
Record a fact-check demo with your AI
You'll have a 3-5 minute video showing how you prompt an AI to catch its own errors on a task relevant to your work (math, dates, claims, etc.), then narrate what you learned about when to trust the o
What you'll make
AI Fact-Check Demo Video
- Open: the task and why fact-checking matters
- Demo: your testing method in real time
- Catch: where AI got it wrong
- Method: the error pattern you found
- Close: when to trust vs verify
Upload your screen recording with narration showing the full workflow from prompt to error discovery
The 3 parts
Each part ends with a check you can see.
Map Your Error Zones
Identify the specific types of factual claims AI gets wrong in your work domain
Done when
You have 5-10 specific examples where AI gave you wrong information, grouped by error type (dates, numbers, citations, procedures)
Design Your Testing Method
Create a repeatable process for catching AI errors before they matter
Done when
You have a step-by-step method that reliably exposes AI errors in your work domain, tested on at least 3 different prompts
Record Your Live Demo
Capture yourself using your testing method to catch AI errors in real time
Done when
You have a 3-5 minute screen recording showing your full fact-checking process, including at least one moment where you catch AI making an error
Capstone: you put together your AI Fact-Check Demo Video.
About this lesson
What you'll be able to do
- Record a working 3-5 minute fact-check demo showing how you catch AI errors on your own work tasks
- Develop a repeatable method for testing when to trust vs verify AI output
- Identify 3-5 specific error patterns where AI consistently fails in your domain
Who it's for
- Anyone who uses AI for research, analysis, or content creation and wants to catch errors before they matter
- People comfortable recording short videos and narrating their screen
What it won't cover
- Advanced prompt engineering techniques
- Comparing different AI models or platforms
- Setting up complex verification workflows
Skills you'll practise
- Error pattern recognition
- Blind testing methodology
- Trust calibration with AI output
- Screencast demonstration