Generative AI, a revolutionary subset of artificial intelligence (AI), is at the forefront of transformative technological advancements. At its core, Generative AI focuses on creating content, data, or even entire experiences rather than just analyzing or processing existing information. This innovative approach relies on complex algorithms and neural networks to generate new, original content that often mirrors human-like creativity and problem-solving. As this technology continues to evolve, it holds immense potential to reshape industries. To get crucial insights about this next-gen technology, DT talked to Ramprakash Ramamoorthy, Director - AI Research, Zoho Corp who briefly elaborated about increasing use and implementation of this technology across different verticals.
The Rise of Generative AI in India
The organizations across India embracing Generative AI to thrive in the fast evolving digital world. The Indian organizations have increasingly recognized the potential of Generative AI for driving progress. As Generative AI continues to gain momentum in India, it may play a pivotal role in the nation's technological evolution. Commenting on the rapid growth of Generative AI, Ramprakash Ramamoorthy, Director - AI Research, Zoho Corp said, “Over the last year, generative AI has created a lot of buzz in India, achieving peak penetration in the usage of consumer-grade large language models (LLMs). Employees also use it in their personal lives. However, when it comes to formal business applications, there are still 'unknown unknowns,' particularly concerning privacy. This ambiguity has led to some hesitancy among Indian companies to adopt generative AI in an official capacity. Additionally, the associated costs for deploying commercial LLMs in business settings are significant. While open-source LLMs are available, they are yet to catch up in terms of accuracy and reliability.”
Skills Shaping Effective Generative AI Implementation
Deploying Generative AI in various sectors necessitates a multifaceted skill set among IT professionals. Additionally, expertise in data privacy and security is paramount to safeguard sensitive information. Putting across his views on the need of skills for effective use of this technology, he stated, “Generative AI presents both opportunities and challenges. One of the critical aspects to understand is that these models can 'hallucinate' or generate incorrect or nonsensical information. There are no built-in confidence indicators, so it's crucial to strike a balance between the productive use of these models and the risk of generating incorrect solutions. Equally important is adhering to the privacy policies associated with using LLMs. Organizations must be cautious not to input sensitive customer or business data into these models, as the consequences of data leakage could be severe. A nuanced understanding of these trade-offs is essential for secure and effective deployment.”
Supporting Organizations Across Verticals to Deploy Generative AI
Supporting organizations across verticals to deploy Generative AI involves a tailored approach to harness the technology's potential effectively. It starts with a comprehensive assessment of specific industry needs and challenges. Talking about their role in helping organizations to leverage Generative AI, he said, “At Zoho, we have over 11 years of experience in AI research and development, which has equipped us with the deep technical know-how required to navigate the complex landscape of AI technologies. We are currently developing our own enterprise-focused LLMs that are built on robust evidence and include safeguards against model hallucinations. While we already offer some generative capabilities, they currently rely on small to medium-sized language models.”
“As we move forward, we intend to integrate the emergent capabilities of LLMs into specific use-cases without imposing unnecessary computational costs, commonly known as the "GPU tax" on our customers. Our unified suite of products also allows for a 360-degree view of customers and employees based off numerous data points across our offerings, making it a fertile ground for AI innovation. Our existing AI stack as well as our ongoing projects are also built keeping in mind our three core tenets of AI as we see it: privacy, value, and user experience,” he concluded.