What is the story about?
As artificial intelligence tools rapidly become a part of everyday software development, a larger question is beginning to surface across the industry: what will it take for engineers to remain relevant?
For Sridhar Vembu, the answer is becoming increasingly clear. Coding, while still essential, is no longer enough on its own.
The Zoho cofounder believes that the future will favour those who understand not just how to build software, but why it needs to be built in the first place. His latest remarks come at a time when AI-powered coding assistants are being adopted widely, promising faster development cycles and greater efficiency. Yet, Vembu’s message cuts through the hype with a more grounded perspective.
According to Vembu, programming remains the foundation of any engineering role, but it is no longer the primary differentiator. Instead, he argues that deep domain expertise is what truly drives value in an AI-driven world.
In a recent post, he explained that customers are not simply paying for code, but for a combination of reliability, security, compliance and a clear understanding of the problem being solved. Engineers who can bring this broader perspective to the table are more likely to thrive, even as AI tools take over repetitive or mechanical aspects of coding.
This shift reflects a broader transition in the industry. As AI systems become more capable of generating code, the emphasis is moving towards higher-level thinking.
Understanding user needs, navigating industry-specific challenges and ensuring robustness are areas where human expertise still plays a critical role.
Vembu’s stance also signals a subtle but important change in how technical talent may be evaluated in the future. Rather than focusing solely on coding proficiency, organisations could increasingly prioritise those who combine technical skills with contextual intelligence and practical insight.
While AI tools such as automated coding assistants are often praised for boosting productivity, Vembu cautions that the reality is more nuanced. He describes the gains as “hotly debated”, noting that while AI can significantly speed up the creation of initial prototypes, the journey from a working model to a polished, production-ready product remains complex.
There is much more to software than just writing code, he suggests, pointing out that several stages in development cannot be easily accelerated by AI. As a result, he advises teams to avoid obsessing over programmer productivity as a standalone metric.
Instead, the focus should shift towards delivering better outcomes for users. Vembu highlights that AI can play a meaningful role in simplifying software by reducing unnecessary complexity, ultimately improving the overall user experience. This, he believes, is where the real opportunity lies.
His comments come in the backdrop of earlier, more controversial statements where he suggested that those solely dependent on coding for a living may need to rethink their career paths. He has also warned that individuals who tie their sense of self-worth purely to economic output could find themselves vulnerable in an AI-driven future.
At the same time, Vembu has pointed out that several human-centric roles, such as caregiving, teaching, mentoring and even creative fields like classical music, are less likely to be disrupted by AI in the near term.
An analysis by Anthropic suggests that the jobs most vulnerable to AI disruption are those built around handling information or performing repetitive cognitive tasks. These typically include roles that rely on reading, writing, analysing or organising data, areas where current AI systems are already highly effective.
The study highlights that professions involving coding, customer interaction and data processing are particularly exposed. For instance, computer programmers rank at the top, with AI capable of covering around 75 per cent of their tasks. Customer service roles also feature prominently, as AI is increasingly being used to handle queries through automated systems. Similarly, data entry positions, which revolve around extracting and inputting information, show a high level of automation potential at roughly 67 per cent.
Other roles identified include legal assistants and paralegals, proofreaders, technical writers, market research analysts, administrative staff, translators and content creators. What ties many of these jobs together is their reliance on structured, repeatable workflows such as drafting documents, editing text, responding to standard queries or interpreting organised datasets. These are precisely the kinds of tasks that AI tools can now perform quickly and at scale.
That said, the findings do not point to an immediate disappearance of these professions. Instead, they suggest a gradual shift in how work is carried out within them. Rather than replacing entire roles, AI is more likely to take over specific tasks. For example, it may generate drafts, summarise large volumes of information or assist with routine communication, while human workers continue to refine outputs, apply judgement and handle more nuanced or complex responsibilities that require context and empathy.
For Sridhar Vembu, the answer is becoming increasingly clear. Coding, while still essential, is no longer enough on its own.
The Zoho cofounder believes that the future will favour those who understand not just how to build software, but why it needs to be built in the first place. His latest remarks come at a time when AI-powered coding assistants are being adopted widely, promising faster development cycles and greater efficiency. Yet, Vembu’s message cuts through the hype with a more grounded perspective.
Domain knowledge over pure coding skills
According to Vembu, programming remains the foundation of any engineering role, but it is no longer the primary differentiator. Instead, he argues that deep domain expertise is what truly drives value in an AI-driven world.
In a recent post, he explained that customers are not simply paying for code, but for a combination of reliability, security, compliance and a clear understanding of the problem being solved. Engineers who can bring this broader perspective to the table are more likely to thrive, even as AI tools take over repetitive or mechanical aspects of coding.
Here is what I tell our software engineers on how to thrive in the AI era: be very good domain experts. Programming skills are the foundation (and we definitely don't want to lose them) but deep domain knowledge is what customers pay for, along with reliability, security, support…
— Sridhar Vembu (@svembu) April 19, 2026
This shift reflects a broader transition in the industry. As AI systems become more capable of generating code, the emphasis is moving towards higher-level thinking.
Understanding user needs, navigating industry-specific challenges and ensuring robustness are areas where human expertise still plays a critical role.
Vembu’s stance also signals a subtle but important change in how technical talent may be evaluated in the future. Rather than focusing solely on coding proficiency, organisations could increasingly prioritise those who combine technical skills with contextual intelligence and practical insight.
Rethinking productivity in the age of AI
While AI tools such as automated coding assistants are often praised for boosting productivity, Vembu cautions that the reality is more nuanced. He describes the gains as “hotly debated”, noting that while AI can significantly speed up the creation of initial prototypes, the journey from a working model to a polished, production-ready product remains complex.
There is much more to software than just writing code, he suggests, pointing out that several stages in development cannot be easily accelerated by AI. As a result, he advises teams to avoid obsessing over programmer productivity as a standalone metric.
Instead, the focus should shift towards delivering better outcomes for users. Vembu highlights that AI can play a meaningful role in simplifying software by reducing unnecessary complexity, ultimately improving the overall user experience. This, he believes, is where the real opportunity lies.
His comments come in the backdrop of earlier, more controversial statements where he suggested that those solely dependent on coding for a living may need to rethink their career paths. He has also warned that individuals who tie their sense of self-worth purely to economic output could find themselves vulnerable in an AI-driven future.
At the same time, Vembu has pointed out that several human-centric roles, such as caregiving, teaching, mentoring and even creative fields like classical music, are less likely to be disrupted by AI in the near term.
Top 10 jobs at highest risk
An analysis by Anthropic suggests that the jobs most vulnerable to AI disruption are those built around handling information or performing repetitive cognitive tasks. These typically include roles that rely on reading, writing, analysing or organising data, areas where current AI systems are already highly effective.
The study highlights that professions involving coding, customer interaction and data processing are particularly exposed. For instance, computer programmers rank at the top, with AI capable of covering around 75 per cent of their tasks. Customer service roles also feature prominently, as AI is increasingly being used to handle queries through automated systems. Similarly, data entry positions, which revolve around extracting and inputting information, show a high level of automation potential at roughly 67 per cent.
Other roles identified include legal assistants and paralegals, proofreaders, technical writers, market research analysts, administrative staff, translators and content creators. What ties many of these jobs together is their reliance on structured, repeatable workflows such as drafting documents, editing text, responding to standard queries or interpreting organised datasets. These are precisely the kinds of tasks that AI tools can now perform quickly and at scale.
That said, the findings do not point to an immediate disappearance of these professions. Instead, they suggest a gradual shift in how work is carried out within them. Rather than replacing entire roles, AI is more likely to take over specific tasks. For example, it may generate drafts, summarise large volumes of information or assist with routine communication, while human workers continue to refine outputs, apply judgement and handle more nuanced or complex responsibilities that require context and empathy.
