Your first paragraph hints that you're more partial to Hanson's idea of having already undergone multiple Singularities that tremendously boost the economy.
> We dont even know what intelligence really is, so we cant predict what super intelligence is or will do.
Paraphrasing from Yudkowsky, an intelligent system is that which implements its goals. If a superintelligent entity has goals even remotely recognizable to us, we can predict what it will do quite easily. Even if you have no idea how Chess works, just by knowing it's a game you can predict both sides will try to win.
I'm with you on brain-machine interfaces that augment our intelligence, though I would definitely say I could think faster, if not better. Just with the example of an internal calculator would help tremendously; somewhere around the order of trillions of synapses fire if you try to multiply two three-digit numbers if your head.
> So the optimist in me sees a future free from meaningless work, where even the average Joe will have an IO of 200 or even 1000.
What do these figures even mean? One of the goals of a Yudkowsky-defined Singularity is that the AI (hopefully quickly) gets around to helping us augment our own intelligence, moving us along the path so that we can become something like how we presently are to ants, but with present us as the new ants. Not just all becoming Einstein.
First, let's do away with this "Rapture of the Nerds" business. It'd be funny if the comparison were at all valid: http://www.acceleratingfuture.com/steven/?p=21
Your first paragraph hints that you're more partial to Hanson's idea of having already undergone multiple Singularities that tremendously boost the economy.
> We dont even know what intelligence really is, so we cant predict what super intelligence is or will do.
Paraphrasing from Yudkowsky, an intelligent system is that which implements its goals. If a superintelligent entity has goals even remotely recognizable to us, we can predict what it will do quite easily. Even if you have no idea how Chess works, just by knowing it's a game you can predict both sides will try to win.
I'm with you on brain-machine interfaces that augment our intelligence, though I would definitely say I could think faster, if not better. Just with the example of an internal calculator would help tremendously; somewhere around the order of trillions of synapses fire if you try to multiply two three-digit numbers if your head.
> So the optimist in me sees a future free from meaningless work, where even the average Joe will have an IO of 200 or even 1000.
What do these figures even mean? One of the goals of a Yudkowsky-defined Singularity is that the AI (hopefully quickly) gets around to helping us augment our own intelligence, moving us along the path so that we can become something like how we presently are to ants, but with present us as the new ants. Not just all becoming Einstein.