AI coding assistants have become a common part of modern software development, helping with code generation, debugging, testing, documentation, and productivity. However, many developers still prefer to handle certain tasks manually to maintain code quality, deepen their understanding, or retain control over critical decisions.

Some examples include:

• System architecture and design decisions
• Learning new frameworks or technologies
• Debugging complex issues
• Security and code reviews
• Performance optimization
• Writing core business logic

Which programming tasks do you intentionally avoid using AI for, and why? Has your approach changed as AI tools have become more capable?

This is a great opportunity to discuss where AI adds the most value and where human judgment still matters most.