Microsoft launches HAX human-AI toolkit
This new toolkit provides a set of practical tools to help teams strategically create and responsibly implement best practices when creating AI technologies. Consisting of four components, the HAX toolkit is designed to help AI creators take a human-centred approach in their day-to-day work.
Microsoft HAX: helping AI creators
Discussing the toolkit, Senior Principal Research Manager at Microsoft Research, Saleema Amershi said: “Human-centeredness is really all about ensuring that what we build and how we build it begins and ends with people in mind. We started the HAX Toolkit to help AI creators take this approach when building AI technologies.”
Coming at a time where AI-infused services are becoming increasingly popular, Microsoft’s HAX toolkit provides support to companies wanting to iron out any inconsistencies and errors in their AI systems.
Microsoft HAX: responsible guidance
Although AI technology is useful to its users, there can be issues with understanding voice commands and/or interpreting pictures correctly. These inconsistencies within AI technology have resulted in AI developers being pushed for responsible guidance.
The new HAX toolkit provides developers with practical tools that translate human-AI interaction knowledge into accountable guidance.
Designed to assist a team throughout the user design process from planning to testing, the four components of the HAX toolkit are:
- The Guidelines for Human-AI Interaction, which provides best practices for how AI applications should interact with people.
- The HAX Workbook helps teams prioritize guidelines and plan the time and resources needed to address high-priority items.
- The HAX Design Patterns offer flexible solutions for addressing common problems that come up when designing human-AI systems. With it comes the HAX Design Library which is a searchable database of the design patterns and implementation examples.
- Teams can also utilize the HAX Playbook to identify and plan for unforeseen errors, such as a transcription error or false positive.
Guidelines for success
Allowing for human collaboration to build better AI systems, the HAX toolkit allows for the creation of fluid and responsible human-AI experiences. This has evolved from a set of 18 guidelines, based on more than 20 years of research.
From its extensive research into human-AI experience, Microsoft recognised the need for companies to bring their guidelines into their workflow early, to have a real impact.
Keen to set themselves apart from other existing human-AI interaction resources, the HAX team created specific guidelines for users to follow. This is opposed to more traditional resources that lean toward tutorials.
As well as the practical tools within the HAX toolkit, the guidelines can also be utilised throughout the development lifecycle.
Divided into four groups, the guidelines are based on when they are most relevant to an interaction with an AI system:
- Initially
- During interaction
- When the AI system gets something wrong and needs to be redirected
- Over time
These guidelines allow for a flexible approach to nurture continuous improvement when curating AI systems. They can help users identify and address issues proactively and systematically rather than troubleshooting when the system encounters a failure situation.
Encouraging users to share their feedback on the toolkit, Amershi said: “We are hoping this can be a trusted resource where people can go to find tools throughout the end-to-end process. We will continue to update and create new tools as we continue to learn and work in this space.