SNU Center for AI ELSI (Director: Hyundeuk Cheon) exploring the ethical, legal, and social issues surrounding AI is hosting a public seminar like below. This seminar will focus on whether AI can possess consciousness, featuring discussions from computer scientists, neuroscientists, and philosophers.
-Topic: <Artificial Intelligence and Consciousness – Scientific, Engineering, and Philosophical Approaches>
-Date & Time: July 1, 2025 (Tuesday), 2:00–4:00 PM KST
-Venue: Mokam Hall, Building 500, Seoul National University
-Pre-registration: https://forms.gle/uqg6uiofyiWTC6Tz6
Speakers & Presentations
1. Yoonseok Choi (Texas A&M University, Computer Science) – "Sensation, Meaning, Time, and AI"
* (Abstract) As artificial intelligence technology rapidly advances, discussions on whether AI can possess consciousness and experience emotions like humans have become increasingly active. Alongside this, an equally important question arises: Can AI go beyond mere information processing and pattern recognition to understand data in a way similar to humans? These two questions are closely interconnected, with the core issue being whether AI—or the brain—has something beyond information processing, such as sensation and meaning. Additionally, both questions share a common factor: the role of sensory-motor aspects and temporal dynamics in AI consciousness. This presentation will provide a concise overview of these topics and present findings from computer simulation experiments
2. Kyumin Moon (Hankuk University of Foreign Studies, Philosophy) – "X for Artificial: the Hidden Conceptual and Biophysical Variables"
* (Abstract) Thanks to the rapid advancement of AI, the discussion on artificial consciousness, once confined to science fiction, has now become a serious academic topic. The question of whether AI can go beyond mere computation to experience has shifted from speculative debate to active research and development. This presentation explores the conceptual and biophysical conditions necessary for artificial consciousness. First, it examines the philosophical debate between realism and illusionism regarding the subjective and qualitative aspects of consciousness, particularly qualia. Through this, it seeks to establish a minimalist and neutral concept of subjectivity that could serve as a reference for artificial consciousness development. Furthermore, by integrating insights from affective neuroscience, predictive processing theory, and biophilosophy, the presentation investigates the biophysical conditions of consciousness. Interestingly, researchers across these fields, despite their differing theoretical foundations, converge on a common skepticism: given the current trajectory of AI development, the likelihood of AI achieving true consciousness or subjective experience remains low. These hidden conceptual and biophysical variables, largely overlooked in AI research, are referred to as "X for Artificial Consciousness." While conceptual variables outline the necessary conditions for artificial consciousness, biophysical variables suggest that beyond embodiment and agency, factors such as energy metabolism and electromagnetic phenomena must be considered in greater depth.
3. Jungkyu Han (Seoul National University, Neuroscience) – "Reverberatory Dynamics: A Cybernetic Framework for Unifying Theories of Consciousness"
* (Abstract) The study of consciousness has led to the development of numerous scientific theories, yet their foundational differences have resulted in ongoing theoretical fragmentation, prompting this paper to take a historical and integrative approach to examine major scientific theories of consciousness to identify shared principles that could contribute to a more unified framework. By tracing their intellectual origins and critically examining the historical trajectory of cybernetics, I demonstrate how cybernetic thinking has shaped our contemporary understanding of consciousness, revealing how fundamental cybernetic principles—such as feedback loops, control mechanisms, and system organization—are embedded within leading modern frameworks, including Global Workspace Theory (GWT), Integrated Information Theory (IIT), and Predictive Processing (PP), highlighting both their points of convergence and divergence. Building on this foundation, I propose Reverberatory Dynamics as a unifying theoretical construct that integrates insights from cybernetics, neuroscience, and cognitive science, referring to the sustained, recurrent activity within neural circuits that enables information integration and persistence over time and, unlike existing models that emphasize information transfer or hierarchical processing, highlights the temporal continuity and self-sustaining nature of neural activity as a core mechanism underlying conscious experience. Consequently, this paper demonstrates that cybernetic principles continue to shape contemporary theories of consciousness and that a reverberatory framework may offer a novel path toward resolving theoretical fragmentation in the field.
-Hosted by: SNU Center for AI ELSI & Institute for Data Innovation in Science SNU
-Supported by: National Research Foundation of Korea
-Contact Information: Ha-Yeon Yoo Assistant Coordinator (snuaielsi@gmail.com)