Skip to main content
Skip to content
Menu

AI Chat Clients Intermediate 4 parts

Create a decision-log of moments when you almost over-relied on AI recommendations

You'll have a 5-10 entry log identifying patterns of when you were tempted to accept AI guidance without pushback, with notes on what independent verification revealed.

What you'll make

AI Over-Reliance Decision Log

  1. Decision Moment: context and AI recommendation
  2. Near-Acceptance: why it felt convincing
  3. Independent Check: what verification revealed
  4. Pattern Recognition: recurring bias types

Upload a screenshot of your completed log showing decision moments, AI recommendations, and what independent verification revealed

The 4 parts

Each part ends with a check you can see.

  1. Test AI's Consensus Pull

    Compare how different AI models respond to the same controversial question from your field.

    Done when

    You have tested one work-relevant controversial question across 3+ AI models and documented where they converge vs diverge

  2. Document Near-Acceptance Moments

    Log real moments over one week when you almost accepted AI recommendations without verification.

    Done when

    You have 3-5 documented moments when AI recommendations felt convincing enough that you almost acted on them without checking

  3. Run Independent Verification

    Take your near-acceptance moments and verify each recommendation against non-AI sources.

    Done when

    You have verified 3+ AI recommendations through independent sources and documented what verification revealed

  4. Identify Your Bias Patterns

    Analyze your decision log to find systematic patterns in when you're most vulnerable to over-reliance.

    Done when

    You can name 2+ specific conditions when you're most likely to accept AI recommendations without verification, with evidence from your log

    Capstone: you put together your AI Over-Reliance Decision Log.

About this lesson

What you'll be able to do

  • Identify 5+ moments when you almost accepted AI recommendations without verification
  • Test AI responses against your own judgment to spot systematic bias
  • Maintain your own perspective when AI outputs feel authoritative
  • Create a reusable process for catching over-reliance before it happens

Who it's for

  • Professionals using AI for decisions or analysis who want to avoid groupthink
  • People who notice AI answers feel convincing but wonder what they're missing
  • Anyone building AI into their workflow who needs to stay independently critical

What it won't cover

  • How to prompt AI better
  • Technical details of model training
  • Political analysis of AI company policies
  • Building custom AI models

Skills you'll practise

  • Critical evaluation of AI recommendations
  • Bias detection through comparative testing
  • Independent verification methods
  • Decision-making under AI influence