REDMOND, Wash. — Microsoft is offering guidance to companies to develop human-centered artificial intelligence (AI) technologies.
Microsoft released the Human-AI eXperience (HAX) Toolkit or a set of tools to create and implement best practices when creating AI that interacts with people, according to a post last week by Leah Culler for Microsoft.
AI products, such as virtual assistants, recommendations and reminders, are growing in popularity.
However, AI applications for people can be unhelpful, such as misunderstanding a voice command or misinterpreting an image, as well as demonstrate “disruptive behaviors” or “even cause harm,” according to Culler.
Responsible AI resources traditionally focus on improving algorithms and models, but they also need to be made accessible to “the practitioners who design the applications people use,” Culler says.
“Human-centeredness is really all about ensuring that what we build and how we build it begins and ends with people in mind,” said Saleema Amershi, senior principal research manager, Microsoft Research.
The toolkit consists of several components from AI planning to testing:
- Guidelines for Human-AI Interaction: best practices for how AI applications should interact with people
- HAX Workbook: helps teams prioritize guidelines and plan the time and resources needed to address high-priority items
- HAX Design Patterns: offers flexible solutions for addressing common problems that come up when designing human-AI systems; HAX Design Library: searchable database of the design patterns and implementation examples
- HAX Playbook: to identify and plan for unforeseen errors, such as a transcription error or false positive