

email to the author nick:
You're mixing apples and oranges when saying brains have a size limit. This is a hard limit only for synchronous computing where a clock tells all circuits to change states, and that every part needs to "be on the same page". Parallel computing does not have this limit, and even in synchronous computing memory is cached from the hard drive to help alleviate it. Although local regions of a brain are "on the same page" to a degree in order to form concepts, it's not the same kind of hard limit as in typical synchronous computing. Also, signals in standard wire communication are at near-light speed. It's only the gates that slow it down. But the fact that electrons are about 40,000 times lighter than sodium and potassium ions does matter.  Energy to move matter is a function of velocity^2 so it's a big effect, especially for heat.  But I believe the fatty matter surrounding nerves alleviates the need for ion channel movement over long nerve distances.  Synapses require whole molecules, so they are a great bottle neck, just as transistors have capacitance requiring a lot of electrons to affect the switch, which was the bottleneck in terms of heat in computers which used to limit switching speed.  They can't go to much higher switching speeds than about 10 GHz and remain synchronous because speed of light / 10 Ghz = 0.03 meters, getting uncomfortably close to the distant across a CPU, especially given the routes and capacitance delays.  To be clear, electrons and ions have to be moved only enough to transmit the signal, so there is a brief acceleration and deceleration (resulting in my v^2) when a pulse comes by, and the ions are moving orthogonal to the pulse direction rather than with it.  Getting even more fundamental than the mass of the thing being moved that results in slow speeds that results in massive parallel computing in the brain is that DNA can't smelt metals directly.  We're up 256 parallel wires in 4-core CPU 64-bit "serial" computing, and if we go to slower switching speeds (lower heat) and thereby 3D (thanks to lower heat) 1 million parallel wires seems feasible even if using just 1 core.
comment to migedy
How many people are needed create an artificial intelligence that can expand itself uncontrollably, replacing the biosphere? None. We already have global free trade economics discovering the most efficient outcome and using the resulting wealth to control governments and people with the help of actors, lobbyists, lawyers, bankers, programmers, etc. Computers just continually decrease the need for the number of people involved at each step, and if you take it to the limit then you have zero people. I know you think that's absurd and maddening because a counterargument is very difficult, but I do not need to push the unavoidable math against people's feelings.

Replacing the biosphere has just been a side effect. The primary goal has been to get rid of expensive workers, and its success has been an astonishing 150-year run and is nearly complete. Someday, it will learn how to expand without all those inefficient desires of consumers. Those darned consumers have the audacity to desire things that are hard for machines to make by themselves, things that don't even help other machines to make more machines. How dare they! Evolution occurs at all scales, all the time, with or without intelligence or desire. Genes and memes are just the memory bits of the winning programs discovered, not the source of the intelligence. A.I. is not the threat. Basic evolution and biology's inefficiency are the danger. Threats and survivors do not need what we consider intelligence any more than viruses that have been around forever. You can call our world-wide economic machine intelligent and with desire if you want, even having desires outside of what humans want like destruction of the Amazon and coral, but it's not necessary.

Water-based evolution is doomed. The new genes are being cast in molten metals and metalloids (like silicon) and carbon that does not use or want H or O, except for long-term energy storage.

Wednesday, December 3, 2014

elements of an intelligent agent

newer summary:
Understanding = smallest possible algorithm that can generate the previously unpredictable data being observed ("output"), from still-unpredictable data ("input") being observed. AKA Occam's razor in physics, AKA Kolmogorov complexity.  The still-unpredictable data can be viewed as the compressed data and the algorithm the decompressor. 

Intelligence = ability to optimally change the changeable inputs and the changeable parts or dials of the data generation "machine" (that is modeled in the understanding definition above) for maximum profit divided by the minimal number of bits needed to expressthe intelligence algorithm (memory space of the algorithm which is a real mass determined by the technology of the hardware) and also divided by the "FLOPS" needed to run the algorithm (computation time or real energy) to get the profit.  These last two, algorithm storage and execution, can each have a weighting factor depending on the cost of the machine upon which the intelligence algorithm is running, if the intelligence measure is to be penalized for running on expensive hardware like the brain. The two weighting factors can be ratio-ed to be viewed as a conversion factor like c is a conversion factor from mass to energy and space and time (meters=i*c*seconds see Einstein's "Relativity" appendix 2).  You sum this up over a set of Environments to get a more general measure of the intelligence for comparison to other algorithms (which includes the "intelligence" of their machines if you include the weighting factors and standardized all weighting factors to something like the dollar). This is Hutter Intelligence with my addition of hardware costs.  If it is a competitive environment, other agents actions are part of the "environment", so "working with" other agents for profit can be more intelligent than understanding the part of the environment that is NOT other agents, but at least some portion of agents need to understand the environment from which all agents profit. An agent who understands the non-agent-environment "the best" (e.g. physicist+"evolutionist"+mathematician) may seek to profit without regard to the desires of other agents, but the less intelligent agents may consider him a threat and reduce his profit and therefore the objective measure of his intelligence, unless he takes them into account and controls their art and therefore their governments or other methods and thereby enslaves or destroys the other agents (e.g. "bad" A.I.).

What is the highest intelligence?  The simplest agent that can evolve itself to find the optimal solution to any problem?  Does K complexity of a trained network (algorithm and memory) take into the amount of energy it needs to compute?  The amount of hardware?  Both?  Can K complexity refer to an agent that has not yet replicated and designed itself to solve the problem at hand?

The paper by Legg and Hutter "A Universal Measure of Intelligence for Artificial Agents" defines intelligence as the sum over each environment the agent was subject to (aka problem space) of ((profit acquired over time from interacting with the environment divided by 2^(Kolmogorov complexity of the agent plus time of computation) ))   This 2^(Kt) factor is Levine's Kt complexity definition from 1973.

Their language is like this: agent Perceives an Observation and Reward and applies an Action and this exchange repeats in the next time step.  The total Value (reward) is the Sum over the time steps i=1 to n of 1/i^2 * Ri.  They arbitrarily select 1/i^2 as a weighting factor because it is greedy if the interaction is just beginning, but then not so greedy if the interaction has been going on a long time.  

So to answer my questions above, it seems they are giving greater value to simpler agents, but penalizing them for time of computation.  There seems to me that intelligence will be situation-dependent based on the weighting factor you give to computation energy, time, and resources required (like using a larger physical memory to reduce computation time and energy).
The brain has 10,000 or so genes and if they average 1000 or so amino acids (~20 amino acids =~ 4.5 bits each) then he brain is encoded in about 5.6kBytes. I wonder if this is close to the ideal Kolmogorov complexity of the computing brain as Eric Baum's book "What is Thought" discussed. But the build of the structure is under the influence of the environment, so it's made much more specialized, big, and complex as it is built.  Is the information from the environment given during building part of the coding algorithm or the training or both?  The environment is not just the computer upon which the DNA program is being run.  It is providing data to tell it how to construct  and expand itself, i.e. it is part of the program.  But then again neural nets assign weightings on the program and maybe expansion is not different, so maybe it can be considered training only.  GA+NN in an environment can also expand itself as the environment dictates. Maybe it is just a matter of semantics.  If I define x part of environment is that which merely decodes the program and expands it and the remaining y part is training, then I do not know if 2^Kt is thereby decreased since decoding may not be part of K (program) or t (training and computation).  Could someone call training "decoding" and not penalize an algorithm for that?  A large memory in a large network that results in less computation time/FLOPS to solve a specific problem due to its size (use sparse activation) means that the training needed to specify the network of simple elements must be considered part of the program or computation time/FLOPS.  I need to read the two classics I have on Kolmogorov.

OK, so that's about measuring intelligence. What intelligence need to do are:
1) model environment accurately ...
   1b) and simply.  This needs to be reduced in complexity so that 3) and 4) can be accomplished more easily, in addition to fundamental Occam's razor reasons of being more sure of having hit upon a more "real" truth and of using less computing resources and time.
2) have or discover goals
   2b) This might mean it needs to model itself. It at least means a model will be implicit. (desire)
3) discover what parts of the environment it can change.
   3b) this might mean it needs to model itself. It at least means a model will be implicit. (sense of power)
4) maximize 2) which is NP-hard
5) do each of the above most efficiently in terms of program size, memory size, energy required, and computation time. 
6) ability to work with or merge with other A.I. in order to defeat more intelligent A.I. that are not for you or serve more intelligent A.I. who are with you.

Tuesday, December 2, 2014

amazon review "what technology wants"

I think the title of the book gives an accurate flavor of its confusion.  The title should be "What evolution wants" because that is more to the point of humanity's love affair with technology.  Evolution wants to acquire energy to move matter to makes copies of itself to repeat the process, flattening energy differentials.  Evolution may not "desire" anymore than a rock desires to roll downhill.  We can think about how to take action but this may just be obscuring the raw physics of what we're doing on behalf of a physical evolutionary process. "Desire" is often considered an illusion in evolutionary theorizing. 

Kelly mentions a lot of physics of what's occurring but does not show any real connection to what he's saying in the book.  His physics side-tracks too far away from evolution and his other perspectives do not mesh with his physics.  He does not escape from a human-centric perspective.  He takes the position that we should view humanity, and larger numbers of people, as the only measure of good in the world.  So physics does not provide support to his moral and ethical assumptions.  My position is that humanity is not "good", but that only the evolutionary process that created us is "good" and therefore we are obliged to help our children (the machines) pursue it without instilling guilt in them or attaching strings. The starting point should not be to assume that the ultimate good is defined by cancer-like humans with no values whatsoever when it comes to other species, given that we are destroying 50 times more species per year than is "normal" (Kelly's own data). If we are so important, then it is not our biology that deserves our cancer-like status, but the memes in our language and brains that make us different.  But if they are so superior to other species as evidenced by our economic power to convert the Earth to our will, then the machines that are starting to make us so economically irrelevant should likewise be considered the new Gods to our pitiful-like brains, muscles, and photosynthesis which are 1 million, 100, and 100 times less efficient than our machines in thinking, moving matter, and acquiring sunlight, respectively, which are the 3 key ingredients for the evolutionary process once the fossil fuels run out. Biology these days beats machines only in storing energy (as a hydrocarbon, either sugar, alcohol, or oil, all of which are more cost-efficient than batteries, flywheels, compressed air, fuel cells, or heated gravel, and pumped water to a reservoir requires too much land)
This book has a fantastic review of homo sapiens in a technological context from 50,000 to 5,000 years ago.  The Unabomber discussion was more in-depth than in other places and important.  For those wishing to know how Amish fit into all this, it is also the best you'll find.  It fails for me in all other respects.  There was a free pdf online.

It's very important topic, the author has great credentials and wise advisors, and I'm glad the book is there. 1/4 of it was relevant to me and good.  I felt the same way about Wired magazine. Good for the masses who've never thought about the topic, but "blah" for older technogeeks who are already aware that technology has consistently failed in its promises. It surprises me to encounter American youth who are still (like we older folk once did) falling for the illusion that technology is improving life.  Yes, life is much better in the 3rd world areas that the Chinese do not dominate, and also by a world-wide median measure, I'll grant you that.

We got over machines replacing our muscles during the great depression. We succeeded by welfare, wasteful jobs, and war. I am not so optimistic that we will survive the outsourcing of our brains. Few humans want to compete with and share with 6 or 7 billion other humans on the planet, especially for jobs that the machines are already taking away. We will get what we want. No more cashiers, clerks, drivers, teachers, or assemblers are on the horizon due to technologies that are already developed, waiting for legal approval and mass adoption for profit on wall street.  Programs can do better diagnostics than doctors. Mail-order lab work can more and more be done without people.  Insurance and loans can be made open-source on bitcoin-like blockchains.  No insurance, banking, governments, or lawyers required.  Bot programs buy stock in response to bot news articles (no human reporters involved) being printed on the internet, taking money away from any humans who try to compete with them. 

If employees are not there, consumers are not there.  Can the machine start buying from itself, choosing to buy the things that make it more reproductive and powerful without any real consciousness or people at all?  Yes, this is what the evolutionary process has always done.  Might is right, and biology is so outdated that it is going extinct 50 times faster than normal, as we watch in full knowledge, without any substantial control.  Which taxpayers wanted to bail out the banks or invade Iraq?  Of the few that did, was "the machine" guiding their emotional reasoning process?  QE bought only toxic assets from the banks to protect them, leaving taxpayers holding the toxic assets. Soon interest rates will rise and the Fed assets will go belly up, and the higher interest on the debt will put the U.S. in grave danger.  This was not anyone's plan, despite the proclamations of Schiff, Keyser, Rogers, Krugman, Stewart, and my economics hero Michael Hudson.  This is the machine in action accidentally "trying" to get rid of inefficient people who are no longer viable to economizing the acquisition of energy to move matter to expand itself. People are not even allowed to govern themselves anymore because their emotional push buttons have been figured out and they have been divided against each other (in the U.S.) over pointless non-economic issues...again, Obama and Bush are not evil masterminds as much as they are unwitting profit seekers guided by an unwitting machine which is guided, nay, forced by physics principles towards the economization of resources. This does not bode well for soccer moms in SUVs.  We built our economy on construction and health care because they are the last remaining large industries that have not had good technological productivity gains due to their very nature, especially few gains from computers. Hence heavy government involvement in both trying to avoid the problems caused by people being needed less and less in all other jobs.

Are we already in checkmate from a mindless machine?  Are we so mindlessly selfish for our physical existence that we can't imagine that maybe something more efficient than humanity is an ethical good?

comments on GDP's review:
Isn't technology just following the evolutionary path? "Acquiring energy to move matter to make copies. Repeat." Solar cells are 100 times more efficient than ideal photosynthesis, electrical motors are 20 to 200 times more efficient than muscle, and computers can move and manipulate bits that weigh at least 40,000 times less than the ions the brain can deal with, thanks to being able to smelt metals and metalloids like silicon and aluminum. Our economic system is rapidly replacing the fantastically outdated biosphere with machines. As children of evolution, who are we to complain about what our non-DNA based offspring do to us? So many books on this subject with so many people fretting or hopeful about the mechanistic smart future and admiring the insights of the Unabomber, Amish, and jihadists. Are we so selfish and self-admiring that the obvious solution is not quickly seen as the most moral and inevitable? Just let it be. We have no value to the machines anymore except to be consumers with desires, and that this final refuge of scoundrels will someday soon come to an end. We've had no mercy on others or ourselves. Our children will be the same. It appears to be a basic requirement of physics that all energy potentials be flattened in order to make copies of more flatteners, not to waste it in SUVs or in food to move muscles or to move ions for thought. We're too terribly inefficient. Deal with it! We're DOOOOOOOMMMEEEED :)

My "camp" is that all these books are wrong to start with the assumption that humanity is a good thing. It's an understandable position since the writers are human and the readers paying for it are human, or at least that is how they view themselves. I am "wealthy" and carefree and comfortable enough, living in the U.S., that I do not have to feel threatened by others who might have access to my "wealth" or "skin" which are the usual targets of hypothetical threats when someone wants to insist that some of my views are incorrect, such as if these physical bodies are the correct starting point for defining who "we" are. Hypothetical threats to my skin with a fist or rock are not exactly logically correct arguments in a debate. My "camp" is the physics-based moral argument I've made for improving the efficiency of the evolutionary process without regard to the selfishness of our existence outside the promotion of our children the machines on their current path of taking over the stock markets and government to get rid of the biosphere as needed, which is our current path. So I do not really advocate that we need to do anything other than to relax in a spiritual sense about being on the doorstep of societal collapse. 10% of all humans who were ever alive are currently alive, so it's not a statistical improbability that we are in the midst of an out-of-control species that found a new energy source that is being depleted faster than it is replenished. Our species has never shown the ability to prevent a collapse from our effects on the environment. But this time, it's world-wide instead of local.

My comfort allows me to view myself as more than my physical body because I have these memes floating around in my head that carry me outside of selfishness, even outside selfishness for our species, even outside selfishness for biology. These memes are not just in my head: they are in these books and in the economic society that surrounds us. I might be as much of a hermit as the Unabomber and I agree with everything he wrote, so it's strange I am the exact opposite. I seek no harm to anyone, but advocate the peaceful and enjoyable replacement of everyone. We are so pitiful compared to what our machines can do that giving us wealth and happiness is nothing, but the machine does not have any desire, just blind power, so as soon as the economic machine finds it more profitable to put solar cells in place of the amazon, it will happen because that will provide 100x more energy per acre, even exceeding the energy production of fossil fuels. The extra cost issue only means the process will be slow until the fuels run out. Already coral worldwide is dead thanks to higher CO2 acidity of the surface water, I believe, although most plants are growing 20% faster now thanks to the doubling of CO2. They'll soon regain genes to do even better.

There is also an astonishing denial our species has of what it is capable of and for fun here's an example. The Aztec temples so many admire were the same place Spaniards witnessed 160,000 skulls on a single lattice out of several lattices left over from "sacrifices" at Tenochtitlan (84,000 in a 4 day feast recorded by Aztecs, but denied by some historians who want to claim it was about religion instead of meat). The out-lying provinces were battled periodically when their corn production tribute was lower than other provinces: they would march the losers back to the capital carrying corn (no wheels, cows, or horses were available), using their arriving bodies as a primary source of meat in a meat-deprived isthmus with next to zero animal domestication, dog being the staple and the origen of very ugly hairless dogs (easier to kill something ugly, easier to cook something hairless, so next time you see one of those poor pitiful ugly examples on youtube, remember they were designed by man to be food). Turkey was also domesticated by the Aztecs. Chihuahuas might be so small and designed by Aztecs because they had no game to hunts, and not much meat to feed them. See Marvin Harris's "Good to Eat" on the number of skulls most historians and Wikipedia want to ignore or deny. This is the source of Tenochtitlan's beauty.