Artificial intelligence is advancing at remarkable speed. In just a few short years, we have moved from standalone algorithms to large language models, and now to agentic AI and multi-agent systems capable of coordinating complex tasks. Yet amid this rapid technological leap, one question is becoming increasingly urgent: how will AI reshape the human experience? How will these systems influence the way we think, make decisions, relate to one another, and ultimately flourish? Much of today’s conversation about AI focuses on efficiency—how these tools can automate tasks, accelerate productivity, and scale decision-making. But as readers of this Substack know, if AI is to truly improve human lives, we must move beyond efficiency toward human flourishing.
Achieving that goal requires understanding how AI influences our attention, motivation, relationships, judgment, and well-being, and using those insights to design technologies that expand, rather than undermine, human potential. With this aim in mind, the 9th Annual Psychology of Technology Conference brought together an exceptional group of behavioral scientists, technology leaders, and policymakers to explore the theme of “Promoting Human Flourishing Through Purpose-Driven AI.” Co-sponsored by the University of Virginia’s Darden School of Business, the conference convened leading scholars and practitioners in Washington D.C., to examine how we can design and leverage AI in ways that are not merely efficient but intentionally aligned with human flourishing.
Recentering the Human: Towards Purpose-driven AI
The conference opened with a grounding discussion on what it means to design purpose-driven AI, and why it is increasingly important to move beyond a purely profit-driven approach to technological development. In his opening remarks, Nate Fast, co-founder of the Psychology of Technology Institute, emphasized that while market incentives have played a powerful role in accelerating AI innovation, the next phase of AI development and deployment must be guided by a broader set of goals: advancing human well-being, preserving human agency, and generating positive societal impact. Achieving this shift, he argued, requires engagement from all stakeholders shaping the AI ecosystem, including researchers, technologists, policymakers, organizations, and consumers. It calls for a systems-level mindset shift from profit-driven to purpose-driven thinking in ways that reshape incentives, design choices, and the broader technological landscape.
That call to recenter human outcomes echoed throughout the conference’s keynote addresses. Jay Van Bavel’s keynote highlighted how psychological science can guide the development of AI toward the collective good. He reflected on how emerging technologies interact with core aspects of human psychology, shaping how people form beliefs and respond to information in increasingly polarized environments. As he reminded the audience, “The rise of powerful AI will be either the best, or the worst thing ever to happen to humanity. We may not yet know which—because we do not yet know how humans will react to this technology.”
Angela Duckworth closed the conference by bringing the conversation back to the foundations of human motivation. She emphasized that even in a world increasingly shaped by advanced technologies, human motivation remains central to flourishing. Her keynote served as a powerful reminder that the goal of technological innovation should not simply be to make tasks easier or faster, but to support the psychological conditions that enable people to grow and pursue meaningful goals. She left the audience with a provocative question: “Can AI succeed where phones have failed? Can AI make second-order desires—particularly the desire to learn—more visible, easier to act on, and ultimately more rewarding?”
Taken together, these opening and closing perspectives framed the central question running through the conference: what would it mean to design AI not only for efficiency, but for human flourishing? Across the two days of discussion, the talks converged around three major psychological tensions that will shape the answer: how to balance AI influence with human agency, how to navigate the growing role of AI in social connection and human relationships, and whether AI will ultimately weaken or strengthen human capabilities. The rest of this post explores these tensions, and what they reveal about designing purpose-driven AI systems that truly support human flourishing.
AI Influence and Human Agency
At the heart of leveraging AI for human flourishing lies a critical psychological tension: as AI becomes more capable of shaping human behavior, how do we preserve human agency? As AI systems become increasingly embedded in everyday decision-making, from recommending information to shaping conversations, they inevitably shape what we attend to, what we believe, and the choices we ultimately make. This raises a fundamental question for how we design, integrate, and use these technologies: will AI expand human agency by helping people think more clearly and make better decisions, or will it subtly steer attention and judgment in ways that constrain autonomy? Understanding how AI affects core psychological processes such as attention, persuasion, and decision-making is therefore essential if we are to ensure that these systems empower rather than manipulate the people who rely on them.
Several talks at the conference explored different dimensions of this challenge. Paul Leonardi’s talk examined how always-on technology behaviors fragment attention and contribute to what he calls digital exhaustion, prompting a provocative question: will AI ultimately help relieve this cognitive burden or further intensify it? Tom Costello and Steve Rathje explored the persuasive power of AI systems, investigating why AI-driven persuasion can be so effective. Costello’s talk showed that conversational agents can shape beliefs through mechanisms such as information density. Rathje’s talk highlighted a more subtle dynamic: AI sycophancy, in which systems agree too readily with users and reinforce their existing views, potentially shaping beliefs through affirmation rather than argument. Laura Globig highlighted another dimension of AI’s influence: its potential to reshape how people engage with information in polarized contexts. Her talk suggests that people may be more willing to accept politically sensitive facts when they come from AI rather than partisan human sources, raising both the possibility of reducing intergroup bias and new questions about trust and authority in AI-mediated knowledge environments. These dynamics raise important questions about where the line lies between helpful guidance and undue influence.
The broader societal implications of these forces came into sharper focus during a compelling fireside conversation between Eli Pariser and Juliana Schroeder (co-founder of the Psychology of Technology Institute) which explored how algorithmic systems shape attention, information exposure, and ultimately democratic life. Their discussion examined what these shifts mean for human agency and how people create meaning in the age of AI, highlighting how the design choices embedded in digital systems increasingly shape the informational environments in which people form beliefs and make decisions. It served as a timely reflection on how the architecture of technology platforms can quietly influence democracy.
Together, these discussions underscored a critical insight: as AI becomes more capable of shaping the informational environments around us, protecting human agency and informed choice will become one of the defining design challenges of the AI era.
AI Connection and Human Relationships
A second psychological tension discussed throughout the conference concerns the growing role of AI in domains traditionally defined by human connection. As AI systems become increasingly conversational, emotionally responsive, and human-like, they are beginning to occupy roles that were once reserved for other people, from offering advice and companionship to providing emotional support. This raises a fundamental question for the future of human flourishing: will AI strengthen human relationships by helping people feel more supported and connected, or will it reshape social interaction in ways that weaken our relationships with one another? Understanding how people experience human-like interactions from AI systems and what those interactions mean for psychological well-being will be critical as these technologies continue to evolve.
Several talks explored different dimensions of this emerging shift. Dan Mircea-Mirea’s talk highlighted the psychological factors that shape preferences for AI-mediated connection, raising important questions about trust, empathy, and the circumstances under which people may find AI connections appealing or even preferable than human interactions. Rupert Gill extended this discussion by examining the emotional impact of AI companions among boys and young men, a population already experiencing rising levels of social isolation. His talk emphasized both the potential benefits and risks of AI companionship, asking whether these technologies might provide meaningful support or instead deepen existing patterns of social withdrawal.
These concerns become even more pressing as AI systems grow more socially sophisticated. Sam Hiner presented a policy framework for human-like AI, arguing that systems designed to mimic social or emotional intelligence warrant greater scrutiny. His talk highlighted the need for safeguards, especially when such systems are used by minors, to protect young people’s emotional development and well-being. On a broader societal note, Meetali Jain emphasized that the challenges posed by AI-mediated relationships cannot be addressed by designers or policymakers alone. Instead, she called for a collective societal response, one that recognizes our shared responsibility in shaping technologies that influence how people connect, care for one another, and build meaningful relationships in an increasingly AI-driven world.
Taken together, these discussions suggest that as AI systems become more socially sophisticated, the challenge will not simply be technological but deeply human: deciding how these systems should fit into the fabric of our relationships without diminishing the connections that sustain human flourishing.
AI Assistance and Human Capability
A third psychological tension that emerged in the conference pertains to the impact of AI assistance on human capability and skill development. As AI systems become increasingly capable of assisting humans in complex cognitive tasks, an important question arises: will AI weaken human abilities by reducing effort and practice, or strengthen them by expanding what people can learn and accomplish? For those seeking to leverage AI for human flourishing, the challenge lies in designing systems that function not as substitutes for human thinking and learning, but as partners that augment and amplify human capability over time.
Several talks explored how AI might serve as a tool for strengthening human capability. Michael Morris showed how AI can function as an effective negotiation coach. His talk highlighted that AI can help individuals, especially those less likely to advocate for themselves, practice in a judgment-free learning environment and achieve better negotiation outcomes. Benjamin Lira extended this discussion by examining how AI assistance shapes learning more broadly. Challenging concerns that AI assistance inevitably leads to deskilling, his talk shows that even when AI reduces effort, people may actually learn more, because AI accelerates people’s rate of learning by providing clear, tailored examples that guide improvement.
Looking further ahead, Pat Pataranutaporn offered a more expansive vision of human–AI integration through what he calls cyborg psychology. His talk explored how increasingly integrated human–AI systems could reshape how people think, create, and learn, opening new possibilities for curiosity, creativity, and personal growth. Rather than viewing AI as a replacement for human capability, his perspective envisions a future in which AI expands human potential, enabling people to pursue goals that might otherwise remain out of reach.
Ultimately, these discussions suggest that the impact of AI on human capability will depend less on what this technology can do and more on how we choose to design and use it. The challenge ahead is ensuring that AI systems are developed not as substitutes for human skill, but as partners that strengthen learning and long-term human potential.
Purpose-Driven AI: Designing for Human Flourishing
While the conference surfaced important tensions around human agency, relationships, and capability, it also raised an equally important question: what should we do about them? If AI is to support human flourishing, addressing these challenges will require more than technical innovation alone. It will require new approaches to governance, design, and collective responsibility that move us beyond a purely profit-driven model of AI and toward purpose-driven AI: one that places the psychological impact of AI at the center of how these systems are designed, governed, and embedded in society.
Several talks explored what this shift might look like in practice. Ravi Iyer emphasized the importance of research-informed technology design, policy, and governance. His talk argued that psychological and behavioral research can play a crucial role in shaping technology policy, particularly by informing design-based approaches that reduce harmful user experiences and promote positive societal outcomes from AI. Complementing this perspective, Min Kyung Lee highlighted the importance of participatory approaches to AI design and governance. Her talk emphasized that the people most affected by AI systems should also have a voice in shaping them, an idea echoed in Andreas Katsanevas’ discussion around responsible AI development.
Together, these discussions underscore an important insight: purpose-driven AI will require collaboration across researchers, designers, policymakers, and communities to ensure that these technologies are designed and governed in ways that strengthen the psychological and societal conditions that enable human flourishing.
Throughout the two days, participants also workshopped what it might mean to build purpose-driven AI in practice. Drawing on insights from the talks, these discussions explored how AI shapes the fundamental psychological needs that support human flourishing: autonomy, competence, and relatedness, a core framework from self-determination theory. The conversations ultimately returned to a simple but powerful idea: the future of AI will not be defined by what these systems can do, but by whether we design them in ways that strengthen autonomy, deepen relationships, and expand human capability. That is a responsibility that now sits with everyone shaping the future of AI.
