Balancing hype and reality of AI’s potential in India's digital transformation journey

When people talk about “AI hype,” we really need to ask what sort of Artificial Intelligence (AI) they are talking about. 

Generative AI (Gen AI) is new, while other types have been around for decades. Some types of AI are familiar while others are obscure or planned for the future, like artificial general intelligence. A decade ago, certain types of AI were novel technologies used by researchers and enterprises to play and win games against humans and achieve significant advances. With new Gen AI technologies such as ChatGPT, GitHub Copilot for developers, Sora and Midjourney, AI is now a major topic of discussion, and enterprises are utilizing these new capabilities. Business executives rank AI and Gen AI among the top three tech priorities for 2024. In fact, 59% of Indian enterprises have already deployed AI, and 60% of them plan to do so in the next 12 to 24 months.

Deep learning neural networks are powerful AI tools that mimic the human brain, specifically, convolutional neural networks in machine vision, where connectivity is inspired by the brain’s visual cortex. However, they are not infallible. Sometimes engineering teams expect them to perform flawlessly. Therefore, it is important to educate stakeholders on the capabilities and limitations of neural networks. Neural networks can achieve remarkable results but need to be applied judiciously. 

However, neural networks can also fail in unexpected ways, such as misclassifying objects due to adversarial attacks or dataset bias. By understanding both the strengths and limitations of neural networks, engineering teams can leverage this powerful technology to drive innovation while mitigating risks. Realistic expectations should be based on the areas where neural networks excel such as the detection of surface defects, detecting or counting objects, reading difficult characters, or detecting anomalies.

Businesses need guidance to cut through the hype and secure AI-driven value for their operations. This involves understanding the practical applications of AI within their specific industry, navigating the implementation complexities, ensuring data privacy and ethics, and continuously assessing the return on investment (ROI) to ensure sustainable growth and innovation. Over 81% of senior Indian business leaders in a survey conducted by Tata Consultancy Services have called for more 'global' regulations and standards on AI, as companies strive to balance security risks and opportunities in their AI strategies.

Currently, it is not so much AI creating and eliminating jobs, although many headlines suggest this. As with the car, telephone, and internet, new jobs and industries will emerge thanks to the growth of AI. According to Nasscom, India has an installed talent base of 416,000 professionals in Data Science and AI, with a current demand of approximately 629,000 professionals, and this is predicted to rise by 2026. Further, sectors like manufacturing amongst others are equipping their front-line workers with new and better AI tools to work faster and more efficiently, while handing certain tasks over to AI-driven automation. 

Moving forward, it is likely that companies will set the democratization of AI and machine learning as a strategic priority. Some tools – like deep learning optical character recognition (OCR) – can be low orno code, meaning they are ready out-of-the-box and do not require specialist training. Other tools are higher end and operate more like ready-made environments for programmers and data scientists to create solutions using provided platforms, tools, and libraries. Whether an engineer, data scientist or developer, the workforce will be upskilled and provided with learning resources and ready-made, easier-to-use AI tools.

Subramaniam Thiruppathi

---

Subramaniam Thiruppathi is Director Indian Subcontinent (ISC) at Zebra Technologies.