To realize this new technology, we must learn to engineer and control molecular self-assembly and reproduction as cells do. Present research has achieved important capabilities in specific areas; the most difficult challenge is integrating these together. In practical terms, this general challenge can be resolved into several strongly linked component challenges:
- 1. Realizing electronically programmable information chemistry: We require revolutionary advances in techniques for engineering and controlling whole-system chemistry at the micro-level (through means such as MEMS technology, synthetic biology, and so on). Such advances would give us an unprecedented ability to analyze, control and guide the synthesis of complex bio- and bio-mimetic chemical systems, thereby enabling radical new ways of information processing. Key technologies are required for the analysis, control, synthesis, adaptation, proliferation and delivery of myriad chemical systems, including complex microelectronics, micro-fluidics and microscopy, constructive information processing architectures, and fine-grained feedback control programming architectures.
- 2. Achieving evolvable and self-assembling molecular information processing: Traditional ICT generally exploits solid structures built into inflexible and unchangeable patterns, while living technology instead relies on soft, self-organizing structures, this flexibility being at the core of its advanced information processing capabilities. Realizing ICT based on artificial living technology therefore requires evolving self-assembled digital circuitry able to bridge the gap between these two physical settings. Nano- and microscale self-assembly using informational polymers (as in biological systems) already lies at the basis for new paradigms for molecular computing. When replicable informational molecules are employed, such systems can be synthesized economically, evolved and optimized. One challenge is to integrate coarser scale placement programming with self-assembly to reach higher complexities and information densities.
Key technologies/capabilities that must be developed include learning to control the evolvability of self-assembling systems, and to develop effective robust algorithms for dealing with both periodic and aperiodic self-assembly. We need to demonstrate examples of evolvable micro- and nanoscale components that self-assemble into functional collectives, and test/develop architectures for programming evolvable self-assembling systems.
- 3. Unraveling bio-molecular production and delivery: The bio-chemical system that supports the spatial packaging, sorting and deployment of proteins and collections of biomolecules in eukaryotic cells represents an extraordinary natural engineering achievement. Specific molecular recognition is complemented by a complex hierarchy of spatially self-organizing structures (from transport vesicles to endosomes and larger cell organelles such as the endoplasmic reticulum). While systems biology is currently exposing the intricacies of this system and its regulation, there is a need for an analysis of the information processing principles in this system and an abstraction of the architecture as a basis for synthetic information processing. This is an exciting challenge and one that will open the path to building ICTs with the capability to reconfigure their physical structures adaptively and on demand, thereby achieving information processing capabilities well beyond current ICT systems.
Key technologies/capabilities that must be developed include techniques for accurate spatial analysis of protein sorting and compartmental dynamics, for membrane computing and its systematic extension to self-assembling systems, and for multiscale simulation of complex soft-matter systems.
- 4. Radically new multi-scale programming tools: Existing hardware has its own machine language for which assembly language represents the addition of “syntactic sugar” that further enables the hierarchical establishment of high-level programming languages such as Pascal, C++, Java, etc. Living technology will require similar multiscale programming tools. That is, if we are to obtain the holy grail of “dial a pattern” or “dial a behavior” at the nanoscale we should be able to describe in a high level language what is it that the matter compiler or the living technology should achieve. This will require a hierarchy of new computational theories and programming languages that start close to the way humans think and move closer and closer to the way a living technology entity operates. This represents a fundamental challenge to the computer science community as a whole because the level of parallelism, uncertainty, noise and computing capacity at the bottom of the scale is enormously different than at the macroscale. Hence fundamentally new programming languages, compilers, simulators and theories will be required.