Generative AI is projected to increase annual global GDP to the amount of trillions if implemented properly, according to analysts. What is India’s stake in this new global phenomenon? In today's podcast, we look at generative AI and LLMs, the issues they pose, the effect they have on Indian society and industry, and how we can look forward to regulating AI while allowing it to grow and provide value to the greatest extent.

Our guest, Nidhi Singh, is a Programme Officer, with CCG Delhi. She is an expert in AI, privacy, internet governance, and emerging technologies. She is hosted by Varsha Radhakrishnan, a Fellow at the Centre for Public Policy Research. Varsha is an expert in emerging technologies and AI, having industrial expertise in this sector.

1. The training data used in AI is heavily open sourced, and based on data from the open internet. This poses problems both in terms of the accuracy of the data as well as privacy, as the data scraped off might not be consented to by parties.

2. There is a lot of interest in India to build LLMs that are usable for citizens. Language poses an issue, as there may not be a vast tract of trainable data in particular languages.

3. There is a global north-south divide in the implementation of these technologies, as many new technologies are being created in the global north and implemented in the global south, leading to a large number of problems.

4. While it is true that policy catches up slower to new technologies, it is important to remember that the problems caused by the implementation and usage of these technologies are fundamental issues of access, equality, and privacy and can be looked at from a constitutional lens.

Key Takeaways: