Who’s who of Stock Market and Investment World log into K J Somaiya Institute of Management!

Just a decade ago, we would have dismissed as ridiculous if somebody had said Artificial Intelligence (AI) and Machine Learning (ML) technologies are going to disrupt every business.  Can we dare do it today? World Economic Forum’s “The Future of Jobs 2018” reports that machines and algorithms could create a net 58 million new jobs in the next couple of years.  As per the industry report cited by The Economic Times, over 50,000 AI related jobs are vacant in India for want of skilled professionals.  Top executives of global companies say that they need futuristic technologies namely, Artificial intelligence (AI), Block chain, Virtual and Augmented reality and Quantum computing, to provide best personalised experience to customers.

Is AI already working for us?

World over, Governments are engaged in creating smart cities, corporates are offering their customers smarter gadgets, militaries are deploying smart and intelligent defence systems, and society in general, is looking forward to embracing smart assistants to enjoy better quality life.  The whole world is looking towards AI to help lead smarter and better life.  In fact, AI has already made inroads into our daily lives – Google assistant, Tesla, Netflix, Apple’s SIRI, Alexa, Cortana, Watson – do they sound familiar? AI is going to greatly influence the way we think, live, experience anything in this world, and even how we govern ourselves. Of course, the great enhancement in our life that comes with AI technology could also throw up ‘Future shock’ as foresaw by Alwyn Toffler.  Are we prepared for that?

Know more about AI

AI systems are endowed with the intellectual processes typical of humans such as ability to reason, interpret, learn and generalise from experience.  The idea of training computers to think like humans is at least 60 years old. However, it has become a reality today, thanks to today’s business ecosystem comprised of systems that generate massive amount of data, huge but inexpensive data storages, high performance computers, sophisticated algorithms, and widespread high-speed internet. AI should assist us address complex problems that need analysis of massive data with hidden patterns. Depending on the complexity of problems being addressed, the AI needed is classified as narrow intelligence, general intelligence and super intelligence. Narrow AI is the intelligence required to solve a specific class of problems and is already seen in various applications such as separation of spam from genuine mails, and autonomous vehicles. General intelligence is the intelligence required to solve a variety of problems. For instance, the intelligence to predict the behaviour of a person in different situations.  This is still not any closer to us.  Super intelligent machines are imagined to be self-aware, introspecting, and self-correcting systems. They are expected to write their own code and source inputs from anywhere on the network. This is still far away from realization.  However, many renowned technologists are apprehensive of the super intelligence stage and foresee the danger of humans becoming subservient to machines.

How AI works?

Today, unlike the past, computers are being trained to learn. They enhance their learning by accessing massive data available on the internet and help us make better decisions. Having lived with your parents for long, your brain’s neural network would have learned the pattern of their responses to a given situation. This learning would help you predict their response in similar other situations.  In AI systems, we try to mimic this natural process of learning by creating and training an artificial neural network (ANN) using a large set of data on animate or inanimate objects and patterns therein.  AI systems thus trained can classify the objects, provide their descriptions, analytics and predictions.  However, ANN’s assessment is probabilistic and always with a certain degree of accuracy.  With greater quantity and quality of data, and valid algorithms, a trained ANN can yield more accurate results.  The training methods used may be supervised or unsupervised. In the former method, algorithms are trained explicitly by telling what a pattern implies, whereas in the latter, algorithms are engaged to identify patterns in the data set on their own.  Advanced mathematical and statistical techniques work in the background. This machine learning approach is at the core of realising AI.

Where AI is heading to?

AI systems can understand the context, tone and tenor of a story, and extend it such that it is almost impossible to distinguish it from that of the original author. They can listen to a piece of music, understand the mood and create music on their own.  They advise us on various legal issues too.  AI-built robots are already providing domestic services including care for elders. AI-built image processing systems have proved to be better than specialist doctors in diagnosing tumours, and pathological and radiological investigations.  AI applications range from health care to banking, fraud detection to genomics, agriculture to entertainment, manufacturing to archaeology and so on.  Hence, their applications in our homes, industry and commerce, government, military, or anywhere are boundless.

Corporates and governments have understood that these technologies have a great potential to disrupt everything being done by them and are in a hurry to embrace them. This points to the future job market. The NITI Aayog, NASSCOM, and many reputed IT companies have foreseen a serious shortage of skilled manpower in AI & ML in India.  AICTE’s expert committee has urged the Indian universities to offer UG programs in AI to meet the challenges ahead of the nation.

What is your Take now?

The question now is not “Whether we need to get acquainted with AI?” but it is “How early we need to do so?”, especially for those who want to be future-ready.