Using AI to deliver better governance – trends in India and abroad:
A comparative study
Assistant Professor, Department of Computer Science",
Mrinalini Datta Mahavidyapith, Birati, Kolkata - 700 051, West Bengal, India.
Email: [email protected]
Abstract: Public administration is to human society what the soil is to plant. It is ubiquitous, diffused into every layer of our lives and living. Ubiquitous also is Artificial Intelligence (AI) that has disrupted or is threatening to disrupt every man-made system. Machine Learning, Big Data, Deep Learning, Artificial Intelligence - whatever you call it, is forcing us to re-imagine, re-invent, re-think, re-conceptualize, re-organize, re-structure every human industry, be it education, healthcare, defense or even entertainment.
How then has AI impacted the nature of public administration in nations across the globe? While many innovative proposals are available on how AI can be applied in various fields of public administration to deliver citizen friendly, transparent and accountable governance, the question arises as to how many of them are pragmatic and viable, more so in the Indian context.
This paper makes a comparative study of how far various nations in general and India, in particular, have incorporated or are planning to incorporate AI in public administrative activities like defense, healthcare, education, law and order, e-governance, etc.
Keywords: Public administration, Artificial Intelligence, E-governance
Section I. Introduction
“Redefining old ideas” is the 21st century mantra. Be it fashion, food, art, education or public administration, old ideas are being continually redefined or augmented by incorporating new technological/socio-economic innovations. Both Public Administration and Artificial Intelligence (AI) are old ideas. What is new is how AI is being utilized or can be leveraged to reinvent public administration of the 21st century so that governments can deliver improved citizen services.
Public administration has been defined as “it is concerned with the implementation of public policy, as laid down by the competent authority economically and efficiently for the benefit of the people”.1 For the man to form and function as a society, public administration is vital. It provides the bedrock of our social lives and living. Whether traveling in a bus to work or disposing household waste or enrolling in a primary school or cooking or leading a peaceful life, public administration is omnipresent, every minute and everywhere. Surely something as vital as public administration must undergo continual revision to line itself with the demands of the changing times.
While on one hand public administration in the 21st century is undergoing a conceptual and structural shift (eg. New Public Management, Alternative Development, Sustainable Development), at the same time it is also getting remolded by technological forces of information and communication technology and now AI.
Section II. Artificial Intelligence – brief overview
There are multiple definitions of the term Artificial Intelligence (AI). It is used varyingly in multiple contexts. Some definitions are reproduced here:
John McCarthy, who had coined the term in 1956, defined it as "the science and engineering of making intelligent machines." 2 The modern definition of AI is "the study and design of intelligent agents" where an intelligent agent is a system that perceives its environment and takes actions which maximize its chances of success.3
Accenture defines AI for business executives in the following way – “AI is a collection of technologies that can enable a machine or system to sense, comprehend, act, and learn. Training a system through machine learning or deep learning is a core part of what makes it intelligent.”4
The Association for the Advancement of Artificial Intelligence defines AI as “the scientific understanding of the mechanisms underlying thought and intelligent behavior and their embodiment in machines.”5
Thus computer systems that are programmed to think, decide, act and outperform human beings are “intelligent”. AI includes machine learning, data mining, natural language processing, intelligent tutoring, image processing, computer vision, virtual reality, etc. Some application areas are: web page ranking, recommendation systems on e-commerce platforms like Amazon, Flipkart, etc., automatic translation of documents, voice assistants like Apple's Siri, Amazon's Alexa, Microsoft's Cortana, and Google's Assistant.
AI is ubiquitous today. Like public administration, it is present almost everywhere hidden in plain sight. With the availability of huge amount of data and cloud-based applications and services the AI industry is expected to cross 191 Billion USD by 2024 at a compound annual growth rate of 37%.6 The Emerging jobs report 2018 by LinkedIn, USA finds that 6 out of 15 emerging jobs are related in some way to AI, and AI skills are being heavily sought after in all industries and not just IT. AI skills globally rose to 190% from 2015 to 2017.7 On the issue of whether AI will improve human living conditions or not, 63% of AI experts have agreed that AI will have a positive impact whereas 37% have opined otherwise.8 These figures underline the breadth and depth of the impact that AI has created on human society.
Section III. Application of AI to Public Administration
Let us now look at how AI can be applied to various sectors of government activities.
10% of the global gross domestic product is spent on health care, 51% of this expenditure is borne by the governments.9 This is one sector that is expecting a shot in the arm by AI-driven solutions. The global AI market in healthcare is expected to grow at a CAGR of 43.5% from 2018 to reach USD 27.60 billion by 2025.10
AI can be applied to electronic health records (collected from different sources and present in heterogeneous formats) to supplement clinical practice by doctors. Machine Learning algorithms can analyze scan or test results (like mammograms) to more accurately predict the disease, its severity and the most probable eventual outcome (curable, not-curable, manageable, etc.). Of course, the doctor must add his or her expertise and experience to the outcome and decide the next course of action. This will have two clear benefits: prevent wrong diagnosis caused by stress, fatigue or haste which in turn will reduce health cost and time to recovery of patients and secondly free up doctor's time to concentrate more on the further line of treatment rather than spending excessive time on interpreting test results. 11",12",13",14
Several organizations including Google are developing ML algorithms that scan vast amounts of electronic health data to detect medical conditions like hypertension, heart diseases, osteoporosis, diabetes, etc. Similar algorithms are already available for detecting eye diseases like diabetic retinopathy from eye scans and MRI.15
AI technology is also being explored in drug research and discovery. Chatbot applications are available that interact with patients to provide health assessment or companionship especially to elderly patients.16 Company like Siemens Healthineers17 have used deep learning neural network to create "digital twin" of organs especially the heart. This virtual organ can be subjected to virtual surgery, medication, etc. to learn how the actual organ and consequently the patient will respond to the treatment before it is actually done. This technique promises time saving and improved treatment.
Education is the best investment anyone can make. It leads to better employment opportunities, higher wages, and economic growth. Around the world, an average of 3.94% of GDP was spent on education in the year 2017. Public or government spending on education was 15.96% in 2016.18 MarketsandMarkets have forecasted that AI in the global education market will grow from USD 373.1 Million in 2017 to USD 3",683.5 Million by 2023, at a Compound Annual Growth Rate of 47% during the forecast period.19
Today’s job market demands newer skills viz. creativity, social perceptiveness, teamwork. AI promises to benefit the education sector by offering personalized learning, Intelligent Tutoring Systems, predictive analytics, etc.
It is common to find students of diverse profiles in the same classroom. One student’s strength and weakness, learning pace, grasping power, level of concentration, likes, dislikes are different from the others. AI has enabled personalized learning and personalized assessment. Each student can learn at his or her own pace and follow different trajectories to reach a common educational goal. Every individual has his or her own learning pattern which can be analyzed by AI to offer the right kind of assessment to the learner. Examples are companies like Content Technologies and Carnegie Learning that have developed such AI enabled intelligent instruction design systems.20 Presentation Translator is a free plug-in for PowerPoint that creates subtitles in real time for what the teacher is saying.21 Thus students who have missed schools need not worry. Predictive analytics using machine learning have been proposed to identify students who need more support or who find it difficult to follow the language being used in the classroom, eg. Microsoft’s Learning Tools and Skype Translator.22",23
GDP share of agriculture for the year 2017 was 11.25% globally.24 AI investment in the agriculture market is projected to reach 21.8% compound annual growth rate in terms of revenue, the global market size is expected to grow from USD 240 million in 2019to touch USD 790 million by 2024.25
AI trends in this sector are: Agricultural robots, smart sensors, precision agriculture, and predictive analytics.
Several key farming regions face a labour shortage, especially during the harvest season. This leads to loss of revenue. Work is underway in developing AI enabled robots that can harvest and pack the yield faster. Example: robot by Harvest CROO Robotics to help strawberry farmers pick and pack their crops.26 Herbicide resistance is another area of concern in agriculture. Companies are using computer vision to streamline where and how much herbicide must be sprayed so that both crop yield and herbicide use are optimized. Example: See & Spray drones by the company Blue River Technology.27 Smart sensors laid out in farms measure soil and climate information which is then fed to a deep learning algorithm for soil and crop health monitoring. Nutrition deficiency in soil, plant diseases, soil defects, plant pests, and diseases can be identified by the algorithm. In some novel instances, even sensors are not needed. A smartphone camera and an AI app can also accomplish the task of soil and crop health monitoring. Example: Plantix, a deep learning algorithm developed by Berlin-based agricultural technology startup company named PEAT.28
Defense/Law and order
Globally, Robotics is the main vehicle for incorporating AI in defense. AI is being used in intelligence, surveillance, and reconnaissance activities in the following ways:
Unmanned autonomous vehicles (on land, sea, and air) and drones are being developed for detection of enemy entities like pirate ships or mines, for monitoring enemy movement, for remotely controlled air strikes, and for reaching areas too risky for humans to go like war zones. Robots are being used in search and rescue missions in wars, terrorist attacks or natural calamities like an earthquake. AI based training simulators are being used for pilots. Bipedal humanoid robots have been developed for ground operations that can function as a unit of soldiers to communicate and collaborate with each using a suite of AI based algorithms. In the law industry, AI is helping to speed up the legal system by sifting through thousands of legal documents to tag and identify those documents that are of relevance to the case at hand. Example: ROSS Intelligence system29 uses natural language processing (NLP) for legal research activities.
Inside the country borders, AI enabled systems are being explored in policing and public security. Predictive policing techniques based on data mining and analytics are already in place to track and predict criminal activities. Analysis of satellite imagery, online content, CCTV footages, and social media posts are revealing hot-spots of the crime landscape, i.e. probable breeding grounds of criminal and terrorist activities. Example: software PredPol30 used by Delhi Police, India, for crime forecasting and advance deployment of police forces to such sensitive areas.
Delivery of citizen services by governments is transforming from the conventional mode to the digital and smart mode. In countries like the US, Singapore, Mexico, governments are making use of AI systems to automate citizen services. These applications can identify, locate, search, tag, auto-fill, translate and help in writing documents. Chatbot applications are in place to answer citizen queries regarding visa, passport and emigration services. AI, data mining and analytics are being combined together to detect cases of tax evasion, to track de-registered firms. The Indian government has reportedly used data analytics to gauge public sentiment from views, opinions, debates, and comments posted by people on crowdsourcing platforms and other social media platforms on various social and political issues. Banks are also putting AI-powered chat-bot assistants in place to handle customer queries automatically, replacing the need of human assistants as far as practicable. Example: SIA launched by SBI bank in India.31
Section IV. How nations across the globe have incorporated AI in administrative activities/ AI inclusion or incorporation in governance: trends in India and abroad – a comparison.
After getting a hint of the prospects and promises of AI vis-a-vis government sectors we now make a comparative study of selected nations and understand their outlook towards the adoption of AI-powered technology in administrative activities.
Among the nations, the US and China are the global leaders in AI.
Leading companies involved in AI, based in these countries
IBM, Microsoft, Alphabet, Google, Apple
Baidu, Alibaba, Tencent, DiDi Chuxing
Top patent holders*
IBM and Microsoft
(top two companies of the world with largest patent applications)
Grid Corporation of China
Universities/public organizations producing AI related scientific outputs*
6 out of top 20 organizations with scientific publications in AI
10 out of top 20 organizations with scientific publications in AI
universities dominate in AI research
Estimated GDP gain by 203035
Venture Capital funding in startup companies
USD 99.5 billion in 201836
USD 7 trillion in GDP gains by 2030
USD 529.52 Million in 201837
Total annual global funding USD 207 billion in 2018.
size of funding received in India is substantially smaller than by those in the US and China, reflecting the limited success of the country's AI startups.
Time of release of AI plans by governments
the focus area of each nation
(1) maintain American leadership in AI, (2) support the American worker, (3) promote public R&D; and (4) remove barriers to innovation.
(1) focus on developing intelligent and networked products such as vehicles, service robots, and identification systems, (2) emphasize the development AI’s support system, including intelligent sensors and neural network chips, (3) encourage the development of intelligent manufacturing, and (4) improve the environment for the development of AI by investing in industry training resources, standard testing, and cybersecurity.
(1) enhance and empower Indians with the skills to find quality jobs; (2) invest in research and sectors that can maximize economic growth and social impact; and (3) scale Indian-made AI solutions to the rest of the developing world.
*Data and deductions obtained from: WIPO (2019). WIPO Technology Trends 2019: Artificial Intelligence. Geneva: World Intellectual Property Organization.
Section VI. Conclusion
– this very aptly captures the government of India's outlook towards AI – one of the greater good –.
From the NITI Aayog’ s discussion paper on national strategy for AI, it is clear that the government's spirit or outlook towards using AI in governance is for societal good. The short and sweet hashtag #AIforALL beautifully captures the government’s primary focus or main motivation for using AI- social and inclusive growth. In contrast, China's approach is purely that of techno-utilitarianism.
1. Siuli Sarkar, Public Administration in India, PHI Learning Private Limited, New Delhi, 2010.
15. Times of India, How AI may assist docs in diagnosing rare conditions, pg 12, Feb 21, 2019, Kolkata
30. predpol----insert ref here