The amount of data being created every second is breathtaking (five trillion bits per second, as George Dyson tells us at ege.org).
Some take this as proof for a tremendous increase in knowledge, others spot only chatter and pocket litter - (and yes, there are estimates that half of the storage capacity - and hence half of the power consumed for 'the internet' - is used for porn).
Some see much more. Already in 1955, when the Norwegian-Italian mathematical biologist Nils Aaall Barricelli had heard about the machines operating on digital code in Princeton and finally managed to go there, he experimented with self-replication and cross-breeding of random strings on that machine. Being an experimentalist he observed carefully and studied patterns that reminded him of biological systems. Couldn't computer-code be treated similarly? Shouldn't there be the possibility of evolution in code?
Roger Dyson highlights the fact of universal operating systems across multiple hardware-platforms facilitating the occurence of multi-cellular operating systems that would be able to cross-breed, mutate, develop ... in the background.
While there is only speculation about the possibility that this could, might or would occur in an uncontrolled way there definitely is the option if not the clear reality of human-initiated 'meta-zoan' activity on the net. Probably most visible in the stealth-features of self-adapting malware.
But couldn't there be some damp, sparsely lit corners of the web where hitherto unidentified code-creatures linger and replicate - just like the amazing worms that curl up under those socks that you lost behind the washing-machine - years ago?