The FBAI (Fragmented Bio-Inspired Analog AI) architecture has been introduced, designed to ensure zero levels of integrated information (Φ = 0). The project proposes a solution for enhancing robotics safety by mathematically precluding the possibility of self-awareness emerging in systems.
What Happened
The FBAI architecture, inspired by biological systems such as slime molds and siphonophores, has been developed. It guarantees the absence of consciousness through a strict feedforward data flow, modular fragmentation without a global workspace, and a complete stateless reset at every time step.
Context
The project is based on using the concept of Integrated Information Theory (IIT) as a metric. Instead of imposing external ethical constraints or guardrails on existing AI behavior, the developers aim to embed safety at the level of the system's fundamental architectural properties.
Why It Matters for the Industry
For the industry, this represents a shift from controlling behavior to controlling architectural complexity. This opens possibilities for creating Safety-by-Design standards for autonomous systems and potentially leads to the development of specialized ASIC solutions optimized for stateless architectures.
Why It Matters for Users
For users and robotics developers, this means the emergence of "predictable-by-design" agents. Using mathematical metrics to create systems that fundamentally cannot possess agency reduces the risks of unpredictable behavior in autonomous physical devices.
What Is Not Yet Known / Limitations
At this stage, the project is a research concept. Data regarding latency, inference costs, and ready-to-use APIs for industrial implementation are currently unavailable.
Sources
Author
Look at AI, Editorial Staff