Before proceeding with any AI project, take a step back to evaluate its broader implications by creating a scorecard.
Assess whether it ultimately benefits society or causes harm.
Document the following:
- Benefits: Outline the positive outcomes and value it could bring to individuals, communities, or the world at large.
- Risks: Identify potential direct negative effects, such as privacy invasions or discriminatory impacts.
- Unintended consequences: Consider indirect or unforeseen issues that might arise, even from well intentioned features.
Also, think through worst case scenarios: What could happen if things go wrong, and how severe might the fallout be?
Stay mindful of these factors throughout development.
If, after weighing everything, the project appears to cause more harm than good on balance—pivot to a different approach or redesign it.