The AI Experiments Canvas is a structured framework that helps you explore artificial intelligence safely and methodically. It provides clear boundaries and guidelines for testing AI applications while considering potential impacts on both your work and beneficiaries.
What is it good for?
- Provides a systematic approach to AI experimentation
- Helps evaluate and mitigate potential risks
- Enables data-led testing of AI applications
When to use it
Get the tool
Here’s a workbook you can work on in Google Drive or download
Alternatively, you can use this Miro version, where you can start to add post-it notes either on your own or as part of a team workshop or exercise.
How to use it
While AI offers powerful capabilities for improving efficiency, many aspects of its functioning and impact remain uncertain. This canvas helps you navigate these uncertainties by providing a structured approach to experimentation.
The framework will help you to think through:
- The outline of your experiment with AI
- The tools you might use
- The risks and mitigation strategies you can put in place (our favourite is ensuring there is a ‘human in the loop’ - someone that is on hand to ensure the responses an AI tool is giving, are accurate.)
- Data and privacy
- And a way to capture the results of your experiment - whether a success or a failure