Imitation of Life: How Biology is Inspiring Computing

  • Nancy Forbes
MIT Press: 2004. 176 pp. $25.95, £16.95 0262062410 | ISBN: 0-262-06241-0

Generations of engineers have recognized that, in many respects, biology does it better. Imitation of Life is a whirlwind history, richer even than its subtitle suggests, through various computational disciplines inspired by biology. This is an ambitious undertaking for such a short book but, although it ignores some important unifying principles, its brevity is also a virtue. The inspirations from biology are scattered throughout the book, and their collective impact is felt best when the book is digested whole, at one sitting. The early chapters on biology as a metaphor are the least satisfying, and any reader who stops there may never return for the genuine delights that follow.

Credit: ILLUSTRATIONS BY CHRISTIAN DARKIN

Some historians argue that the inspiration provided by avian feathers and flapping may have delayed heavier-than-air flight for centuries. Critics would argue that artificial neural nets, genetic algorithms, cellular automata and artificial life, which are covered in the first four chapters, are the modern equivalents of flapping, whereas advocates see them as mature fields no longer in need of review. Nancy Forbes is generous about their successes, but does little to resolve the issue. Only later are there brief discussions of hierarchy, modularity, layers of control, and system architecture — key concepts in computing that could help to inform a more thorough analysis.

In contrast to the earlier sections, which use biology as a metaphor, the chapters on DNA computing and biomolecular self-assembly describe the direct use of biological chemistry or materials to create technological artefacts that have little or nothing to do with biology. Forbes is clearly more interested in these topics, and this enthusiasm may well spread to readers; the section on the intriguing computational power in the organization of DNA is particularly well presented. These chapters start to make it clear that understanding biological principles in some depth is an essential part of profitable imitation. Making DNA computing work requires a firm grasp of the principles and careful design, but now anyone can download cellular automata or genetic algorithm software and run laptop artificial-life experiments.

The chapters on amorphous computing, computer immune systems and biologically inspired hardware further underscore the idea that, as biology is better understood, inspiration can proceed more from mimicry than metaphor, and contribute more directly to solving difficult computational tasks. What makes this work (and these chapters) more compelling is the fact that engineers in these fields have crossed disciplinary lines to gather a deep and practical understanding of biology and biological experimentation. The recent explosion in our detailed knowledge of biology, and the glimpses this provides of its organizing principles, has considerably enriched biologicaly inspired computing.

The final chapter reverses direction and looks at biology through the lens of computer science and electrical engineering. This is conceptually the deepest chapter but its brevity limits it to a few well-chosen examples. The technologies of any age have always provided metaphors for biology — from myths about our origin involving dust and clay to the industrial revolution's hydraulic and, later, electromagnetic imagery, from telegraphs to telephones and finally to computers. The book contains almost no explicit discussion of complexity, and this omission is particularly noticeable here. Much of computer science is about organizing complexity, from ‘very large-scale integrated’ circuit design to object-oriented programming and the layered protocols of the Internet.

The history of flight is again instructive. By the nineteenth century, engineers had realized that lift, drag and propulsion were the key fundamental mechanisms, and toy gliders became commonplace. Yet only with the Wright brothers' insight that active control was needed for steering and compensation for uncertainties did flying literally take off. By the 1940s, unpiloted aircraft had demonstrated fully automatic transatlantic flight and landing, and some engineers now argue that flight would generally be safer without pilots.

Similarly, the vast majority of computers now are ‘embedded’, with automated sensing, control and actuation, all entirely hidden during normal operation. Computer control systems such as these represent both the main use of computers and the main source of complexity in technological systems, but are barely mentioned either in this book or elsewhere. This is a pity, as they are the points of greatest contact between engineering and biology. Biologists would have benefited from a discussion of sensing and adaptation in computation and networking, such as Internet routing and congestion control, because sensing and adaptation are widespread in biology, for example in the immune system. Perhaps biologically inspired computing is not yet at a ‘Wright brothers’ stage, with many fundamental mechanisms just emerging, old superficial metaphors being set aside, and systems-level integrating concepts remaining murky. Hopefully, the Wright brothers of biologically inspired computing are among the many fascinating characters described by Forbes, and their subjects will take off as promised.

This book is easily accessible but is probably most suited to, and beneficial for, biologists, as a clearly written, non-technical primer describing activities on the other side of campus. Biologists will have to ignore some unfortunate simple errors (such as neurons called axons, AIDS as an autoimmune disease and vaccines made from weakened antigens) but can easily read around them. Nonetheless, this text helps to bridge a daunting technical language barrier and should facilitate further dialogue between biologists and computer scientists.