Will AI Assistant make developers redundant? by Marit van Dijk

AI assists with code comprehension and maintenance, which takes up to 58-80% of developer time, rather than writing new code

Key features of JetBrains AI Assistant include:

• Code explanation and documentation generation
• Refactoring suggestions
• Test case generation
• Bug finding and fixing
• Code review assistance
• Version control integration

Research shows productivity gains:

• JetBrains found up to 8 hours saved per week
• GitHub study showed 78% task completion vs 70% without AI
• Developer experience improved through reduced cognitive load

Important limitations and considerations:

• AI models can hallucinate or generate incorrect code
• Critical to understand and verify generated code
• Models may be outdated or have knowledge cutoff dates
• Security and data privacy concerns with cloud-based solutions

Jobs will evolve rather than disappear:

• Developer demand is increasing despite AI adoption
• Skills and roles will adapt like previous tech transitions
• More focus on higher-value work as routine tasks are automated
• Still need humans for critical thinking and complex problem-solving

Best practices for AI usage:

• Verify and understand generated code
• Use AI as an assistant, not a replacement
• Be aware of automation bias
• Maintain good development practices
• Focus on leveraging AI for comprehension and maintenance tasks

AI excels at:

• Language processing and code explanation
• Documentation generation
• Pattern recognition
• Summarizing information
• Automating repetitive tasks

AI struggles with:

• Mathematical operations
• Randomness
• Domain-specific knowledge
• Complex business logic
• Learning from corrections