"An Enactive AI? Computing and Sense-Making Beyond the Data-Driven Approach"
Feb 13, 2025
auto_awesome
Luc Steels is a Professor Emeritus of Artificial Intelligence with a focus on enactive approaches, while Takashi Ikegami is a Professor at the University of Tokyo specializing in complex systems. They explore the intricacies of AI versus human cognition. Questions arise about participatory sense-making and the potential for enactive AI. The discussion critiques data-driven models, emphasizing the ethical dangers of AI and the human misuse of technology. They also highlight creativity's connection to humanity and reflect on experimental robotics that encourage interaction and understanding.
The conversation emphasizes that the self is a dynamic process rather than a static concept, impacting our understanding of identity in AI systems.
The speakers caution against uncritical AI adoption, highlighting how human biases influence algorithms and the need for ethical technology use.
Participatory sensemaking is vital for meaningful interactions with AI, advocating for collaborative approaches that enhance creativity and community dynamics.
Deep dives
Exploring the Concept of Self
The discussion revolves around the philosophical inquiry into the nature of self, arguing that self is not a static entity but an ongoing process. The speakers emphasize the importance of keeping questions about self open and not rushing to find definitive answers, suggesting that this approach leads to greater inner peace. The notion of oneness among humans adds to this understanding, highlighting how interconnectedness shapes our identity. The episode positions the concept of self as a dynamic inquiry relevant to both artificial intelligence and human cognition.
The Role of AI and Robotics in Understanding Cognition
The conversation introduces two esteemed researchers in AI and robotics, who diverge in their perspectives yet converge on core themes regarding participatory sensemaking. Key questions arise regarding the capabilities of robots in mimicking human-like cognition and the potential for AI to engage in sensemaking. The discussions delve into the nuances of participatory sensemaking and how it distinguishes between human activities and robotic processes. This exploration emphasizes the need to understand the implications of AI research without losing sight of the significant differences in cognition between humans and machines.
The Dangers and Biases in AI Systems
The speakers caution against the uncritical embrace of AI systems, highlighting the potential dangers posed by predictive algorithms, including the propagation of biases and misinformation. Emphasis is placed on the human influences that shape these algorithms, urging awareness of the ethical implications of technology use. Moreover, they draw parallels between AI challenges and historical issues, such as pesticide use, and advocate for a more cautious approach to deploying these technologies. The conversation stresses the necessity for responsible AI development that acknowledges and mitigates inherent biases present in the data.
The Future of AI and Participatory Sensemaking
In discussing the future of AI, the speakers underscore the importance of participatory sensemaking as a driving force in building meaningful interactions. They highlight the potential of AI to not merely replicate human intelligence but to offer a new kind of intelligent engagement that enhances creativity. The conversation reveals that understanding sensemaking requires an active participation that can yield evolving insights into community dynamics. By recognizing the importance of co-creating meaning in the development of AI, they advocate for a collaborative approach to technology.
Art, AI, and Experiential Learning at the Biennale
The speakers express excitement about their upcoming collaboration at the Venice Biennale, where they plan to present a robot integrated with participatory AI functionalities. This installation, titled 'Am I a Strange Loop?', aims to evoke discussions around consciousness, self-awareness, and the interplay between human and machine cognition. The goal is to showcase how experiential interactions can reshape perceptions of AI and challenge traditional narratives around technology. Through their artistic and scientific approach, they hope to inspire audiences to contemplate not just the functions of AI but its potential for deeper societal engagement.
With few conversations did I feel the stakes to be so high, so thorny and complex. For this conversation on “Computing Differently,” I sat down with Dr Luc Steels and Dr Takashi Ikegami, two of the world’s preeminent researchers in the fields of Artificial Intelligence, Artificial Life, and robotics — but researchers who come at the question of AI from a decidedly divergent perspective, that of the enactive approach and participatory sense-making. The conversation was one of not just defining the current stakes of AI research, but of considering the outer reaches of each guest’s thinking about AI and defining some of the intractable questions in the field today. How close does the apparent sense-making of a robot come to human sense-making? What defines participatory sense-making as a distinctly human activity? Can there be such a thing as an “enactive AI”? If so, what insights might it afford us about human cognition, and about AI itself? How are we to apply appropriate caution when discussing the current frontiers of AI research? Where should its priorities be? How can we grapple with the very real dangers of AI already at hand, such as the hypernormativity of predictive systems which propagate harmful biases and drive information pollution? As Luc Steels points out, it is not so much the AI systems that we ought to fear, but the human uses and misuses of them, and the exponential looping effect that takes hold between human and machine.
From our discussion of the basics of robotics and large language models emerged some of the most limpid definitions of participatory sense-making I’ve heard yet, and both speakers took great care in clarifying some of the basic terms of the discussion, which have too often been obscured by the popular media. Whether or not you feel you have a stake in the ongoing AI conversation, this conversation sheds light on so many of the fundamental questions of what it means to be a creative, enactive, and participatory being in the world today — in short, what it means to be human.