The World Wide Web is nothing like a brain

In the current issue of Wired, Kevin Kelly argues that the web and every computer gadget combined make up something called the ‘One Machine’, which is comparable to a ‘very large brain’. Stuart Derbyshire and Anand Raja of the psychology department at the University of Birmingham think he must be out of his mind.

Occasionally, an idea comes along that is so catastrophically wrong that it usefully highlights a truth that is otherwise obscured by half-truths.

A brilliant example of this is Kevin Kelly’s recent suggestion that the interconnection of the world’s computing power through the internet somehow constitutes a thinking entity that rivals the human brain (1). According to Kelly, writing in the current issue of Wired, all the worlds computing processors (including PCs, of course, but also cellphones, webcams, mp3 players and cameras) constitute the hardware of what he calls the ‘One Machine’. The One Machine is programmed by us, through the World Wide Web, and the operation of the One Machine resembles ‘the thought patterns of a very large brain’, says Kelly.

Except, of course, the activity of the One Machine, even if we accept it as an entity, is not the slightest bit like thought. As Charles Arthur rightly points out in his response to Kelly on the Guardian technology blog, the human brain is biological and comparisons between transistors and neurons, not to mention between hyperlinks and neurons, are not comparing like with like (2). The human brain does not work like an electronic computer, even a massive, super-interconnected computer.

All very true, but, unfortunately, psychologists and neuroscientists seem all too ready to argue that the brain is, at least a bit, like a computer, and that the mind is, at least a bit, like the software that runs on a computer (3). A typical neuropsychological account of experience involves parts of the brain pointing to other parts of the brain with various functions associated with each part. Vision, for example, is something that occurs at the back of the brain, where the basic features of a scene (shapes, colours, motion) are picked out. Information then flows from the back to the front of the brain along two pathways.

The pathway along the bottom of the brain elaborates what is being seen, while the pathway along the top of the brain elaborates where it is. A recent review in the academic journal Trends in Cognitive Sciences argued that the brain ‘converts information about spatiotemporal sequences into meaningful actions through interactions between early and higher visual areas processing form and motion and frontal-parietal circuits involved in the understanding of actions’ (4). What the authors are saying is that the brain somehow produces meaningful action by integrating lots of information, which is not a million miles from what Kelly is describing in his theory about the brain-like One Machine.

Instead of explaining perception, thought and behaviour, the neuropsychological understanding provides us with a sophisticated description. That description might be factually correct, and may have application, but it is not an explanation. Explanations identify causal relations; our current models of how the brain works merely pull the conscious being that perceives, thinks and behaves out of the magic hat of the brain.

There is a temptation to think that the inadequacy of our neuropsychological explanations is because of a lack of knowledge about how the brain works. What we need is more technology and more experiments. There is some notable truth to this, of course; we really don’t know how the brain works and more technology and more experiments are likely to be valuable. Yet merely providing more information about the brain (and merely bringing together ever-increasing computing power) will not by itself yield an understanding of perception, thought and behaviour.

While raw processing power is a necessary condition for perception, thought and behaviour, it is not sufficient. An understanding of thought must, at the very least, also include an account of the contents of thought. We have argued elsewhere that an understanding of thought based on an interrogation of neuronal (or electrical) activity will fail to describe thought because thought is not the firing of neurons (5). We believe that any understanding of thought should feel and sound like what it is.

The remarkable thing about human beings is not our brains per se, but the way we make knowledge explicit (6). We don’t just ‘see’ something and react to it, as a computer might; we place ourselves into a relationship with the thing that we are seeing. Consequently, while many animals might be said to ‘see’ the sky, only human beings will try to see more. The sky’s the limit, or not, depending on what we are trying to do.

Human vision sets off a series of reactions, located within an environment of meaning, and not limited by the properties of the stimulus interacting with a brain. For humans, patterns of action and interaction during normal development introduce this meaning (7). Commonly, an event or object will become the focus not just of the infant but of the infant and a primary caregiver; the infant then learns about the event or object through the other person. Early interactions of this type will not amount to thought by the infant, but rather are a discovery in action and feeling with meaning gradually acquired according to someone else.

Thought emerges within the contours of these discoveries and is not something that is a direct product of the brain itself. Thought is not substantive, it is lived, and firing neurons or sparking silicon cannot explain thought.

Both a human brain and a computer might, very loosely, share processing properties. But all of the processing power in all of the world’s computer gadgets and all of the world’s brains combined and jumbled together in any formation desired will never yield even a single thought – because thought is not the product of raw processing power. Modern computers locked together within the World Wide Web are brilliant at yielding information and connecting people, but this is in no way emulates a mind. The information and exchanges are massed but are not collective; the mass of exchange is not aimed at anything in particular and has no centre or particular point where experience might occur or distill. All human beings are the centre of experience, and particularly brilliant human beings – Mozart, Lenin, Einstein – forcefully distill collective intent and understanding. No computer can do this.

Essentially random exchanges of information cannot have an intentional existence and cannot have intelligence, as Kelly asserts, because intelligence, by its very nature, has to be directed and intentional and has to have a particularity where existence resides. Human beings have experiences because they are a particular centre within a world of symbols and meaning. Information meets resistance within us and is experienced; it doesn’t just flow through or across us and somehow generate experience as an accidental residue. We are grateful to Kevin Kelly – because against his ridiculous proposal for the One Machine, the exceptional qualities of human beings appear as obvious.

Stuart Derbyshire is a senior lecturer in psychology at the University of Birmingham. Anand Raja is reading psychology at the University of Birmingham.

(1) Infoporn: Tap into the 12-million-teraflop handheld megacomputer, Wired, 23 June 2008.

(2) No, Mr Kelly, I’m afraid the internet is not as clever as a single (human) brain, Guardian, 30 June 2008

(3) Minds, Brains, Computers: The Foundations of Cognitive Science – An Historical Introduction, RM Harnish, WileyBlackwell, 2001; From Computer to Brain: Foundations of Computational Neuroscience, WW Lytton, Springer-Verlag New York, 2002

(4) ‘Linking form and motion in the primate brain’, Trends in Cognitive Sciences, 2008;12:230-236

(5) What makes humans special?, by Stuart Derbyshire and Anand Raja

(6) The Knowing Animal: A Philosophical Inquiry Into Knowledge and Truth, by Raymond Tallis, Edinburgh University Press, 2005

(7) The cradle of thought: Exploring the origins of thinking, by P Hobson, Macmillan, 2002.