Irving Wladawsky-Berger

Why Do We Need Data Science when We’ve Had Statistics for Centuries?

Data Science is emerging as as one of the hottest new professions and academic disciplines in these early years of the 21st century.  A number of articles have noted that the demand for data scientists is racing ahead of supply.  People with the necessary skills are scarce, primarily because the discipline is so new.  But, the situation is rapidly changing, as universities around the world have started to offer different kinds of graduate programs in data science.  This year, for example, NYU is offering two new degrees, - a general Master in Data Science, and a more domain-specific Master in Applied Urban Science and Informatics. 

It’s very exciting to contemplate the emergence of a major new discipline.  It reminds me of the advent of computer science in the 1960s and 1970s.  Like data science, computer science had its roots in a number of related areas, including math, engineering and management.  In its early years, the field attracted people from a variety of other disciplines who started out using computers in their work or studies, and eventually switched to computer science from their original field.

This was the case with me.  I used computers extensively while a student at the University of Chicago, where I worked closely with Professor Clemens Roothaan, - one of the pioneers in the use of computers in physics and chemistry.  As an undergraduate, I worked part-time at the university’s supercomputing center which he founded, and later he was my thesis advisor as a graduate student in physics.  When the time came to look for a job, I realized that I enjoyed the computing side of my work more than the physics.  I decided to switch fields and in 1970 joined the computer science department at IBM’s Watson Research Center.

Not unlike data science today, computing had to overcome the initial resistance of some prominent academics.  I still remember a meeting in 1965 with a very eminent physicist from whom I was taking a graduate course.  He asked me what I planned to do research on for my degree, and I told him that I was already working with Professor Roothaan on atomic and molecular calculations.  He just said that good theoretical physics should require no more than pencil and paper, rather than these elaborate new computers.  In his mind, this wasn’t real physics.  A number of the physics faculty felt the same way.  Change does not come easy, even for brilliant physicists.

Sexting Apps Versus Helping Cure Cancer

I recently read a very interesting article in the NY Times Magazine - Silicon Valley’s Youth Problem.  “In start-up land, the young barely talk to the old (and vice versa).  That makes for a lot of cool apps.  But great technology?  Not so much,” says its tagline.  Its author, Yiren Lu, graduated from Harvard last year with a degree in math and is currently a masters student in computer science at Columbia. 

I really enjoyed the article.  Ms Lu is a very good writer, storyteller and ethnographer, i.e., a kind of cultural anthropologist.  Her article chronicles what she calls the Smart Kids and their Sexting Apps culture of Silicon Valley, a youth bubble combined with “a frenzied bubble of app-making and an even vaguer dread that what we are making might not be that meaningful.”   

“There is a sense among them of manifest destiny, of This is our time,” she writes.  What do people in Silicon Valley plan to do once they hit 35 and are officially over the hill?, asks a member of Quora, a popular question-answer site.  “Despite its breathtaking arrogance, the question resonates; it articulates concerns about tech being, if not ageist, then at least increasingly youth-fetishizing.” 

This question brought to mind don’t trust anyone over the age of 30, a phrase that embodied the generation gap of another era, the sixties, when societal and cultural battles were fought in a number of areas, including civil rights, the Vietnam War, and the feminist movement.  While the generation gap Ms Lu writes about deals with very different, less weighty issues, there are a few similarities.

An Old Friend, - the IBM Mainframe, - Turns 50

On April 7, 1964, IBM announced the Systems/360 family of mainframes.  “System/360 represents a sharp departure from concepts of the past in designing and building computers,” said IBM’s then chairman and CEO Thomas J. Watson Jr.  “It is the product of an international effort in IBM’s laboratories and plants and is the first time IBM has redesigned the basic internal architecture of its computers in a decade.  The result will be more computer productivity at lower cost than ever before.  This is the beginning of a new generation - not only of computers - but of their application in business, science and government.” 

In April of 1964 I was a second year college student at the University of Chicago, working part time at the university’s computation center, which used IBM computers.  I still remember attending a presentation on the announcement given by a visiting IBM executive.  Over the next several years I used high-end models of S/360 for the physics calculations I was doing as part of my doctorate studies.  My thesis sponsor, - Chicago professor Clemens Roothaan, one of the early leaders in computational sciences, - was consulting with IBM on the design of future versions of S/360, and I was also involved in some of this work.  This relationship with IBM researchers led to my joining the computer sciences department at IBM’s Watson Research Center once I finished my studies in 1970.   

I was closely associated with mainframes through most of my 37 years in IBM.  I was involved in a number of research initiatives on the future of large systems.  After moving to the large systems products divisions in the mid 1980s, I worked on the evolution of mainframes to CMOS-based microprocessors and parallel architectures.  Later in the 1990s and 2000s I worked closely with the mainframe teams as they supported the Internet, Linux and other emerging initiatives I was involved in.   

Having lived through the near demise of mainframes in the early 1990s, - which would have inevitably led to IBM’s own demise, - their ability to have survived after all these years is truly impressive.  A look at the IT industry over the past several decades will reveal the large number of once great products and companies that are no longer around.  Few computer families could trace their vintage to the 1980s, let alone the 1960s.  There is something pretty unique about the mainframe being not only alive but doing so well after all these years.

Why is the mainframe still around, celebrating its 50th birthday last week?  What enables it to keep reinventing itself while embracing the latest technologies, including the Mobile Internet, Cloud Computing and Big Data?  In a world where product life-cycles are measured in web years, what can we learn from the mainframe’s rather unique longevity?

Data Science: from Half-Baked Ideas to Data-Driven Insights

In a Big Data presentation last year, MIT professor Erik Brynjolfsson pointed out that throughout history new tools beget revolutions.  Scientific revolutions are launched when new tools make possible all kinds of new measurements and observations.

Big data is leading to such a measurement-driven revolution, brought about by the new digital tools all around us, including our mobile phones; searches and web links; social media interactions; payments and transactions; and the myriads of smart sensors keeping track of the physical world.  These digital tools are enabling us to collect massive amounts of information on who we are, what we do and how we interact as individuals, communities and institutions.

The explosive growth of big data is in turn giving rise to data science, one of the most exciting new professions and academic disciplines.  Data science is a mashup of several different fields.  Its data part deals with acquiring, ingesting, transforming, storing and retrieving vast volumes and varieties of information.  Its science part seeks to extract insights from the data by applying tried-and-true scientific methods, that is, empirical and measurable evidence subject to testable explanations and predictions. 

One of the most exciting part of data science is that it can be applied to many domains of knowledge, given our newfound ability to gather valuable data on almost any topic. But, doing so effectively requires domain expertise to identify the important problems to solve in a given area, the kinds of questions we should be asking and the kinds of answers we should be looking for, as well as how to best present whatever insights are discovered so they can be understood by domain practitioners in their own terms.

Innovation Hubs in the Global Digital Economy

Last week I wrote about Cloud, Services and the Transformation of Production, based on Part I of Escape from the Commodity Trap: Will the Production Transformation Sustain Productivity, Growth and Jobs?, a paper by UC Berkeley professor John Zysman.  Part II of the paper deals with the economic and policy implication of this ICT-based production transformation. 

“For the advanced industrial democracies to expand the real income of the citizens and sustain growth in employment and productivity, their economies will have to escape from the commodity trap,” writes Zysman in the paper’s abstract.  “The commodity trap is the price-based competition throughout markets for standard goods and services, which puts pressure on wages and profit margins alike.  Clearly, the way out of this trap is to create distinctive high value added products  - both goods and services.  The emerging transformation of the production of goods and services is dramatically altering what is produced, where, how, and who captures the value.  It creates opportunities and challenges.” 

Early in the Industrial Revolution , the so-called Luddites smashed the new textile machines that were threatening their jobs.  Zysman reminds us that while many of the Luddites lost their jobs, the overall English economy prospered and the average standard of living went up. “It is easier to identify the jobs that are destroyed,. . .  than to imagine the ways new jobs will be created.  Of course, the character and location of those jobs may advantage some communities and disadvantage others.” 

Disruptive innovations, - like the power looms allegedly smashed by protesting Luddites, - have been displacing human labor for the past two centuries.  These periods of creative destruction, - when new technology-based innovations led to painful transitions in once-dominant companies and jobs, - eventually worked themselves out.  Over time, these same disruptive innovations rejuvenated the economy, and led to the creation of new industries and jobs.