Tuesday 11 February 2014

The Information Age and the myth of anthropomorphism


If that opening hasn't put you off reading further, then welcome to this short intellectual coffee-break!

How we use words is important. “The limits of my language mean the limits of my world.” (Wittgenstein, Philosophical Investigations 1953) You can trace the evolution of ascribing mechanisms with human characteristics back through history to certain philosophical movements. Personally, I blame the mathematicians and logicians, those fearful meddlers and “Gradgrinds” of history, for this dichotomous mess we find ourselves in. I suppose there is a certain mentality in humanity that would like to assume that people’s behaviour and the machines we build are synonymous. If only we were all perfectly logical. These days we have a habit of talking about inanimate and stupid objects as if they are sentient beings. Regularly you will come across irate people standing over a machine with a mixture of frustration and animosity because the bloody thing is not doing what it should and won’t act within reason - like say humans would. To be clear: Computers do not have “memory” that is in any sense similar to human memory, they are not intelligent and the notion of Artificial Intelligence is just that, a notion, that many argue will never be a reality.

If you haven’t already been told yet, you live in a world that is sometimes called the Information Age and this nonsensical nomenclature has its roots stretching back to Claude Shannon, and his influential paper: A Mathematical Theory of Communication (1948). Since then we have allowed the idea of information to be understood in terms of electrical blips and signal transmission, which is a branch of engineering and applied mathematics. It is important to remind ourselves that Shannon paper is entitled: “A Mathematical Theory of Communication” and was principally a paper on compressing data operations and reliably storing and communicating data. He explicitly states that: “Frequently the messages have meaning; that is they refer to or are correlated according to some system with certain physical or conceptual entities. These semantic aspects of communication are irrelevant to the engineering problem.” He doesn’t care what “meaning” the messages or transmissions have. (Nor, by the way, does Google care about meaning.) Everything is simply algorithmic!

Maybe living in the world that information theorists constructed, we have forgotten how to fit our human understanding to the machines. Wittgenstein says that: “if we construe the grammar of the expression of sensation on the model of ‘object and designation’ the object drops out of consideration as irrelevant.” (Wittgenstein, Philosophical Investigations 1953) We have designed systems that systemically take meaning or “object” from any consideration of “information”. So if you are wondering why your collection of information that you paste into a thesis/assignment/project hasn’t yielded you an “A” remember that that machine hasn’t done the understanding part. You should remember “information” is something to be engaged with and understood and isn’t just its binary form: ”01101001 01101110 01100110 01101111 01110010 01101101 01100001 01110100 01101001 01101111 01101110”.