The rise of the machines? Not any time soon

At a recent anthropology conference, my colleagues and I were sitting around a table discussing our session when a wasp decided to come and interfere with our proceedings. Here I am talking about the insect creature, not the White Anglo Saxon Protestant variety. As one friend tried to shoo the wasp away another told us it would be better to ignore, rather than kill it, because when under attack, wasps will emit a danger pheromone and a search party will be sent from the nest. Rather than dealing with one wasp, we might have to contend with several more. Pretty impressive I thought, and it made me wonder if wasps were displaying a kind of consciousness? It definitely appears that wasps are interconnected, they have a social awareness of their own species and support systems to help and to assist each other. As a view of consciousness is based on awareness and interdependent interconnectivity, it would appear that wasps seem to possess a kind of consciousness.

Humanists tend to disagree with the view that nonhuman animals and machines are capable of consciousness and are very possessive about it as an exclusively human characteristic. John Gray’s book Straw Dogs argues (quite depressingly) that human uniqueness, of which consciousness and a transcendent mind are important components, is a delusion. But do humanists “‘throw the baby out with the bathwater’, so to speak, by rejecting the idea that nonhuman animals and machines might possess or come to possess consciousness? And what might anthropology contribute to this discussion?

Artificial Intelligence as a discipline is influenced by the humanist philosophy of René Descartes (otherwise known as Cartesian) and he is celebrated and attacked in kind for his dualist philosophy. Descartes theorised the human mind is transcendent whilst bodies, animals and machines are not. It may come as a surprise then to think a founding father of humanism and human uniqueness also inspired and repelled a generation of technologists interested in designing intelligent systems.

Traditional AI, as it referred to in the field, uses the model of a rational and logical mind as the basis in the development of AI systems. Other types of AI research (particularly that of Rodney Brooks) such as behaviour-based robotics, or embodied AI are anti-Cartesian in the sense that they do not see the mind as the site of all knowledge and believe that bodies are essential for developing artificial systems. Traditional and embodied based robotic researchers are interested in developing autonomous artificial machine systems. Cybernetician Kevin Warwick argues that such systems are possible, but additionally that humans and machines can become interconnected systems and as such the boundary between the human being and the machine is dissolved. We all know that machines are excellent calculators and logicians, but how are machines in distinguishing a wall from a door? Or how effective are they in telling the difference between an eye twitch and a wink? No!

Getting to the heart of consciousness means working out just what consciousness is and there are many different kinds of consciousness. A standard dictionary definition implies that consciousness is a ‘awareness of one’s environment’. Are machines aware of their environment? No! But they are dependent on their environment along with all humans, animals and things.

Consciousness is more than just awareness and environmental interdependence. It is about transformation because invariably the contents of consciousness are obscured by psychic, cultural or political processes. As an anthropologist I am interested in the importance that consciousness has played in the social imagination. I cannot explore them all here, so I will briefly summarise a few here. Arguably the most important theories of consciousness are derived from politics, psychology and religion. Carl Jung’s theory of collective unconsciousness is about making conscious ancestral archetypes as a crucial part of an individual’s development. Class consciousness is a Marxist idea of change whereby the masses realise they are capable of overthrowing a small and wealthy but powerful elite. Achieving consciousness requires effort. This is crucial for uncovering the contents of the Freudian unconscious – the idea that the mind has a hidden storage department for memories, thoughts, feelings and sensations that can interfere with the conscious processes of an individual. These visions of consciousness are inherently transformative.

With all the developments in today’s technology it is easy to get carried away by science and technologists who argue we are on a precipice of a machine revolution with human beings on the way out. I am not of the view that we have anything to fear from machines and think machines would be more useful if they were autonomous, intelligent and possessed levels of consciousness (as similar for example to the social operations exercised by wasps).  One benefit might result in the way we interact with machines, it might then be possible to use natural spoken language to operate machines. Disappointingly many of us know how unmanageable our computers become once we activate the voice-control option. Moreover, understanding how to combine organic nerves with machines is also crucial to the development of sophisticated prosthetic devices. There is exciting research in this area showing this technology’s potential in giving independence and freedom of movement to individuals with physical disabilities. At present though much of the talk about a revolution in AI consciousness is really fantasy dressed up as science.

Throughout October and November, The Independent Online is partnering with the Institute of Ideas’ Battle of Ideas festival to present a series of guest blogs from festival speakers on the key questions of our time.

Dr Kathleen Richardson is a British Academy postdoctoral fellow at the department of anthropology, University College London.  She is speaking at the Battle Satellite event Artificial intelligence, bionic men and human consciousness on Tuesday 25 October, organised in association with the Manchester Salon, Manchester Science Festival and The John Rylands University Library, University of Manchester.

Recent Posts on Battle of Ideas

• An open letter to Melanie Phillips
• Privacy challenges ahead 2013
• Anders Breivik: Narcissist. Murderer. Insane?
• Free speech is the real loser in Terry v Ferdinand
• 'Abortion industry': A slur on the women who support their choice

• davidbb

  It is hilarious to find an incandescent and piercing intellect such as Marx’s described as ‘rambling’. Such vanity.

• http://openid-provider.appspot.com/bjgates Brian

  Perhaps “human exceptionalism” would be a better term to describe those who believe that  we are fundamentally different from animals, and “biological exceptionalism” to describe those who believe that there is a fundamental difference between living and non-living systems?

• JamesChambers123

  DavidBB:

  I’d go further and say not only did he ramble but his polemic ramblings were incorrect and unoriginal and the only reason anyone thinks anything of “his philosophy” (which was a blatant rip-off of Hegel’s dialectic idealism but lacking the subtly) or “his economics” (which was a blatant rip-off of Ricardo’s but taken to implausibly ridiculous levels) is because the USSR ascribed the man god status and actively promoted him and his teachings to a population of poor uneducated mugs via an extensive propaganda campaign which lasted nearly a hundred years. His popularity today is just a residual effect of this propaganda combined with the public’s ignorance of philosophy and economic theory.  

  Leaving aside for a moment the second-rate nature of his teachings, as a person Marx was an obnoxious, misanthropic, pro-violence racist and eugenics enthusiast.  I have nothing positive to say about him or “his philosophy and economics”, except that I’m glad they’ve been resigned to the dustbin of history where they belong. 

• rabbitlug

  Yeah, that does capture it a little better. The reason I suggested anthropocentrism over humanism is that the former would imply that we put ourselves exclusively at the heart of everything whereas the latter would imply that we put humans *first*, but not necessarily to the exclusion of all others.

  However, I’d go with human exceptionalism, or perhaps radical humanism.

  With regard to consciousness, as always with any abstract concept, any question about it rests firmly on the definition; for me it is blindly obvious that insects have consciousness as I define it as “awareness”, not necessarily “self awareness”. Having said that, the self is also an abstract concept and as Hume argued, look inside yourself, try to grasp the “I” and you end up with a handful of nothingness, so perhaps animals are also self aware in that they are independent, perceiving and experiencing beings.

  Nagel wrote a good paper on subjectivity in his “What’s it Like to be a Bat?”

• Terrence Zellers

  As far as I’m concerned the bat is dead.  http://machineintheshell.blogspot.com/2010/05/what-it-is-like-to-be-bat-in.html

  The mysterians and narciscists won’t admit it until the machines achieve artificial stupidity, but consciousness is really a rather low bar.      Dennett answered most of the outstanding questions of consciouness twenty years ago in “Consciousness Explained”.    Dennett shotgunned, unfortunately, making some mistakes along the way, but what he said in CE is fairly straightforwardly, for some of us, respooled with a little work into a quite coherent picture.

  The real problem is that contrary to what most people think, consciousness is not intelligence.  The “really hard problem” in contrast to Chalmer’s “hard probem” (actually plural) is AI.   Organizing the data flow of a machine to be conscious is not a problem, but without the intelligence that connects ideas across the breadth and layers that humans are used to in their mutual interactions, it will never “seem” conscious.  

  (I’ve fairly recently lightbulbed how the functional organization of consciousness might subtly solve one very small but significant piece of the AI problem – as said, it certainly is not the whole enchilada.  But getting code in place to demonstrate it – if it turns out to actually work – is going to take me a while.  Otherwise a long, long and tedious history of discussion on the-brights.net, in the forums most recently in The Philosopher’s Brain thread … if you are really, really enjoy pain).