AI Transformation: Impact on India’s Service-Based Economy

AI transformation is not just evolving; it is advancing at an alarming rate that poses significant risks to India’s service-based economy. Recent information from Zerodha co-founder Nithin Kamath illustrates a concerning scenario: artificial intelligence can now undertake longer and more intricate tasks than ever before, progressing so swiftly that numerous human roles might become obsolete.

AI’s Task Execution Capabilities

One graph showcases that the duration of tasks AI can complete with 50% success is doubling every seven months. Only a few years ago, models such as GPT-2 could manage brief, basic inquiries. Currently, GPT-4o can effectively maintain a conversation for nearly an hour.

The most sophisticated models, such as Sonnet 3.7, are already capable of managing multi-hour assignments like training image classifiers. These developments are far beyond mere experimentation; they are emerging as real alternatives.

AI Accuracy and Task Complexity

A second graph paints an even more alarming picture. It charts AI’s success rates based on task complexity and duration. On straightforward, brief tasks, leading models are now achieving over 90% accuracy. Even complex and ambiguous challenges, once deemed impossible for AI, are witnessing notable improvements. The only area that still struggles is the category of long and complex tasks, which remains below 30% success.

However, even this threshold is beginning to diminish.

Faced with this information, Kamath posed a thought-provoking question on X: “How long do we have left?”

Concerns from Industry Leaders

He is not the sole voice expressing concern. Atomberg founder Arindam Paul forecasts that 40–50% of white-collar roles in India, particularly in IT and BPO, could disappear.

A recent study from IIM Ahmedabad reveals that 68% of white-collar professionals anticipate that AI will automate their jobs, at least in part, within the next five years. Jefferies projects that half of all entry-level white-collar positions—from marketing to sales to software—may be eliminated in just 1–5 years.

This situation is not hypothetical. It is unfolding now. India, which heavily depends on services and has limited manufacturing capabilities, is particularly exposed to these rapid changes.