"The development of full artificial intelligence could spell the end of the human race", Stephen Hawking famously said. I don’t know whether Hawking will be proven right or wrong. But I do know that artificial intelligence (AI) is a misunderstood term.
Firstly, Hawking was referring to AGI (artificial general intelligence), not functional AI, which is the subject of most discussions today. When it comes to AI, we are barely scratching the surface. Gloomy headlines and provocative pictures of robots usurping jobs result in widespread panic and misunderstanding. They make us believe that we have no control over our future.
Technology is never deterministic. It is a function of our choices today. Since AI will impact all our lives, it is vital that we democratize opinion and invite everyone to make an informed choice. To that end, I am enthused by the national AI strategy released by the National Institute for Transforming India (NITI Aayog) in June.
The document’s theme is "AI for All", focusing on optimizing social good. It acknowledges that India, despite being an IT powerhouse, has not consistently delivered pioneering AI technology. Its strategy is designed to maximize India’s advantage as a late-mover. It encourages the development of technology to solve India’s unique set of challenges and explores opportunities to leapfrog and build the foundational R&D capability necessary for global competitiveness.
China estimates that by 2030, AI-related activities will generate 26% of its GDP. The UAE has set up a ministry of AI. The US has created a strong ecosystem comprising advanced research institutions, universities, labs, start-up hubs and institutional capital. The French government will spend €1.5 billion over five years to support research in the field, encourage start-ups, and collect data that can be used and shared by engineers. The UK plans to nurture 1000 government-supported PhD researchers by 2025. It has also set up the Turing Fellowship, inspired by the work of Alan Turing, who designed an experiment to test a machine's ability to exhibit intelligent behaviour equivalent to, or indistinguishable from, a human.
It is no secret that India missed reaping the benefits of the first, second and third industrial revolutions. If it is to embrace the Fourth Industrial Revolution equitably, inclusively and sustainably, taking a leadership position on AI is critical.
In order to achieve the goals of "AI for All", India must address certain barriers and bottlenecks, such as the availability of broad-based research expertise, the absence of an enabling data ecosystem, low awareness around AI adoption, issues of privacy and lack of stakeholder collaboration.
Have you read?
While every industry can potentially be disrupted by AI, I am glad that India’s AI strategy will focus on five sectors in which its application can create quantifiable social impact in the medium to long-term: healthcare, agriculture, education, smart cities and smart mobility.
NITI Aayog has adopted a three-pronged approach to address the challenges in these sectors. First, it will undertake exploratory proof of concept AI projects. Second, it will strive to build a thriving ecosystem, in collaboration with start-ups and mature enterprises. Third, it will create a multi-stakeholder, multinational approach. An integral part of India’s national AI strategy will be to share best practices and innovative operating models with other developing countries, thereby making it an innovation hub for emerging economies.
Realizing India’s potential to shape the global technological landscape, the World Economic Forum has partnered with the Government of India to set up the Center for the Fourth Industrial Revolution India in Mumbai, Maharashtra. According to the Forum President Børge Brende, the Center will "work to accelerate the development and implementation of governance protocols for emerging science and technology to best serve citizens, society and the public at large. It is worth noting that the Center for the Fourth Industrial Revolution will adopt a multi-stakeholder approach and pilot practical tools for the agile governance of specific technologies."
We live in a world of global challenges that cannot be solved with a local mindset. India’s efforts to cross-pollinate best practices will go a long way in setting an example for the rest of the world. In an era obsessed with building walls and firewalls, India is building bridges across borders and boundaries.
To unlock the true potential of its vision, the report has concrete proposals. Research in India is still in its infancy. The two-tiered programme to set up CORE (Centre of Research Excellence) and ICTAI (International Centres of Transformational AI) has the potential to push the frontiers of technology development, by fostering institutional and peer-to-peer collaboration. The proposed umbrella organization along the lines of CERN will propel "moonshot research projects". The "CERN for AI" model is well-suited to India’s approach. It involves directing research efforts, studying global advancements and encouraging international collaboration.
Strong research capability needs to be supplemented by wide adoption across start-ups, the private sector, PSUs and government. The NITI report posits the creation of a National AI Marketplace (NAIM) to overcome the barriers of development and deployment. The marketplace will enable optimum price discovery and, more importantly, bring to bear different approaches to data collection, aggregation and annotation - vital steps for deriving intelligence through the poetry of data.
For the marketplace to blossom, India needs a future-ready workforce. According to the World Economic Forum, 85% of 2030’s jobs do not exist today. We need to change the way we teach in schools, focusing on "why" instead of "what". To that end, I believe that the Atal Innovation Mission and Atal Tinkering Labs can nurture innovators who will shape the future of India. In May, I met a 14-year-old at a tinkering lab in Delhi. She had built a prototype of "Lord Ganesha" who offered blessings and prasāda using motion sensors.
In colleges and vocational training institutes, we need to create a stronger connection with industry, to manage the unemployability challenge that many Indian engineering graduates face. We live in times of tectonic change. The curriculum that incoming college students are introduced to may become irrelevant by the time they graduate. This can be deeply unsettling, and we need to arm our students with emotional resilience. Unfortunately, no university in the world has a course that teaches AI and emotional resilience. Maybe India can take leadership in creating such a course. It will be much needed at a time when the media constantly reports that robots are coming for our jobs.
We need a larger discourse to focus on creating awareness about AI and its impact. Instead of sowing panic with sensational headlines, we should highlight lesser-known facts. For example, each job loss due to automation will lead to five new jobs, requiring different skill sets. Re-skilling is easier said than done, but it is possible, with a combination of a national mentoring program, decentralized content-delivery mechanisms and community-based learning models.
Last, but by no means least, is the issue of ethics. Consider the driverless car conundrum explained beautifully by the MIT Moral Machine. If a driverless car meets with an accident and has a choice between killing two passengers or five pedestrians, what should it do? This is a very hard question. Most people answer that the car should kill the two passengers, as long as they are not one of them.
Exploring the contours of debates like this requires our artists, philosophers, lawyers, activists, politicians, technologists and business leaders to come together. As the saying goes, technology is far too important to be left to the technologists alone.
As we march into the Fourth Industrial Revolution, I hope we approach AI with cautious optimism, take time to define design principles, and remember that the future is a function of our choices today. It is not preordained.