Session Highlights

Shri J Satyanarayana, IAS (Retd), Former Secy, MeitY & Former Chairman Aadhar Auth

An insightful session on "Leveraging AI in the Intelligent Age" was delivered by our esteemed speaker, Shri J. Satyanarayana, IAS (Retd), Former Secretary of MeitY and Former Chairman of UIDAI (Aadhaar Authority). The session offered participants a deeper understanding of how the rapid advancements in Artificial Intelligence can be strategically utilized for effective governance and national development, while underscoring the critical need for responsible implementation and human oversight.

The session began by defining Artificial Intelligence (AI) as the "simulation of human intelligence by computers, by providing them the ability to LEARN, REASON, DECIDE & SELF-CORRECT," noting that AI "mimics human neuron system". The mechanics of AI were explained through a child development analogy, involving continuous observation, training with feedback, validation, and correction. Foundational capabilities like Natural Language Processing, Computer Vision, Machine Learning, and Deep Learning were highlighted.

A significant focus was placed on Agentic AI, defined as a "software system that can autonomously perceive its environment, make rational decisions, and take actions to achieve a specific goal." A key distinction was drawn: " While LLMs provide answers, agentic AI can both think and act. ". A crucial design principle for AI agents is to "build human-AI collaboration" for strategic decision-making, quality assurance, and exception handling, with a firm emphasis that "human intervention is necessary" and users must "draw a line (lakshman rekha)".

The pervasive "Practical uses of AI" were explored across sectors like healthcare (e.g., improved diagnostics, early disease detection), agriculture (e.g., crop advisories, yield prediction), finance (e.g., fraud detection), transport (e.g., autonomous vehicles), and education (e.g., adaptive learning). The discussion also addressed AI's profound "impact on Management Science," necessitating its redefinition across areas such as decision science, strategy, and risk management.

The session thoroughly examined "Potential harms of AI," underscoring that "There is as much fear about the harm AI can do as the hope of the benefits it can provide us". Risks include harm due to mistakes, misuse, bias, privacy infringement, and intellectual property rights violations. The quote from Geoffrey Hinton highlighted this concern: "These things could become more intelligent than us and could decide to take over, and we need to worry NOW about how we prevent that happening". Mitigation strategies include promoting ethical principles, implementing "Responsible-by-Design", establishing robust regulatory frameworks (like the NIST AI Risk Management Framework and the EU AI Act), fostering education, multi-stakeholder collaboration, and responsible data governance. The speaker stressed the need for "implementing AI Act also not merely guidelines" for AI in governance.

Takeaways for AI in Governance, Digital Infrastructure, and the Future of Intelligent Systems in India:

AI in Governance: India's potential for AI transformation was demonstrated by the Government of Meghalaya's FIRST Cancer Care program, which utilizes an FDA-cleared handheld device for early breast cancer detection.

Digital Infrastructure: Concerns were raised about the high energy consumption of Large Language Models (LLMs), potentially leading to "energy less" scenarios by 2050 due to resource depletion. This prompted a recommendation for India to focus on Small Language Models (SLMs) and develop "Public Process Models (PPMs)" with fewer parameters (e.g., less than 1 billion) for government functions, offering a more sustainable approach than global LLMs.

Future of Intelligent Systems in India: India is currently Rank 40 in the Oxford AI Maturity Index 2024. Discussion happened at the end of session - while AI may displace certain jobs, it will also "create new job’s", emphasizing the importance of "skill upgradation". AI's application in legal processes, such as summarizing judgments, etc was also mentioned. The session concluded that India's intelligent future depends on strategic development, robust governance, and ethical considerations, ensuring human oversight remains paramount.