About Us Contact Us Help


Archives

Contribute

 

US And India: Building The Joint Future Of Artificial Intelligence

Roopesh Mathur
09/12/2018

IIT AGNE, the association of IIT alumni in New England and TiE-Boston, hosted a spirited discussion on the pitfalls and potentials of using AI in the future on September 11th 2018 at The Hubspot in Cambridge MA. This event featured Prof. P.P. Chakrabarti, Director, IIT Kharagpur, Mr. Partha Ghosh, former Partner at McKinsey, Dr. David Cox, Director of MIT-IBM Watson AI Lab and Dr. Venkat Srinivasan, ex-academic, serial entrepreneur, author and former CEO of RAGE Frameworks.


In his opening remarks, Prof. Chakrabarti, outlined the significant investments being made at the IITs in research into technology that can scale to be a public good to provide services and opportunities to all segments of the large Indian population. Artificial Intelligence or AI, promises to be a chief technology that will empower large sections of Indian society, especially those at the bottom of the pyramid and in remote areas. IIT KGP has started Center of Excellence in Artificial Intelligence Research this year, that focuses on creating frameworks for AI in large areas of importance, such as healthcare and education.


Mr. Partha Ghosh, then broadened the discussion into the implications for society as whole and the evolution of humans, society and culture. There are significant issues at the core of widespread adoption with AI, such as implicit biases inherent in the data that AI algorithms uses learn from the real world and exploitation of AI for control and suppression of dissent by governments and private entities. AI can exacerbate inequality, since it depends on the application of large scale computing power to learn extremely large sets of data, which grants enormous power to entities that control both. Hence it is critical that AI is regulated and leveraged to empower human society.


Dr. Cox, a computational neuroscientist and AI researcher who recently left Harvard University to head the MIT-IBM Watson AI Lab, which is a unique collaborative endeavor between academia and industry, delved further into these issues. While acknowledging the pitfalls, AI offers a unique opportunity to create a better and more egalitarian world. For example, knowing and recognizing biases in real world data can allow the creation of “fair AI”. Fundamental breakthroughs in AI can be made that take away the need for massive computing power. AI can be regulated like a public good, such as electricity, that allows access and use to every segment of the population.

Dr. Venkat Srinivasan took the discussion in a more practical direction, where AI can be applied to current problems facing the world today, as a proponent of “Operational AI”. He outlined a framework that can be used to guide the development of new technologies based on AI, such as solving problems that cause inefficiency and bottlenecks and ensuring impact on target populations. He described his work at EnglishHelper, which is using AI to read and write English to large numbers of children all over India and the world, who would not have the opportunity otherwise.


This was followed by a lively discussion with a number of important questions being posed by the audience. AI is a vital part of the future for human societies, and while it reflects the inherent biases in human nature, it should become an enabler that empowers humans to rise above their circumstances, solve critical societal problems and discover new potential.  AI needs to evolve beyond statistical techniques and work with text and symbols, in the same way that children can learn about the world from a few examples.


Mr. Raj Melville and Ms. Nilanjana Bhowmick, President of IIT AGNE and TiE introduced the speakers. Mr. Ron Gupta of IIT Kharagpur foundation and Mr. Puran Dang, who is from the first batch of IIT KGP, welcomed the speakers and the audience.




Bookmark and Share |

You may also access this article through our web-site http://www.lokvani.com/



















Home | About Us | Contact Us | Copyrights Help