Anthropic CEO reveals career options for Indian students entering software and engineering roles

As AI systems take over more coding tasks, Anthropic CEO Dario Amodei says software and engineering careers are changing rather than disappearing. He advises Indian students to focus on design, judgement and human-centred work that AI is slower to replace.

As AI systems grow more capable of writing code and handling routine technical work, students planning careers in software and engineering need to reassess where long-term opportunities may lie. That was the central theme of a recent conversation between Dario Amodei, CEO of Anthropic, and entrepreneur Nikhil Kamath, where Amodei spoke at length about how AI could change technical careers.

Amodei said AI models are likely to take over coding tasks before they replace more of the software engineering roles. "I think coding is going away first, or coding is being done by the AI models first,” he said, drawing a distinction between writing code and the larger engineering task of deciding what to build and how systems should work. According to him, full end-to-end automation of software engineering will take longer.

For students in India deciding what skills to focus on, Amodei suggested looking beyond pure coding. He pointed out that design, understanding user demand, and system-level thinking will continue to matter. “The elements of design or making something that’s useful to users, or knowing what the demand is those things may still be present,” he said, even as AI handles much of the execution.

Amodei also spoke about the idea of “comparative advantage” in an AI-driven workplace. Even a small human contribution can become valuable when paired with powerful AI systems. “Even if you’re only doing five per cent of the task, that five per cent gets super-amplified,” he said, explaining how people could become far more productive by focusing on what humans do better.

Beyond software roles, the Anthropic CEO revealed other areas that may offer more stability. He said students should consider human-centred roles and work that involves the physical world, where context, regulation, and real-world interaction matter. He also pointed to the semiconductor space as an example of an industry with strong tailwinds, noting that it combines physical engineering with long-term demand and is harder to automate quickly.

For aspiring founders and developers, Amodei offered a clear warning against building shallow products on top of AI models. “You shouldn’t be just a wrapper,” he said, warning against apps that simply offer a basic interface over an existing model. Instead, he advised building businesses with strong "moats," something like areas of expertise or depth that large AI labs may not want or be able to replicate.