Explanation: A UX Designer should take the following steps to create and deliver responsible and transparent AI technology12:
- C. Document model cards to clarify intended context and use cases. Model cards are standardized documents that provide essential information about a machine learning model, such as its purpose, performance, limitations, and ethical considerations3. Model cards can help UX Designers to communicate the design choices and trade-offs of their AI systems, and to ensure that they are aligned with the user needs and expectations4.
- D. Provide clear explanations of AI predictions or recommendations. Explanations are user-facing descriptions of how and why an AI system produces a certain output, such as a prediction, a recommendation, or a decision5. Explanations can help UX Designers to increase the transparency and trustworthiness of their AI systems, and to empower users to understand, control, and evaluate the AI outcomes6.
References: Salesforce Debuts AI Ethics Model: How Ethical Practices Further Responsible Artificial Intelligence, Generative AI: 5 Guidelines for Responsible Development - Salesforce, Model Cards for Model Reporting, Model Cards: A Framework for Increasing Trust in AI Systems, Explainable AI: A Guide for Making Black Box Machine Learning Models Explainable, Salesforce Supports AI Regulation Advancing Digital Trust and Innovation - Salesforce
C. Document model cards to clarify intended context and use cases.
Model cards are documents that describe the intended use, performance, and limitations of AI models. They help ensure that the AI technology is being used responsibly and transparently, as they provide clear information about the model's context, data, and assumptions. This can help reduce the risk of unintended consequences and build trust with users.
D. Provide clear explanations of AI predictions or recommendations.
Clear explanations of AI predictions or recommendations help build trust with users and increase understanding of how the AI technology works. By providing an understandable explanation of how a prediction or recommendation was made, users can gain a better understanding of the technology and how it is intended to be used. This can also help reduce the risk of unintended consequences and improve accountability.