관심 있는 학생 여러분들의 많은 참석 바랍니다.

■ ABSTRACT
In the new age of AI, the open research challenges lie in exploring unknown to far-out ambitions and ideas that aim at shrinking the significant gaps between AI and other types of intelligences – most notably, human intelligence (e.g., common sense), social intelligence (e.g., reciprocity), and natural intelligence found in living organisms (e.g., plants). Futher, with the emergence of artificial general intelligence (AGI)-native 6G networks, AI embodiment will allow AI agents to transcend their purely computational nature and enable them to interact with their surrounding physical environment via sensory grounding through 6G, giving rise to autonomous goal-directed agentic AI and novel types of agent-environment interactions, e.g., human-AI interaction. Next-generation beyond 6G networks are anticipated to orchestrate the development of AGI systems in order to unleash a new “G” of mobile networks that will operate on an AGI-native telecom brain. Clearly, the use of an AGI-native telecom brain in next-generation networks opens the door for exploring insights from neuroscience to expand the creative potential of AGI network brains, especially in light of the nascent area of integrating AI and psychedelic medicine, two fields that may not have historically shared any common ground.
In this research seminar, we will elaborate on why some argue that the acronym AI might actually not stand for artificial intelligence, but active inference. Active inference is considered the key to true AI: Less artificial, more intelligent. Although active inference is still relatively young, it has a growing impact across various disciplines, e.g., human-AI interaction and HCI in general. Premised on first principles of physics, active inference reconciles and unites embodied, enactive, and cybernetic theories of brain, mind, and life in general. It is an ideal methodology for developing more advanced AGI systems by biomimicking the way living intelligent systems work, while overcoming the limitations of today’s AI and its profound lack of key elements found in living intelligent systems in terms of perception, memory, reasoning, and generating actions. Towards this end, we will first introduce the main concepts of the biomimetic active inference framework and show how to leverage it to move AI research beyond the human brain toward the emerging 6G World Brain. We will then delve into active inference’s dark-room problem, a recurrent puzzle faced by biological agents like us humans, and how to resolve it by advancing active inference via widening its scope from biological agents to psychoactive agents. We show that active inference points in the exactly opposite direction of today’s envisioned AGI-native 6G trajectory. The journey beyond 6G might not be outward via AI embodiment, but inward into what neuroscientists coined the brain’s dark energy, thereby not only resolving the dark-room problem but also helping us penetrate ever deeper into cognitive space.

■ HOST: 임완수 교수 (전자전기컴퓨터 공학과)