Day Two of the IA summit expanded on the concept of making AI and humans work together and become adaptable with each other instead of fighting for dominance in the design process. These are some of the sessions that developed this concept further and shone a light into how human knowledge must be applied to machine capabilities to make solutions work more effectively for users.
Don’t forget to human
Elizabeth Churchill, Director of User Experience at Google, opened the day by taking a personal look into what she considers the challenges of applying humanity into the development of artificially intelligent tools. She focuses on the concept of “weak” AI, and how we should use it in context to create trust and mutual understanding that can enhance intelligence and expand knowledge compared to a “strong” approach, where we try to emulate the consciousness and self-reflection that humans possess.
There are three challenges Churchill sees when integrating humans and AI effectively. The first problem tackles the lack of consensus of what we mean by AI and how it can assist human development. John McCarthy, Mark Weiser, Danny Bobrow and Margaret Boden are quoted to show the variety of opinions and perspectives that have risen in the last 70 years. They comment on how AI leads to cooperative intelligence, how tools and systems need to be embedded in the environment operating for us and how autonomy is an ever-flowing property. She urged the audience to see Douglas Engerbart’s The Mother of All Demos videos to reflect on technology being used as an enhancer of human performance.
The second challenge is understanding human ingenuity. Churchill talked about research and observations of people performing everyday things that help her team determine how they interact, react and use the technology prevalent in their environment. She says that “technologists sometimes forget to be human” and that we need to see the dynamic between people’s attitude and their reflection about their daily rituals as a signal that we’re starting to think about life more mechanically.
Finally, challenge three focuses on creating new tools that account for biases, ethics and the cultural differences in the uptake of AI. Churchill urges us to study technology and establish guidelines so that tools can be sustainable in helping all people in the long run. Like Alan Cooper’s presentation yesterday, she advices caution between expediency without thinking of long term consequences.
A taxonomist, a software engineer and a UX researcher make IAs want to buy
Jenny Benevento, Gio Fernandez and Jill Fruchter gave an amazing and engaging talk about how Etsy benefited from the power of taxonomy, software engineering and user research to allow the (previously) artificial intelligence and machine learning driven company to tackle the limitations of data driven decisions. Throughout the presentation, the speakers embrace the notion of wanting a system that adapts with the passage of time to take care of its user evolving needs.
The session started with a hypnotizing and dynamic slide, used as the company’s overall screensaver, reflecting insightful words and concepts of actual user generated search queries. These themes showed the complexity of the wants and needs specific to Etsy users and the challenge that the company faced to leverage them in designing an experience that was applicable to their unique search behaviors.
To approach this challenge, the team decided to separate their user base between buyers and sellers, groups with differing goals in terms of finding products and portraying their merchandise. The project started as the user researcher studied behaviors on users finding jewelry online. It showed that Etsy users have different ways of browsing and searching for content, focusing on visual components within a page and having concerns on not seeing everything that may be available based on their search. This concept, more colloquially known as FOMO (Fear of Missing Out), is what may overwhelm the users in their continued efforts to find the right items. In concluding this facet of the project, Fruchter highlighted the biggest insight gathered from the research: “search is the task, finding is the goal.”
Next, Benevento explains that Etsy users have the tendency to follow the berry picking technique, described by Marcia Bates in the late 1980s, which describes how users query for a specific concept and update subsequent queries based on the results gathered until they feel satisfied that they’ve discovered or found what they were looking for. Fernandez followed by saying that query classification was used in a manual process to categorize documents and listings to feed them into a model. However, this alone didn’t help mitigate the problem. Instead, it led to an understanding of the user knowledge as an asset and a mindset of trusting that the users were the experts. Therefore, fancy data models was not be only or best solution to this problem.
To conclude the session, the team shared their lessons learned, which tie well to the overall themes of the conference and the way humans contribute the most to providing solutions. They start by acknowledging that “data is queen” and AI is not the solution if there’s no consideration for the user needs, the data value proposition and the insight it provides on user behaviors. They support and encourage people to look at multidisciplinary thinking to approach holistic problems and have confidence that users may be the most qualified to provide answers.
During the question portion of the talk, the audience had a pleasant surprised that validated the team’s approach and conclusions about their project. As if by magic, Marcia Bates (previously referenced by Benevento) walked up to the microphone, praising the Etsy team for having the right approach to solving the problem of catering for two sets of users, as the experience for the indexer and the searcher is fundamentally different.
The session was a case study that showcased the value and contributions that different perspectives and various skillsets bring to the development of successful solutions to user challenges. It was encouraging to see collaboration amongst team members in studying and understanding the situation being presented leading to the determination of the processes, concepts and tools that were needed to realize that technology is not always the solution to technology-driven problems.
“The humanness is being taken away”
I had the opportunity to see Amber Case at a UX meetup in Seattle at the beginning of this month and I was entranced by the way she discusses topics, explains concepts and imparts knowledge. Today, she presented “Designing Calm Technology” as the end-of-day keynote session, to share the eight principles of calm technology and how they can deter people from the potential and unneeded problems that our reliance and expectations of technology can create.
The calm technology principles revolve mainly around the concept of designing technology to become minimally disruptive, helpful and valuable for our everyday activities but not the center of attention and applicable to the social norms that we use to guide our actions. She emphasized the notion that people shouldn’t strive to be machines and machines shouldn’t be built to emulate people. Instead, technology should be curated to be applicable to situations that need it and get value from its use. As Case continued with this fun and extremely engaging presentation, she cautions that technologies can crumble into chaos, so we must be aware of their limitations and the effects they have on users when they rely heavily on it.
This presentation helped me reflect about the overall theme presented at day two of the 2017 AI Summit and the gadgets and technologies that are part of mundane activities during the day. I’ve become introspective in the application and effects that the heavy use of technology has on my own life. Now, I feel the responsibility to study and apply guidelines explained by Churchill and Case to make sure that if the technology I extensively rely on fails, I have the mental fortitude to remain calm and carry on.
For more information in the principles of calm technology, visit calmtech.com.
Come back tomorrow and for a last day of live blogging from our on-the-scene author, Julieta Sanchez.
Information architecture is an often misunderstood job title. Are they designers? developers? managers? All of the above? In this article we'll discuss what information architecture is, why it's related to usability, and what are the common tools/programs used in information architecture.