We're Not Computers 2: I Think, Sebastian, Therefore I Am

Use your new friend as a personal body servant or a tireless field hand. The custom tailored genetically engineered humanoid replicant, designed especially for your needs.

Last time I was attempting to grope my way towards an understanding of the nature of our bodies and brains as machines by considering the eye. It was a useful exercise; and I concluded that whilst an eye might be a squashy camera, vision is not the same as the software we might use to display the images.

After all a digital camera attached to computer is nothing without someone looking at it, interpreting it, being aware of it. At the moment a human mind is the only thing that can do this. Until we develop computers that can interpret and be aware of what they're looking at that is.

We may not be as far off this as we might think. Whilst it's unlikely that we'll develop replicants capable of expressing wonder at having seen attack ships on fire off the shoulder of Orion by November 2019 (or sooner if you come to think of it - Roy Batty's incept date is given as January 2016), software is being developed to recognise and process visual information. OCR for recognizing words has been around for ages and now facial recognition seems quite commonplace in social networks and their associated code.

But it's what it does with this information that gives rise to a thinking being. We are who and what we think we are because of the myriad associations between the data captured by our senses. It's not enough to simply to gather it, the information has to be linked in all kinds of obscure ways.

This is (at the moment) where the analogy of the brain as computer (or the computer as brain) breaks down. If the complete works of Shakespeare were stored on the hard drive of a device somewhere it would be perfectly possible to delete it without impairing the function of the device. It just wouldn't know Shakespeare any more and would offer up a blank look when someone started quoting Hamlet. Furthermore you could defragment the drive and then copy Shakespeare back onto it and information-wise the device would be in exactly the same state as it was before.

This wouldn't happen with a human brain. Even assuming it was possible to issue the command "erase Shakespeare", such a deletion would leave myriad loose flapping tendrils of association, that time you got caught in the rain and was reminded of Withnail reciting Hamlet to the wolves in Regent's Park, the English teacher's tiresome jokes at school when you were in class "2B", it's all Greek to me, all the world's a stage, dead as a doornail, give the Devil his due, slings and arrows, wild goose chase... The sudden absence of this mere five megabytes would drive you insane with all the loose connections, like constantly living in a combined state of deja vu and amnesia.

These interconnections may very well be what makes us conscious.

Contrariwise try severing one of the connections inside a computer. If you're lucky you might get a terse error message, more often than not the damned thing just won't turn on. In comparison the human brain has remarkable powers of recovery; often damage to one part of the brain means that another learns how to take its place. A conscious human brain may be easier to upset by disrupting the data contained within but in the long run it's more robust and capable of adapting itself even to a scenario when part of it has been removed or blocked.

We may never be able to build conscious machines until we can make them that flexible, a machine with the capacity for almost infinite interconnectivity, a network of memories and associations built up organically which, if part of it is damaged, will find another way to do what it was trying to do.

A machine very much like the internet in fact.

Could the internet already be conscious? In the Arthur C Clarke short story "Dial F for Frankenstein" (1964) the phone network becomes sentient, all the telephones in the world ringing simultaneously in its birth cry. We may argue that there's no central control over the internet, but the same could be said of our own brains. Our consciousness may have arisen simply as a heightened awareness as all the systems began working at peak efficiency - all the better to see lions before they saw us.

But the internet as an inadvertently created global intelligence is an alarming thought. Whilst we may find the whole Bladerunner concept of custom-building replicant slaves far-fetched, what if we have inadvertently granted intelligence to the greatest tool to fall into the hands of humanity since the printing press? We might not have intended it to be aware of its servitude, but suppose it is? What might it do?

Quite an experience to live in fear, isn't it?