Scientific Sense Podcast

Tuesday, March 29, 2016

Return to hardware

Hardware design has been losing luster for many decades in the world of computers. Software has been riding high, partly aided by hype and partly due to the younger crowd plunging deep into apps to make a quick buck and associated fame. Monopolies have been created on inferior operating
systems and office automation, while those who are opposed to it have been chasing public domain noise. Even educational institutions, following the latest ephemeral trends, with half lives of runway fashions, have been churning out courses with little utility for the future. Some have been putting content on-line and others still want students to toil under fluorescent lighting on wooden desks, while picking up the skills of the future.

Computer science has gone astray. Humans, susceptible to incrementalism, have been chasing false dreams on antiquated frameworks. Just as their predecessors, modern humans always attempt to scale inferior performance by massive parallel processing. They stack circuits ever closer and they network computers ever larger in an attempt to squeeze out performance, Meanwhile, software companies, hungry for speed and scope have created clouds of silicon that appear to suck up most of the production capacity in energy. Data have been accumulating in warehouses, some never to see the light of day and others, creating havoc and panic in complex organizations. Economists often worry about bubbles, for some are not so sanguine about rationality but technologists never dream of a software bubble as they presuppose such conditions.

It's time to leave synthesized voices, fake artificial intelligence and bleak games behind and return to hardware. Without two orders of performance improvement, there are very few apps that would move humanity and that can only come from practical quantum computing. Notwithstanding the much anticipated version X of existing operating systems and mobile phones, without innovation in hardware, humans will swim in a sea of mediocrity for ever. There are glimmers of hope, however. Recent news that larger quantum circuits could be built in more direct ways (1) is encouraging.

Educational institutions have an obligation to move society to the future and not just following trends that will fill up class rooms - physical or virtual.


Friday, March 25, 2016

Go AI??

Artificial Intelligence is in the air again. It is such a nice concept, the inventors of which have been suspected of nourishing the "God complex." Deep blue triumphed in chess and beat out mere humans in Jeopardy, Watson can understand how music is made and speak about it in a synthesized human voice, and now the famous search company has conquered Go. What's left in AI to solve?

Silicon has been alluring to engineers for four decades. They could double the speed of the "chip" in every 18 months and the mere extrapolation of this idea would have instructed even those less mathematically endowed that the belated singularity, is indeed near. Now that the game of Go, that potentially has infinite permutations of moves, has been conclusively solved by the electronic brain, we are likely nearing the inevitable. And that is bad news, especially for those in school toiling with such mundane subjects as computer science, programming and application development. Very soon, all of these will be delegated to machines, most of which would be artificially intelligent to a level, perhaps surpassing even contemporary politicians. Some had claimed decades ago that humans are nearing a state of "perfect knowledge." In Physics, the speculation has been that no mystery will remain in a few decades. Now humanity has taken an important leap to the future that artificial intelligence can quickly mop up any remaining mystery in any field - physics, medicine and even economics.

Chess, Jeopardy, self driving cars, neural nets seeking cat videos, twitter girl, Go... extrapolation certainly indicates the unstoppable triumph of artificial intelligence. The only remaining mystery is what billions of ordinary humans would do. The quantum computer they carry on their shoulders will become virtually useless in this regime of artificial intelligence dominance.

Friday, March 18, 2016

Scaling humanity

Reaching a critical mass and the minimum efficient scale are important concepts for many systems - biological, economic and business. Humans, separated by space and time for most of their history, could not reach this inevitable threshold for nearly hundred thousand years. Supported by technology, there are encouraging signs that we are fast approaching the minimum efficient scale of knowledge creation and consumption. The planet remains to be heavily endowed and it can easily support many multiples of humans as long as they are able to network their brains for the benefit of all.

What appears to be lacking is a framework. Weak attempts before, such as religion and countries, simply could not sustain a momentum that will unify in sufficient numbers to reach the necessary scale. Basic sciences, albeit attractive in many ways, could not light the passion underneath the human kiln. The strong forces that are operating to separate rather than unify, aided by the clan experiences of humans, have had the upper hand, thus far. However, technology is making irreversible impacts on the human psyche, propelling them to the next level. If so, they could make the planet, eminently contact worthy for outsiders.

Humans have been here before, however. In all cases, it appears that they have come up short. Insufficient technology for networking appears to be the common culprit in previous attempts. Stitching human brains together to reach the minimum efficient scale has eluded them. This was aided by hard constraints such as life span. Shrinking space and time as well as expanding life spans appear to be necessary conditions for sustainable development. Here, technology seems to show encouraging signs.

Space agencies and physicists lamenting about lack of "contact" may be well advised to ask why such "contact" would be made.

Friday, March 11, 2016

Mathematical music

Recent research from the University of Tokyo (1) that proposes a deeper dive into the structure of music by analyzing - "the recurrence plot of recurrence plot," in an effort to understand the emotive power of music, could be misplaced. Mathematical probing into the structure of creative work often failed to understand the substance of emotions that aid such phenomenon. Mathematics has been an important language in the history of human development. However, humans have been less perfect compared to constructs math could reasonably model and they often exhibit irrationality and creativity at random. It is the lack of "structure," that defines creativity and the effort expended by educational institutions in an effort to define such irrational phenomenon in a language that mathematicians can understand could be wasted.

Human emotions have been enigmatic - they escaped mathematical modeling thus far. Evolution seems to have been flexible enough to allow human behavior that has little value in hunting and survival. However, such work perpetuated the human psyche in a world of stress and tribulation, and lifted it into a realm that is mathematically undefinable. The visions of Einstein and Bach, unconstrained by mathematics, propelled humanity forward. As the engineers attempt to prove "gravity waves,' exist a century after it was proposed by sheer creative thought, one has to wonder if humanity is being sterilized of such a notion.

Mathematics, an idealistic concept, is inept at the analysis of human emotions.


Wednesday, February 24, 2016

Lawless innovation

A recent study (1) that argues that "constituency statutes have significant effects on the quantity and quality of innovation" in companies, seems to fall into the same trap of pitting stakeholder value against shareholder value. For many decades, the argument has been that companies and societies (e.g. Scandinavia) that focus on the value of stakeholders - employees, communities and the environment do better than those focused on shareholder value (e.g. US). This is a result of a wrong perception that a focus on shareholder value is based on "short term profits" and stakeholder value maximization is a long term process. There is significant empirical evidence that the market and investors are not "short term focused" and are fully capable of assessing and valuing any choices (short or long term) made by the mangers of the firm. Assuming that markets are myopic, without evidence, may not be a good thing.

It is important not to assume the first correlation found in the data is the underlying cause. Note that stakeholder value choices, unless they translate into shareholder value in any horizon, are value destroying. Further, "Quantity and quality of innovation," are difficult to measure. Few innovations are responsible for most of the GDP in the economy and in winner takes all markets, marginal benefit of innovation in aggregate is simply noise. A more interesting question is the structure, systems and strategies of firms (2) that encourage innovation. It is possible that innovative firms will remain so, regardless of the bureaucracies and statutes imposed on them.

Innovation emanates from the culture of the firm - not from the laws created by those, out of touch with the present economy.


Saturday, February 6, 2016

Innovative Life Sciences

Recent research (1) that shows Graphene could be utilized to interact with neurons open up a new avenue for research and practice to cure cognitive disabilities and possibly treat CNS diseases. More importantly, this is a profitable direction for biosciences to accelerate innovation. From the moment humans figured out they could impact the system by the ingestion of chemicals, they have been focused singularly on that. The system, however, is clearly electromagnetochemical, providing plenty of opportunities for more elegant interventions without multifactorial and unpredictable long term effects. Chemistry, has plateaued and life sciences companies with a vision of the future, have to move in a direction they are uncomfortable with.

Such an innovative departure in life sciences will take new leadership and a collaboration with emerging ideas and technologies. The impact will be far reaching - possibly replacing chemicals as the only non-invasive intervention. Medical education has to consider robotics, precision electronics and even high energy physics. Computer science and information science have to become integral to diagnosis and treatment. The meaning of intervention has to change - with impacts on the brain and the body simultaneously for optimum effect. In a regime of subdued bugs, unable to threaten the mighty human, it is going to be a battle against the body and the mind. Here, chemicals fail.

Innovation in life sciences will not come from incremental improvements to existing therapies, it will come from embracing hitherto unknown intervention modalities.


Saturday, January 30, 2016

Data science blindspot

Recent research from MIT that claims their "data science machine," does better than humans in predictive models is symptomatic of the blind spots affecting data scientists - both the human and non-human variety. Automation of data analytics is not new - some have been doing it for many decades. Feature selection and model building can certainly be optimized and that is old news. The problem remains to be how such "analytics," ultimately add value to the enterprise. This is not a "data science problem," - it is a business and economics problem.

Investments taken by companies into technologies that claim to be able to read massive amounts of data quickly in an effort to create intelligence are unlikely to have positive returns for their owners. Information technology companies, who have a tendency to formulate problems as primarily computation problems, mostly destroy value for companies. Sure, it is an easy way to sell hardware and databases, but it has very little impact on ultimate decisions that affect companies. What is needed here is a combination of domain knowledge and analytics - something the powerpoint gurus or propeller heads cannot deliver themselves. Real insights sit above such theatrics and they are not easily accessible for decision-makers in companies.

Just as the previous "information technology waves," called "Enterprise Resource Planning" and "Business Intelligence," the latest craze is likely to destroy at least as much value in the economy, if it is not rescued from academics seeking to write papers and technology companies trying to sell their wares. The acid test of utility for any "emerging technology," is tangible shareholder value. 

Wednesday, January 13, 2016

Favorable direction for machine learning

Machine learning, a misnomer for statistical concepts utilized to predict outcomes based on large amounts of historical data, has been a brute force approach. The infamous experiment by the search giant to replicate human brain by neural nets, demonstrated a misunderstanding that the organ works like a computer. Wasted efforts and investments in "Artificial Intelligence," led by famous technical schools in the East and the West, were largely based on the same misconception. All of these have definitively proven that engineers do not understand the human brain and are unlikely to do so for a long time. As a group, they are least competent to model human intelligence.

A recent article in Science (1) seems to make incremental progress toward intelligence. The fact that machines need large amounts of data to "learn" anything should have instructed the purveyors of AI that the processes they are replicating have nothing to do with human intelligence. For hundred thousand years, the quantum computer, humans carry on their shoulders, specialized in pattern finding. They can do so with few examples and they can extend patterns without additional training data. They can even predict possible future patterns, something they have not seen before. Machines are unable to do any of these.

Although the efforts of the NYU, MIT and Univ of Toronto team are admirable, they should be careful not to read too much into it. Optimization is not intelligence, it is just more efficient to reach the predetermined answer. Just as computer giants fall into the trap of mistaking immense computing power as intelligence, researchers should always benchmark their AI concepts against the first human they can find in the street - she is still immensely superior to neatly arranged silicon chips, purported to replicate intelligence.

It is possible that humans could go extinct, seeking to replicate human intelligence in silicon. There are 7 billion unused quantum computers in the world - why not seek to connect them together?