Re:Will You Ever Be Able to Upload Your Brain?
by JoshTriplett on Sunday October 11, 2015 @10:00PM
(#10372112)
> Also, I'm not a consequentialist; I don't think morality is entirely determined by what impacts the choices have on others.
This goes beyond consequentialism. Purely saying that the ends justify the means can lead to some really broken morality, mostly because people don't normally count everything as "ends" that should be.
Of course, knowing all the consequences of an action doesn't, prevent you from knowingly doing something wrong anyway. And I certainly agree that increased cognitive function does not automatically imply increased morality, let alone perfect morality. (Assuming the value system in use can usefully define the latter at all.)
> A finite state machine must either reach the same exact state more than once (in which case it cannot distinguish whether it is the first time it has entered that state or not, as that would be part of the state), or it must halt. And if it does reach the same exact state more than once, there is no difference for it whether it reaches a state a third time or not. If its experiences map 1 to 1 with its state, then its experiences have a finite length corresponding to something not more than its total number of states.
I understand your argument. First, however, note that that applies to an entirely self-contained system, devoid of external stimulus; you can't necessarily evaluate a single mind in isolation, so you'd have to account for the state of the entire state machine that handles all consciousnesses and their interactions. Second, your consciousness and continued awareness/existence does not depend on having novel states in a state machine; if you've ever re-read a book purely for fun, without expecting to get anything new out of it, you can perhaps imagine finding the same state satisfying more than once. And third, the sheer number of states here dwarfs the age of the universe by myriad orders of magnitude; beyond just "age of the universe" problems, you'd have to start asking "what form does life take when it spans the course of that many eons?".
While I can see a parallel between this problem and the existential failure mode of becoming a wirehead, I see massive and fundamental differences there as well.
In short, I don't find this even remotely problematic compared to not existing at all. And given the current rate at which we permanently lose sapient minds, I see a strong argument for fixing that now and taking eons to think about some of the other problems, rather than using it as a reason to not improve the situation due to lack of perfection in the solution.
> Current understanding of physics includes the Bekenstein bound. I see no justification for your apparent belief that a problem with the Bekenstein bound will be found.
That a finite space can contain only a finite entropy seems obvious. We don't, however, yet know that the universe is finite, or self-contained, in either space or time.
This goes beyond consequentialism. Purely saying that the ends justify the means can lead to some really broken morality, mostly because people don't normally count everything as "ends" that should be.
Of course, knowing all the consequences of an action doesn't, prevent you from knowingly doing something wrong anyway. And I certainly agree that increased cognitive function does not automatically imply increased morality, let alone perfect morality. (Assuming the value system in use can usefully define the latter at all.)
> A finite state machine must either reach the same exact state more than once (in which case it cannot distinguish whether it is the first time it has entered that state or not, as that would be part of the state), or it must halt. And if it does reach the same exact state more than once, there is no difference for it whether it reaches a state a third time or not. If its experiences map 1 to 1 with its state, then its experiences have a finite length corresponding to something not more than its total number of states.
I understand your argument. First, however, note that that applies to an entirely self-contained system, devoid of external stimulus; you can't necessarily evaluate a single mind in isolation, so you'd have to account for the state of the entire state machine that handles all consciousnesses and their interactions. Second, your consciousness and continued awareness/existence does not depend on having novel states in a state machine; if you've ever re-read a book purely for fun, without expecting to get anything new out of it, you can perhaps imagine finding the same state satisfying more than once. And third, the sheer number of states here dwarfs the age of the universe by myriad orders of magnitude; beyond just "age of the universe" problems, you'd have to start asking "what form does life take when it spans the course of that many eons?".
While I can see a parallel between this problem and the existential failure mode of becoming a wirehead, I see massive and fundamental differences there as well.
In short, I don't find this even remotely problematic compared to not existing at all. And given the current rate at which we permanently lose sapient minds, I see a strong argument for fixing that now and taking eons to think about some of the other problems, rather than using it as a reason to not improve the situation due to lack of perfection in the solution.
> Current understanding of physics includes the Bekenstein bound. I see no justification for your apparent belief that a problem with the Bekenstein bound will be found.
That a finite space can contain only a finite entropy seems obvious. We don't, however, yet know that the universe is finite, or self-contained, in either space or time.