Governing AI: Indian Dimension

When the Philosophical Transactions of the Royal Society decides to devote an entire issue to ‘Governing artificial intelligence: ethical, legal, and technical opportunities and challenges’, one can be sure that a. the topic is important, and b. the discussion will be of the highest quality.  One is not disappointed by the November 2018 issue, edited by Corinne Cath et al.

I am still wading through the issue and will visit a few articles when I get time. Meanwhile, the one article that is relevant and interesting in this issue is on ‘Artificial intelligence policy in India: a framework for engaging the limits of data-driven decision-making’ by Vidushi Marda. Her affiliation is listed as a Lawyer and Digital Programme Officer at ARTICLE 19 and a research analyst at Carnegie India. The article focuses on potential limitations and risks that arise from data-driven decisions in general, and in the Indian context in particular. She rightly argues that “AI applications operate in societies that are chaotic, biased, unequal and steeped in historical discrimination and disparity”. Treating such issues as purely statistical or programming artifacts wouldn’t be enough and social/ethical dimensions must be considered in the process of building AI systems and policy. The author also argues forcefully that  India is an important jurisdiction to consider for a number of reasons for discussions related to AI policy and governance. Overall, the article is well articulated and very relevant.

Reference:

  1. Philosophical Transactions of the Royal Society A, 15 November 2018 Volume 376 Issue 2133

About the author

Ardhendu Pathak, Ph.D.

I have spent more than 3 decades, first as a researcher in academia and then as a technology executive with industrial giants GE and AIRBUS. During this time I have traversed several industries from ocean engineering to aerospace, with appliances and electrical distribution in between. As an engineer and also as a student of philosophy I have an abiding interest in technology and its ethical use. This blog is my attempt to highlight and clarify some of the key issues at the intersection of technology and ethics, and to some extent, law and corporate governance as well.
If you are an educator, policymaker, corporate advisor, or a technology student, then you will find this blog interesting. Your comments on the posts or shares can help me and others as well.

Ardhendu G PATHAK, Ph.D.

View all posts

4 Comments

  • Useful and relevant like your earlier posts. For physical products we have specs. Is there an analogue for digital (AI) products?

    • I like your thought of comparing AI solutions to physical products

      For AI productive we have objective functions . The objective is to be achieved in an environment of constraints and rewards.

      If constraints are well defined many of the fears critics have can be taken care off.

  • India has so much abundance of natural intelligence and a land of overwhelmingly large population of highly opinionated philosophers! Basically everyone is a philosopher. Even when a road digger is instructed by engineer to dig a 2 feet deep trench, he thinks! Pipe is only 4″ diameter why 2′ trench? And he digs only 1 foot! So can’t agree more, India needs for sure a considered policy on AI.

  • Thank you Ardhendu,
    Very relevant article, and of course, it applies the world over! Deep learning technology can make sense of a galaxy of data and reveal underlying chaotic and dysfunctional realities; the risk is that the sheer volume of the data and the repetition of situations may lead to a sense of acceptability of these realities as normal. If used to predict and propose outcomes, AI must be guided by external objectives given by the policymaker/owner of the algorithm.

Leave a Reply

Your email address will not be published. Required fields are marked *