Full Breakdown
AI Reshapes Software Engineering: Skills, Jobs, and Collective Action
7/12/2026, 8:48:34 PM
Core Shift in Software Engineering Roles
In 2022, OpenAI released ChatGPT, and within a few years AI-generated code has become routine. Companies such as Google now report that roughly three-quarters of their code is written by AI tools. For many engineers, the daily work that once centered on writing, designing, and problem-solving has moved toward reviewing code produced by machines. Matt, a software engineer who commutes to Pawling, New York, says his responsibilities have “shifted away from coding, problem solving and software architecture towards reviewing code generated by artificial intelligence.” The change has prompted engineers to protect their “axe” by avoiding AI where possible and to consider new career paths.
Background & Context
Software engineering was a flagship, high-paying U.S. profession in 2022, with 1.5 million practitioners earning about twice the national median salary, according to the U.S. Bureau of Labor Statistics. The field expanded after President Barack Obama’s 2013 “Computer Science for All” initiative and a surge in coding bootcamps, which produced millions of new developers worldwide. The rapid adoption of generative AI, however, has accelerated layoffs: more than 600,000 U.S. tech workers lost jobs after ChatGPT’s debut, and the unemployment rate for computer-science graduates rose to 7 % in 2024.
Data & Statistics
- 75 % of Google’s code now written by AI (company statement).
- 600,000+ U.S. tech workers laid off since 2022 (Layoff.fyi).
- Unemployment for CS graduates: 7 % (2024) vs. 6 % (2023).
- Underemployment rate for tech workers: >19 % (New York Fed).
- Indeed tech job postings fell 36 % from 2020 to 2025.
- University enrollment in computer-science majors dropped 8.1 % and graduate enrollment fell 14 % in 2025-2026 (National Student Clearinghouse).
Key Voices
- Matt – refuses to use AI on personal projects, fearing skill erosion.
- George Dover – after a 2024 layoff, used AI to generate and audit code, eventually securing an AI-focused engineering role.
- Sam – laments that AI has taken “the creative, fun part of the job” and left him “reviewing code I didn’t write.”
- Kaitlin Cort – former engineer who founded the “What We Will” resource center to aid displaced tech workers and promote unionization.
Official Statements & Responses
Bouke Klein Teeselink, assistant professor of economics at King’s College London, says the profession’s future hinges on the ability to evaluate AI-written code rather than write code itself. Ethan Mollick, associate professor at the University of Pennsylvania’s Wharton School, adds that success now depends on “defining problems, designing systems and directing AI tools effectively.” Harvard professor David Malan warns that the high cost of AI models will keep human engineers essential for code review and security. Brown University’s Shriram Krishnamurthi notes that while AI-generated code quality improved only last year, the growing demand for validation will “weed out” engineers who have not re-tooled.
Criticism & Opposition
Many engineers describe a “dark” outlook, fearing layoffs to AI-augmented talent at firms like Google, Amazon, and Netflix. The shift has spurred collective action: Cort’s organization now receives “at least 10 new applications each day” from workers seeking union support, and groups such as the Alphabet Workers Union are being consulted for guidance.
Verbatim Quotes
- “I am actively trying to keep my axe sharp,” — Matt, software engineer
- “It’s hard to say what exactly the profession will look like in two years, but it’s clear that the skill of writing code is over,” — Bouke Klein Teeselink, assistant professor, King’s College London
- “Now it’s not about who can write the most code,” — Ethan Mollick, associate professor, University of Pennsylvania
- “He hates that AI has taken over “the creative, fun part of the job”, and reduced it down to the worst part: “reviewing code I didn’t write”, said Sam.” — Sam, software engineer
- “There’s a lot of need right now,” — Kaitlin Cort, founder, What We Will
Why It Matters
The AI transition reshapes labor demand, emphasizing code-review, security, and system design over raw programming. Companies face rising AI infrastructure costs—OpenAI reportedly spent $8 billion and Anthropic $3 billion last year—costs that may eventually be passed to customers. Simultaneously, the profession’s cultural identity is in flux, prompting new forms of worker organization and calls for industry standards.
What’s Next
Cort’s “What We Will” continues to expand support services, partnering with existing tech unions to develop collective-bargaining frameworks as AI adoption accelerates.
