Findings

The Rise of AI

Over the years, artificial intelligence (AI) has undergone remarkable advancements, revolutionising how we experience various facets of our lives, whether in healthcare, entertainment, education, or transportation. A captivating trend within AI is emotional AI, also known as affective computing, which aims to grant machines the ability to comprehend, interpret, and respond to human emotions. This infusion of emotional intelligence into AI systems holds the exciting potential to make our interactions with machines more natural and personalised.
Emotional AI has found applications in diverse areas, including customer service, mental health, and even gaming and entertainment. In the field of mental health, Emotional AI is a rapidly evolving domain with the potential to reshape how we interact with technology.

Emotions and AI

Empathy is an intricate concept with various aspects, but at its core, it involves grasping someone else's emotional states. In essence, if machines can attain this understanding, they have the potential to enhance their assistance to us. This also raises the question of how we can integrate design with Emotion AI, ensuring that machines not only understand but also respond to human emotions in a thoughtful manner.

Insights we gained

People turning to AI for Companion

Research has shown that users often develop a profound attachment to their AI companions, with certain studies noting how they alleviate feelings of loneliness. AI companions are designed to provide emotional support and companionship during moments of solitude. These companions can supposedly be programmed with emotional intelligence, placing a strong emphasis on empathy and support. Not only do they possess the ability to recall past conversations, but they also have the capacity to ask follow-up questions and offer guidance without any hint of judgement. Moreover, AI companions have the potential to empower users by helping them overcome social anxiety through the art of initiating communication or by encouraging them to open up about their personal experiences and emotional vulnerabilities.

Through our findings and exploration of AI, mental health and emotions, we were driven to delve deeper into the evolving relationship between users and AI companions. Our research showed a significant reliance on AI for emotional support, particularly in the realm of combating loneliness. Motivated by these insights, we aimed to fuse generative aesthetics along with AI in our design decisions. By integrating these elements, our objective was to create an artefact that not only understands and responds to users' emotional states but also provides a comforting and visually pleasing experience. Through Coded Comfort, we intend to communicate the potential of AI companions, coupled with thoughtful design, to serve as valuable sources of comfort and companionship for users navigating the complexities of their emotions and mental well-being.

Design decisions

TAs we wanted to merge generative aesthetics along with AI, we contemplated on how the visuals could contribute to an engaging and an open-ended intimate experience for viewers. For the visuals, our primary goal was to evoke a sense of comfort. Initially we tried to use StableDiffusion to integrate them into visuals. However, we found that this approach did not accurately capture what we truly aimed to convey. Drawing inspiration from Refik Anadol, a new media artist and designer, we observed his project, which utilises data-driven machine learning algorithms to create abstract, dream-like environments. Using that part which we liked, we wanted to make our visuals capture dynamic, swirling particles, conveying the various emotions we experience—sometimes, those feelings that are challenging to articulate when we communicate with each other.

Challenges

When we initially began our discussion, our goal was to uncover the meaning and context of human emotions. This exploration led us through various concepts, occasionally adding complexity to our discussed topics. Along the way, we encountered challenges in grasping certain concepts and ideas, resulting in a disparity and gap between knowledge and execution. We were initially unsure of how to proceed with executing our plans. Through our ongoing discussions, we fortunately reached compromises and simplified our ideas and concepts to align with each other's strengths and weaknesses. We made sure that each group member was 1) interested in the chosen topic and 2) had a clear understanding of it.

Feedback

After receiving two rounds of feedback—first during our proposal presentation and then after showcasing our artefact—we gained valuable insights to enhance our project. Initially, we realised the importance of emphasising our project's intention, prompting us to question whether it aimed for a commentary, a poetic experience, visualising a gap in human-AI connections, or creating a new AI.

Subsequently, we engaged in discussions to clarify our project's purpose. In the second feedback round, the focus shifted to improving documentation, urging us to provide a clearer expression of our intended message. These feedback iterations prompted thoughtful discussions, guiding our project to improve both conceptual clarity and documentation precision.

Achievements

Our team engaged in numerous discussions and research to develop our final concept, Coded Comfort. Throughout these discussions, we were not only receptive but also confrontational enough to voice our opinions whenever possible, even though this initially posed a challenge. With each small step, we ensured that our ideas aligned with the concept, emphasising intentions and objectives. Additionally, we were open to experimenting with new AI bots, incorporating ideas that aligned with our concept. We are all highly satisfied with the outcomes and experiments.