The Digital Labyrinth

Why are airliners, nuclear reactors and suspension bridges so dependable?
Why are large computer systems so unreliable and expensive to build?
Why do successful businesses grow to become unmanageable?
Why are some complex systems intrinsically easier to design than others?

The Digital Labyrinth is a forthcoming book which attempts to answer these fundamental questions about the nature of effective design.

Design is about the problem of searching a vast space of possibilities - a 'fitness landscape' - to find the optimal solutions occupying the peaks of fitness. However, the ruggedness of the fitness landscape is radically different for different problem domains - mechanical engineering and computer programming for example. The book explores how the character of design is determined by the underlying search space, and why some design problems are inherently easier to solve - irrespective of their apparent complexity.

Good design is about building systems able to adapt to changing requirements. Darwinian selection has built biological systems that are resilient and adapt fluidly and dynamically. Computer systems on the other hand are notoriously brittle and may collapse if a single bit of the program is mutated.

Recent results in complexity theory and evolutionary biology provide key insights into how nature has solved the design problem. Evolutionary pressure drives biological systems to the very edge of chaos, where their ability to adapt is maximised. The structure of such optimal systems provides vital insight into how more robust and adaptive solutions for many problems could be built - from software architectures, business organisational structures and even the political structure of our society.

Click here to see a detailed chapter list.