This research examines the growing use of AI technologies in social connectivity through the lens of responsible AI. The project aimed to explore the future development of AI technologies in social contexts and investigate how these technologies can be designed responsibly for young adults, with a focus on their values. The primary objective was to generate design implications that would guide designers in the responsible creation of AI technologies that foster social connectivity.

To achieve this, the research undertook a thorough review of theoretical data to, (1) define social connectivity and its significance, (2) investigate current and emerging trends in AI technologies within social contexts, and (3) examine responsibility and the relevant guidelines for responsible AI. Based on the insights gained, the project focus was refined to concentrate on the responsible development of personal AI virtual assistants that enhance interpersonal social connectivity amongst young adults. The in-depth theoretical analysis culminated in the development of the “Framework for Designing Responsible AI Virtual Assistants for Social Connectivity,” which connected the explored topics and contextual values to the trends observed in AI virtual assistants and their potential social effects.

Building on these insights, the values of young adults pertinent to AI in the social domain were identified and elicited through design experiments. The research employed a combination of Speculative Critical Design and Value Sensitive Design approaches to effectively design, plan, and execute these experiments.

Ultimately, by identifying, eliciting, and defining relevant values, the research generated design implications in the form of nine value pyramids to guide responsible AI development. The research discovered nine values that young adults uphold regarding AI-fostered interpersonal connectivity: Human Agency and Oversight, Privacy and Data Governance, Transparency, Societal and Environmental Wellbeing, Personal Effort, Human-Human Connectivity, Genuity and Authenticity, Personalization, and Emotional Connectivity. These values were crucial in generating "Implications for Design," where each value was contextually defined and expanded into value principles and design norms, following Van De Poel’s (2013) Hierarchy of Values. This study ultimately generated design implications intended to guide designers in the responsible creation of future AI virtual assistants that foster interpersonal social connectivity.

This research examines the growing use of AI technologies in social connectivity through the lens of responsible AI. The project aimed to explore the future development of AI technologies in social contexts and investigate how these technologies can be designed responsibly for young adults, with a focus on their values. The primary objective was to generate design implications that would guide designers in the responsible creation of AI technologies that foster social connectivity.

To achieve this, the research undertook a thorough review of theoretical data to, (1) define social connectivity and its significance, (2) investigate current and emerging trends in AI technologies within social contexts, and (3) examine responsibility and the relevant guidelines for responsible AI. Based on the insights gained, the project focus was refined to concentrate on the responsible development of personal AI virtual assistants that enhance interpersonal social connectivity amongst young adults. The in-depth theoretical analysis culminated in the development of the “Framework for Designing Responsible AI Virtual Assistants for Social Connectivity,” which connected the explored topics and contextual values to the trends observed in AI virtual assistants and their potential social effects.

Building on these insights, the values of young adults pertinent to AI in the social domain were identified and elicited through design experiments. The research employed a combination of Speculative Critical Design and Value Sensitive Design approaches to effectively design, plan, and execute these experiments.

Ultimately, by identifying, eliciting, and defining relevant values, the research generated design implications in the form of nine value pyramids to guide responsible AI development. The research discovered nine values that young adults uphold regarding AI-fostered interpersonal connectivity: Human Agency and Oversight, Privacy and Data Governance, Transparency, Societal and Environmental Wellbeing, Personal Effort, Human-Human Connectivity, Genuity and Authenticity, Personalization, and Emotional Connectivity. These values were crucial in generating "Implications for Design," where each value was contextually defined and expanded into value principles and design norms, following Van De Poel’s (2013) Hierarchy of Values. This study ultimately generated design implications intended to guide designers in the responsible creation of future AI virtual assistants that foster interpersonal social connectivity.

Social media has revolutionized how we access and share vast amounts of information, fundamentally altering the ways we interact, debate, and form opinions. Social media platforms have become predominant channels for information with personalization algorithms significantly shaping the content we encounter, thus, have a big impact on society. Initially, none of the social media platforms were created with the aim of delivering news. However, as their user bases expanded and their features diversified, a significant portion of their users began perceiving and using them as a news source. The research specifically focuses on the platform X (formerly Twitter), which is selected for its mission to promote and protect public conversation, positioning itself as the town square of the internet.

This thesis explores the design interventions to disrupt incidental news consumption and foster healthy discourse on social media platforms, particularly X (formerly Twitter). The literature review incorporates interdisciplinary elements, such as recommender systems, echo chambers, EU regulations, and democracy models, providing a comprehensive framework for the study. Insights from stakeholders, including politicians, non-profit organizations, and policy advisors, revealed overlooked aspects and guided the exploration of potential changes to the social media landscape. An in-depth analysis of X's features and issues informed the development of design proposals to introduce frictions in news consumption, aiming to increase exposure diversity and facilitate healthy online discourse. By analyzing qualitative data from stakeholder interviews, provotypes, and the user evaluation session, the research identifies challenges and opportunities in designing interventions.

This thesis provides design recommendations to introduce friction to incidental news consumption on social media and uncovers users’ preferences and concerns about online discussion spaces which aim to foster healthy discourses. In the end, the thesis uses these design recommendations and redesigns the initial design proposals to be able to provide a concept and solidify the recommendations for the future research.

Finally, this thesis advocates for the introduction of frictions into endless social media feeds to bridge echo chambers and enhance the diversity of viewpoints encountered. In other words, this research demonstrates that social media experiences do not always need to be seamless. Thoughtfully introduced frictions can provide moments for reflection and encourage users to engage with a broader range of perspectives, ultimately supporting a more informed and democratic society.

This work represents an initial step towards a more thoughtful and informed social media experience, contributing to a healthier democracy and a better-informed public. While the thesis acknowledges that influencing regulatory change is a long-term endeavor, it hopes to go beyond the scope of the thesis and be an influence to the future regulation practices.