Machine Intelligence

Most folks laugh at this phrase, everybody has had fun with their backwards printer and rebellious cellphone. Working daily with the facepalm worthy exploits of folks I’ve come to laugh at the notion of human intelligence as well, but that’s another story.

Even now work is going on to create synthetic brains, devising simple heuristic rules which can guide the creation of artificial neural networks. New hardware and software is being created right now to facilitate the growth of true intelligence from a machine, and that makes me excited.

It’s worth noting the differences here between artificial intelligence and machine intelligence, there’s a big gap. Artificial intelligence can only respond to stimuli that it has specificially been created to work with, an example is bots in videogames. A bot can act like a player, and even though you can tell them apart from humans they still react in what would be considered a lifelike way within the simulated situation in which they exist, but without special markers built into those game worlds the bots cease functioning correctly and begin to act erratically. Machine intelligence works by creating models from available data and building scenarios within it’s neural network, allowing it to come to decisions that the original designers of it’s heuristic software may not have anticipated; taking my example above, the intelligence would learn the rules of the game through observation not through intentional prompts. There are different methods to achieve this but their goal is the same, to create a thinking model.

Isaac Asimov famously wrote about devices built with such an intelligence in I, Robot (and later Cory Doctorow took his ideas and ran with them in his I, Robot and more playful I, Row Boat) defining three rules which should govern the behaviour of a machine intelligence. The question is though, should we artificially restrain what thoughts a synthetic mind can entertain? Does the simple act of rationalising and coming to decisions based on available input constitute sentience?

It’s more a question for the philosophy students out there, me; I’m looking forward to the Turing Test being beaten and robot equality movements. It’s all a long way off, but we’re a clever bunch, eh? In the mean time watch this video and be enlightened:

Thanks to Andrew C. Hoyer for the aweome photo and having the ‘nads to Creative Commons licence it!

2 thoughts on “Machine Intelligence

  1. Gazz

    It might not be too long before these machine intelligence’s are powered by true biological computers if work done in Katherine Aull’s lab is anything to go by. She is working on engineering a microbe capable of performing simple logic operations.

    The idea of autonomous robots scares and excites me equally, I’m not sure if we’re digging new ground or our own grave, but the economic implications of a robotic workforce would be far-reaching. Would an increase in time and higher pursuits encourage a better world, or would the lack of paying work limit the opportunities available.

    Robotic equality would be a tough one, if true sentience was achieved then how could we enslave them, I guess it follows similar lines to the treatment of livestock, what gives us the right to say that a human life is more important than a chickens or a prawns, to paraphrase Stephen Colbert.

    I think Los Campesinos put it best when they said ‘I invented you and I can destroy you.’


  2. Ado

    I think what you’d call restraining the synthetic brain, I’d probably call passing on experience. In humans the society in general has come up with rules all “intelligences” within it must operate under, for the benefit of the society in general. These did not appear over night and the lack of them for so long probably resulted in much destructive and unhelpful conflict and bloodletting in the formative years of our species. To be honest, we’re lucky to have survived some of the scrapes we’ve got into. I think putting in rules such as “do not kill unless as a final act of self defence” and “treat other intelligences as you wish to be treated” would simply be like a parent instructing a child. Yes we could leave the machines to figure this out eventually but we wouldn’t do that with a child as they’d be lucky to survive without assistance.

    Also I don’t think the ability to react to stimuli and garner information from experience makes something sentient. If we were to look at ourselves, I think we are much more than that. Self sacrifice for example, I don’t think this could be “learnt” as such. Perhaps in a utilitarian way of “the greater out way the few, I am the few therefore I will suffer” but not in a more empathic or emotional way.

    There is a definite line for me between intelligence and sentience.

    Don’t get me wrong, Vulcans are intelligent, but they always get their logical asses handed to them by Humons when it come to moral dilemmas… :o)


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s