Five-Pointer AI Recommendations For India Govt

artificial intelligence

The designing of ‘smart machines’ — that can replicate human intelligence and thinking prowess — is indeed one of the greatest inventions by mankind, till date. While it has made the scenarios depicted in the pages of futuristic and dystopian science fictions possible, it has also raised the significance of closely monitoring the impact of this emerging technology on people, and society at large. Today, it has become very critical to understand aspects like verification, validity, security and control, to ensure effective use of AI, and to prevent the occurrence of any fraud within the economic or financial systems.

In order to ensure constructive implementation and utilization of AI, the government needs to quickly and smartly lay down its application for the betterment of the society. Some of the areas wherein the application of this new technology can be explored, include – precision agriculture, predictive health care, cognitive education, targeted employment, optimized utility, and so on. Others, where we are already seeing its effects are Autonomous Cars, Unmanned Aerial Vehicles, Conversational Search & Activation, Chatbots, Facial Recognition, Recommendation Engines while e-shopping etc.

While the Government of India (GoI) is already investing in a wide array of programs like Make in India, Startup India, Skill India, Power for All, Smart Cities and, above all, Digital India, now it is time to focus specifically on the use of AI and its implications. This can be achieved by creating new chapters to these already existing programs, and extend their reach to the world of AI. Firstly, the GoI needs to invest in an active public outreach program that helps in crystallizing the AI applications in all areas at the moment, and build a concrete but dynamic “repository of AI applications”.

According to a recent survey of AI researchers, it was found that 80 per cent of the respondents firmly believed that human-level AI can eventually be achieved. Half of the respondents further believed that it is at least 50 per cent likely that this target can be achieved by the year 2040. Given this context, it is clear that research in AI will continue at all frontiers in the commercial and academic labs, one feeding the other, and the government will increasingly need to support such initiatives in future, within an appropriate policy framework. Suggestion two, is for the government to encapsulate a “strategic R&D plan for AI”. This will further require the government to, firstly, actively fund in AI research for the development of public and social applications by the private enterprises and academia, and secondly, continuously monitor the progress and implications of this plan.

Consider a scenario, wherein, the GoI wants to pass a fertilizer subsidy directly to farmers based on the area of cultivation, soil type, season, fertilizer type, crops grown, total land ownership etc., while also basing the subsidy on additional variable metrics like crop sales by the farmer, weather forecasts, global supply of inputs for fertilizer manufacture, new innovations in manufacturing practices and so on. The use of AI technology can ensure that all this happens seamlessly and automatically, with the help of “machines” (for instance, by using drones and intelligent machines for dynamic data collection) and “big data”. Additionally, machines and computers can be made to learn and revise the models only by churning huge volumes and variety of data with very high velocity. As a result, here AI is completely dependent on the availability of big data in an open way, nationally and globally. The good news is that several countries, including India, have established open government data platforms. However, given the evolutionary nature of machine learning and artificial intelligence, it is critical that nationally and globally linked open government data set (which is also dynamic) is made available for government AI. The third suggestion, therefore, is that the GoI should strengthen the national open data platform through an ever-evolving linked open data (LOD), and data models and international agreements are replicated (perhaps within the UN OGD initiatives) for the use of open data and data models of other governments.

The other big question that often becomes the center of attention in the AI debate is the implication of emerging technologies on human jobs. However, despite this constant fear of machines replacing humans, effective and sustainable technologies are being adopted at a large scale. Given the scenario, it is important that the GoI actively intervenes to assess the AI workforce and policies, and invests in upskilling the human employees. As a result, the fourth suggestion for the GoI is to monitor R&D investments for AI, and develop an assessment study or a policy framework that can drive actions towards facilitation of skill upgrades for those impacted by AI. This would require the GoI to actively invest in incentives to create the right mix and quality of AI workforce.

Finally, while AI seems to have substituted human intervention with “intelligent machines” in many areas, like anything “artificial”, it is important to govern its implications and reach to ensure positive and controlled outcomes. A machine-learned model identifying incorrect subjects, a Bot wrecking an individual’s privacy, a drone making an illegitimate entry in controlled areas, are all a possibility even if you are using the best-in-class technology. Hence, it is imperative for governments to be sensitive about choices of technology, quality of resources, safety, security and governance of AI. In the light of the same, the final suggestion for the GoI is to devise a specific “safety, security and governance” roadmap for use of AI. Every organization needs to exclusively create this architecture that charts out a framework pertaining to AI and related subjects.

To conclude, even though technological advancements are better formulated in a free environment, the measure of a government’s success can only be determined when the effects of such innovations are utilized for public good and all the side-effects are minimized. At a time when AI and other emerging technologies are increasingly gaining momentum, it is all the more necessary for the GoI to be able to look past the short-term accomplishments and lay down a proper framework for AI and its applications, to ensure long-term gains for citizens and societies.