How Elizabeth Churchill is putting people and purpose at the heart of AI

From personal devices to platform infrastructures, technology is ever more ubiquitous in our lives. Indeed, it is part of the fabric of our daily existence. This is why it is so crucial that we understand more about the relationship and interaction between people, technologies, and technical infrastructures – especially with so much emerging in the era of AI.

To do this, Elizabeth Churchill, department chair and professor of human-computer interaction (HCI) at Mohamed bin Zayed University of Artificial Intelligence, believes we can draw inspiration from the long-established fields of psychology, sociology, anthropology, and human factors and ergonomics – as well as the computer sciences. 

The social sciences have helped us understand people and cultures, while the engineering-focused world of human factors and ergonomics has helped us optimize the design of objects and tools for human use. Building on these fields, the comparatively new discipline of HCI is perfectly placed to help us understand the why of people’s need for and use of interactive technologies. As a result, we can understand what we should design and build, and how to make  those things usable, useful, delightful, and maybe even magical. This is especially the case with technologies that utilize the ‘smarts’ of artificial intelligence techniques. 

This has been the work of Professor Churchill for several decades. Since her undergraduate studies in psychology, she has sought to understand how people process information, learn, create, and interact with technologies and – through technologies – interact with each other. 

Her focus has been on helping people, and she sees technologies as having huge potential for augmenting human potential and experience in very positive ways.

“HCI is about helping people realize their potential,” she says. “It’s about building technologies that expand people’s experiences. How do we support people in getting things done? How do we make people productive in ways that are meaningful and important to them? How do we design something so it doesn’t get in the way but helps? How do we help people connect with each other and collaborate more effectively? HCI is about putting a human into a useful, productive, satisfying, and delightful relationship with technology, and through technology, with each other. 

“But while I try to focus on the positive, we also need to understand the motivations and actions of ‘bad actors’– people who wish to use technologies to exploit people or to do others harm for their own benefit. Many HCI scholars are working on this, focusing on understanding, predicting, and designing for safety from harm.” 

Facilitating impact

With people using AI technologies more frequently (just look at the popularity of ChatGPT, Midjourney, and the like) these types of questions and considerations have become even more essential for academics and industry players to address. 

In this, Churchill has long been ahead of the curve – thinking about the importance of intentionally designing not just the everyday experiences we have as consumers, but also designing for the creators of those experiences. 

“When it comes to AI, this is no different”, she says, distinguishing three areas of design when it comes to creating new AI experiences: 

• Defining and designing what AI is and could be, focused on AI researchers and engineers.
• Designing with AI, which means creating meaningful and useful experiences in products and platforms, focused on the HCI, UX and product/platform engineering professionals.
• Designing for AI, which means making what AI is doing more transparent and malleable, focused on AI systems themselves, and designing for these systems to communicate what they are doing with us.

“When designing AI, we need to look at what is the potential of any algorithm or approach,” Churchill explains. “How can we create long-term value for people and avoid potential harms? 

“When designing with AI, we need to recognise that AI is a malleable material, and we can use it to design incredibly helpful tools like navigation aids, personal assistants, creativity and productivity tools, decision making platforms, and entertainment experiences. In this area, I’m particularly excited by possibilities in healthcare, and in creating better cultural connections – for example with improved language translation. 

“And when we’re designing for AI, we can think about how we build better interfaces and tools so people can better understand what AI is contributing, aimed at more transparency and dialogue, a true collaboration between people and AI-based systems.”

Having taken her first class in AI and symbolic modeling / reasoning during her bachelor’s degree in the 1980s, she went on to do a Master’s degree in AI — then known as ‘knowledge based systems’ — at the University of Sussex in the UK, where her Masters thesis was on the development of an adaptive intelligent tutoring system to teach complex programming concepts. From there she went on to be awarded a Ph.D. from the University of Cambridge in ‘user modeling’ based on understanding how humans reason and problem-solve. 

Moving into industry in the late 1990s with a move to the US, she drew on social, computer, engineering, and data sciences to take research ideas into prototypes and products. Building teams of researchers and engineers at some of the world’s leading technology companies, she and her teams created innovative end-user applications and services. For example, at Fuji Xerox’s Palo Alto Research Center (FXPAL) she built a team called the Social Computing Group. 

“One of the first products that I was involved in designing, developing, and launching was an interactive public display called Collaboposter,” she recalls, “which was based on our research into how public display technologies can improve information sharing in public spaces — moving away from the paradigm of one-person-to-one-personal-computer, and focusing instead on community, social, interactive, public computing.” 

Later, at Yahoo!, she formed, staffed and directed research at the Yahoo! Labs’ Internet Experiences Group, again prototyping and launching several social media products. Following that, she joined eBay as Director of Human Computer Interaction at eBay Research Labs, creating a number of innovations in personalization for shopping recommendations.

It was as a senior director at Google, however, that Churchill turned her eye to designing creativity tools for those who are responsible for the everyday experiences we all have with our personal devices; using human-centered design to understand developer and designer needs.

“I didn’t work on products that get launched to consumers, such as search or navigation aids,” she explains. “I worked on designer and developer tooling and productivity. That was my passion. And that was why I went into Google, because I was excited about providing tools for developers and for designers that are really good, easy to use, that have great documentation and use cases.”

 Among the projects  Churchill worked on at Google with her teams are open-source UI software development kit, Flutter, Google’s design system, Material Design (now called Google Design Platform), and a next-generation operating system, Fuchsia.

“My belief is that if you give designers and developers better tools, and address what they need to make great products, then you give them the space to be more creative and to spend more time thinking about who they are designing and developing for,” she says.“They can then be focused on the impact and outcome of what they are building.”

“For example, if I can have use cases and the hooks in a design system or developer tooling system for accessibility to be a first thought, not a last thought, then those designers and developers will build things that have accessibility hooks that others can build upon. So accessibility becomes a prior, not a ‘nice to have’ tacked on at the end of the development cycle.

“If we can give developers tools that help them produce infrastructure in an easier way, then their personal creativity headspace gets unleashed and they have more time to think about what they want the tool to do, and what the experience should be for the users they care about.”

For the love of people

Underpinning all of this work is a deep love for and fascination with people, and a commitment to creating positive interactions with technology – a key driver for Churchill’s research and career path as she continues to strive for better solutions.

“I am a big fan of understanding people and I’d say every step along [my career] has been motivated by understanding more deeply how people act and interact. I have been called the David Attenborough of technology design” she laughs. 

“Career-wise, while many opportunities have been serendipitous, I have always either explicitly chosen or shaped opportunities focused on understanding the relationship between people, existing or emerging technologies, and technological infrastructures. The current wave of AI and the capabilities and potential pitfalls of GenAI (generative AI) including large language models (LLMs) is no exception.”

Churchill argues there is a socio-technical aspect to AI and other technologies that may be considered by researchers and developers, but too often the technology production process is weighted in favor of the ‘technical’ side, without much consideration for the ‘people’ side – the human element. 

“I think we see negative results of technology because not enough weight was put on the potential social impacts when evaluating the technology’s potential,” she says. “A challenge, in the positive sense, is trying to get people to understand that any technology that’s designed and then released is a socially impactful technology, as well as a deep technical achievement.

“This is one of the reasons I’m at MBZUAI. Our President, Eric Xing, isn’t just investing in AI. He wants a human-centered focus so that we explicitly think about the impact of AI in all its formats on people. This will assure we do the best research and innovation in service of people and sustainable AI-powered experiences. This is a perspective I am fully signed up for.

Churchill highlights another aspect of MBZUAI that she believes is crucial to achieving this vision: Culture.

“I am a big fan of community culture, and we are dedicated to that at MBZUAI. I want to cultivate curiosity, not certainty, and humility not arrogance. This is how we find the most effective processes and the right outcomes for people. We want students to have a beginner’s mind, a collaborative attitude, and to feel so comfortable in asking hard questions about the why, as well as the what and the how.” 

“MBZUAI has established itself in terms of driving technology and infrastructure innovation in AI,” she adds. “It aims to empower students, businesses, and governments to advance AI as a global force for positive progress. It concentrates its research resources and talented minds on AI for good and is committed to finding solutions to the world’s most pressing challenges with transformative research in areas such as healthcare, education, and climate. 

“The University is definitely [also] a champion of responsible AI and teaching such values to the next generation of change-makers. I am so honored, excited, and inspired to be part of shaping this conversation at MBZUAI and, with my colleagues at MBZUAI, in the world.”