However, any time the power is down -- (and we're approaching the 10th anniversary of the great blackout of August 14, 2003) -- it is obvious to everyone that our society is balanced delicately on toothpicks and the Singularity will go stumbling into the dark with the rest of us. Meanwhile , it looks like this project has almost reached its funding goal.As author and mathematician Vernor Vinge first suggested - if we create an intelligence greater than our own, it can develop an intelligence even greater than itself, perhaps in a matter of only days. This resulting exponential intelligence escalation has come to be known as the Singularity. Once we start the process, it is very unlikely that we can stop it, and we have no idea what a hyper intelligence could - or would - do.
This idea, for me, is just not convincing -- fun to speculate about and a great plot for science fiction. Once Deep Blue beat Kasparov, we knew for sure that the computers we make are smarter than us.
First thing is first: current A.I. (even Deep Blue and Watson) is dumb. They are not intelligent. Why? Because they did not earn their information. They were given it by humans. Everything A.I. currently "knows" is of human origin, ultimately. In order for something to be intelligent, they have to evolve. They have to earn the information for themselves. They also have to be autonomous. This means that they cannot be dependent on us for energy and maintenance. This is not the world we find ourselves in today. But it will be the world we find ourselves in by the 2040s (I hypothesize). Everyone in robotics is now well aware that intelligence must evolve. Evolutionary robotics is in its early stages, but it will advance quickly. And even if evolutionary robotics fails. The Global Brain won't. Higher distributed intelligence will emerge this century (via the Internet). The actual patterns of the Internet mirror the development of a nervous system and centralized brain in early life forms. The Internet is essentially making our environment intelligent. This will reach a critical threshold this century. Whether you call it the Singularity or something else... it will be a higher level of intelligence that our planet has never known before.
It will be a higher level of intelligence that our planet has never known before.
I hope so. I hope the intelligent robots understand in a mathematical way that collaboration, empathy, and kindness are more likely to lead to the survival than aggression. (or am I wrong?) (p.s. typo in my first note has been fixed. -- I hope the typo didn't make you lose respect for me, although maybe you missed it.)
Haha I missed the typo... and even if I did, I wouldn't have lost respect in you. All the models show that a higher distributed intelligence can only emerge from higher levels of cooperation (and consequently reduced competition - it is a factor dependent on resource scarcity/abundance). This is how multi-cellular organisms evolved and this is how centralized brains evolved. Individual agents (i.e., cells, neurons, or people) realize that there is "more to be gained" by joining and collaborating with a collective, than it is to be a selfish individual. How this is done is interesting to study, because we are individually selfish and composed of "selfish genes". But cooperation and altruism can (and do) evolve from that base of selfishness. In short, the Global Brain will be the by-product of infinite cooperation and the marginalization (and/or elimination) of competition at the level of us, the agents that bring it into existence.