Intent: the New Critical Success Factor in Creating AI

Share this post:

AI is a transformative and exciting technology that will no doubt reshape the way we live and work for decades to come. That said, we need to be extremely thoughtful and intentional about how we design these solutions.

Those who design, develop, and deploy AI are responsible for ensuring that every solution they create has a clear intent and provides consistent value to users, while accounting for the ethical implications of its existence.

For the past three years, my team and I have been researching and creating a formal approach to designing AI solutions that are intent-focused rather than tech-led. When it comes to the creation process, defining and aligning on the intent of the AI is the single biggest key to success. A team needs to clearly state and understand the purpose behind what it’s planning in order to responsibly deliver user and client value. This begins with shifting the team’s focus from technical capabilities and limitations to the human perspective.

In the early days of our effort, we worked with teams from across IBM to understand their processes and approaches to AI. We found that while most teams had one or two AI experts, the other team members often lacked the domain knowledge necessary to be creative about how AI might be applied to their work.

There was a pressing need to gear up AI designers and practitioners for a new way of working — specifically, applying a human-centric approach to the development of an AI solution while also accounting for technical details like data preparation and training. We saw this as an opportunity to build deeper AI domain knowledge for our designers, while also helping our AI experts frame their efforts through a more user-focused lens.

Ultimately, we needed a way to deliver quality outcomes consistently at scale. We had deep domain expertise in AI but needed a reliable framework to turn that expertise into exceptional experiences.

The Essentials

To address this critical need, we have created Team Essentials for AI, a nine-lesson course for teams eager to design and build AI applications that emphasize user needs over technology. It’s integrated into Enterprise Design Thinking, IBM’s framework for collaborating, aligning teams, and solving users’ problems. The new Team Essentials for AI expands on these ideas and applies them directly to the AI space.

This holistic framework provides a repeatable workflow for people designing impactful, responsible, and human-centric AI. It gives designers, developers, and everyone in between a common language and framework for thinking through the development of an AI solution. When teams apply these scalable methods, they’re able to move faster and deliver differentiated outcomes over time.

Participants in the course will learn our AI Essentials Framework: a set of activities, tools, and principles that enable teams to:

• create thoughtful, human-centered AI solutions,
• work through a fictional case study to see how a team could use this framework,
• and collect resources that they can apply on their team immediately.

We also had deep concerns about the ethical implications of the AI solutions being created across the industry and wanted to embed ethical responsibility and accountability directly into the design and development process.

To put it simply, we saw a business need – but also a societal responsibility — to create tools and resources that help people be more creative and practical in their approach to data and AI, while also helping them identify and account for the ethical implications of the technology to ensure that the solutions we create are aligned with the values of the people who use them.

We are on the cusp of a phase shift in how we interact with machines. The more that machines are able to simulate and mimic human behaviors and characteristics, the more it’s on us, as the designers of these systems, to build systems that are emotionally and psychologically responsible.

To dig deeper into how we’ve been thinking about designing for AI, stop by our companion site to read about AI ethics, designing for conversation, and human/machine relationships.