India gears up to grab a share of the AI market backed by govt support

However, many of the country's achievements in this space, such as development of the world's first neuromorphic chip that emulates the neurons and synapses in the brains, aren't well publicised

After the ban on some 59 apps of Chinese origin a few months ago, all eyes are on Indian start-ups to fill the void. In the social media space, just like in the broader technology landscape, several bespoke solutions have been born out of artificial intelligence. And this is where the government is seen pushing the pedal lately.

In May, Minister for Electronics and Information Technology Ravi Shankar Prasad launched India’s first AI portal. IndiaAI, as it is called, is meant to showcase the latest in AI-related research, projects, people and schemes, and also position itself as a platform for top minds to come together.

“This is a national AI portal, which will document AI (related works) happening across the segments--government, academia and start-ups. If I am a student and I want to pursue an AI career, (this portal will tell me) what are some of the learning courses for me, and so on,” said Sangeeta Gupta, senior vice president at Nasscom.

The platform hosts regular webinars and publishes columns from tech experts. The team has also painstaking profiled about 200 'AI' start-ups in India. In time, it plans to offer courses for things like basics of AI and how to code, said a senior member at IndiaAI team.

While India is not in the same league as the US or Canada, says Gupta of Nasscom, whatever being done is also not adequately publicised. For instance, India ranks third in terms of peer-reviewed AI papers, or that an IIT-Delhi professor made the world’s first neuromorphic chip, which emulate the neurons and synapses in our brains.

The IndiaAI platform is part of a unified vision for AI, which was documented in Niti Aayog’s ‘National Strategy on Artificial Intelligence’ paper in June 2018. Based on recommendations of two expert committees, the document laid out that AI research and adoption should focus on five key areas: healthcare, agriculture, education, smart cities and transportation.

In agriculture, various pilots currently underway aim at using AI and data science in soil monitoring, predictive crop analytics, supply chain and weather forecasts. On the infrastructure side, traffic patterns are being studied to advise decongestion strategies. The biggest stride, however, is in the education sector.

Under Atal Innovation Mission programme, the government is pushing for students to take up AI and other computer technologies early on. This is being done by funding schools to set up ‘Tinkering Labs’ with do-it-yourself kits on science, electronics, robotics, sensors and 3D printers. Besides, incubation centres are being set up in colleges. Since April 2018, over 3,500 tinkering labs and about 70 incubation centres have been set up across the country. Further, the government recently put AI in the curriculum in some 5,000 schools.

On the health side, the government has taken several bold strides. Through Ayushman Bharat Yojna, which aims to provide insurance cover to some 100 million families, a repository of health data of citizens is being envisaged. Once the exercise is complete, the powerful dataset will enable researchers to conduct studies and find answers to larger questions on gene pools, susceptibility to diseases for people of certain genetic makeup, and precision healthcare.

“With regard to education in AI, a considerable amount of investment and efforts are in place in our country,” said V Kamakoti, a Professor in the Department of Computer Science and Engineering at IIT Madras. “Almost every institute across India has AI courses, and many have a centre of excellence or laboratories in AI. But it doesn't stop in just teaching courses. A lot of research work is also being done. But the important thing is the adoption of AI by the industry which has to come up in a big way,” added Kamakoti, who served as the chairman of one of the national taskforces on AI.

In February, Niti Aayog CEO Amitabh Kant said India could add 1.3 per cent of its gross domestic product (GDP)--about Rs 3,500 crore--annually through the use of AI and ML. The figure means that companies and solutions could one day be worth this much.

Since the definition of an AI company is not absolute, it is typically understood that firms that do social media, enterprise software, analytics, chatbots, healthcare and IT services use AI in meaningful ways. Applying this lens throws up names like Freshworks, Sharechat, Reliance-owned Haptik, Niramai, InMobi, and even IT stalwarts like Infosys and TCS. Barring the IT services firms, only two companies--Freshworks and InMobi--are valued at over $1 billion while there is no data on the collective value of all AI firms in India.

One of the ways the government has supported AI is by making funds available in the VC market. VCs typically invest in high-growth companies working on cutting-edge technologies. The government has so far contributed Rs 3,798 crore towards venture capital support to start-ups, some of which has made their way to AI development.

Experts however believe that what is missing is a coherent strategy to finance AI start-ups. For instance, in Canada, a handful of prominent AI institutions  including Vector Institute for Artificial Intelligence, Montreal Institute for Learning Algorithms and Alberta Machine Intelligence Institute get a 125-million Canadian dollar  grant from the government over five years, as part of the government’s national AI strategy.

In India, in comparison, most of the R&D budget is skewed towards defence. During 2017-18, the Defence Research and Development Organisation (DRDO) accounted for 31.6 per cent of India’s total R&D spent (Rs 1.13 lakh crore), according to official figures. Department of Science and Electronics received 19 per cent and 0.3 respectively.

“The next phase of AI will depend on data, and data is something India is rich in. We are trying to somehow unlock this data that is available with various ministries and departments and make it available as a coherent training data,” said an official at MeitY.  A dataset of self-drive cars – the largest such data set present in India – is available at IIIT-Hyderabad and not a lot of people know about it, the person added.

According to an Accenture report, AI solutions could potentially be worth $957 billion in India by 2035.