Human Supremacy Alert : Urgently Repeating Myself for Slow Cats...,

PNAS |  Most technologies are made from steel, concrete, chemicals, and plastics, which degrade over time and can produce harmful ecological and health side effects. It would thus be useful to build technologies using self-renewing and biocompatible materials, of which the ideal candidates are living systems themselves. Thus, we here present a method that designs completely biological machines from the ground up: computers automatically design new machines in simulation, and the best designs are then built by combining together different biological tissues. This suggests others may use this approach to design a variety of living machines to safely deliver drugs inside the human body, help with environmental remediation, or further broaden our understanding of the diverse forms and functions life may adopt.  

ABSTRACT

Living systems are more robust, diverse, complex, and supportive of human life than any technology yet created. However, our ability to create novel lifeforms is currently limited to varying existing organisms or bioengineering organoids in vitro. Here we show a scalable pipeline for creating functional novel lifeforms: AI methods automatically design diverse candidate lifeforms in silico to perform some desired function, and transferable designs are then created using a cell-based construction toolkit to realize living systems with the predicted behaviors. Although some steps in this pipeline still require manual intervention, complete automation in future would pave the way to designing and deploying unique, bespoke living systems for a wide range of functions.

Most modern technologies are constructed from synthetic rather than living materials because the former have proved easier to design, manufacture, and maintain; living systems exhibit robustness of structure and function and thus tend to resist adopting the new behaviors imposed on them. However, if living systems could be continuously and rapidly designed ab initio and deployed to serve novel functions, their innate ability to resist entropy might enable them to far surpass the useful lifetimes of our strongest yet static technologies. As examples of this resistance, embryonic development and regeneration reveal remarkable plasticity, enabling cells or whole organ systems to self-organize adaptive functionality despite drastic deformation (1, 2). Exploiting the computational capacity of cells to function in novel configurations suggests the possibility of creating synthetic morphology that achieves complex novel anatomies via the benefits of both emergence and guided self-assembly (3).

Currently, there are several methods underway to design and build bespoke living systems. Single-cell organisms have been modified by refactored genomes, but such methods are not yet scalable to rational control of multicellular shape or behavior (4). Synthetic organoids can be made by exposing cells to specific culture conditions but very limited control is available over their structure (and thus function) because the outcome is largely emergent and not under the experimenter’s control (5). Conversely, bioengineering efforts with 3D scaffolds provide improved control (6⇓–8), but the inability to predict behavioral impacts of arbitrary biological construction has restricted assembly to biological machines that resemble existing organisms, rather than discovering novel forms through automatic design.

Meanwhile, advances in computational search and 3D printing have yielded scalable methods for designing and training machines in silico (9, 10) and then manufacturing physical instances of them (11⇓–13). Most of these approaches employ an evolutionary search method (14) that, unlike learning methods, enables the design of the machine’s physical structure along with its behavior. These evolutionary design methods continually generate diverse solutions to a given problem, which proves useful as some designs can be instantiated physically better than others. Moreover, they are agnostic to the kind of artifact being designed and the function it should provide: the same evolutionary algorithm can be reconfigured to design drugs (15), autonomous machines (11, 13), metamaterials (16), or architecture (17).

Here, we demonstrate a scalable approach for designing living systems in silico using an evolutionary algorithm, and we show how the evolved designs can be rapidly manufactured using a cell-based construction toolkit. The approach is organized as a linear pipeline that takes as input a description of the biological building blocks to be used and the desired behavior the manufactured system should exhibit (Fig. 1). The pipeline continuously outputs performant living systems that embody that behavior in different ways. The resulting living systems are novel aggregates of cells that yield novel functions: above the cellular level, they bear little resemblance to existing organs or organisms.