The Indicator from Planet Money We read your mail on AI-proof jobs and how to fix crime labs
23 snips
Sep 11, 2025 Listener feedback sparks a lively discussion on how seasoned software developers are adapting to the challenges posed by AI, highlighting their vital problem-solving skills. The conversation shifts to the alarming state of U.S. crime labs, where inadequate accreditation standards threaten forensic integrity. The hosts delve into innovative solutions for improving quality assurance, advocating for universal accreditation and national standards that could revolutionize forensic science practices. Tune in for insights on job security and the future of forensic evidence!
AI Snips
Chapters
Transcript
Episode notes
Experience Makes Engineers AI-Resistant
- Older software engineers are more valuable because they can diagnose and fix faulty AI-generated code.
- Experience lets them rescue systems that less-experienced "vibe" coders and AI tools produce.
Combine Models With Deep Domain Knowledge
- Using large language models increases a senior engineer's productivity when coupled with deep understanding.
- Senior engineers add value by precisely prompting tools and repairing their mistakes.
Risk Of Treating Software As Disposable
- Widespread reliance on throwaway AI-generated software could degrade quality and security.
- That shift might create a lot of prevalent but unstable and insecure software across the economy.