Reveal the informational nature of conscious self and sense of agency under the information closure theory of consciousness

揭示意识信息闭合理论下意识自我和能动感的信息本质

• 批准号: 22K20679
• 负责人: チャン ユーチャン
• 金额: $ 1.75万
• 依托单位: The University of Tokyo
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Research Activity Start-up
• 财政年份: 2022
• 资助国家: 日本
• 起止时间: 2022-08-31 至 2024-03-31
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-22K20679/
• 关键词: Consciousness; ICT; Mathematical Model; Neural Systems; Sense of Agency; Artificial Intelligence; Scale Problem; Simulation; information closure; Conscious Action

Research Update: We have successfully navigated the first phase of our project, establishing mathematical foundations of conscious self and agency within the Information Closure Theory (ICT) framework. By hypothesizing informational closure (IC) and non-trivial informational closure (NTIC), we've begun to understand how consciousness might be scale-dependent within the neural system. The second phase, presently underway, involves small-scale simulations. Early results are promising, with artificial agents displaying potential NTIC processes post-training. We're working diligently to determine if these processes encode not only environmental data but also information about the system's partition and action states. If successful, these results will provide a critical step towards applying our refined ICT framework to leading artificial systems in phase three, thereby contributing significantly to both neuroscience and ethical discussions in AI research.In addition to our progress in the first two phases, we've also made several significant breakthroughs in terms of the practical implications of our research. We've identified key neuroscientific insights that could have a profound impact on our understanding of the human brain and consciousness, and we've also uncovered potential ethical implications of our work with AI systems.

研究更新：我们已经成功地导航了项目的第一阶段，在信息封闭理论（ICT）框架内建立了有意识的自我和代理的数学基础。通过假设信息封闭（IC）和非平凡的信息封闭（NTIC），我们已经开始了解意识如何在神经系统中依赖于规模。目前正在进行的第二阶段涉及小规模的模拟。早期结果是有希望的，人造药物在训练后显示了潜在的NTIC过程。我们正在努力确定这些过程不仅编码环境数据，还要编码有关系统分区和操作状态的信息。如果成功的话，这些结果将为将我们精致的ICT框架应用于第三阶段的领先人造系统的关键步骤，从而在AI研究中对神经科学和道德讨论都有很大的贡献。除了我们在前两个阶段的进步外，我们还对我们的研究实践含义进行了一些重要的突破。我们已经确定了关键的神经科学见解，这些见解可能会对我们对人的大脑和意识的理解产生深远的影响，并且我们还发现了我们使用AI系统的工作的潜在道德含义。

项目成果

The Sense of Agency as active causal inference at an abstract action plan level.

代理意识是抽象行动计划层面的主动因果推理。

• 发表时间: 2022
• 作者: 松葉史紗子;西嶋翔太;西嶋 翔太;Acer Chang
• 通讯作者: Acer Chang