AI Philosophy

How I use AI as a development tool.

I use AI to speed up parts of development, but I treat every output as draft work that still needs engineering review.

Where AI Does The Heavy Lifting

  • Researching unfamiliar APIs, libraries, and implementation patterns.
  • Drafting boilerplate, test scaffolding, documentation, and examples.
  • Comparing refactoring options before final implementation decisions.
  • Handling repetitive tasks so I can spend more time on higher-risk work.

Where I Step In

  • Security-sensitive code, authentication, permissions, and data handling.
  • Business rules that depend on real stakeholder context and constraints.
  • Architecture decisions with long-term maintenance costs.
  • Ensuring compliance and regulatory requirements such as GDPR, HIPAA, and others.

My review standard

AI-generated code gets the same scrutiny as code written by any developer on the team. It is, effectively, advanced code-completion.

Correctness

Does it actually solve the requirement, handle edge cases, and behave correctly under realistic conditions?

Maintainability

Will another developer understand it later? Does it fit the project’s existing patterns and constraints?

Risk

Does it introduce security, performance, licensing, privacy, or operational problems?

The principle

AI is a force multiplier for capable developers. It accelerates research, drafting, and iteration, but it cannot replace accountability. I use it to move faster on research and drafting, then I own the final architecture, code quality, and production outcome.