
Soft Skills Engineering Episode 506: I hate my job with AI and my team-mate thinks I suck
21 snips
Mar 30, 2026 They tackle mourning the loss of hands-on coding as AI shifts work toward reviews, maintenance, and higher-leverage tasks. They explore ways to find new satisfaction by shaping standards, architecture, and product outcomes. They also dig into dealing with a perfectionist teammate who undermines leadership and how candid conversations and exposing people to broader responsibilities can help.
AI Snips
Chapters
Transcript
Episode notes
Encode Team Standards Into Rules
- Encode standards into deterministic rules like linting so AI and juniors produce fewer errors.
- Dave recommends investing time in meta work (lint rules, processes) because it improves AI outputs and reduces review time.
Treat Reviews As The Next Engineering Puzzle
- Treat code review and QA as solvable bottlenecks by redesigning process and tooling.
- Jameson suggests both technical and organizational fixes to remove the new review bottleneck created by AI-generated code.
Big System Problems Still Need Humans
- Hard system problems remain uniquely human: fault tolerance, security, scalability, and UX.
- Dave says these architecture and product problems resist simple prompt-based fixes and reward deep thinking.
