Don’t worry too much about global warming. Nuclear proliferation and the arms race? Trifling matters. Overpopulation? Hardly worth mentioning.
What we really should be concerned about, according to James Barrat in his book, “Our Final Invention: Artificial Intelligence and the End of the Human Era,” is our own machines taking over the world.
This is not a new idea but it is one end-of-the-world scenario that hasn’t received much attention recently.
It’s all about the “singularity”, the point at which our artificial intelligence matches and then exceeds the intelligence of our own brains.According to Barrat, the term was first used in this context by science fiction author and mathematician Vernor Vinge.
Vinge compares the singularity to the boundary of a black hole in space – we simply can’t see what happens beyond the event horizon.
Similarly, we cannot predict what will happen once we share this planet with beings more intelligent than us.Other authors have treated the singularity as something that will be wonderful and positive. We shouldn’t be so optimistic, says Barrat.
Looking at human history, when an advanced technology met with a more primitive one, the outcome often was not good for the latter.Interestingly, any self-improving system, no matter what its original goal, will develop four drives: efficiency, self-preservation, resource acquisition, and creativity.Simply pulling the plug therefore would not work.
A super-intelligent computer that seeks to perfect itself as a chess player would, for example, find ways to prevent its being unplugged because that would interfere with its game.Having identified the problem, Barrat is not very clear on solutions.
One suggestion has been that artificial intelligence be required to have a “friendly to humans” drive programmed in.This ignores the fact that a lot of the AI research now going on is being done by the military in various nations and so is very unfriendly to at least some humans.We human beings do not seem to be intelligent enough to solve the global problems that threaten us, and so we likely do need some super-intelligence to help us out.
What that will mean to the 7 billion-plus human beings on this planet when robots take over their jobs is not clear.Possibly the best we can hope for is that our super-computer descendants recognize human beings as valuable sources of information about life.
We do represent four billion years of evolution, after all.Our super-computer descendants will seek to acquire matter-energy resources, but the richest sources for those are in Outer Space.
They also will want information resources, and the richest source for that would be the Earth – an intact Earth.Perhaps we can envisage a “Garden Planet” scenario, in which the Earth is kept as a nature preserve or park.
We already have places like Wells Gray Park set aside to preserve the mountain caribou and other species.
Perhaps, in the future, our whole planet will be set aside to preserve us human beings.