Organoid Array Computing: The Design Space of Organoid Intelligence

In this paper, we explore the artificialization and networking of biological via brain organoids—three-dimensional, stem-cell-derived structures that recapitulate aspects of human brain architecture and function. These organoids serve as a platform for investigating the emergent properties of biological neural networks and the potential for developing an in-vitro to in-silico cognitive architecture. Our research addresses the burgeoning field of organoid intelligence (OI), wherein biological substrates are interfaced with computational systems, providing an adaptive framework for embodied computation. A common distinction between software and hardware in the field of biocomputing assumes DNA as software and cells as hardware. By evolving through biochemical and physical signaling feedback, organoids challenge this dichotomy. OI integrates both, enabling biological systems to move along the continuum from software to hardware into a multiscale machine. We begin by examining the current interfacing technologies that enable the connection between organoids and digital systems, evaluating the proof-of-concept studies that have laid the groundwork for OI applications. This analysis includes a critical assessment of the existing practical and technical limitations that hinder the realization of scalable OI. We then propose design strategies aimed at overcoming these obstacles, emphasizing the need for a nested approach to experimental design. New permutations enable the iterative development of OI modules, facilitating the integration and application of polycomputational neural assemblies. The design space of OI focuses on the growing dimensions and analysis of inputs, outputs, interfaces, and frameworks across multiple scales. We posit that the design of OI is less an act of top-down design and more a process of guided evolution, wherein higher-order cognitive functions emerge organically from the intricate interplay of lower-level biochemical substrates. Through this, we speculate on how higher-order functions can emerge from networking biological matter from embedded substrates “downstream”. Our research aims to uncover new dimensions in the information-processing capabilities of OI, positioning OI as a novel form of AI.