AI to End Coding-Centric Era: Infosys Chairman Forecasts 170M New Roles

The era of the traditional software developer, defined primarily by the ability to write manual code, is nearing its end. Speaking at Infosys' Investor Day on February 17, 2026, Chairman Nandan Nilekani delivered a stark assessment of the technology landscape, characterizing the rise of Artificial Intelligence as a 'root-and-branch' transformation. Unlike previous technological shifts such as the move to mobile or cloud computing—which Nilekani described as additive layers—AI represents a fundamental rewrite of how businesses operate, from customer journeys to organizational structures. This shift marks a departure from deterministic systems, where a specific input always yields the same output, to a non-deterministic paradigm where AI agents and models produce varying outcomes based on the same prompt.

Nilekani’s central thesis is that writing code will no longer be the primary goal of technology professionals. Instead, the industry is pivoting toward 'AI engineering' and 'agent orchestration.' This transition requires a workforce capable of managing complex, autonomous agents that can execute tasks, rather than simply writing the syntax to perform them. While this disruption may seem threatening to the existing IT services model, Nilekani remains optimistic about the scale of opportunity, forecasting the emergence of approximately 170 million new roles globally. These positions will focus on 'forward deployment'—the bridge between powerful AI models and the specific, often messy, realities of enterprise operations.

“

Speaking at Infosys' Investor Day on February 17, 2026, Chairman Nandan Nilekani delivered a stark assessment of the technology landscape, characterizing the rise of Artificial Intelligence as a 'root-and-branch' transformation.

A significant bottleneck identified during the briefing is the 'deployment gap.' While the performance of large language models (LLMs) and specialized AI continues to accelerate, the actual implementation within large enterprises remains sluggish. Nilekani noted that technology is currently far ahead of its deployment because the 'hard stuff' of implementation involves deep organizational change, the breaking down of data silos, and the modernization of legacy systems. For many Global 2000 companies, the challenge is not just acquiring AI, but restructuring their entire operating model to accommodate it. This creates a massive demand for IT services firms to act as consultants and integrators who can navigate this complexity, rather than just providing 'bodies' for coding projects.

The shift to non-deterministic systems also introduces a new layer of technical risk and management. In traditional software, a bug is a logic error that can be traced and fixed; in an AI-driven system, the same prompt might yield different results depending on the model's state or context. Managing this variability requires a new skillset in AI governance and quality assurance that the current workforce is only beginning to develop. Nilekani emphasized that the transition will be difficult for talent, requiring a complete rethink of what it means to be a 'tech worker.'

Looking ahead, the success of major IT players like Infosys will depend on their ability to retrain hundreds of thousands of employees for this new reality. The focus is shifting from 'how to code' to 'how to make AI work.' This involves not just technical proficiency but a deep understanding of business processes and the ability to orchestrate multiple AI agents to achieve a specific business outcome. As AI models become more commoditized, the value will increasingly reside in the ability to deploy them effectively within complex enterprise environments, making the 'last mile' of AI integration the most valuable segment of the market.