See all 'Engineer In Wonderland' postsI am fascinated by the potential of biomimetics.
Years ago I fell into conversation with an academic biologist.
He told me that if you want to make stuff, find an enzyme to do it because enzymes, which are nature's catalysts, do things very well - generally far more effectively than simple chemicals or thermal processes.
They work by pushing just the right electro-chemical buttons to make reactions happen - putting a key in the lock rather than breaking the door down.
Enzymes cannot be designed from the ground up for specific tasks at the moment because most of them are proteins and 'the protein folding problem' - what shape any given protein will end up, and where its charges will be - has yet to be solved.
This is not because no one knows how to do it, but because there isn't enough free computer time in the world to have much of a stab at it.
IBM's original
Blue Gene supercomputer was specifically designed to model protein folding.
Destined to be the most powerful computer ever made, it seems to be one of the great IBM 'hey, this would be interesting and might do a lot of good' projects.