|
Size: 440
Comment:
|
Size: 2893
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 5: | Line 5: |
| It takes place in | Basic problems with a hard sci-fi space opera: 1. The universe is too big to have multi-system civilizations. * No true FTL; there is no hard scifi FTL drive that seems plausible. * Wormholes mostly solve this, but artificial wormholes are probably something that only can be made by extremely dramatically-boring civilizations for whom all problems not like "the heat death of the universe" are trivial. * Naturally occurring wormholes are not likely to be navigable or go anywhere interesting to people. * Preferred solution, then, is that the wormhole network was created by a now non-interacting "dramatically boring civilization" for their own inscrutable reasons. 2. Individual biological life will probably become irrelevant before a large multi-system could arise. * In the 2020s, the most likely mid-term technological scenarios favor a transition away from organic life as the species that would be sent to explore the universe. Humans would follow far behind a friendly AI colonization effort preparing the way for them, or be absent entirely because of unfriendly AI. * If the current AI Spring proves to be a false start on human-level AI (for example, because stochastic gradient descent is too training-data inefficient), this probably still only buys a century or so before brute force approaches like artificial evolution become viable. If artificial evolution isn't viable, then human consciousness is probably an extremely rare fluke in naturally-occurring life as well, which is unfavorable for a setting with non-human sophonts. * A reactionary movement rejecting AI, in the presence of a false-start 2020s AI Spring that leads to AI which threatens mass social upheaval but not superhuman intelligence, could suppress AI development for a long time only in the presence of a general technological contraction. Otherwise, it will probably be too easy for a rogue actor to develop AI anyway, given the potential rewards. * An interstellar civilization arising from the scenario above is likely to be crushed by the first rival which has/is strong AI that it encounters. * The civilization in above would need independent reasons to reject pursuing human mind-uploading. This is plausible but a civilization that would categorically reject human mind uploading is probably highly regressive socially. If they don't reject it, why wouldn't uploads be doing the highly-dangerous galaxy exploration? |
Connected Space is a spiritual prequel to Beyond the Twelve Worlds, taking some of its concepts (particular, backstory concepts) and using them to develop a hard sci-fi "space opera" with appropriate setup to avoid adventure-destroying technologies such as characters who can be backed up (either because they are natively digital intelligences or because they are digitized human minds.)
Basic problems with a hard sci-fi space opera:
- The universe is too big to have multi-system civilizations.
- No true FTL; there is no hard scifi FTL drive that seems plausible.
- Wormholes mostly solve this, but artificial wormholes are probably something that only can be made by extremely dramatically-boring civilizations for whom all problems not like "the heat death of the universe" are trivial.
- Naturally occurring wormholes are not likely to be navigable or go anywhere interesting to people.
- Preferred solution, then, is that the wormhole network was created by a now non-interacting "dramatically boring civilization" for their own inscrutable reasons.
- Individual biological life will probably become irrelevant before a large multi-system could arise.
- In the 2020s, the most likely mid-term technological scenarios favor a transition away from organic life as the species that would be sent to explore the universe. Humans would follow far behind a friendly AI colonization effort preparing the way for them, or be absent entirely because of unfriendly AI.
- If the current AI Spring proves to be a false start on human-level AI (for example, because stochastic gradient descent is too training-data inefficient), this probably still only buys a century or so before brute force approaches like artificial evolution become viable. If artificial evolution isn't viable, then human consciousness is probably an extremely rare fluke in naturally-occurring life as well, which is unfavorable for a setting with non-human sophonts.
- A reactionary movement rejecting AI, in the presence of a false-start 2020s AI Spring that leads to AI which threatens mass social upheaval but not superhuman intelligence, could suppress AI development for a long time only in the presence of a general technological contraction. Otherwise, it will probably be too easy for a rogue actor to develop AI anyway, given the potential rewards.
- An interstellar civilization arising from the scenario above is likely to be crushed by the first rival which has/is strong AI that it encounters.
- The civilization in above would need independent reasons to reject pursuing human mind-uploading. This is plausible but a civilization that would categorically reject human mind uploading is probably highly regressive socially. If they don't reject it, why wouldn't uploads be doing the highly-dangerous galaxy exploration?
