HumAIn Dialogues lecture: Mahault Albarracin

HumAIn Dialogues has invited Mahault Albarracin to give a presentation on Thursday April 25, 2024.

Mahault Albarracin is a distinguished figure in the fields of cognitive science, psychology, and technology, with a particular interest in active inference, gender fluidity, affordances, and scripts. Currently pursuing a Doctorate in Cognitive Computing at the Université du Québec à Montréal (UQAM), Albarracin has made significant contributions to research and product development, intertwining academic insight with practical applications in technology and product management.

Before embarking on her doctorate, Albarracin completed a Bachelor's degree in Sexology (2011-2015) and a Master of Arts in Sexology (2018) at UQAM. Her academic journey is complemented by professional certifications, including a React Bootcamp at Collabo Academy and a Web Development Bootcamp at Udemy Alumni. This educational foundation has paved the way for Albarracin to explore and contribute to the understanding of human cognition, behavior, and societal impacts through technology.

Albarracin's career trajectory reflects a blend of research prowess and leadership in technology and product development. She began her career with roles that spanned web development and scientific research, progressively taking on more significant responsibilities. In 2021, Albarracin was appointed Head of Product Innovations & Research and Research Scientist on computational phenomenology at VERSES. 

Her work in computational phenomenology and cognitive computing stands out for its innovative approach to integrating theoretical knowledge with practical applications. Albarracin's research has been published in peer-reviewed journals and she has been involved in discussions and developments at the intersection of technology, cognitive science, and social impact.

Abstract: "From Consciousness to Social Groups: Bridging Scales through Active Inference"

This talk presents a comprehensive journey from the foundations of consciousness to the intricate dynamics of social groups, unified through the lens of active inference. We begin with the "inner screen model" of consciousness, exploring how the brain's hierarchical structure underpins our conscious experience through models such as the Bayesian global workspace, attention schema, and self-world models. This framework offers a novel perspective on consciousness as an emergent property of active inference processes, emphasizing the recursive nature of agent-environment interactions.

Transitioning to the social domain, we explore how these principles extend to the formation of epistemic communities within social networks. Here, confirmation bias and the active sampling of information lead to the emergence of echo chambers, demonstrating the role of active inference in shaping social cohesion and ideological segregation. This segment highlights the interplay between individual cognitive mechanisms and social structures, elucidating how beliefs are formed, reinforced, and polarized within groups.

Finally, we envision a future where ecosystems of intelligence, comprising both natural and artificial agents, coalesce around shared narratives and goals. Grounded in the active inference framework, this vision suggests a blueprint for collective intelligence that transcends the boundaries of individual agents. We propose a path towards designing intelligent ecosystems that harmonize human and artificial intelligences, fostering communication protocols that enable belief sharing and curiosity-driven exploration.

Throughout this talk, we weave a narrative that showcases active inference as a fundamental principle connecting individual cognition with collective behaviors. By examining consciousness, social dynamics, and the potential for integrated ecosystems of intelligence, we illuminate a pathway towards understanding and designing complex systems where agents, whether biological or artificial, form specialized sub-clusters and engage in higher-level goals as a unified ecology.