Experts Emphasise Robust Planning For Successful Gen AI Implementation

Artificial intelligence, particularly in the form of generative AI or GenAI, which produces diverse data outputs from text to 3D models, requires meticulous planning to ensure data consistency and integrity, experts assert.

Experts delved on this at a panel discussion at the BW Businessworld Gen AI Summit. Santanu Bhattacharya, Scientist at MIT Media Lab, cautioned against the hype surrounding Gen AI, highlighting the importance of robust planning amidst the evolving landscape. "Today we are in the middle of a hype cycle in Gen AI. When you have a tool called a hammer, everything looks like a nail," Bhattacharya remarked. He noted the shift where problems once classified as regression now fall under the umbrella of AI, and now, Gen AI.

Talking about Gen AI at a higher level, Nishith Pathak, Global Lead- Architecture & Innovation, DXC Technology said that organisations need a clear roadmap about how they want to utilise the investment in Gen AI. Talent is always a scarcity during the kind of transformation that is happening. 

"Having a robust framework for doing a data governance and having an ethical mode of utilising AI is always been a criteria," he stated.

When asked about ethical concerns in Gen AI Lily Prasad, Chief Technology Officer of Food Safety and Standards Authority of India about addressing ethical concerns in Gen AI implemsaid, "It is the most hyped term in the IT industry nowadays. As far as ethical business is concerned, we need to have very robust planning on how I am going to use [Gen AI], what is the method adopted or the model," Prasad emphasised. She underscored the importance of data consistency, integrity, and a single source of truth in ethical Gen AI usage.

On the success story of Gen AI Upendran Nandakumar, Founder & CEO AyatiWorks said, in healthcare segment instead of going to Google, patients go to the website  to ask questions related to their health and it helps them to analyse their overall health. 

He added, this is going to be a continuous learning process and we need to keep learning.

Madhav Bissa, Program Director at NASSCOM, outlined challenges faced during Gen AI implementation, categorising them into data, compute, models, talent, regulation, and intellectual property (IP) realms. "The challenges can be categorised as data, compute, models, talent, regulation, and IP. So what happens is that these challenges remain the same and the categories remain the same. It's just the form and structure change," Bissa elucidated.

The insights provided by experts underscore the critical need for meticulous planning to overcome challenges and ensure ethical and successful implementation of Gen AI technologies. As the field continues to evolve, a comprehensive approach to address data, computational, regulatory, and ethical aspects is imperative for harnessing the transformative potential of generative AI.