Scientific Sense Podcast

Monday, October 26, 2015

Machines do not learn!

Artificial Intelligence is fashionable again, aided by cheap computing and cheaper memory, computer behemoths have been diving head first into the abyss. It has happened before, albeit, with much limited computing power. Just as in the previous iteration in the 80s, the current excitement may die down in a few years. Simulated machine voice and jeopardy are easier problems to solve but finding a cure for cancer is a lot harder. Similarly, churning through limitless text - “search”, is a mechanistic process but “curing death” is not. Soon, hopefully, the “leaders” will realize the naked truth – machines do not learn and that the “singularity” they envision is many centuries away.

Machines have captured the human imagination from the start. When they trained animals to drive farm equipment many thousands of years ago, they understood how machines could be built and powered to enhance productivity. More recently, the silicon chip, a conventional and mechanistic processor, has raised their ambitions to a level that may not be realistic. In the initial going, it was focused on productivity, very similar to farm equipment but now some believe their machines can learn. If so, this is a departure for humans from sustenance to imagination, from mediocrity to advancement, from tactics to strategy. But alas, machines do not learn!

“Machine learning,” a term used too liberally by information technology and consulting firms, may need further definition. One can clearly demonstrate machines do not learn by themselves and if so, we are back to using machines to enhance productivity – a posture humans have been in for ten thousand years.

Saturday, October 24, 2015

Real finance

Recent observations by a previous Fed chairman that it is puzzling why one would advocate a reinstatement of Glass-Steagall act because he could not see how the crisis in 2008 would have been avoided had it been in effect, is symptomatic of regulators who are good tacticians but not strategists. From Princeton to Wall street, financial institutions and their regulators still believe intermediation is “God’s work.” Academics, slaves to economics text books from last century, still believe financial intermediation is integral to markets and the economy. If the former chairman is intellectually honest, he may want to revisit the transcripts from Jackson Hole in 2005, before spewing wisdom. The one who came before him, who spewed infinite wisdom for many decades seem to have finally realized, “something was wrong.” He is still selling wisdom at $75K per pop. Meanwhile financial malls, that do everything but intermediation, with a skewed incentive to take risks for their bonuses that depend on the upside potential emanating from their actions and they are generally not accountable for the downside risk, invest into fooling both the academics and the regulators.

Although the chairman may be an expert of the depression era, what he may be missing is that the modern economy needs very little finance as we know it. Any financial activity, that does not have a direct connection to real assets and real investments, adds no value to the economy. Real assets today comprise of only two things – real estate and intellectual property. Financial intermediation in the former can now be done by the internet and it does not require “God’s men,” behind oak desks in the penthouses. And financing needed to develop and nourish IP is not something the mega banks have any clue about. They are dinosaurs with a ridiculous combination of disparate businesses with conflicting incentives that still pose a systemic risk to the entire population.

Glass-Steagall is a necessary condition to eliminate the greed of the ignorant and I am sure “God” will approve.

Wednesday, October 21, 2015

Hilbert hotel

Recent news that scientists have realized a quantum Hilbert hotel using a beam of light is indicative of the power of thought experiments, validated by simple experiments. Both interesting thought experiments as well as simple validation techniques are on the decline not only in Physics but also in other fields. In the Hilbert thought experiment, a hotel with infinite rooms, fully occupied, could make an additional room by shifting occupants one room up. More generally, the same hotel could make infinite empty rooms by shifting people to even number rooms leaving an infinite number of odd number rooms. The human brain that evolved through practical needs of Sapiens, has been lured into thoughts it was never designed to have.

Thought experiments on Infinity and zero (absolute nothing), such abstract concepts, are in a favorable direction to nourish the brain and perhaps move it into the next quantum state of knowledge. It takes no investment, no heavy machinery and not even high mathematics. It, however, takes imagination, something that has been draining out of humanity. Humans have been unable to internalize infinity and zero, perhaps because of the limitation of the brain but it is more likely because the brain gets inundated with the finite continuously. The thoughts that envelope the finite are mediocre at best and they are not extensible to the abstract nature of infinity or zero.

Thought experiments are possibly the best path to fundamental shifts in human knowledge.

Tuesday, October 20, 2015

ET’s very large SUV

Recent discovery of an object by NASA’s Kepler telescope around a star, some 1500 light years away that cannot be explained by typical exoplanet properties, has raised speculation that it may be an artificial structure, created by advanced intelligence to garner energy from the star. A simpler explanation, such as a fragmented comet could be more valid, but certainly less interesting. Astrophysics, quickly degrading into bedtime stories of fancy, may yet have a long way to sink in conventionalism, before it could make a contribution.

Humans, seeking intelligence elsewhere, with attributes similar to theirs, seem to forget that their own experiences clearly show that intelligence is a network process. The mighty bacterium, able to dance with billions in unison, appears much more intelligent as a system than humans could ever be. By substantially increasing the diversity and the probability of favorable mutations at the micro-level, the system is able to evolve and adapt faster in a universe that appears hostile in every direction. Harnessing energy in such systems will be an internal process and may not require structure-building or star-domination, only phenomena humans are able to understand fully. Granted, finding such intelligence will remain outside human capabilities for the foreseeable future, if not for ever.

The recent fervor in seeking extra-terrestrials, is symptomatic of the limited understanding of humans of the formation, development and propagation of intelligence.

Tuesday, October 6, 2015

Training Schrödinger’s Cat

A recent article in Nature describes a possible break-through in quantum computing. Engineers at the University of New South Wales seem to have built a quantum logic gate in Silicon, possibly opening up a pathway toward practical quantum computing in Silicon based processors. In theory, the technology could turn conventional bits into quantum bits, allowing traditional chip architectures to morph into those supporting quantum computing.

This is good news. As many philosophers worry about singularity, for no apparent reason, creative engineers have been moving the goal posts closer to practical quantum computing. It is a technology discontinuity, something if shown to be practical, could substantially change the economic prospects of a whole generation, currently stuck in a plateau of no innovation. They are beginning to realize that calling a phone 6, 7, 18 or 6S is not really innovation – it is packaging. They are realizing that creating fake computer generated voices in marketing announcing Sherlock Holmes and Watson, is not really innovation, it is bad sales technique. They are realizing that skipping version 9 and going straight to 10 is not innovation, it is a cheap gimmick. They are realizing that names that are difficult to pronounce are not really technologies but ways to siphon money out of companies who want to jump on the latest band wagons.

The next discontinuity is not in software – it is in hardware. Unfortunately, world's largest hardware companies do not seem to understand it and the world is now replete with incompetent software companies – creating operating systems, applications, "artificial intelligence" and databases – with no innovation content.

Saturday, October 3, 2015

Mirror, mirror…

There is an audible buzz in the air ever since NASA finally announced the “well kept secret” - there could be water – the flowing variety – on Mars. This is exciting news, and it brings back visions of Green men (I wonder if they were women) walking out of shiny space crafts – the “contact,” humanity has been waiting for a long time now. There have been more good news – subtle but nonetheless important – related to the same. A recent article in “Astrobiology,” shows “Earth like planets,” orbiting close to stars likely have magnetic fields – making them even more “Earth like,” allowing life to flourish like bacteria on a dirty towel.

It seems like we are nearing the historic occasion – predicted by NASA to be 2020, when life would absolutely be discovered outside Earth. All efforts are focused on finding the “twin Earth,” with Oxygen, flowing water, a magnetic field and a hard surface to stand on – where humanity will ultimately shake hands with an organism – 6 ft tall with two hands and legs, cone shaped head and (hopefully) with a sense of humor. Some of the greatest living physicists don’t think ET will be in a playful mood at all, for they argue ET will make contact only to extinguish humanity. Not sure who is right, but how could one blame ET to take such an extreme stance.

Mirror, mirror…. who is the best organism in the universe? Don’t tell me there is nothing else – the right answer is – humans who are superior to a vast array of sub intelligent creatures across the universe.

Sunday, September 27, 2015

The multiplicative effect

With world population at 7.5 Billion with only 40% having an internet connection, the information multiplicative effect on global economies is significant if complete saturation can be achieved. At the current rate, saturation is unlikely to arrive for another two decades. Accelerating this has beneficial effects on all economies and the UN could possibly lead an effort in this direction with investments from all participants. Countries such as the US that will benefit disproportionately, with a dominant presence in search and social media, may be willing to invest their fair share in this direction.

Information has become a basic necessity for humans – an indication that they are slowly maturing into a level 1 society. For 100 thousand years they struggled with food and sex, attributes of a primitive society and although half the world population is still locked into the same objective function for survival, there are reasons for optimism. In a fully connected world, there will be billions of brains – quantum computers in their own right – analyzing and interacting with newly emerging information at any point in time. With that, humanity could possibly escape from the trivialities of politics, clan conflicts and ego – and look outward in concert for the first time. It is a powerful notion – a brain storm with a billion brains is possible now and one can escape dreary conference rooms, conventions and the status-quo. We could reach a slope where only innovation matters and not tenure, color, elitism and even skills and know-how. Those who make new and valuable things will be the leaders and the next generation will look back in astonishment how a society with similar intelligence content as they could behave so inexplicably in the past.

The information multiplicative effect, so powerful that anybody ignorant of it should not be leading anything – countries, companies, localities and even households.

Friday, September 18, 2015

Elite efficiency

A recent paper (1) studies the “distributional preferences of the elite” and makes a somewhat obvious (startling to them) conclusion - “the elite” prefers not to distribute compared to the general population. There are many problems with the study including the premise and the definition of the “elite,”  comprising of the graduates of the Yale Law School (YLS). The assertion is that these graduates are destined to power and influence (and presumably wealth) and hence the “branding” of “elite.” Institutions and researchers looking backwards and frozen in time, may be in for a shock when they look outside their theoretical and historically adorned windows and see the future.

How do we test a hypothesis that YLS students prefer not to redistribute compared to a random graduate? Could we use the same study and data? Suppose we prove that YLS students are not “elite” at all because the probability of a YLS graduate to have any influence on society is roughly equal to a random graduate, what would it imply for the study? Economists tend to use fancy words and create complexity so that they can live within their secluded ivory towers, contributing nothing to society. In this context, what exactly does equality-efficiency trade-offs mean? Does it mean that the “elite” like to keep the money for themselves and phenomena that cannot be understood in this framework is assumed to maximize efficiency?

Practical educators and researchers should stop wasting time assigning labels and useless observations in a label prone, segregated and tiring society.

(1) The distributional preferences of an elite
Raymond Fisman1,*, Pamela Jakiela2, Shachar Kariv3, Daniel Markovits4
1Department of Economics, Boston University, Boston, MA, USA.
2Department of Agricultural and Resource Economics, University of Maryland, College Park, MD, USA.
3Department of Economics, University of California, Berkeley, Berkely, CA, USA.
4Yale Law School, Yale University, New Haven, CT, USA.