Introduction
AWS CEO Matt Garman, speaking with Matthew Berman, called the idea of firing junior developers because "AI can do their jobs" "the dumbest thing I've ever heard" and argued AI should augment training and productivity instead.
Quick definition
The core point: replacing junior developers with AI risks depleting future human expertise and learning capacity.
Context
The exchange referenced Kiro, AWS's AI-assisted coding tool, and criticized simplistic metrics such as the share of lines of code produced by AI as misleading measures of value.
The Problem / Challenge
Garman says the argument to cut junior hires overlooks that juniors are often low-cost and early adopters of AI; removing them sacrifices people who learn to decompose problems and build long-term technical judgment.
"The dumbest thing I've ever heard."
Matt Garman, CEO / AWS
Solution / Approach
His recommendation is to keep hiring college graduates and use AI as an educational aid: writing unit tests, improving documentation, and enabling agentic workflows that pair developers with AI assistants.
- Use AI as on-the-job mentoring for juniors
- Measure code quality and maintainability, not volume
- Emphasize meta-skills: critical thinking, problem decomposition, continuous learning
Why code-volume metrics fail
Garman notes that AI can produce many lines of code, but more lines often mean more complexity; better outcomes come from fewer, clearer, well-tested lines.
Implications for organizations
Firms should avoid headcount cuts justified solely by automation, and instead combine junior hiring with AI tools to accelerate skill development and sustain future capabilities.
Conclusion
The practical takeaway: deploy AI to train and augment junior developers rather than to replace them, preserving long-term talent pipelines and adaptive skills.
FAQ
Short answer: AWS CEO Matt Garman warns that replacing junior developers with AI undermines long-term talent development and should be avoided.
-
Why did the AWS CEO call firing juniors "the dumbest thing"?
Because juniors are cost-effective, adopt AI quickly, and their removal would erode future technical learning and capacity.
-
How can AI help junior developers according to the AWS CEO?
AI can assist by generating unit tests, improving documentation, and supporting workflows that teach through collaboration.
-
Is measuring AI impact by lines of code useful?
No; Garman warns that counting lines favors quantity over quality and can reward bad code.
-
What skills should juniors focus on in the AI era?
Critical reasoning, problem decomposition, creativity, and a learning mindset to adapt to new technologies.