AI pair programming
AI pair programming is working alongside an AI assistant the way you would with a human pair: it suggests, you steer, and you review continuously rather than delegating a task and checking back later. The reviewing is the part that makes it pairing rather than outsourcing.
The analogy is imperfect in a useful way. A human pair has judgement about the project, remembers yesterday's decision, and pushes back on a bad idea. A model has none of that unless the context carries it, which is why rules files and attached context do the work a colleague's memory would.
Where the analogy holds is the rhythm: a tight loop of suggestion, review, and correction, with attention on both sides.
What this means in practice
The mode fits inline completion and scoped edits, where you see everything as it happens. It breaks down with autonomous agents on large tasks, which are closer to delegation — and delegation needs different habits, mainly explicit verification, because nobody was watching.
Prompts for the tools this applies to
Related terms
- AI code completion
AI code completion predicts and suggests the code you are about to write, usually as inline grey text you accept with Tab.
- Agentic coding
Agentic coding is using an AI tool that plans and executes a whole task in your codebase rather than completing one line at a time.
- AI code review
AI code review is using a model to examine a diff or a file for bugs, security problems, and quality issues before a human reviews it.