Scientific Sense Podcast

Saturday, April 8, 2017

Living on luck

In a couple of weeks, an asteroid of nearly a kilometer wide will pass by the blue planet at just a million miles away, a mere whisker by any measurement. As the "leaders" throw hand grenades at each other, few have understood the precarious position their habitat is in. Reducing emissions to protect the environment is great, attempting to save it from instantly evaporating could be a bit better. Humans, carrying a heavy organ on their shoulders, seem to have wasted hundred thousand years, proving to themselves they are great and if not, thinking how to make them great again.
It is a miracle that the world survived in this active shooting gallery. Nothing much has happened since the first rulers of the planet were wiped out 65 million years ago by a massive event. And, the probability of that happening again is close to 1. As space agencies clamor to travel to nearby planets, just to demonstrate they are capable or design rockets that can take large and diverse payloads, they have to understand that none of these ego boosting triviality, will save humanity. As the most "intelligent" private investors, design, redesign and reuse rockets, as if that is the end game, they have to understand that there is a higher order problem to solve. Moving deck chairs on the titanic, optimizing how many chairs to move from one side to another or even creating a "game," that allows such movements to be automatic, is not going to be helpful at the face of calamity.
It will be ironic if the "greatest species," get wiped out by a similar event that happened before. At least the previous greats did not know it was coming.

Friday, April 7, 2017

Quantum optimization

Classical annealing has been a method of optimization in many areas, for ever. Recently, quantum annealing, based on quantum tunneling, was proposed as an improved process. Challenges remain (1), however, as qbits are immensely limited to make anything practical. The most profitable path to knowledge, at the intersection of quantum mechanics and optimization, may be breaking open. Flux quantization, as the authors argue, may allow us to tackle complex optimization problems.

It is clear that humans have been stagnant in the advancement of mathematics. At the intersection of theoretical physics and mathematics, there is a beautiful avenue to knowledge, something that has been less travelled in. There, determinism takes a back seat and probabilistic speculation dominates. Those who want interplanetary travel are simply leaving the movie theater before the show is over and it is possible that the few brain cells left to humanity could, indeed, take a quantum leap.

Humans wanted to always optimize as it has been ingrained in their structure from the moment they started to explore. They wanted to minimize distances and maximize the probability of a killing or analogously, minimize the probability of being killed. Ingrained in their quantum limited brain is the concept of optimization and now we could possibly take it further. In this context, it is important to remain humble as many before us have thought they indeed optimized, only to realize that there were many attributes beyond the equation that was used.

Flux quantization could provide a path forward to optimization but it is important to keep in mind that we have been here many times before.


Saturday, April 1, 2017

Hubble Trouble (1)

Hubble, from the University of Chicago, observed for the first time in 1929 that galaxies are flying away from us. Many candles later, the constant attributed to his name, is still being refined. Now, it appears that there is a revision to the constant, perhaps as much as 8% downward. This is not the first time physicists realized that they are yet to understand how things work. Complexity typically portends lack of understanding and the current state of affairs in high energy physics, albeit data emanating from big steel in Geneva like water from a faulty fire hydrant, is stagnant. Physicists and engineers are a dangerous combination, the former dreaming up complexity and the latter standing ready to prove anything thrown at them by steel and concrete. But this is unlikely to lead anywhere. It appears that one does have to have 1.5 brains to break out of the stalemate - and we have not had a specimen for 100 years and it is unclear if we will ever do.

Predicting the behavior of complex non-linear systems is dangerous. Making uncertain observations and attempting to prove those by creating dark stuff is even more dangerous, for this will lead the next generation down rabbit holes to the wonderland. Hanging mirrors to measure reverberations to the diameter of a proton, indeed, is a great accomplishment. But one has to note that engineers have always been good at proving hypotheses but replicability could be more challenging. The best way to proceed when one cannot understand over 94% of the observations is to ask if the underlying theories are correct. That's what we do in most fields.

As they revise the Hubble constant, perhaps, it is time to rethink how research is done in high energy Physics.


Friday, March 31, 2017

Open research

Recent news (1) that the European Commission will soon join the Bill & Melinda Gates foundation to launch an open research publication platform is encouraging. Increasingly, university based research, peer reviewed or not, are shown to be suffering from inherent biases - both of the author and her peers. Academics, tenured or not, seem to have a clean objective function - driven largely by volume and in the vast publication enterprise, replicability is certainly an afterthought. Open research, thus, has to expand beyond conventional research if it has to capture useful insights from those who never climbed the high mountain.

Research and manufacturing have become synonymous. The latter takes raw materials and create commoditized widgets and the former takes raw data and create research papers with no new insights and if there were, they are typically not replicatable. The peers, who approve of the production, seem to form a colony of conformists, ready to approve papers that support the highly subscribed ideas and marginalize anything new. Academics and politicians share many common characteristics as both of them live in a bubble with their "peers," and have little interest to ever look outside the window to see how the world actually works.

Open research, if it encompasses the world at large, providing an avenue for new ideas to float through, could be a game changer. But then, those behind it are all controlled by their "peers."


Saturday, March 11, 2017

The economics of health

The economics of health has been complex. Diverse stakeholders - patients, providers, manufacturers and payors - who all seem to have a partial picture of the whole problem, have made it more difficult to solve. To top it all off, policy makers, with little understanding of this complex picture, seem to trade in photo ops and healthcare acts.

The value accruing to society is an important and possibly the only consideration of policy. The evaluation of policy choices will require an understanding of the value of the foundational units - the individuals, themselves. Unfortunately (or fortunately), the value of an individual is not the discounted stream of utility to society and thus healthcare systems around the world, making decisions on a fixed value of a human, are misguided. Since the individual is perfectly capable of changing the future stream of utility, she can only be considered to be a basket of interacting options. Thus, any policy that forces the individual to prematurely exercise such options is unambiguously suboptimal.  Additionally, policies that reduce the flexibility afforded to the individual to pursue new ideas will devalue the options held by the individual. This automatically reduces the value of the individual, both to herself and to society.

For example, any policy that can be shown to reduce the incentive of an individual to move from one job to a more attractive one, will have a deleterious effect on aggregate utility. Any policy that does not allow an individual to climb out of a low value position, either due to a capital constraint or due to failing health, will reduce aggregate utility. Any policy that does not provide incentives to prevent health problems for individuals will automatically increase the expected future costs in the system. Any policy that does not allow market forces to move the complex system to an optimal position by sharing the economics among the various participants will have a negative value to society.

Politicians are likely the worst people to create and implement healthcare policies. It has been known for ever that "healthcare is this complex." It will take a bit more than theatrics and politics.

Sunday, March 5, 2017

Natural Intelligence

Recent news that researchers could now encode 215 petabytes of data in a single gram of human DNA (1) and possibly all the information created from human inception in a container of DNA that weighs couple of pickup trucks, is a constant reminder that Nature works in a dimension, possibly incomprehensible to humans. Recently, humans have been worrying about the growth of noise at an accelerating clip with over a zettabyte a year. Although some call this data and even information, an apt description of it is noise, as much of it is useless to advance knowledge.

More importantly, the capacity of data storage revealed by the DNA structure tells us that the status-quo technologies that some believe will propel us to the proverbial singularity are mere toys of insignificance. The only possible answer to the Fermi paradox is that extra-terrestrials find our level of development too premature to make contact. Evolution has been slow and biological systems seem to have taken almost 4 billion years to cobble together something that could start to think about knowledge replication and perpetuation. But the technology afforded to humans is too crude and likely in the wrong direction to make a significant leap toward harnessing high dimensional energy and tunneling through spacetime.

Those who are actively seeking "artificial intelligence" in Silicon may be well advised to return to studying natural storage and processing that appear to be fully integrated and freely exhibiting quantum states.


Wednesday, March 1, 2017

Pipe-climbing bacteria

Recent news (1) that bacteria can climb up a pipe at an astonishing speed of 2.5 cm/day is a constant reminder that the first inhabitants of the Earth is highly capable of eradicating those who appeared late. As they happily climb out of every hospital sink across the world, infecting their distant cousins, it appears inevitable that advanced biological entities will eventually succumb to potent simplicity. The information slope in the first cell has been so astronomically high that it seems obvious that incremental additions since then, have been trivial. Meanwhile, humans have been chasing extra-terrestrials and artificial intelligence as if they have conquered everything else.

Ignorance has been a constant companion for humans from inception. Their observations of the heavens always resulted in wrong hypotheses. First they saw the Sun going around the Earth and now they see dark energy and dark matter. In medicine, first they saw inexplicable spirits affecting the human body to create diseases and now they see pathogens and the body attacking itself. Alchemists have given way to modern chemists, manufacturing concoctions they don't understand to solve both biological and engineering puzzles. And in economics, they have created theories that seem to work only in regimes and those diving head first into the financial abyss have turned into TV personalities. But complexity does not advance knowledge - it just adds to ignorance.

As the ignorance in chief leads the world to the brink of disaster, it may be worthwhile to keep an eye out for the little things, that creep up the pipes.


Thursday, February 23, 2017

Forecasting society's prospects

The future prospects of human societies could be dependent on two important metrics - per capita knowledge and per capita compassion. It is difficult to define either precisely but proxies could be on offer to measure the slope of both. If outcomes of system behavior point to a decline in either of these metrics, human societies could be in trouble from a long term viability perspective. Even moderate increase in these metrics may not be sufficient to propel societies to the next level.

There have been troubling signs for many decades. Scientific pursuits impacted by high uncertainty often produce results that are less compelling to move aggregate knowledge, let alone per capita knowledge, on a positive slope. To make matters worse, there has not been a jump in knowledge progression for over a century, dominated by applications of existing knowledge. On the compassion axis, society seems to have bifurcated - some exhibiting high sensitivity and others falling victims to materialism. This is a potent combination and a perfect storm that may move humans back, perhaps for the first time in their history.

Education systems seem to have failed to improve per capita knowledge and segmented societies seem to have arrested compassion. And this poison cocktail has paved the way for ignorance to rise to power and that can initiate a spiral down in both directions - knowledge and compassion. They are correlated and that indicates that there are no obvious solutions in the horizon. To feel compassion, one has to deeply understand others, even those who appear foreign. And, to gain knowledge, one has to understand how the universe works, even the corners that appear exotic.

A break in the dam is sudden, an earthquake unanticipated and death arrives unannounced. Are there sufficient numbers among the 7.5 billion who could right the wrong? It will be a shame to throw it all away after hundred thousand years of progress.