I believe that computing shall be even more pervasive and ubiquitous than it is today

and will evolve into a high-impact, use-inspired basic science1. Over the last seven decades,

we have seen the beginning of a major technological revolution, the Information Revolution.

Many of us who are middle-aged or part of the baby-boomer generation have been witness to

this growth. The last decade has especially seen computing emerge from its strongholds in

engineering and the mathematical sciences and has diffused not only into all corners of

academia (biology, social sciences, humanities, etc.) but also more impressively into quotidian

life. It is the latter aspect that will drive computer science and provide even more growth

opportunities. Witness the growth of companies like Google, which have made search and to

google an overloaded and essential verb. The social impact of computing is at its all time high

and its ubiquity and pervasiveness will only grow. In my opinion, growth will be in democratized

iComputing: computing for everybody and by everybody.

We are at the confluence of many perfect storms of opportunity. Processing and storage

hardware is cheap and is increasingly available in the cloud. Mobile and traditional

communication systems and the Internet connect us and our embedded devices, our vehicles,

our homes, and our offices into a global matrix. Yet, I believe we have seen only the tip of the

iceberg.

Therefore, computer science will be better served if it considers itself to be a broad

discipline with a myriad of interacting components and not just a slew of vertical silos wherein

practitioners design and build compilers, operating systems, architectures, and graphics

systems, while some propound algorithms and theory. It must also rely on disciplines other than

mathematics (e.g. physics, statistics) for intellectual succor. I increasingly believe that computer

science is a use-inspired science with a high-impact footprint where many sub-disciplines are

applied to solve problems. Use and value will drive the discipline of computer science. Further,

in the age of austerity and growth in the developing world, there will be an emphasis on use-

based research and development. Still, there will be a place for fundamental and basic work.

Consider this. Louis Pasteur, a founding father of modern microbiology, began his quest with

the more basic need of preserving food2. I believe that computer science will also head this

way, use-driven and use-inspired yet innate to the human experience.

In marked contrast to the early formative years of computer science, which overlapped

with the ebb and flow of cold war funding and patronages, the eco-system of today is

dramatically different. Computer science will not be just driven by the next new hammer that is

invented at the behest of high priests or driven by plain curiosity. However, there will be

incentives in finding nails everywhere around us, like geocachers, to construct new hammers or

adopt existing hammers. Consider Google’s PageRank algorithm. It should be noted that the

original and well-touted PageRank algorithm or hammer had a purported use. With the passage

of a decade, the venerable algorithm has to also act as an honest broker, thus bringing in

additional constraints to search and a rethinking of the mathematical and algorithmic

underpinnings.

I have followed this mantra in my own work. I embedded myself increasingly in the

physical and biological sciences by learning the subtleties and nuances of the other science and

realizing scalable, and robust methods and workflows. The interpretable and actionable end-

result – vortices in unsteady flow, distinction between dyscalculics from normal subjects, robust

subtypes of triple negative breast cancer – is paramount rather than the singular method. Often

new basic methods or foundational principles had to be discovered or re-invented within the

constraints of the user-science. Robust engineering practices already achieve this two-step

tango; computer science will increasingly adopt the same tango as it increasingly addresses the

needs of the larger population.

Computational modeling and data analysis will play an important role in defining salient

processes and associations that can be stored and processed in portable, scalable, and robust

prototypes and systems. These systems will be deployed in a multitude of human-centric

applications that in turn parlay human sensory perception and more usefully cognition. This is

my utopian albeit utilitarian view of computer science.

The role of mathematics and statistics especially in the age of Big Data cannot be

understated. Foundational work in theoretical regimes will continue; the increased emphasis on

high-dimensional and probabilistic learning algorithms offers one example. Similarly, 3D printing

and novel manufacturing processes could not have become a reality without the pioneering

work of geometer Prof. Herbert Edelsbrunner and many others. New systems and programming

languages will also be required. The growth of graphical processing units and probability

processors (Lyric Labs) will only usher expedited growth in hitherto unexplored application

areas. There will always be a need for compilers for all these specialized solutions; the

demands of a probabilistic language where a variable carries the semantics of a distribution will

require some deliberation. Human-computer interaction will be even more stressed and

text/speech/image/video/ processing will be eventually woven into a single tapestry of a user

interface.

Applications will form the inter-disciplinary bridges to others in social sciences,

engineering, and medicine and it is through applications that the user plays an important role.

Personalized medicine and all the associated advancements in the physical and biological

sciences will require a bigger role from computing software and infrastructure, while traditional

consumers in enterprise computing, finance, weather prediction, etc. will continue to place even

more demands on real-time services and storage.

In academia, there will be more branches and rivulets. Cybersecurity, systems biology,

finance, fluid dynamics, medical imaging, neuroscience, etc. already vie for attention from

academics in various departments of computer science. E-governance, social media, social

innovation, and smart cities will follow, especially spurred by the availability of Big Data3. Each

of these inter-disciplinary growth opportunities will lead to a reexamination of cross-disciplinary

fault lines and new flurry of research will ensue. The exact shape and form of these changes will

depend on the local academic, business and cultural environment.

The delivery of curriculum will also experience a sea change. There is certainly an

interest in all things computing; we have seen this in burgeoning enrollments at both university

and K-12 levels. Some of the new improvements will be of vocational nature. It is therefore not

surprising that one speaks of imparting the three Rs and C to the uninitiated4. Similarly, there

will be an increased emphasis on active learning in flipped classroom settings where students

actually engage in problem solving while being part of collaborative teams. This pedagogical

approach is increasingly shown to be effective in more general settings5. In many settings,

computer science is already taught in this manner and is likely to become the norm.

In closing, it is my opinion that computer science will adopt a utilitarian face and will

make even greater forays into interdisciplinary ventures while strengthening all components in a

cross-disciplinary fashion. By placing computer science squarely in Pasteur’s quadrant2,

practitioners can engage in a rigorous albeit a very relevant science and in essence enable

computer science to play an even more important role in modern life and human civilization.

This is a guest post by Raghu Machiraju, CEO, Abiobot (IndieBio Alumni). Raghu also serves on the faculty at Ohio State University in the Department of Computer Science and Engineering.  He also has an appointment in the College of Medicine at OSU.  His interests include image analysis and visualization especially as they apply to topics in biology, medicine and engineering. Over the years he also been increasingly working on problems of computationally biology and bioinformatics.

1 Marc Snir, Communications of the ACM, ViewPoints, Vol. 54, No. 3, March 2011.

2 Stuart Cantrill, Speaking Frankly: The allure of Pasteur’s quadrant, The Sceptical Chymist, July 7, 2013.

3 Data Science for Social Good, 20th ACM SIGKDD Conference, http://www.kdd.org/kdd2014.

4 Leah Hoffman, Computer Science and the Three Rs, Communications of the ACM, Vol. 55, No. 12, October 2012.

5 Freeman et al. Active learning increases student performance in science, engineering, and mathematics, PNAS,

doi: 10.1073/pnas.1319030111