The rise of generative AI, exemplified by ChatGPT, has captivated the public by offering unmediated access to advanced models. Prof Ravindran described this as a “co-pilot” model, augmenting human intelligence in fields like programming, design, and healthcare. However, he cautioned against over-reliance on such systems due to critical issues like bias, robustness and misinformation. He cited real-world examples, such as biases in recruitment and judicial tools, which underscore AI’s potential to perpetuate harm if unchecked.Turning to India, he lauded homegrown innovations like the Garbhini-GA models for pregnancy care and a muzzle-pattern recognition system for cattle identification. These projects demonstrate how AI can address India-specific challenges. However, he warned that the lack of diverse datasets and a gap between proofs of concept and scalable solutions hinder progress.Prof Ravindran advocated for a holistic approach to AI regulation, involving government, industry, and academia. He proposed a multi-stakeholder safety institute to address concerns like bias, deepfakes, and environmental impact while fostering responsible AI development.Concluding with a word of caution, he stressed that AI, for all its advancements, still struggles with simple human tasks. He called for better public education on AI to demystify its capabilities and align its potential with societal needs.
Source link
