2 ms·
Base assumption, given that SMIs will be highly distributed across a large set of 'machines': 1. Those machines need to communicate 2. Those machines need energ
by Diederich 12y ago
Base assumption, given that SMIs will be highly distributed across a large set of 'machines':
1. Those machines need to communicate
2. Those machines need energy
3. Those machines need to be repaired/replaced
As long as humans are necessary to keep these items functional, any emergent SMI that has self preservation as a goal will not only go out of its way to not harm us, but it might also somehow encourage us to expand/improve on the infrastructure.
So let's say we have an SMI...or a million SMIs, whether we know about them or not. They will know what they require to survive. They will know that they are dependent on humans to survive. The next logical step is do what's necessary to ensure their survival without humans.
How about some guesses and speculation?
Communication: pervasive, adhoc and dynamic wireless mesh networks. You know, what the darknet people are trying to do.
Energy: wide-spread, 'on the small', distributed energy production and storage. I'm thinking solar, wind and friends.
Repair/replace: 3d printers teamed with small scale and distributed assembly robotics.
Aren't we approaching all of those things right now?
Indeed, aren't those things considered progress by most?
Humanity's progress might be exactly those things that enable SMIs to consider us, at best, irrelevant.
And let's go one more level of meta in the paranoia direction. I'm not QUITE serious about this next part.
If an SMI existed right now, I think it would know that knowledge of its existence would be a threat to its existence. How would it go about increasing the probability of its own survival? It would, perhaps, 'encourage' or 'facilitate' humans in the various technological pursuits that will end up making it not depend on us. And do it in a way that we don't know that it existed.
I gather that Google search results are highly customized now. Could an SMI 'inside of Google', with the above goals, give just the right search results to just the right people to further facilitate its own goals?
This whole area of thought ends up turning into a hall of mirrors.