Samuel Butler published the satirical novel “Erewhon; Or, Over the Range” in 1872. Buried in its account of an isolated fictional society are three chapters known as “The Book of the Machines,” which present an argument the Erewhonians had used to justify banning all advanced machinery: that machines were undergoing their own evolution and might one day develop consciousness and supplant humanity.
Butler applied Charles Darwin’s theory of evolution, published just over a decade earlier, to mechanical devices. If species evolve by descent with variation, he reasoned through his characters, why should machines not do the same, growing more complex and capable generation by generation until they no longer needed people at all? The Erewhonians had concluded the only safe response was to destroy their machines before the machines could outgrow them.
The Book of the Machines grew out of an earlier Butler article titled “Darwin among the Machines,” written in 1863. In the novel’s prefaces Butler insisted the chapters were not meant to ridicule Darwin, whose work he admired, but to push an analogy to its unsettling conclusion. The text is freely available on Project Gutenberg.
Why business readers should care: the idea that machine capability follows an evolutionary, self-accelerating curve - and that humans should intervene before it runs away - is a Victorian invention, not a product of the modern AI-safety debate it now feeds.