Artifical Intelligence is rewriting the grammar of software development. And writing codes will no longer be the central role for tech professionals, Infosys co-founder Nandan Nilekani said Tuesday.

AI, he said, is being adopted faster than any previous technological transition, from the internet to smartphones, and is poised to fundamentally reshape how businesses operate.

"Talent will have to deal with a world where writing code will not be the goal. It'll actually be making AI work, orchestration, and those kinds of things," Nilekani said at Infosys' Investor Day. "Customer journeys, operating models, and mental models all have to change. Every enterprise must rethink how it operates."

While coding may end, new jobs will be created.

The talent transformation is huge, he said adding there will be a need for AI engineers, forward deployment engineers, forensic analysts - roles that didn't exist a few years ago.

Greenfield coding productivity is not the real challenge. The real world is managing trillions of dollars in legacy systems with undocumented dependencies, he said.

He highlighted the unprecedented speed of AI adoption, noting that while the internet took over a decade to reach a billion users and smartphones five years, AI adoption is happening in just a couple of years, enabled by the infrastructure of prior tech eras.

"While the internet took over 10 years to reach a billion users and smartphones five years, AI adoption is occurring in just a couple of years," he said. "Each technology transition has had implications, but this time, it's a fundamental change in the way businesses operate. Customer journeys, operating models, talent - everything has to change."

Nilekani stressed that legacy system modernization can no longer be deferred, citing high maintenance costs, siloed data, and rising security risks.

"Accumulated tech debt over decades must be paid. You no longer have the option to defer this," he said, noting that AI provides tools to modernize faster and more economically, though implementation remains complex.

He also emphasized the shift from buying software to building AI solutions, with enterprises layering agentic AI interfaces on existing systems to simplify customer journeys. "Foundational systems will increasingly become systems of record, but the interface will be agentic," Nilekani said.

On legacy systems and modernization, he said accumulated tech debt over decades must be paid. Many large companies spend 60 per cent to 80 per cent of their IT budget on maintenance. That has zero business value. It’s time to flip the ratio.

AI gives us the tools to modernize fast, economically, and at scale - but implementation is the hard part, he said.

Highlighting the deployment gap between AI capabilities and enterprise implementation, he noted the urgent need for talent transformation, change management, and cleanup of technical debt. "It’s not about using AI tools. It's about productivity out of those tools. Otherwise, you'll get false productivity," he said.

Nilekani concluded that enterprises must adopt first-principle thinking, agnostic system design, and cross-firm collaboration to harness AI successfully. "The opportunity is bigger than ever before, but the risk is in execution, not in the opportunity itself," he said.