People of ACM - Michelle Zhou

March 16, 2021

Why is this an exciting time to be working at the intersection of AI and human-centered AI?

Three main reasons. First, AI has advanced faster than ever before. Such advances impact both individuals and organizations in unexpected ways (e.g., delivering/receiving AI-empowered healthcare)—for good or bad, but certainly fast. It is important to investigate and predict both beneficial and adversarial AI impacts from a human-centered perspective.

Second, because of rapid AI advances, many organizations want to adopt AI to scale their operations while containing costs. However, not every organization has the required human skills or financial resources to create their own AI solutions. It is, therefore, a perfect time to investigate how to democratize AI adoptions especially for "less privileged" organizations. This requires us to understand who will build AI, who will interact with AI, and how to lower the barriers of entry to the AI world. By addressing these questions, we will help to prevent a potential “AI divide.”

Third, despite AI advances, AI is far from perfect. It is important to study how humans could better leverage today's resources (e.g., computing power and rich data) to help advance AI faster.

Juji claims to offer the world’s only AI assistants with cognitive intelligence. Will you tell us what you mean by “cognitive intelligence,“ and could you give us some examples of its practical benefits?

In our human world, "cognitive intelligence" refers to our abilities to understand, remember, solve problems, learn, and make decisions. In the machine world, "cognitive intelligence" refers to machines that possess an advanced, human-like version of artificial intelligence, including the capacity to actively listen to users and respond to users empathetically, and read between the lines to learn about a user deeply and personalize user guidance.

Cognitive AI assistants have several unique advantages over typical AI assistants. For example, to help a university’s recruitment/enrollment effort, a Juji AI assistant can successfully multitask, interviewing prospective students to elicit students’ needs and interests while answering students’ questions during the interview. More important, Juji’s pre-built cognitive intelligence enables an Admissions Office staff to quickly set up such an AI assistant without coding, a key step toward democratizing AI.

What do you envision as some exciting new applications for AI assistants in the next five to 10 years?

I can see advanced AI assistants being used in diverse areas to help humans in multiple aspects of our lives, especially as long-term companions. For example, AI personal wellness assistants will know about our wellbeing better than we know ourselves. They can check in on us daily and provide deeply personalized wellness tips based on our unique characteristics to keep us healthy, both physically and mentally.

Another advance will be in lifelong learning, where AI personal learning assistants will understand our individual learning needs and styles and help us to learn more, better, and faster.

Everyone would like an AI companion like the AI assistant depicted in the popular movie "Her". Perhaps an even more advanced version of “Her” will know us better than we know ourselves, and will listen to our complaints, cheer us up, and offer life guidance whenever possible.

What role does TiiS play in advancing the field?

To the best of my knowledge, TiiS is the only scientific journal that solely focuses on publishing cutting-edge research at the intersection of AI and HCI. Given the increasing importance of human-centered AI, TiiS plays a critical role in advancing this whole field. First of all, it presents a great venue for researchers to publish truly interdisciplinary work, fostering interdisciplinary collaborations across multiple fields to advance both AI and HCI. Second, TiiS serves as a guide for both academia and industries to not only learn the cutting-edge technologies and solutions in human-centered AI but also the applications and guiding principles of these technologies. In 2020, TiiS established a special column that invites leaders of AI and HCI to provide their unique perspectives about human-centered AI, with an inaugural article on AI ethics and practice by Ben Schneiderman.

What advice would you offer a younger colleague who is interested in pursuing a career in intelligent user interaction/human-centered AI?

First, it is critical to learn the fundamentals of both AI and HCI, since human-centered AI is an interdisciplinary field and requires interdisciplinary knowledge and skills to excel.

Second, it is important to participate in AI and HCI research communities. For example, attending the ACM International Conference Series on Intelligent User Interfaces (IUI) would be a great start. It’s an opportunity to learn more about human-centered AI advances, and, more important, to get to know others in the community and build partnerships.

Third, it is important to get hands-on experience, for example starting a human-centered AI research project. Find a practical problem that you are interested in solving (e.g., building a conversational AI agent to help students find suitable jobs) and try to solve it yourself. If you can’t think of a problem on your own, check out the "future work" section of a scientific publication. You can then iteratively build, test, and improve your solution. If you believe your solution is novel, evaluate it, write it up and submit it to the right conference (e.g., IUI) or journal (e.g., TiiS) to share your experience with others.

Michelle Zhou is Co-founder and CEO of Juji, Inc., a California-based company that develops cognitive artificial intelligence assistants in the form of chatbots. She is an expert in human-centered AI, an interdisciplinary area that intersects AI and human-computer interaction (HCI). Zhou has authored more than 100 scientific publications on subjects including conversational AI, personality analytics, and interactive visual analytics of big data. Prior to founding Juji, she spent 15 years at IBM Research and the Watson Group, where she managed the research and development of human-centered AI technologies and solutions, including IBM Watson Personality Insights.

Zhou serves as Editor-in-Chief of ACM Transactions on Interactive Intelligent Systems (TiiS) and an Associate Editor of ACM Transactions on Intelligent Systems and Technology (TIST), and was the Steering Committee Chair for the ACM International Conference Series on Intelligent User Interfaces (IUI). She is an ACM Distinguished Member.