Author: R M Ramanathan
A Perspective
What we commonly label as “disruptive” is often the result of breaking the status quo. Disruptions seem to have their cycles and a few important ones come to my mind. In 1800s, we invented steam engines, cracked the math of thermodynamics, giving way to a number of derived industries: factories, locomotives, mills, etc. The lifecycle of this innovation reached its destiny sixty years later with the first ever depression. People could have been smarter and learned their lesson. But they became smarter by inventing Electricity and Magnetism just eight decades later (1880s) than the invention of steam engines. That disrupted status quo and rapid electrification took place across the world. Another eight decades later (1960s) we got transistors, computers and the beginnings of the internet. Six decades since then (2020s), we are in the cusp of another cycle. The sure candidate for next disruption is AI along with bio/nano technology.
I notice a pattern here. All disruptions happen when the previous one is mastered, done and dusted. Today we are within arms length of reaching the limits of Moore’s Law. The situation is primed for the next disruption. Amidst the evolution on doubling computing power every year and a half, the silent winner has been an exponential increase of data over the same period. Every computerized establishment saved cost, did more of the same (old) business in the same business year. Well, we homo sapiens, are too smart to run same (old) businesses. We leveraged data for intelligence and built new markets, new products and new strategies. In the last 2-3 decades of the recent technological cycle, we have progressed enough on big data technologies to enable the next disruption.
We have had statisticians and mathematicians working in fancy n-dimensional spaces, which did not have any utility whatsoever. We have had techies who sounded cool knowing how to play with a new breed of toys. They took pride in providing solutions for age-old tasks such as billing, inventory or queuing. They were content being glorified calculators, or were they? Then we had database operators donning data analyst roles due to their proximity and expertise in handling data. This is when statisticians started learning programming, and programmers picked up interest in statistical science. Bang! Bang! Data Scientists were born (to rule next set of grandiose things).
Publications that were once little more than academic, have suddenly found their utility in industry. Many courses in Data Science and Machine Learning have mushroomed both online and offline. Today’s college kids scramble to do Data Science projects. Companies need them the most. Those who fail to recognize this risk labelled not being adaptive.
In India, I notice a peculiar attribute among those who attend basic knowledge sessions. They do not find these simple enough! They are quantitatively smart. Yet they find these basic sessions not so basic. This could mean that they have old ideas of jobs, salaries, titles and promotions. They must be educated that disruption in technology is going to disrupt how people are going to get hired and fired. Let me explain.
The jobs of the future are not going to look like anything that we have today. The market is changing fast. There are Data Science competitions happening all over the web. The Kaggles, TechGigs, HackerEarths and HackerRanks are all looking for talent. This is the new recruitment process rather than resumes and interviews alone. If you are a job seeker, these platforms will push your limits and test your skills at an international level. They will also expose knowledge and skill gaps.
Companies are becoming more and more open to pay-per-performance rather than pay-per-month. Data Scientists cannot be evaluated from a series of interviews. Knowledge may not mean action. Knowledge and action take years to build. College kids will need to understand this. There is no free lunch, forget about being spoon-fed. Let’s be honest. Opportunity to excel is very much available, but take help from self and open resources on the web. Educational institutions in India will have to prepare students to face circumstances rather than to score in exams. Managers in companies will have to enable delivery by individual employees rather than managing delivery for the team.
Data Science is an applied science. It has application in varied fields. What is important is to choose and focus on a particular field. Let me draw an analogy here. A calculator can be used by an accountant for simple arithmetic. The same accountant will have limitations when using it to work out problems in aerodynamics or quantum mechanics. Likewise, one must make the choice to apply data science in their field of expertise; or to apply in a field where they can quickly pick up domain knowledge. Acquiring an array of machine learning techniques and their applications, mastering big data technologies, becoming domain expert to provide solutions that create impact will take considerable time and effort. Giving oneself that time and putting in diligent efforts will make one competitive.
There are a lot of start-ups working on cool Data Science problems. The problems range from autonomous driving to healthcare, from weather forecasting to IoT. They require specific skills such as image/speech/video processing, domain expertise or product development. Government organisations are sitting on a lot of data that can be leveraged for better governance. There are also opportunities to work on Data Science projects as a freelancer or independent consultant: stock market, BitCoins, open source projects to serve the SME market. One can even think of working for two small companies at same time, solving both their problems; or even a 4-day workweek with a company.
Finally, a few words on how-to. The problems are open-ended with little or no mentoring. One must be a go-getter. The use of mere jargon such as machine learning, robotics or AI attract attention immediately. The higher pay packet only highlights the current supply-demand gap. I say this so that young engineers do not getting carried away. But Data Scientists contribute to the bottom line. Best talents have a place for the next decade or so. Attending a mere 3-hour workshop will only increase curiosity and insecurity. It is important to take up hands-on projects and gain a deeper understanding. There are a lot of platforms where you can prove your skills and build reputation. Get down to work and you will get noticed. Good Luck.
Author: R M Ramanathan
Ramanathan is a Data Evangelist with over a decade of experience contributing to Analytics functions in Retail/FMCG/Finance. He has extensively worked on various statistical modelling and machine learning models. A vivid follower of business strategies and geopolitics. He cherishes knowledge sharing, building organisations from scratch and mentoring ambitious individuals in various platforms.
The IEDF Blog 