Why AI hasn’t replaced software engineers, and won’t
If you are a developer wondering whether your role is next on the chopping block, the answer is no. Arvind Narayanan and Sayash Kappor have tackled the question of AI-induced job losses through the lens of a profession uniquely exposed to such disruption: software engineering.
We argue that the evidence is sufficient to reject the narrative that once AI capabilities hit a certain threshold, it will cause mass layoffs. Given that this is true even in a sector with very few regulatory barriers, most other professions are likely to be even more cushioned.
The first reassuring fact is that the data simply does not support the idea that AI is driving mass unemployment.
In March 2025, New York became the first U.S. state to add an AI disclosure checkbox to WARN Act filings. During the full first year, more than 160 companies filed WARN notices. Not a single one checked the AI box.
While AI accelerates the phase of typing code into a computer, software engineering is about a great deal more than that:
If writing code isn’t the bottleneck, what is? Surveys on task breakdown point to things like meetings or debugging. This just leads to more questions: what are developers actually doing in those meetings, and why can’t it be done by AI? Won’t debugging get automated as capabilities improve? To understand the real bottlenecks, we have to get qualitative, and dig into software engineers’ own understanding of what it is they do that resists automation.
When we did this analysis, it revealed three things as the real bottlenecks: (1) deciding and specifying what to build, (2) verifying and being accountable for what is delivered, and (3) the deep human understanding, of the codebase, the business, and the environment, required to carry out both of these.
I find AI assistance also helps me with the deciding and verifying steps, but it is the “deep human understanding” that remains key to the value I provide. Give me all of the AI assistance in the world and the value I produce will still be reliant on how deeply I understand both the problems and the solutions that the agents are building for them.
Key takeaways
- Current data shows zero evidence of AI-related mass layoffs, highlighted by New York’s 2025 WARN Act filings where not a single company flagged AI as a reason for redundancy.
- The core bottleneck for engineers is not typing code, but rather the strategic decision-making, accountability, and deep contextual understanding of the business environment.
- AI acts as a tool that accelerates execution, but the human capacity to specify problems and verify solutions remains the primary source of professional value.
