At the Conversational Human-AI Interactions (CHAI) Lab, we design AI that enhances human capabilities rather than replacing them. We believe conversational AI should listen deeply, understanding not just words, but context, intent, and the person behind each interaction. Our research focuses on creating systems that help people navigate complexity across domains—from healthcare decisions to creative exploration to educational challenges—while preserving human agency. By handling routine cognitive tasks, well-designed AI creates space for what humans do best: nuanced judgment, meaningful connection, and creative problem-solving. We design for the full spectrum of human experience—moments that require empathy, careful decision-making, and genuine understanding. Our approach recognizes that people seek more than efficient transactions; they want technology that respects their intelligence and supports their goals through personalized, collaborative, and supportive interactions. Like sharing tea, meaningful human-AI interaction requires patience, attention, and mutual respect. We envision conversational AI as a collaborative tool that listens carefully, responds appropriately, and creates space for human wisdom to flourish.
We approach conversational AI through three interconnected dimensions that form the foundation of all of our research:
<aside> 💭
Understanding the Human Experience We investigate how people naturally think about, interact with, and make sense of conversational systems. By deeply understanding human cognition, communication patterns, and needs, we can design AI that feels intuitive and supportive rather than alien or frustrating. This foundational research reveals not just what people do, but why they do it—insights that inform every aspect of our design process.
</aside>
<aside> ✍️
Creating Human-Centered Solutions Armed with a deep understanding, we develop conversational AI that genuinely serves human goals. Our systems are designed to enhance learning, democratize access to information, support complex decision-making, and facilitate meaningful connections. We create tools that extend human capabilities rather than replace human judgment, ensuring that technology remains in service of human agency and creativity.
</aside>
<aside> 💡
Refining Through Continuous Learning We believe great conversational AI emerges through iterative refinement informed by real human experiences. We continuously evolve our design methodologies, testing our assumptions against lived realities and adapting our approaches based on what we learn. This commitment to refinement ensures our systems grow more helpful, more inclusive, and more aligned with human values over time.
</aside>
We envision a future where conversational interfaces will serve as the primary modality for human-computer interaction, designed to be personalized, context-aware, and fundamentally human-centered.
These future systems will augment human cognition, support complex decision-making, and reduce cognitive overhead, allowing people to focus on tasks that require uniquely human capacities such as creative thinking, ethical reasoning, and interpersonal connection. As conversational interfaces become ubiquitous across personal and professional domains, they will be designed with principles of user autonomy, privacy preservation, and cultural sensitivity as foundational requirements. This vision extends beyond technological capability to encompass the relational dynamics between humans and AI systems. Future conversational AI will be structured to support human flourishing by creating interactions that respect individual agency, accommodate diverse communication styles, and enhance rather than diminish human capacity for meaningful choice and self-determination. The realization of this vision requires sustained research into the social, cognitive, and ethical dimensions of human-AI interaction, ensuring that as these technologies become more powerful and pervasive, they remain aligned with human values and supportive of human potential.