Man I wish I could get a Carnegie Melon sponsorship to talk about bullshit extrapolations. Jeesh you would have thought econometric modeling based on the past trend rate of growth to make a prediction about the future trend rate of growth was well dead. Here it is again in all its mechanistic glory.
Two Three flaws. First-off past rates of increases in the number of computational turns a processor can achieve do not translate into a sure, certain, or even likely predictor of the future rates of growth. This is not some natural law folks!
Second and perhaps more important is the question of what and how processors process when they process? What would a 1 billion tetra-flop processor process with asteroids as the software? That is, even if we accept the idea that in X years micro-processors will be 1 billion time more powerful (speakers claim) the question is will the software be 1 billion times more powerful? Video games are not a bad test of this reasoning. Surely FPS are vastly improved then when the little Doom 2.9 gigabyte shareware version was released in in 1995. But a billion times better?
But let us cut to the quick of it. For some war between human and robots we would need robots to have some sentient qualities which gave them the capacity to develop an identity as “robots” and yet as an equal but exploited underclass to humans. Think about that for an evening or two minutes and then laugh.
Let us descend even further into the absurdity and assume for argument sake that robots did develop consciousness, that they were capable of identifying as robots in some general sense. Surely there would be different robots occupying different levels in the social division of labour: there would be automotive robots, cleaning robots, pilot robots, bouncer robots, maybe prostitute robots, grad student robots, professor robots, lawyer robots…ok well you get the picture: to some extent all these locations in the social division of labour could be explained as “workers” subservient to capital. But what we also know is that this is not sufficient to provoke a war against capital. So why should robots be any smarter?
Third, and perhaps more importantly (here I had to think back to Donna Haraway in a senior undergraduate seminar and then fast forward to Philip Mirowski’s Machine Dreams ) there is the issue of the relation of robots to humans. Robots are in fact human prosthetics. They are not and will not be something that stands apart from and against human beings. They are, and will always be something enlisted, tied up with the human condition…with our creative capacity to transform nature and ourselves. To the extent that robots go to war with humans they will do so in the way machines always have: as an extension of our (in)humanity. These prosthetics will not go to war against humanity but in the service of some elements of humanity against other elements of humanity. And it is for that reason we ought to debate and evaluate the role of technological development in our society not in the sense of oh god “it is alive” but in the sense that they are we.
Here is the video enjoy: