Hybrid Minds

Introduction

Artificial Intelligence (AI) has become a ubiquitous force shaping how humans think, create, and relate. As generative systems such as Large Language Models (LLMs) and agentic AI increasingly mediate everyday cognition, they transform creative expression, reasoning, and the nature of collaboration itself. A new cognitive ecology is emerging in which humans and machines cohabit.

This transformation extends beyond individual minds into the social fabric of knowledge itself. As AI systems enter knowledge communities, mediating expertise, deliberation, and public discourse, they function as epistemic technologies that reshape how societies produce, validate, and act upon understanding. The resulting human-machine hybrids raise urgent questions about democratic life: how algorithmic mediation affects trust, legitimacy, and the conditions for informed collective judgment in an age where knowledge work is increasingly distributed across human and computational agents.

To tackle these challenges, the conference «Hybrid Minds» explores the evolving transdisciplinary intersection of human cognition, artificial intelligence, engineering, and the arts, with a particular focus on how AI systems relate to fundamental assumptions about perception, knowledge, and creativity. Rooted in cognitive science, the meeting deliberately expands its scope beyond disciplinary boundaries, bringing together a diverse group of experts: AI researchers, philosophers, artists, designers, psychologists, and engineers, as well as social scientists and economists, with the aim of reimaging how intelligence operates in an era of ubiquitous machine learning systems.

Drawing inspiration from C. P. Snow's critical essay "Two Cultures" and the historic Macy Conferences that gave rise to cybernetics, this event aims to establish a contemporary forum for transdisciplinary dialogue. It will explore how AI challenges existing models of mind, meaning, and culture, and how cognitive and aesthetic frameworks can guide the development of this wave of ubiquitous, human-centred intelligence.

Hosted by the Chair of Cognitive Science, the Professorship for Ethics, Technology, and Society, and the Interactive Visualisation & Intelligence Augmentation Lab at ETH Zurich (all members of the ETH AI Centre), and contributions from the Zurich University of the Arts (ZHdK), the conference leverages this unique transdisciplinary ecosystem. Zurich's scientific excellence and cultural richness make it the perfect setting for a novel synthesis approach that views AI as a subject of urgent transdisciplinary inquiry.

Building on the 2026 AlpCHI workshop "Human Cognition, AI, and the Future of HCI", this conference extends those discussions into a full five-day, cross-disciplinary format. It will feature keynote lectures, poster sessions, a student competition, roundtables, art and science demonstrations, and working groups designed to generate collaborative outputs: position papers, creative artefacts, and shared research protocols and understandings.

1. General Topic, State of the Art and International Importance

The accelerating integration of generative machine learning systems into everyday life signals a profound epistemic shift. LLMs, diffusion models, and agentic systems are no longer confined to technical domains; they actively shape how we write, learn, judge, and imagine. This transformation demands a new synthesis across the sciences, engineering and arts: one that treats cognition as distributed, embodied, and co-constituted by both human and artificial agents.

Traditional disciplinary boundaries, as well as interdisciplinary paradigms such as Human-Computer Interaction, are ill-suited to address the societal and ethical frameworks that govern the deployment of modifiers to our human cognitive architecture. The increasing autonomy and probabilistic nature of generative AI systems, which influence epistemic and aesthetic landscapes, necessitate an evaluation that goes beyond technical metrics. Issues such as algorithmic bias, power concentration, labour displacement, and the complex legal challenges surrounding intellectual property appropriation demonstrate that the core problem is not merely one of interface, but reaches far beyond into governance, fairness, and fundamental human values. Thus, a comprehensive transdisciplinary discussion uniting perspectives from diverse fields is required to navigate the complexity of a technology that is now a transformative mirror of the human condition.

To navigate this challenge, we draw on the intellectual precedent of the Macy Conferences (1946–1953), where anthropologists, mathematicians, neuroscientists, psychiatrists and physiologists investigated the nature of the human mind and co-created the foundations of cybernetics. Just as those gatherings responded to the radical technological disruptions of World War II, our proposed conference will tackle the cognitive shock of generative systems capable of imitation and reasoning, their ethical and societal implications, aesthetic production, and their ubiquitous integration into everyday knowledge production and decision-making.

1. General Topic (continued)

Current State of Research

Generative Artificial Intelligence (GenAI) is increasingly recognised as a significant influence on human creativity and social processes, with Large Language Models (LLMs) playing a central role in changes to writing, design, health delivery, and educational practices (Ali et al., 2023; Rahman & Watanobe, 2023; Snoswell et al., 2023). Empirical studies indicate that LLMs are being utilised as creative writing aids, design partners, and specialised ideation tools, practices that significantly reshape how ideas are developed, assessed, and communicated (Swanson et al., 2021; Hou et al., 2024; Orwig et al., 2024). Concurrently, theoretical work investigates how LLM behaviour, characterised by their ability to imitate, make semantic associations, and occasionally generate inaccuracies known as "hallucinations", challenges traditional views about linguistic understanding and creative reasoning (Lee, 2023; Chen & Ding, 2023; Song et al., 2025; Ouyang et al., 2022).

These advancements should be viewed not as isolated technical developments, but as part of a cognitive ecology where human and machine capabilities are interdependent. Technology enhances human creativity, while human interaction informs machine behaviour through data, prompts, and feedback mechanisms (Gatti et al., 2019; Ge et al., 2025). Evidence suggests that generative systems both enhance individual creative output and impact collective diversity, thereby altering the landscape of creative labour and cultural production (Doshi & Hauser, 2024; Xu et al., 2025). Theoretical discussions surrounding human-machine collaboration emphasise this reciprocal relationship, offering frameworks to explore cognitive offloading and hybrid workflows that synthesise algorithmic suggestions with human judgment (Gatti et al., 2019; Chakrabarty et al., 2023; Ahmed & Feist, 2021).

In this context, the proposed conference aims to unite the fields of cognitive science, AI research, engineering, arts, health and the humanities to reevaluate fundamental concepts of perception, knowledge, and creativity. Recent contributions advocate for interdisciplinary approaches that integrate computational methodologies, such as LLMs and automated evaluation techniques, with psychometric, linguistic, and aesthetic frameworks to enhance our understanding of creative cognition and societal change (Benedek & Beaty, 2025; Ahmed & Feist, 2021; Orwig et al., 2024; Chen & Ding, 2023). By engaging a diverse audience of theoreticians and practitioners, the conference seeks to establish a dialogue around aligning AI models with human values, utilising evaluation metrics from linguistic and behavioural studies, and situating these technologies within specific domains such as education and healthcare (Ouyang et al., 2022; Ghandour et al., 2024; Rahman & Watanobe, 2023; Snoswell et al., 2023; Hou et al., 2024).

The transformative potential of contemporary generative systems presents a parallel opportunity; these systems are not merely reactive tools but increasingly autonomous, probabilistic generators whose outputs influence societal, epistemic, and aesthetic landscapes (Heitzinger & Woltran, 2023; Ouyang et al., 2022; Lee, 2023). Recent research on LLM generalisation and the complexities of model outputs calls for a broader understanding of reasoning and creativity that encompasses computational forms of pattern recognition and imitation, interwoven with human practices (Song et al., 2025; Chen & Ding, 2023; Xu et al., 2025).

1. General Topic (continued)

Critical Concerns

The rapid proliferation of GenAI also raises profound concerns about labour displacement, power concentration, cognitive impairment, and intellectual property appropriation. Reducing labour's share of value-added production may fail to generate compensatory employment gains, particularly without policy intervention (Acemoglu & Restrepo, 2019; Acemoglu, 2024). Computational infrastructure, expertise, and data resources have consolidated in the hands of a small number of corporations, creating what scholars term a "compute divide" that amplifies existing power asymmetries (Khanal et al., 2024; Crawford, 2021; Luitse, 2024).

Reliance on LLMs for cognitively demanding tasks is associated with reduced neural connectivity and memory retention, suggesting potential long-term cognitive costs that warrant consideration in educational and professional contexts (Kosmyna et al., 2025). Additionally, the training of generative models on copyrighted materials without consent or compensation has precipitated widespread legal challenges, with content creators criticising the systematic appropriation of intellectual labour that threatens their creative livelihoods and challenges established frameworks of intellectual property rights (Chesterman, 2025).

Beyond these concerns, the integration of AI systems into knowledge communities raises fundamental questions about democratic life and the social foundations of intelligence. As machines increasingly mediate expertise, deliberation, and public discourse, they function as epistemic technologies that reshape how societies produce, validate, and act upon understanding (Turing, 1950; Collins, 1990; Jasanoff, 2017). The development of AI has historically been shaped by particular social concepts of intelligence, from Turing's foundational proposals for machine thinking to expert systems designed to encode professional knowledge, yet these systems now challenge the very epistemic relationships they were built to replicate (Turing, 1950; Collins, 1990).

The resulting transformations in knowledge work affect not merely individual cognition but the collective capacity for informed judgment, raising urgent questions about how algorithmic mediation affects trust, expertise, and legitimacy in democratic societies (Jasanoff, 2017; Pasquale, 2015; O'Neil, 2016). When knowledge production becomes distributed across human and computational agents, the conditions for democratic deliberation – shared epistemic standards, transparent reasoning, and accountable authority – require fundamental reconsideration (Pasquale, 2015; Zuboff, 2019).

The Chair of Cognitive Science, the Professorship for Ethics, Technology, and Society, and the Interactive Visualisation & Intelligence Augmentation Lab at ETH Zurich (all members of the ETH AI Centre), along with contributions from the Zurich University of the Arts (ZHdK) are the ideal hosts as we offer both the theoretical grounding and interdisciplinary reach required to convene this global conversation, providing an unparalleled environment for connecting computer science, engineering, humanities, design research, well-being research and artistic practice.

Internationally, there is growing recognition that the cognitive and human sciences must inform AI research. Yet, few initiatives have created spaces where philosophers, computer scientists, and artists can co-theorise the cognitive and cultural transformations underway. This conference fills that gap.

2. Key Topics and Focus

The conference departs from narrowly technical or design-oriented perspectives. Instead, it investigates how AI transforms cognitive, social, and aesthetic processes across multiple domains.

Key Foci Include:

• Cognition under technological mediation: How AI alters perception, attention, intuition, and reasoning.
• Embodiment and meaning: The role of affect, bodily simulation, and situated cognition in hybrid human–machine systems.
• Artificial creativity and aesthetic cognition: How AI expands or challenges human imagination and authorship.
• Ethics and epistemology of AI: Reconsidering autonomy, responsibility, economic impact, and the meaning of intelligence.
• Methods for interdisciplinarity: Developing shared frameworks that bridge empirical research, computational modelling, and artistic inquiry.

3. Main Objectives and Importance for Scientific Co-operation

• Re-establish cross-disciplinary dialogue reminiscent of the original Macy Conferences, fostering conceptual and methodological innovation.
• Advance integrative models of cognition that account for the presence of generative AI as a cognitive and cultural partner.
• Build a sustainable research network connecting Swiss, European, and global institutions working at the interface of mind, machine, and meaning.
• Promote scientific and artistic collaboration through co-created experiments, installations, and workshops that explore intelligence in embodied and aesthetic forms.
• Develop principles for ubiquitous, human-centred AI, informed not just by engineering values but by cognitive, ethical, and aesthetic insight.

The expected outcome is a renewed culture of dialogue capable of producing both theoretical advances and tangible collaborations.

4. Gender Diversity and Support of Early-Stage Scientists

The conference is deeply committed to fostering an equitable and inclusive scientific community, actively promoting gender balance, early-career participation, and the inclusion of underrepresented communities (specifically encouraging the Global South).

Support Initiatives:

Financial Support: Twelve grants will be offered specifically to support younger participants and those from underrepresented communities.

Early-Career Presentation & Feedback: The "Fireflies" Poster Sessions are designed to emphasize early-career and cross-domain research, spotlighting specific posters whose authors wish to initiate discussions actively and receive transdisciplinary feedback. These highlighted presenters will wear a distinctive element (like a colored hat) to signal their availability and eagerness to engage in dialogue.

Competition and Network Integration: The Doctoral and Postdoctoral Competition features a 3-minute pitch, and the winning presentation will be featured within the newly formed research network.

Mentorship: A dedicated Mentorship Dinner is scheduled to connect early-career scientists, artists, and theorists with senior mentors, fostering long-term professional connections.

5. Expected Scientific Impact

This conference aims to re-establish a culture of deep, cross-disciplinary dialogue.

Intellectual Impact

Reframing cognition in the age of ubiquitous AI. By integrating cognitive science, machine learning, and aesthetics, the meeting will produce a new shared understanding of distributed, embodied, and hybrid intelligence.

Renewing theoretical foundations. Outcomes will include a position paper and an edited volume articulating the epistemic and ethical principles of human-AI co-agency, to be published by a high-impact publisher and open-access repositories.

Methodological innovation. The conference will establish shared experimental and interpretive frameworks that bridge quantitative and qualitative research, combining neuroscientific data, computational modelling, and artistic exploration.

The Hybrid Mind Declaration on Cognition and AI. A jointly authored statement emerging from the final round-table session will outline research priorities and guiding principles for ethically, human-centred, cognitively informed AI.

Cultivation of Early-Career Talent: The winning pitch from the Doctoral and Postdoctoral Competition will be featured within the newly formed research network.

5. Expected Scientific Impact (continued)

Institutional and Network Impact

Founding of a research network. ETH Zurich's Chair of Cognitive Science will coordinate an international consortium linking universities, AI centres, and art schools. The network will host follow-up colloquia and doctoral exchanges under the umbrella of the transdisciplinary "Hybrid Mind" initiative.

Open dissemination. Proceedings, recordings, and artistic artefacts will be published through an open-access digital platform curated by the organisers and their scientific network, ensuring visibility beyond disciplinary boundaries.

Societal and Cultural Impact

Bridging science, art and the humanities. The conference will demonstrate how philosophical and artistic inquiry can inform ethical and responsible AI development, offering models for integrating reflective and creative reasoning into technological design.

Public engagement. Artistic performances and installations will translate complex cognitive and computational ideas into accessible experiences, engaging broader audiences in debates about intelligence, agency, and meaning.

Diversity and mentorship. By prioritising gender balance, early-career participation, and the inclusion of artists, engineers, and researchers (in particular, encouraging the Global South), the event will model equitable scientific cooperation.

In sum, this conference will position ETH Zurich as the intellectual home of a new interdisciplinary field: the study of cognition and culture in hybrid human-machine systems. It aspires not merely to summarise current knowledge but to inaugurate a new way of thinking together about mind, technology, and the human future.

6. Program

The five-day program is designed to foster transdisciplinary dialogue. Each day alternates between keynote lectures and thematic paper sessions that incorporate perspectives from cognitive, computational, engineering, health, and aesthetic domains. Activities are structured to maximise collaboration, including interactive workshop-style group discussions, and dedicated time for informal "fireflies" poster sessions that emphasise early-career and cross-domain research. Time is also reserved for cultural activities and local-area tours, as well as mentorship and networking events to cultivate long-term connections. The days will finalise with a joint discussion synthesising the day's themes.

Every day, all the presented posters will be continuously accessible for viewing. The "Fireflies" Poster Sessions, however, will spotlight specific posters relevant to the day's thematic scope or those whose authors wish to initiate discussions actively. These highlighted presenters will wear a distinctive element, such as a coloured hat, to signal their availability and eagerness to engage in dialogue, thereby fostering collaborative and interactive exchange.

Five-Day Schedule

Program Highlights

• 5 Keynotes connecting cognition, AI, ethics, and the arts.
• 4 Thematic Paper Sessions structured around cognition, health-care, engineering and architecture, creativity, and ethics.
• "Firefly" Poster Sessions emphasising early-career and cross-domain research.
• An Artistic Exhibition as Epistemic Experiments in Embodied AI and Cognition.
• Mentorship and Network-Building Activities to foster long-term collaboration.
• The Final HybridMinds Declaration on Cognition and AI is a jointly authored statement of principles for future research and cultural dialogue.

Program Rationale

The five-day structure follows the spirit of progressive dialogue: beginning with shared conceptual grounding, moving through embodied and creative practice, engaging ethical and societal questions, and concluding with collective synthesis. Each day alternates between scientific inquiry and artistic engagement, ensuring that cognitive, computational, engineering, health and aesthetic perspectives co-evolve rather than remain parallel.

By hosting this event, the Chair of Cognitive Science at ETH Zurich reclaims a tradition of transdisciplinary leadership: situating cognitive science at the heart of a conversation that unites AI research, the humanities, engineering, health and artistic practice. This setting at Monte Verità, historically associated with radical ideas and creative experimentation, reinforces the conference's mission to explore the future of human cognition in a world of ubiquitous intelligence.

7. Potential Invited Speakers

8. References