AI has lowered the barrier to building software, but it has not removed the need for engineers. The code itself is becoming cheaper to produce, forcing a bigger question into focus: who decides what should be built, how it fits, and whether it creates a useful result?
That shift is already visible in hiring data and developer workflows. AI agents now appear in one of every three pull requests on GitHub, compared with fewer than one in 10 a year ago, while software development postings have risen even as overall job postings fell.
The Code Is Cheaper, But Software Is Bigger Than Code
Most conversations about AI-assisted software development focus on productivity. That focus makes sense because AI is lowering the cost of producing code, helping more people move from an idea toward a working implementation.
Yet code is only one input into successful software. Someone still has to decide what to build, understand how it fits into the existing environment, verify it, secure it, and determine whether it delivers the intended business result. Lowering the barrier to building does not erase those responsibilities; it makes them more important to the final outcome.
As implementation gets cheaper, the value of engineering shifts away from raw labor input and toward outcomes. Those outcomes include increasing revenue, lowering cost, reducing risk, and improving reliability. An engineer who connects technical decisions to those results brings value that code generation alone cannot provide.
That change also reshapes the skills needed for programming. Being able to use AI in the work and improve efficiency is a strong starting point, but communication, interpersonal skills, and leadership also matter. The work moves toward judgment, coordination, and oversight as more implementation flows through automated systems.
Hiring Data Points Toward Senior Engineers and AI Roles
U.S. software development job postings rose almost 15% from late February 2025, the launch date for Anthropic’s Claude Code, while overall job postings fell 7%. That contrast complicates predictions that AI would rapidly erase demand for developers.
Data from Indeed Hiring Lab shows that U.S. software-development postings have risen nearly 15% since February 2025. The rebound is concentrated in experienced AI-oriented positions, with senior roles accounting for 71% of the increase between May 2025 and May 2026 and job titles mentioning AI accounting for 37%.
The broader picture still carries a warning. Software-development job postings remain about 27.5% below their pre-pandemic level, and the increase is not spread evenly across the profession. The strongest growth is appearing in senior-level roles and positions tied to AI.
That pattern supports the view from labor market researchers that the profession is evolving rather than disappearing. The market is placing a premium on people who can guide AI-assisted work, connect systems to existing environments, and judge whether the result meets its purpose.
AI Agents Are Moving Into the Main Development Flow
GitHub’s pull-request data shows how fast the workflow is changing. One in three pull requests now involves an AI agent, compared with fewer than one in 10 a year ago. If that pace holds, within the next two years, most of the code pushed to GitHub could be written by an agent.
GitHub leaders have said that more than 80% of code could eventually be written by AI. That future would not remove humans from the process, because people are still needed to oversee automation and provide user input.
The most striking part of that forecast is not only the amount of machine-written code. It is the possibility that much of it may never be fully read by a human. That makes decisions about verification, security, reliability, and fit within the existing environment central parts of engineering rather than optional checks at the end.
The pressure is already visible among younger workers. Stanford University research found that employment in AI-exposed fields for people aged 22 to 25 has fallen 16% since the rise of generative AI, while early-career software developers saw a larger employment decline of 20%.
At the same time, the long-term outlook for software developers remains positive. The U.S. Bureau of Labor Statistics projects employment to grow 10% from 2025 to 2035, rising from about 1.72 million to 1.89 million jobs.
The Next Engineering Advantage Is Measured in Outcomes
These numbers point to a profession under pressure, not a profession vanishing. AI can generate more of the implementation, but engineering still requires people to choose the right problem, manage risk, validate the result, and connect the finished system to a business goal.
That is why the future belongs to engineers who can work across both sides of the change: using AI agents to produce software while bringing the communication, interpersonal skills, leadership, and technical judgment needed to make that software matter.
The barrier to building is falling. The bar for building something valuable is not.
Based on
- AI has lowered the barrier to building. Here’s what that means for engineers. — thenewstack.io
- AI Is Changing The Economics Of Software Engineering — forbes.com
- AI taking job of software developer? What labor market data shows — cnbc.com
- In the AI era, coding means thinking like a systems architect | Nature — nature.com
