5 ways Charles Darwin influenced tech
5th Feb 2009 | 16:51
200 years on, Darwin's legacy is in the tech we use daily
With 12 February marking 200 years since the birth of Darwin, much is rightly being written about his profound impact on biology.
Less appreciated has been his impact on the world of digital technology, particularly computing and robotics. Here are five ways Darwin's influence extends into tech:
1. Genetic algorithms
Dating back to the 1950s – about the time computers became fast enough to generate useful results – these sought to model real world evolution on a computer.
It was soon realised that they could be applied to a far wider range of problems.
A genetic algorithm works in a way analogous to natural selection. The short version: a bunch of possible solutions to a problem is randomly generated. Successful (fit) solutions are chosen and allowed to reproduce, while the unsuccessful are discarded.
And so on, until a copy is found that comes closest to solving the problem.
This has turned out to be a great way of solving difficult problems or creating entirely original designs, across a huge number of applications, from stocks and shares to videogames.
2. Videogames design
Darwin's influence on gaming goes right back to the 1960s: one of the very first programming games was called Darwin, and involved victorious programs copying themselves into vacant memory.
Sounds dull, but this was the inspiration for the later and more influential hacker game Core War.
More recently, there have been influences both in the object of a game (Spore) and in the techniques used to generate a game's world (Creatures).
Spore doesn't always let scientific accuracy stand in the way of great gameplay, of course, but Creatures deliberately modelled itself on biological reproduction:
As Steve Grand, Creatures' designer, puts it, "Creatures doesn't use a Genetic Algorithm - GA's are a rather abstract thing; Creatures uses something far more similar to real genetics in real organisms."
Programming robots to perform useful tasks can often be a slow, error-prone process – even more so should the robot be physically modified or upgraded, requiring reprogramming.
This approach enables robots to adapt to upgrades by evolving new control systems on the fly. In this example, a robot previously equipped with peg legs learned how to use its new jointed limbs without being taught.
- Circuits can be evolved that outperform conventional circuits by up to 100%.
- USB memory sticks have been evolved that are far more efficient than human designed versions, boosting the life of flash memory by up to 350%.
- Cochlear implants have been developed where the correct configuration of electrodes is evolved rather than done by hand - returning hearing to one patient whose doctors had not succeeded in 10 years.
- And even police photofits are getting the Darwin treatment by generating facial variations the witness can then select and fine tune. In tests, witnesses were twice as likely to recognise faces produced in this way, than via conventional mugshots.
5. Manufacturing and aerospace
Volvo has used an evolutionary program to schedule complex manufacturing jobs, reducing the time taken to create the schedule from four days a week, to just one.
In Scotland, a similar program has been used to manage the storage and supply of 7 million barrels of whisky, a task previously requiring five people.
And software based on genetic algorithms now helps manage the US air traffic control system.
Each new application is a further crushing of the idea that evolution is unproven or can't be tested – unequivocal proof of evolution in action and the continuing power of Darwin's ideas. Happy 200th birthday.