Artificial Intelligence and the Knowledge question

​

Working Group: Rajeev Sangal (Convener), Avinash Jha, J K Suresh, Sunil Sahasrabudhey, M Sasikumar, Shatrunjay Rawat , Nachiket Udupa

​

AI or Artificial Intelligence is much talked about these days. There are predictions that it will enter into all aspects of our lives - from production to education and health, from distribution of products to information and entertainment, from social affairs and justice to politics and societal decision making.

There are fears of centralization of power, disempowerment of users, bias (unintended as well as planned), breach of privacy, loss of personal data, malevolent uses, malfunction of technology, and finally going all the way to takeover of human kind. Justifications for technology come from benevolent uses of technology - producing better health, mitigating disasters, performing of hazardous tasks by machines, to providing comfort, and indeed life without any work!

What about AI and Knowledge? What knowledge do AI systems possess today? Is it explicit or tacit (including intuition)? If it includes both types of knowledge can the machine reason with it transparently? Are systems explainable? Can they carry out a dialogue? Or the systems are completely opaque? If so, what are the implications - what new problems does this raise?

How do we deal with bias that might come into AI due to machine learning based on data? Who takes responsibility, if an AI system has bias?

Existing data reflects inequities and injustices of the past. Would AI systems, which use the same data for machine learning, end up propagating and promoting these? Human society has been evolving, at least yearning for equity and justice, non-violence, truth, etc. Valiant attempts have been made in history to build better societies. Would AI mean the end of evolution of society, leading to staying stagnant in status quo?

AI technology, like any other technology, would get used by the powerful and the privileged first. Would it lead to greater, even unprecedented, centralization of power?

Today, the big technology companies already possess private data on hundreds of millions of users, which they use to manipulate them. This manipulation is not limited to their buying habits and showing of directed advertisements, but is moving towards control over their thoughts. Control is exercised on circulation of certain types of messages, especially containing negative opinions and emotions. AI embedded in algorithms, is already playing a role. With the future developments of AI, such manipulative power would get deployed with even greater "ferocity", but quietly.

AI promises comforts, and a life without any work. In this scenario, the human beings become totally dependent on machines. It raises the ultimate question about what is the human "purpose"? Life with food, clothing and shelter, including play and entertainment, but bereft of emotions and relationships and without any meaning - would it lead to happy human beings and a life worth living?

The discussion on the theme “ AI and the Knowledge question” in the Convention will be in the form of a Panel , and would seek to touch upon all these issues and offer insights into them.

​

​