Takshashila hosts a Bangalore Expert Group on Artificial Intelligence

The Takshashila Institution organised a conclave titled ‘Putting India on the road to becoming an AI superpower’ at its premises on 3 August 2019. The discussion had participation from 15​ ​stakeholders from academia, industry, and civil society.

The backdrop for this conclave was the Union government’s push for the adoption of Artificial Intelligence (AI) across the full spectrum of political, economic, defence operations, and many more. Over recent years, NITI Aayog came up with a National Strategy for Artificial Intelligence[1], the Department of Industrial Policy and Promotion designated an AI task force[2], the interim budget earlier this year allocated money to set up a national portal for Artificial Intelligence[3]. More recently, NITI Aayog has circulated a cabinet note asking the government to allocate ₹7,500 crores over the next three years[4]. According to NITI Aayog, the funding will be used to set up five institutes or centres of research excellence (CORE), 20 international centres for transformational AI (ICTAI), along with AIRAWAT, a national cloud computing platform[5]. These developments raise the following questions: What goals can be achieved using these funds? And what an optimum AI ecosystem for India should look like?

There is a need to outline how the government should engage in AI development and deployment. This includes what sectors to focus on and what measures to take. To that end, The Takshashila Institution hosted a one-day workshop at its office which went into the granular aspects of achieving optimum allocation of funds in the AI space.

The conclave was attended by 16 experts from across academia, industry, and civil society. The participants were divided into groups for brainstorming ideas. This was followed by plenary sessions where the groups came together to discuss their ideas.

In the first session, the groups were tasked with identifying national goals for India in AI. In the plenary session, the group reached a consensus that AI, for India, should be a catalyst for national development. This further implies by 2030, AI should contribute to 5% of India’s GDP, servicing the export and the domestic sector. AI should be leveraged to achieve New India 2022 goals, i.e., reducing poverty, achieve sustainability, increase wellness, quality of life, raise rural incomes, enhance national security, propel innovation, and economic prosperity. It is in India’s interest to use AI to transform governance. AI applications should be used for highly efficient and equitable public service delivery, tax administration, judicial system, law enforcement, public procurement, project management, transparency, and integrity.

The breakout groups were then given two sessions and one plenary session in which they were asked to define high-level strategies to achieve the goals they identified. The broad conclusions they arrived at were as follows:

  • Attracting world-class faculty; India should target attracting 500 leading researchers to work domestically over the next three years. Each of these researchers/practitioners should train 5-10 PhD/Master’s graduates.
  • Attracting top AI companies; India should be the preferred location for both homegrown and foreign companies to carry out research, development and, commercialisation of AI products and services. The Global AI top 100 should be targeted to set up facilities in India and train/hire domestic talent.
  • Creating new Centres of Excellence; India should aim to develop and maintain 20 centres of excellence across cities in partnership with industry and academia.
  • Setting up a national authority for data governance; A new nodal body tasked with oversight of rules governing data ownership, privacy, commerce, standards, and algorithms.
  • Setting up an AI promotion board; India needs a nodal body that looks at the creation and management of the Centres of Excellence, set up data repositories, common platforms, and administer challenge grants and provide funding to entrepreneurs

The final round for the breakout groups was to look at possible projects and initiatives that India should aim for to meet the goals and strategies identified.

The group proposed 6 major initiative areas at a total budget of ₹15,200 crores over three years. They are as follows:

  • Simultaneous real-time translation of the Prime Minister’s 75th Independence Day Speech into 30 Indian languages; A town-hall held by the Prime Minister on India’s 75th Independence day that translates questions and answers in 30 languages in real-time using AI could serve as a moonshot to inspire investments, research, and education in AI. Estimated cost: ₹1000 crores over 3 years.
  • Set up 20 Centres of Excellence; India should set up 20 challenges-focused Centres of Excellence to aid research and deployment of AI applications. They would be set up through the AI promotion board and will be mandated with challenge-focused tasks.  Estimated cost: ₹700 crores (Infrastructure cost), Annual operating cost: ₹2600 crores (₹130 crores x 20).
  • Deploy AI for Health through a federated data platform and assistive primary healthcare where AI should be used to accelerate universal healthcare coverage. AI applications should be leveraged for medical screening, improving diagnosis, medical decision support, and data-based recommendations for public health policymaking and interventions. Estimated cost: ₹1900 crores over 3 years; Set up a federated data platform for health (₹1500 crores) and Support the technology development for assistive primary healthcare (₹400 crores).
  • Challenge Grants; A challenge grant should be awarded as a reward to teams that achieve extraordinary breakthroughs in solving an identified problem. Estimated cost: ₹1800 crores over 3 years.
  • Atal Tinkering Labs v2; To set up equipment, training, and sandbox facilities across schools in India. Estimated cost: ₹1000 crores over 3 years.
  • AI for Expedited Justice; To eliminate backlogs of certain types of legal cases in 3 years. Estimated cost: ₹1000 crores over 3 years.

Read the event report here.

Workshop Report

The recommendations of the Expert Group have been summarised here.

[1] “National Strategy for Artificial Intelligence – NITI Aayog.” https://niti.gov.in/writereaddata/files/document_publication/NationalStrategy-for-AI-Discussion-Paper.pdf. Accessed 8 Aug. 2019.

[2] “Artificial Intelligence Task Force.” https://www.aitf.org.in/. Accessed 8 Aug. 2019.

[3] “Budget 2019: India serious about investing in Artificial Intelligence.” 1 Feb. 2019, https://www.livemint.com/budget/news/budget-2019-india-serious-about-investing-in-artificial-intelligence-1549004513204.html. Accessed 8 Aug. 2019.

[4] “Niti Aayog proposes Rs 7500-crore plan for Artificial Intelligence push.” 20 May. 2019, https://economictimes.indiatimes.com/news/economy/policy/niti-aayog-proposes-rs-7500-crore-plan-for-artificial-intelligence-push/articleshow/69403255.cms. Accessed 8 Aug. 2019.

[5] “Niti Aayog pushes for Rs 7500-crore proposal for AI platform, research ….” https://www.businesstoday.in/current/economy-politics/niti-aayog-pushes-rs-7500-crore-proposal-ai-platform-research-institutes/story/348175.html. Accessed 8 Aug. 2019.