AI was supposed to kill engineering jobs: why new data shows the opposite

For the past few years a confident prediction has circulated through the technology industry: that artificial intelligence, by writing code on demand, would soon make software engineers redundant. New data covered by TechCrunch suggests the opposite has happened, with engineering roles emerging as among the most resilient rather than the most threatened.

The forecast was intuitive on its surface. If a model can generate working code from a plain-language description, the reasoning went, then the people who write code by hand should be the first to be replaced. That logic drove headlines, investment theses and a good deal of anxiety among early-career developers.

The reality has proven more complicated. Rather than eliminating engineers, AI coding tools have mostly changed how they work, automating routine boilerplate and accelerating tasks that once consumed hours. The result, the data suggests, is engineers who can do more, not fewer engineers overall.

Part of the explanation lies in what coding actually involves. Writing the lines of code is only one piece of software engineering; understanding requirements, designing systems, debugging subtle failures, and judging trade-offs all demand context and accountability that current AI tools do not supply on their own. Those higher-order skills have, if anything, become more valuable.

There is also a phenomenon economists call induced demand. When a task becomes cheaper and faster, organisations often respond by doing more of it rather than less. Cheaper software development can mean more software gets built, expanding rather than shrinking the total amount of engineering work to be done.

The data does carry nuance. The experience of junior developers, whose early tasks overlap most with what AI handles well, may differ from that of seniors, and some routine roles could see pressure. But the broad pattern points to augmentation, with engineers using AI as a tool, rather than wholesale replacement.

The finding fits a longer historical record. Earlier waves of automation, from compilers to high-level languages to cloud computing, were each expected to reduce the need for programmers, and each instead expanded the field by making software cheaper and more capable. AI appears, so far, to be following that arc.

None of this means the work is static. Engineers are increasingly expected to be fluent with AI tools, to review and verify machine-generated code, and to focus their attention on the parts of the job that demand human judgement. The skill set is shifting even as the headcount holds up.

There are caveats worth keeping in mind. Labor data can lag, hiring cycles fluctuate with the broader economy, and a single snapshot does not settle a question this large. The resilience observed so far could still change if the capability of the tools advances faster than the demand for new software.

Still, the TechCrunch report suggests the simple story of AI erasing engineering jobs has not matched the evidence. Instead, the early data describes a profession adapting to a powerful new tool, with the work changing shape while remaining, for now, stubbornly in demand.