Tuesday, July 19, 2016

Shining a light on a black hole

Research from the Moscow Institute of Physics and Technology (MIPT) (1) hypothesizes a measurable and elegant difference between a black hole and a compact object. The event horizon of the black hole, defined by the Schwarzschild radius, is significant - anything slightly bigger shows fundamental differences in behavior. A beam of scattered particles shows discrete spectra in the presence of a compact object that escaped collapsing into a black hole. If it were a black hole, it will be in a constant process of collapse, with a complete stoppage of time for an external observer, resulting in a continuous and smooth spectra.

The concept of a black hole has been an enigmatic thought experiment for physicists and amateurs alike. Contemporary theory fails in the singularity and speculates a stoppage of time inside the event horizon, something that cannot be fully envisioned by humans trained in the practical regime of Newtonian Mechanics. A black hole will never stop collapsing from an external perspective and so there cannot be any ex.post question on a black hole. Theories that attempt more detailed explanation beyond the event horizon is fantasy - just as the mathematically elegant string theory that cannot be tested. In spite of all the engineering progress in the last hundred years, fundamental understanding has remained akin to a black hole - in suspended animation. A handful of men and women from the turn of last century remain to be responsible for most of the abstract knowledge that humans have accumulated. The reasons for this is unclear but lack of imagination appears to be the prime suspect.

Fooling around with mathematics may give contemporary scientists satisfaction but explaining the stoppage of time will require more than that.

(1) http://esciencenews.com/articles/2016/07/01/the.energy.spectrum.particles.will.help.make.out.black.holes




Thursday, July 14, 2016

You are what you learn

Recent research from MIT (1) shows that the underlying attributes of music – consonance and dissonance – are not hard wired. Contrasting preferences of tribes with little exposure to Western music such as some Amazon tribes and those with gradually increasing exposure, culminating in accomplished American musicians, they prove that the preference toward consonance over dissonance is learned. Music, thus, appears to be personal and preferences largely generated by experience rather than an innate mechanism in the brain.

In the contemporary regime of accelerating data, the brain is bombarded with an information stream, it was never designed to tackle. An intricate quantum computer, specialized in pattern finding but with rather limited memory, the brain has been stretched to undervalue its advantages and it has been struggling to keep large swaths of data in its limited memory banks. The learning processor, however, has been able to efficiently design and store information in heuristics and dump the underlying raw data as fast as it can. As it loses history, the stored heuristics drive function and generate preferences, as if they are part of the original operating system.

The finding has implications for many areas not the least of which is in the treatment of Central Nervous System (CNS) diseases such as racism, alcoholism and ego. Fast discarding of underlying information due to a lack of storage capacity, prevents back testing of learned heuristics. A limited training set of underlying data could have irreversible and dramatic influences on end outcomes. More importantly, a brain that is trained with misguided heuristics, cannot easily be retrained as the neurons become rigid with incoherent cycles.

You are what you listen to, you are what you eat and more importantly, you are what you learn.

(1) http://esciencenews.com/articles/2016/07/13/why.we.music.we.do

Tuesday, July 5, 2016

The failure of finite elements

Engineers and mathematicians, with a core competence in building complex structures from elemental and standardized components, have had a tough time with domains not amenable to prescriptive and deterministic logic. These include high energy physics, biology, economics and artificial intelligence. The idea that the behavior of a system cannot be predicted by its components is foreign to most disciplines and the applications of such hard sciences, supported by engineering and technology.

In complex organisms such as companies, it has long been recognized that outcomes cannot be predicted by an analysis of its components, however standardized they may be. The “rules of engagement,” if not defined in elegant and closed form mathematics, appear to be less relevant for those seeking precision. However, there is almost nothing in today’s world that could be defined so precisely and the recognition of this concept is possibly the first positive step toward embracing reality.

The interplay between physicists wanting to prove century old predictions and engineers standing ready to prove anything by heavy and complex machines, has been costly to society. The interplay between biologists and chemists wanting to influence systems with precise and targeted therapy and engineers standing ready to do so, has been costly to society. The interplay between economists looking to apply statistical precision to the unknown and engineers ready to build models to whatever is needed, has been costly to society.

Complex systems cannot be broken down to finite elements for the behavior of the system does not emanate from its components. 

Thursday, June 30, 2016

Cognitive monopoly

Major discontinuities in human history have often led to monopoly positions in subsequent markets, driven by winner takes all characteristics. In the modern economy - automobiles, airplanes and computers certainly fit this view. In the case of the internet, invented by tax payer investments, attempts by a few to monopolize the flow of electrons, have been averted thus far. But "Net neutrality," is not something that rent seeking behemoths are likely to accept in the long run, even if they did not pay for it.

The nascent wave - machine cognition - has the monopolists scrambling to get the upper hand. In this wave, capital, as measured by megaflops and terabytes, has a significant advantage. The leaders, plush with computing power, seem to believe that there is nothing that may challenge their positions. Their expectations of technology acceleration appear optimistic but nonetheless we appear to be progressing at an interesting enough trajectory. Although many, including the world's leading scientist, are worried about runaway artificial intelligence, one could argue that there are more prosaic worries for the 7 billion around the world.

Monopolies generally destroy societal value. Even those with a charitable frame, acquire the disease of "God complex," as the money begins to flow in. Humans are simple, driven by ego and an objective function, either biased toward basic necessities or irrational attributes that are difficult to tease out. Contemporary humans can be easily classified by intuition, without even the need for the simplest of algorithms - Nearest neighbors - into those with access to information and those who do not. Politicians and policy makers have been perplexed by the fact that such a simple segmentation scheme seems to work in every part of the world population from countries to counties and cities. Cognition monopolists will make it infinitely worse.

Can Mathematics be monopolized? The simple answer is yes. In a regime of brute force over highly available computing power for a few, answers could be found by the blind and the dumb. Perhaps, there is still hope for the rest as we have seen this movie before.



Saturday, June 25, 2016

Clans continue

It appears clear that habits formed over hundred thousand years cannot be changed in a mere hundred years. As homo-sapiens ventured out of the African Savannah, they were still tightly organized as small clans, less than a hundred in strength, with their own unique language, culture, religion and morality. In Europe and Asia, within hundreds of years of arrival, they erased their close relatives with astounding efficiency. They also successfully navigated disease and climatic change that reduced them to a few thousand - emerging out of the bottle neck, with even tighter clan relationships.

Technology - aircrafts, computers and the internet - opened up the modern economy in the blink of an eye. Economists, excited by the possibilities, argued for the opening up of countries, continents and economies, but they did not realize the behavior patterns integrated deeply into the human psyche. Countries cling to their languages and apparent cultural nuances aided by politicians who in autocratic and socialistic regimes seem to have convinced the populace that they can implement strategic policies that will make their countries, "great again." In advanced democracies, a larger percentage of the population, seem to have self taught the same ideas and in some rare cases they have found surrogates, who will sing the same tune as the autocrats, even though he/she does not know the words to the music. A dangerous trend has emerged in clans that profess to be democratic and sophisticated. The question is whether learning from mistakes is possible - something that made humans successful in the past. Ironically, in the complex modern economy, the outcomes are not clearly observable and often has long cycles. Getting mauled by a tiger is immediate feedback but having a stagnant and deteriorating economy has little feedback for the larger population.

The modern economy, still largely driven by the clan instincts of the seven billion that occupy the Earth, cannot be shocked out of its stupor by logic. Perhaps photographs from space that show the little blue spot in the midst of chaos may appeal to the artistic side of humans. Little closer, they will find no demarcations as depicted on maps and globes. After all, humans have shown great capabilities to think abstractly, albeit, such thoughts are not often tested by logic.


Saturday, May 28, 2016

Redefining Intelligence

Intelligence, natural or artificial, has been a difficult concept to define and understand. Methods of measuring intelligence seem to favor the speed and efficiency in pattern finding. "IQ tests," certainly measure the ability to find patterns and artificial intelligence aficionados have spent three decades teaching computers to get better at the same. Standardized tests, following the same template, appear to measure the same attribute but couch the results in "aptitude," - perhaps to make it sound more plausible. And, across all dimensions of education and testing, this notion of intelligence and hence "aptitude," appears prevalent.

However, can the speed of pattern finding be used as the only metric for intelligence? Certainly in prototypical systems and societies, efficiency in finding food (energy) and fast replication are dominant. Pattern finding is likely the most important skill in this context. If so, then, one could argue that the status-quo definition of intelligence is a measurement of a characteristic that is most useful to maximize a simple objective function, governed largely by food and replicability. At the very least, a thought experiment may be in order to imagine intelligence in higher order societies.

If intelligence is redefined as the differential of the speed in pattern finding - an acceleration in pattern finding - then it can incorporate higher order learning. In societies where such a metric is dominant, the speed of finding patterns from historical data, albeit important, may not qualify as intelligence. One could easily see systems that have very slow speed of pattern finding at inception if energy is focused more at the differential, allowing such systems to exponentially gain knowledge at later stages. Sluggish and dumb, such participants would certainly be eradicated quickly in prototypical societies, before they can demonstrate the accelerating phase of knowledge creation.

Intelligence - ill defined and measured, may need to be rethought, if humans were to advance to a level 1 society. It seems unlikely.

Monday, May 23, 2016

Salt water bubbles

Economists, closer to salt water, appear to be prone to thoughts of inefficiency and bubbles in the financial markets, something that can be cured by a single trip to the windy city. A recent study from Columbia University (1) asserts that they could find over 13,000 bubbles in the stock market between 2000 and 2013. Using supercomputers, no less, and "big data," they appear to have "conclusively shown" that stock prices take wild and persistent excursions from their "fair values." Unfortunately, these academics, who profess to be "data scientists," are yet to encounter the phenomenon of "random walk," further evidence that "data scientists" should stay away from financial markets. After all, the “physicists” who descended into Wall Street have had a checkered history of “abnormal returns” wrapped in consistent negative alpha.

The remark from a graduate student from Harvard - "I expected to see lots of bubbles in 2009, after the crash, but there were a lot before and a lot after," is symptomatic of the problem faced by “data scientists,” seeking problems to solve in super-domains they have no clue about, where participants, who determine outcomes are equipped with pattern finding technology. They may have better luck in real markets, for prices in financial markets are determined by a large number of participants, each with her own inefficient algorithms. The most troubling aspect of the study is that the authors of the study believe that “a bubble happens when the price of an asset, be it gold, housing or stocks, is more than what a rational person would be willing to pay based on its expected future cash flows.” In a world, immersed in intellectual property, where future cash flows cannot be forecasted precisely, the value of an asset cannot be determined by such simple constructs that have been rendered invalid for decades.

The lure of financial markets have been problematic for “data scientists” and “physicists.” However, a cure is readily available in academic literature emanating from the sixties.

(1) http://esciencenews.com/articles/2016/05/04/stocks.overvalued.longer.and.more.often.previously.thought.says.study

Monday, May 16, 2016

Small step toward bigger hype

Recent research from the University of Liverpool (1) suggests a method by which computers could learn languages by semantic representation and similarity look-ups. Although this may be in the right direction, it is important to remember that most of the work in teaching computers language or even fancy tricks, is not in the realm of "artificial intelligence," but rather they belong to the age old and somewhat archaic notion of expert systems. Computer giants, while solving grand problems such as Chess, Jeopardy, Go and self driving cars, seem to have forgotten that rules based expert systems have been around from the inception of computers, much before some of these companies were founded. The fact that faster hardware can churn larger set of rules quicker is not advancing intelligence but it is certainly helping efficient computing.

Engineering schools appear to still teach ideas that are already obsolete. Programming languages have been frozen in time, with prescriptive syntax and rigid control flow. Today's high level languages are certainly practical and immensely capable of producing inferior applications. Even those who could have "swiftly," assembled knowledge from previous attempts seem to have concocted together a compiler that borrows from the worst that have gone before it. As they proclaim "3 billion devices already run it," every hour an update is pushed or conduct conferences around the globe dotting and netting, the behemoths don't seem to understand that their technologies have inherent limitations.

Computer scientists, locked behind ivy walls, are given skills that the world does not need anymore.

(1) http://esciencenews.com/articles/2016/05/06/teaching.computers.understand.human.languages