by Lauri Järvilehto
I wrote recently a brief paper on the idea of the social singularity and its relevance to EM. The paper is a quick draft, and therefore somewhat superficial, but it addresses some of the central issues we have been addressing in the think tank.
In the paper I argue first that the classical singularity hypothesis has major problems that arise not only from the technical feasibility of AI, but also philosophical problems concerning the entire notion of AI. In addition to this, regardless of whether we will eventually have an AI singularity, the social intelligence explosion is coming considrable faster so as to render AI singularity of secondary importance.
The social singularity argument, which draws from Vinge’s IA (intelligence amplification) hypothesis is that once we have real-time information networking between individuals, we will be looking at a massive explosion in the utility of information. Imagine that you could get an answer to any of your questions from the best expert in your field in real-time – anytime, anywhere.
It appears to me that the IA explosion is imminent. We already have all the technology we need: a real-time Q&A service (Aardvark and Quora), a real-time social network (Facebook) and internet ubiquity (the iPhone and Android phones). Now what we need is a service that makes this kind of functionality agile enough, and a critical mass of users, with emphasis on the latter. It should not take much more than five years to get there. Five years ago, there was no (public) Facebook, no Twitter – no social media to speak of. At these rates of acceleration in accessibility of information, the singularity may be just a matter of a few years. Like Kurzweil said, the singularity is near. And the singularity, insofar as I can see, is us.
If you’re interested, you can download the paper here.