Humane Chatbots entwerfen: Der interdisziplinäre Ansatz des MIT

In the present-day world of digital information overload, young adults are frequently stuck in a whirl of online distractions like social media and chatbots competing for their attention. But can we envision a digital universe where these platforms evolve from mere distractions to helpful moral companions? This intriguing question sparked a unique venture at MIT, where an anthropologist teamed up with a computer scientist to devise a one-of-a-kind undergrad course.

Named Humane User Experience Design, this course offers an intriguing blend of computer science and anthropology, nudging students towards creating AI chatbots designed to facilitate self-improvement. Not your usual course, it spans the divide between computer science and anthropology. It allows computer science majors to meet their humanities prerequisite while furthering their professional aspirations.

Blurring Boundaries and Bridging Disciplines

The course came into existence thanks to a collaboration between professors Arvind Satyanarayan and Graham Jones. They have leveraged linguistic anthropology techniques to guide students on how human interactional needs can be seamlessly integrated into programming. What’s interesting is that the course even caught the attention of the MIT Morningside Academy for Design (MAD), which funds this innovative project. MAD takes pride in fostering novel classes that obliterate traditional department walls, as Humane UXD perfectly illustrates.

When Friendship Fuels Innovation

Professors Jones and Satyanarayan’s professional partnership, which began when they jointly advised a doctoral student on data visualization for those with visual impairment, has since morphed into a vibrant friendship. This relationship has allowed them to delve deeper into their disciplines’ intersections. Satyanarayan savors the profoundness anthropology lends to human-computer interaction tactics, while Jones is intrigued by how Satyanarayan’s teaching philosophies play out.

Furthermore, Jones finds the design aspect of teaming up with a computer scientist exciting. He notes that our speech and interactions follow specific patterns, reflected in AI models like ChatGPT. These models, trained on human communication, contain inherent patterns poised to be harnessed.

Student Projects

One particular project, dubbed “Pond,” targets newly-fledged college graduates grappling with the challenges of adulthood. Unlike conventional chatbots, Pond offers pragmatic advice on ‘adulting,’ enabling users to transition from college life to independent adulthood more smoothly. Another project, ‘News Nest,’ facilitates engagement with credible news sources in a fun way by using colorful birds representing different news areas. This helps counter doomscrolling and fosters media transparency.

A slightly different approach is adopted by Team M^3, which has developed AI into a social deduction game where users interact with different chatbots to solve a murder mystery. All these projects underscore the transformative power of interdisciplinary education and the diverse skills and insights students gain.

For more information, check out the original article at MIT-Nachrichten.