‘AI will gobble up most low-hanging jobs of coders'

The Hindu

26-07-2025

Tech India is witnessing a transformational shift, with its bellwethers currently busy deploying their AI-first strategy to nurture increased human+AI collaborations at their workplaces.
This means a few unpleasant things for the software developer community(software): most of the low-hanging, also some of the not so low-hanging coding jobs will go to their AI colleagues; AI will make some developers and their current contribution to coding fully redundant; developers will need to adapt and learn how to work with their AI partners. However, AI has the potential to free up developers from boring, repetitive tasks, allowing them to focus on deep-coding, for complex, intelligent and creative projects.
Lead tech players have been quite vocal about their focus on human+AI play, more so in the last couple of months, signalling an impending shift in the tech talent space. For instance, N. Chandrasekaran, Chairman of Tata Sons and Tata Group recently wrote to TCS shareholders, in the company's FY25 annual report, that the rise of automation promised a future of 'dark factories.' IT and business services were moving toward autonomous operations and the path was clear, he told shareholders adding, 'GenAI is not just another tech cycle — it is a civilisational shift. TCS is uniquely positioned to lead this transition.''
Wipro Chairman Rishad Premji, more recently at an annual general meeting said, AI, especially generative and agentic AI, was becoming a game-changer and it was helping his company to rethink how it should work and uncover new growth opportunities.
Interestingly, both Wipro and Infosys recently announced the deployment of 200 AI agents each on certain functions to free up people.
Infosys deployed 200 enterprise AI agents as part of its Infosys Topaz AI offerings and in collaboration with Google Cloud's Vertex AI platform. These AI agents are designed to automate complex workflows, enhance decision-making, and improve efficiency and reduce costs across various industries like healthcare, finance, telecom, retail, agriculture and manufacturing.
Wipro said, agentic AI was already being integrated into core business processes, and it built over 200 agents in partnership with hyperscalers. These agents were meant to independently handle tasks across departments such as HR, finance, and legal, driving scaled efficiencies and outcomes.
These are not coding jobs, however, analysts say these trends will solidify and become secular in time, redefining all job roles, including that of developers, across the industry.
'Given the transformational and disruptive capacity of AI, organisations no longer require an army of coders, most of them with low single digit experience in software development. The immediate impact we are seeing is the reduction in intake of junior level talent, and shrinking in size of project teams,'' said Muthu Kumaran, Operating Partner and Head of India Operations, Recognize, a U.S.-based private equity firm.
Most of the mundane activities, around typical software development lifecycle, such as code generation, code refactoring, integration, debugging, and validation could go to AI counterparts, said Mr. Kumaran who earlier held various roles across technology and global delivery at Cognizant.
According to him, most organisations today are using AI as a complementary tool, a machine acting as a peer to a human/developer. When the shift happens towards using AI as a complete replacement, there is a high probability that coding jobs become restricted to understanding complex systems and debugging non-trivial issues. 'Most of the coding jobs as they exist today could be done by AI and coders will have to move to higher forms of architecture and system design,'' he forecast.
Dr. Vikas Khare, Associate Dean School of Technology, Management and Engineering, NMIMS, at Indore said, many repetitive and template coding chores, including bug fixes, basic HTML/CSS web construction, basic CRUD operations, boilerplate generation, and test case writing, were currently being automated by AI.
Junior-level programming positions involving predictable and well-documented problems have already started to be replaced by tools like GitHub Copilot, Replit AI, and low-code/no-code platforms, he noted.
``As generative AI models advance over the next five years, they will probably be able to manage increasingly complicated front-end development, backend API design, integration chores, and even some data analysis and pipeline building duties. It is also possible to fully automate old code upgrades and routine program maintenance.'' he further said.
Neeti Sharma, CEO, TeamLease Digital, said many AI tools were capable of writing codes, providing real time suggestions and even writing full functions on simple prompts, and therefore many junior developer roles such as web or mobile app development, testing and documentation would be replaced by AI in the next few years.
'The role of a developer will get redefined from pure coding to using complex problem solving skills, judgement and including creativity – skills that AI cannot fully replicate,'' she added.
Flip side
However, she said, any kind of coding that involved creative problem-solving, strategic design, and understanding complex user needs is likely to remain human-centric. Dominic Pereira, VP, Product Management, Automation Anywhere, said AI was already transforming routine coding tasks by automating standard front-end design, back-end operations like server setup and database management, and continuous code optimisation. 'Over the next five years, one will witness this trend accelerating. However, complex problem-solving, creative design, and strategic planning will remain human-driven,'' he added.
Sreekumar Pillai, CTO at Experion Technologies said, Looking ahead, over the next five years, we expect AI to handle more advanced coding tasks that involve pattern recognition across large codebases and even some elements of system design. This could include generating microservices from business rules, automating code migration, and performing AI-driven debugging.
B.S. Murthy, CEO, Leadership Capital, said, wherever human ingenuity, critical thinking & imagination is needed, AI is yet to make a huge practical impact on tasks or applications. AI is not likely to replace coders/system engineers who code in C++, which is used to build operating systems, gaming, graphics and critical secure applications. 'AI will not immediately replace domain competencies like tech architects, dev ops, UI/UX, product management, robotics & embedded systems. Talent high on Math & imagination will rule the roost in this decade,'' added Mr. Murthy.
Developers would evolve into supervisors and collaborators who focus on strategic decisions, ethical considerations, domain-specific logic, security planning and creative problem-solving that AI cannot replicate, said experts.
Ayan Barua, co-founder and CEO of Ampersand, said AI is eating the grunt work of software, not the craft. 'In five years, the machines will write most of the scaffolding, but humans will still design the skyscraper.''