On 8th October, as a part of the RAISE2020 summit, an insightful panel discussion was conducted. The panel consisted of industry experts working in various sectors and trying to make AI use in those sectors and domains more ethical.

The Keynote speaker of the day was Dr. Jassim Haji (President, Artificial Intelligence Society, Bahrain) called the cultural values and ethical nature of nations like India and Bahrain as key to implementing ethics in AI. According to him, eventually, it narrows down to us humans in inducing high ethical standards. Mr. Abhishek Singh as a host and Mr. Shameek Kundu made the session more practical and informational than technical know-how one so that people not even involved right now can get motivated after understanding the value of the whole idea.

The panel gave ideas on how we have to start focussing on operationalizing ethics, and for that, we need to Articulate ethics first and then execute it. Ms. Kay Firth-Butterfield, Head of AI&ML, Centre for the Fourth Industrial Revolution, WEF (Structures for responsible management of AI), said that we aren’t successful with full implementation and execution of AI until we take everybody along with a global perspective and a team as diverse as possible.

To make our AI implementation more ethical, we must have ethically sourced data, consent from the people sharing data, clean and fair data, and proper benefits shall be given to the data providers. These were the views of Mr. Vilas Dhar (President, Patrick J. McGovern Foundation). He also stressed on the importance of feedback loops in achieving higher ethical standards. Adding on to this, Dr. Balaraman Ravindran (Head, Robert Bosch Centre for Data Science and AI at IIT-M) also talked about including “ethical clearance” in most of the projects.

The distinguished speakers also stressed enough on how it is critical to re-train, educate, and then train and motivate organizations, employees, and stakeholders to have a truly ethical AI usage in all our systems.

The discussion also revolved around real-life scenarios where speakers saw AI being executed wonderfully or wherever the take away was that there is a long way to go for us in achieving ethics in AI. In continuation of this, Ms. Arunima Sarkar (AI Lead, for Centre for Fourth Industrial Revolution, India, World Economic Forum) talked about gender bias in recruitment in a reputed organization, which she said was another perspective at ethical decision making. 

This initiated the discussion with Mr. Keshav Dhakad (Legal Head, Microsoft India) on how we even on very large scales, fail in achieving high quality and ethics, which points to the need for more rigorous work in the field. He shared how Microsoft created a chatbot after much research and yet some trolls taught the bot some abusive language that the bot couldn’t differentiate between a good command or bad command or even response for that matter.

Ms. Fernanda Viegas (Co-lead, Google Pair) shared her thoughts on the same and said that in high stakes decisions, ethical AI should be at the forefront with the belief that the whole world is our stakeholder and must adopt the law of humanity more than anything. She said we have also come long ahead in the journey of having people from all sorts of educational, religious, and nationality to make an inclusive, ethical AI experience.

An excellent thought that summarised it all was given by Mr. Will Griffin (Chief Ethics Officer, Hypergiant) “Adhere to the best personal ideals and best self values in decision making.” 

Image by succo from Pixabay