By now, there is little doubt left in our minds that the world we will live in, by 2030, will be heavily influenced by Artificial Intelligence. AI will be as ubiquitous as electricity is these days. Rarely a week passes by without any news items about new applications of AI in areas that have been untouched hitherto. The ensuing threat from this is that many jobs will become redundant. They certainly will. But the jobs it will make redundant are the ones that are not the most effective use of our time and skills. And data science is at the core of driving these AI solutions and applications.
One may think that AI may also make data science, as a job, redundant. It may make some parts of the current form of data science redundant, but not the whole discipline. ‘Why?’ you may ask. I say that because this discipline is not static. It is evolving. Like most disciplines, data science has also changed over the past decades and years. While there may not be a consensus on the exact definition of data science, there is definitely an agreement that it has become more multi-disciplinary, multi-faceted and multi-domain. Boundaries that existed between statistics, programming, data wrangling, visualisation, engineering are blurring within the Data Science discipline. Not only these, other areas like behavioural science and economics are also fast getting subsumed. Experts from newer and varying domains are becoming more receptive to leveraging data science to make better decisions. Along with domains like financial services, banking, e-commerce and retail, other domains like agriculture, medicine or manufacturing are all finding nifty problems to solve with AI-driven solutions.
Certain tasks that are seemingly obvious to humans are often tough for machines to build intelligence within. To overcome this, machines skim through and analyse massive amounts of data, evaluate hundreds and thousands of scenarios and come up with recommendations. This limitation of AI also becomes its strength. A task that may take humans an inordinate amount of time to complete, an AI solution with moderate computing power may be able to accomplish easily. Remember AlphaGo beating Lee Sedol in the game of Go? This type of AI starts to mimic gut-based decisions humans make.
This can be very relevant in Business Strategy Development. Accomplished industry leaders leverage their expertise and experience from the past to make strategic decisions. AI-driven solutions can glean through multitudes of scenarios, moves, and counter-moves, evaluate the outcomes and recommend strategies that are more information based than gut-based. This is where AI can find a seat at the decision table and strategy development.
The students of today need to continuously evolve and experiment along with getting the basics right. At the very core of data science is a lot of Math. Having a quantitative approach and aptitude will definitely help. Building on that, a good and intuitive grasp of the techniques is very crucial. There are various tools and open-source packages available for anyone to experiment, build and try out different solutions. With these, a curiosity to solve real-life problems is key to building a career in data science.
(The author leads the Data Science group at Walmart Labs to leverage big data, data science and technology to enable faster and smarter business decisions. He is leading key initiatives to deploy algorithmic products that consume Walmart scale data and infuse smarter decisions across retail lifecycle to deliver multi-billion-dollar impact. He is an alumnus of Indian Statistical Institute, Kolkata and has over 17 years of extensive experience in the field of predictive analytics and data science. He was recently recognised as one of the top data scientists in the country by Analytics India Magazine.)
