3 ms·
I absolutely agree, it got much more interesting as I started fleshing out my thought process. As for the motives, for a simulation to be helpful, it would hav
by polshaw 14y ago
I absolutely agree, it got much more interesting as I started fleshing out my thought process.
As for the motives, for a simulation to be helpful, it would have to run at greater than real time, no? Which.. is surely impossible, at least for a complete 'perfect' simulation. Right? Imperfections (and the limit of a simulation having to be simpler than its host) would infer a finite chain of simulations, which would vastly reduce the likelihood of us existing in a simulation.
Also if these wars (etc) exist, the simulation is going to be knocked offline at some point, destroying all descendant universes, shrinking the pool of possible existent universes much further still.
So (AFAICT) we have gone from it being almost a certainty we are part of a simulation (only one true universe from infinite similar simulations), to being very unlikely.
I may check out the culture series, but i'm not one for reading much fiction, sadly.. impatient internet world that has formed me and all.
- ep103 14y ago> it would have to run at greater than real time, no? Not for the people in the simulation : ) We'd just be self-conscious byte code. Which actually is the question Iain Banks brings up. They want to run simulations on how a society will respond to a political action, but that would imply creating self-conscious byte-code in the simulation. And wouldn't that be genocide when they turn off the sim? So maybe the only ethical thing is to never turn off the sim...