People + AI Guidebook by Google - useful for anyone wanting to build AI products in a more human-centred way.
User Needs & Defining SuccessData Collection & EvaluationMental ModelsExplainability & TrustFeedback & ControlErrors & Graceful Failure
Highlights some interesting questions and considerations:
Which user problems is AI uniquely positioned to solve?
Does our training dataset have the features and breadth to ensure our AI meets our users’ needs?
How can we ensure that raters aren’t injecting error or bias into datasets when generating labels?
Which aspects of AI should we explain to our users?
What are the pros and cons of introducing our AI as human-like?
How should we show users the confidence associated with an AI prediction?
How will we reliably identify sources of error in a complex AI?