Our Possible Imminent Divinity

We might soon be gods.

John Searle might be right that digital computers could never be conscious. Or the pessimists might be right who say we will blow ourselves up before we ever advance far enough to create real consciousness in computers. But let's assume, for the sake of argument, that Searle and the pessimists are wrong: In a few decades we will be producing genuinely conscious artificial intelligences in substantial quantity.

We will then have at least some features of gods: We will have created a new type of being, perhaps in our image. We will presumably have the power to shape our creations' personalities to suit us, to make them feel blessed or miserable, to hijack their wills to our purposes, to condemn them to looping circuits of pain or reward, to command their worship if we wish.

If consciousness is only possible in fully embodied robots, our powers might stop approximately there, but if we can create conscious beings inside artificial environments, we become even more truly divine. Imagine a simulated world inside a computer with its own laws and containing multiple conscious beings whose sensory inputs all flow in according to the rules of that world and whose actions are all expressed in that world -- The Sims but with conscious AIs.

[image from http://tapirangkasaterbabas.blogspot.com; go ahead and apply feminist critique whenever ready]

Now we can command not only the AI beings themselves but their entire world.

We approach omnipotence: We can create miracles. We can drop in Godzilla, we can revive the dead, we can move a mountain, undo errors, create or end the whole world at a whim. Zeus would be envious.

We approach omniscience: We can look at any part of the world, look inside anyone's mind, see the past if we have properly recorded it -- possibly, too, predict the future, depending on the details of the program.

We stand outside of space and to some extent time: Our created beings can point any direction of the sphere and not point at us -- we are everywhere and nowhere, not on their map, though capable of seeing and reaching anywhere. If the sim has a fast clock relative to our time, we can seem to endure for millenia or longer. We can pause their time and do whatever we like unconstrained by their clock. We can rewind to save points and thus directly view and interact with the past, perhaps sprouting off new worlds from it or rewriting the history of the one world.

But will we be benevolent gods? What duties will we have to our creations, and how well will we execute those duties? Philosophers don't discuss this issue as much as they should. (Nick Bostrom and Eliezer Yudkowsky are exceptions, and there's some terrific science fiction, e.g., Ted Chiang. In this story, R. Scott Bakker and I pit the duty to maximize happiness against the duty to give our creations autonomy and self-knowledge.)

Though to our creations we will literally have the features of divinity and they might rightly call us their gods, from the perspective of this level of reality we might remain very mortal, weak, and flawed. We might even ourselves be the playthings of still higher gods.