Content, AI, and the Human Touch: Applying Design Thinking to the Legal Market

Domain Expertise and Content-Driven Design

Despite the breakneck speed at which technology is advancing, adoption and implementation often tend to lag behind. There are a few reasons why this happens, but most of them boil down to a lack of domain expertise and poor usability, which are important when users are working with significant volumes of content.

There is no shortcut to developing true, deep domain expertise, but designers can facilitate a human-centric approach that informs product design and development through relevant domain knowledge. As I mentioned earlier, this can benefit from an intentional, participatory-design process that is empathetic to customers and users alike.

A second reason many technology products fail—despite their fantastic technology—is that they don’t allow users to interact with the right content in a user-friendly way. Enabling systems to identify content that is relevant to the user relies on baking domain expertise into the product. Creating a truly valuable product for legal professionals requires designers to work closely with subject-matter experts to understand and address the tangible connection between users and the body of knowledge on which they rely to succeed. It also requires a hypothesis-driven approach to product development, optimizing the user experience and human-AI interaction through continuous design and usability testing.

Similar to other knowledge workers, legal professionals have developed particular ways of parsing the vast amounts of data at their disposal. Successful professionals rely on highly skilled, experienced in-house human resources and use tried-and-true methods that have worked for decades. For any technology tool to break into the legacy industry—or others—it must surpass professionals’ existing workflows, which with they’re already comfortable. Accomplishing this requires a simple, elegant user experience that becomes possible only by having a proper understanding of the scope and type of content to which users need access and a thorough grasp of how professionals in the field access and relate to that content in doing their jobs.

Human-AI Collaboration

The shift toward AI and ML already has and will continue to revolutionize how people create technology and interact with information. What I find fascinating, as a product designer, is the research on and definition of systems that support human-AI collaboration. User experience already has quite a history and a solid foundation in fields such as human factors, cognitive ergonomics, and usability engineering. So it only makes sense to apply the same holistic, human-centered approach to AI systems.

We must understand the user’s perspective and needs, then create a process that weaves this understanding into every element of a product or service. While it is tempting to think only of ways in which AI could help improve workflows for professionals, it is beneficial to think about human-AI interaction not as a one-way street but as a dialogue between entities with different perspectives and understandings. By using AI to automate mundane tasks such as search and discovery, we can help legal professionals access data faster and more accurately than ever before. This, in turn, would allow legal professionals to look at bigger-picture questions and work on more challenging and, ultimately, more rewarding tasks.

Interactions between humans and AI systems shouldn’t feel transactional. Instead, the experience should be more like a dialogue and exchange of perspectives and insights from which both sides can benefit. Let’s re-envision the AI-assisted search example I mentioned earlier through this lens.

It is easy enough to enter a search query into a database and wait for the results. But consider what would happen if users could share their domain knowledge with the machine. They could inform the direction of a search by fine tuning key parameters, according to their knowledge and past experience. In our legal example, should they be looking for cases, content, guidance, or annotations?

The machine would then use this better-informed query, leveraging its processing speed and accuracy to identify and share patterns or trends that they either might otherwise have overlooked—or that would have taken significantly longer to find. Users would then be able to confirm or update their initial beliefs and move on to the next phase in their workflow. Plus, the next time users encounter the same situation, the system might be able to assist them in an even more professional and knowledgeable way and provide finer-tuned results.

From strategy and engineering to the production and delivery of the product, as designers, we need to constantly be thinking about how we could usefully apply new technologies such as AI or ML to a product, in ways that would help our users solve their problems.

Upending the Status Quo

Our broader research into the legal market shows that a common stumbling block to adopting new technology might ultimately be the person who could benefit most from it. Top leadership might fear the disruption that implementing a new technology could bring into the existing workflow, and they often choose the status quo over tangible change.

Every day, getting into the hearts and minds of our users challenges product designers. However, we must also understand the goals and motivations of their organization and the customers who have the purchasing power. When we can envision, then build products that are simple to use and integrate with existing processes, are well thought out and intentional, and deliver greater satisfaction and efficiency and a higher return on investment (ROI), the implementation and adoption of legal technology will become inevitable.