-
Creating test suites for every stage of your ML workflow
Testing is a crucial part of machine learning (ML) workflows. A robust test suite ensures that your ML pipeline works as expected at every stage, from data collection to model deployment. By building test suites at each stage of your ML process, you can quickly detect issues, improve the reliability of your systems, and save
-
Creating time-aware AI systems that slow decision making
Time-aware AI systems are designed to incorporate the dimension of time into their decision-making processes. Unlike conventional AI systems that operate in real-time, time-aware AI systems slow down decision-making to allow for more thoughtful, reflective, and deliberate choices. This concept is crucial in contexts where speed could lead to suboptimal decisions, and instead, more measured,
-
Creating tools that empower users to override algorithmic bias
To develop tools that empower users to override algorithmic bias, the focus must be on transparency, user control, and fostering an understanding of how algorithms work. Here’s a breakdown of how these tools can be built: 1. Transparency and Awareness Empowering users starts with transparency. They need to be able to see what data is
-
Creating user flows that reward emotional intelligence
Creating user flows that reward emotional intelligence involves designing interactions within a digital interface that acknowledge, enhance, and guide users through emotionally intelligent responses. The idea is to create spaces where users not only solve problems but are also encouraged to engage with content in ways that promote emotional understanding and regulation. Here’s how you
-
Creating user research protocols tailored to AI products
Creating user research protocols for AI products requires a thoughtful approach, as AI systems often involve complex, unseen processes that users interact with in a very different way compared to traditional products. Below are the essential components of a research protocol specifically tailored for AI-driven products. 1. Define Research Goals Before anything, the research goals
-
Creating user-controlled override tools in ML-powered products
In machine learning-powered products, creating user-controlled override tools offers a powerful way for users to intervene in the model’s decision-making process, giving them greater control over the system’s outputs. These tools are particularly useful in industries where outcomes are high-stakes, like finance, healthcare, and autonomous systems. Here’s how to approach building user-controlled override tools in
-
Creating systems that help AI and humans learn empathy together
Empathy has long been recognized as a cornerstone of human interaction, but in the context of artificial intelligence (AI), it presents both challenges and opportunities. As AI continues to develop, the ability to create systems that help both humans and machines learn empathy together can significantly improve user experience, collaboration, and emotional intelligence in digital
-
Creating systems that value transparency over persuasion
When designing systems—especially those involving AI, data, or decision-making processes—the balance between transparency and persuasion is critical. Systems that prioritize transparency over persuasion allow users to make informed decisions based on clear, unbiased information. In contrast, persuasive systems often aim to influence or direct user behavior, sometimes in subtle or overt ways. Here’s how you
-
Creating team dashboards for live model quality monitoring
Creating team dashboards for live model quality monitoring involves several key steps to ensure that your models’ performance is actively tracked, with actionable insights available for your team. A well-structured dashboard should provide a real-time overview of your model’s health, allowing teams to quickly detect issues and respond effectively. 1. Identify Key Metrics Start by
-
Creating space for ethical pause in AI system flows
Designing AI systems that allow space for ethical pause is crucial for fostering responsible decision-making and reducing harm. As AI systems become more integrated into daily life, it’s essential to build deliberate moments where users or even the system itself can reflect on the ethical implications of actions or choices. Here’s how to approach this: