Lesson Content

Three modes of AI relationship.

Collaboration: You and AI work together on a task, with your expertise and judgment actively shaping the output at multiple points. AI contributes generation, synthesis, or structure; you contribute domain knowledge, judgment, and quality control. Neither could produce the result alone – or the human alone would produce a lesser result.

Delegation: You hand a task to AI and review the output. Your role is assignment and review, not active participation in the work. Delegation is appropriate when the task is well-defined, the scope is bounded, and the output can be fully evaluated before use.

Exclusion: You do the task without AI involvement. Some tasks require complete human judgment, carry professional or ethical obligations that AI cannot bear, or would be diminished by AI involvement in ways that matter.

Choosing the right mode – a decision framework.

The right AI mode depends on three factors:

Stake and consequence: High-stakes decisions involving real-world consequences lean toward collaboration or exclusion. Low-stakes, easily-reversible tasks can be delegated.

Uniqueness requirement: Tasks requiring your specific expertise, voice, relationship, or professional judgment lean toward collaboration or exclusion. Generic tasks that could be done by anyone with the right information can be delegated.

AI capability match: Tasks that fall squarely in AI's capability strengths (synthesis, generation, analysis of text) can be collaborated on or delegated. Tasks that require real-time data, physical judgment, interpersonal skill, or domain expertise AI does not have should exclude AI or use it minimally.

The risk of reflexive delegation.

The biggest AI fluency failure mode is reflexive delegation: assigning tasks to AI without evaluating whether delegation, collaboration, or exclusion is actually the right mode. Signs you may be over-delegating:

• AI is making judgment calls you should be making
• You are using AI output without substantive review
• The quality or voice of your work has drifted from your professional standard
• Decisions with real consequences are being driven by AI outputs you have not verified