Artificial Intelligence (AI) innovation has recently caught fire, fuelled by the necessity to make sense of and use the swelling deluge of data. By just 2018, 2.5 quintillion bytes of data were created each day. That pace is only accelerating. India’s current digital profile includes 123 crore Aadhaar users, 118 crore mobile users, 50 crore internet users, 38 crore smartphones and about 25 crore people on social networks.
Machine Learning (ML), a subfield of AI, is the ability of a machine or program to learn over time when provided with relevant data, and to continuously improve its performance of complex tasks. AI is a constellation of technologies that enables machines to act with higher levels of intelligence and emulate human capabilities of sense, autonomous comprehension and action. Traditionally one had to program or specify an algorithm, in painstaking detail, for a computer to do a limited set of tasks. Thanks to many decades of research, virtually unlimited computing power, decreasing costs of data storage/processing, and the billions of sensors and devices, now machine-learning algorithms, layered over suitable computer architectures, have become self-learners.
This combination of ‘bigdata’ and ML has seismic impacts on society, education, and policy. Dubbed the new ‘elixir’ of the World of Business and Technology, global investments in cognitive and AI solutions are said to rise upwards of 50 per cent CAGR (compound annual growth rate) hitting $57 billion in 2021.
Wittingly or unwittingly, most of us already use systems suffused with AI. Google translation, Amazon’s Alexa, Apple’s Siri, and Netflix’s recommendations are all powered by AI. The emails you receive go through a learning algorithm that intelligently tries to remove spam and unwanted emails, your GPS and map applications provide real-time recommended routes for your commute (if you visit Spokane, Washington in the US, you may see traffic lights controlled by AI to reduce energy and car emissions), perhaps your GMAT exams are graded by a learning algorithm and the credit card you just inserted into an ATM is security-verified by yet another smart system.
As another example, India’s healthcare industry— where AI has many applications—is evolving at a breakneck speed: (1) Data management by compiling and analysing medical records and patient history (2) Automated analysis of tests, X-Rays, CT scans, data entry, and other repetitive tasks by faster, typically more accurate digital agents; that would help experts to spend more time on complicated cases (3) Health monitoring, virtual nurses, medication management, precision medicine, multi-ICU management and drug creation are some other exciting applications.
Meanwhile, banks are using AI algorithms to reduce fraud and loan defaults through deep behaviour analysis. In India, machine learning and computer vision have been applied in quality analysis of rice and pulses, tea testing, aroma measurement of Basmati rice, quality detection for chillies and other vegetables and in sericulture. AI strategy is increasingly deemed as a significant competitive advantage and perhaps crucial to corporate survival and national prominence.
We know that India is facing its worst-ever water crisis, with some 600 million people facing acute water shortage. It’s not an easy challenge, but AI techniques could help in cutting down wastage of water and improving wastewater treatment systems.
Yet, any technology is a double-aged sword. More so for AI as it potentially impacts most human endeavours. Ethics, data-privacy, and job security are all major issues. A McKinsey study suggests that automation will displace some 400 to 800 million jobs by 2030, requiring as many as 375 million people to switch job categories entirely.
Our policymakers must take note of the consequent gigantic shifts that are underway in order to craft appropriate national/global responses. Here are some policy recommendations.
Workers getting displaced and roles changing will be made inevitable and accelerated by AI. Ensure that the workers of the future are equipped with the education, retraining, and skills they will need to become capable ‘digital citizens’.
Concentration of market power and the resulting socio-economic inequity needs to be actively addressed. Introduce significant measures to share the benefits of AI across communities, including by supporting local economic growth.
Diversity of opinions and workforce will be important to reduce the systemic biases that automated AI could unintentionally create. Address concerns over the changing nature of working life, income security, and potential biases from algorithmic systems at work.
What shape AI takes and how beneficial it becomes for humanity depends on ‘Human Intelligence’. What is for sure is that the future of AI is exciting and there is a huge upside to it which must be leveraged and used carefully.
Dean, School of Computer Science & Engineering and Dean of Research, Xavier University Bhubaneswar