Envisioning Card

Summary: Envisioning cards are created to assist the development team in incorporating human values into the design processes of AI systems.

Type of pattern: Process pattern

Type of objective: Trustworthiness

Target users: Product managers

Impacted stakeholders: Developers, data scientists

Lifecycle stages: Design

Relevant AI ethics principles: Human, societal and environmental wellbeing, human-centered values, fairness, privacy protection and security, reliability and safety, transparency and explainability, contestability, accountability

Mapping to AI regulations/standards: EU AI Act, ISO/IEC 42001:2023 Standard.

Context: Responsible AI refers to the development and use of AI systems that adhere to ethical principles, such as Australian AI Ethics Framework. One of these principles is human-centered values. However, developers often find it challenging to implement these high-level principles in practice due to a lack of technical guidance and tools. The ethical principles do not provide concrete guidance and software engineering methods on how to design an AI system that aligns with these principles.

Problem: What are ways to ensure that human-centered values are incorporated in the design of AI systems?

Solution: Envisioning cards are designed to help the development team operationalize human values during the design processes for AI systems. The cards are based on four envisioning criteria, including stakeholder, time, value, and pervasiveness. The stakeholder criterion helps the team consider the effects of the AI system on both direct and indirect holders. The time criterion emphasizes the long-term implication of the AI system on humans, society, and environment. The value criterion guides the team to consider the impact of the AI system on human values. The pervasiveness criterion addresses the challenges encountered if an AI system is widely adopted in terms of geography, culture, demographics, and so on.

Benefits:

  • Low cost: The use of envisioning cards is at a relatively low cost, in terms of both money and time.
  • Communication: Envisioning cards provide a visual representation of ideas; the cards also make these ideas easy to understand and share with others.

Drawbacks:

  • Limited scalability: Scaling the use of envisioning cards can be challenging when working with large teams or complex AI systems.
  • Potential for biases: Envisioning cards can be influenced by the biases of the individuals using them.

Related patterns: N/A

Known uses:

  • To increase awareness of impacts on human values, the VSD Lab created envisioning cards that can be used in the design stage based on five envisioning criteria: stakeholders, time, values, pervasiveness, multi-lifespan.
  • Nathan et. al. propose four envisioning criteria – stakeholders, time, values, and pervasiveness and characterize how the criteria can be considered in interactive system design.