19 ms·

Is it impossible for software to prove it's likely in a VM just because it is 'encoded in it'? It depends on the implementation but it's definitely not impossi

by Tenoke 5y ago

Is it impossible for software to prove it's likely in a VM just because it is 'encoded in it'? It depends on the implementation but it's definitely not impossible a priori and especially not because of Godel.

lettergram 5y ago
I’m not sure that’s the case, because we don’t know the underlying system (assuming true). More importantly, If math is designed to reflect reality and his proof effectively states you can’t prove the system is correct within the system — then effectively aren’t we saying we can’t prove we are in a simulation? Honest question - I have no idea lol
qayxc 5y ago
> it's definitely not impossible a priori and especially not because of Gödel. You forgot something there: you know the underlying system and about its existence. In the absence of this knowledge it's indeed impossible to know. Here's an example from real-life physics: there's no way of telling by means of measurement, whether you're accelerating or at rest but subjected to a gravitational field. A software example would be this: how would a program written and running in Factorio (like Pong [0]) be able to tell that it's implemented by means of video game mechanics running in an emulator executed within a VM on server hardware? If the programming environment itself features no way of interrogating its runtime environment, then there's just no way. [0] https://forums.factorio.com/viewtopic.php?t=76155 https://forums.factorio.com/viewtopic.php?t=76155
- Tenoke 5y ago
  If we assume that this is running in a truly huge Factorio instance then monitoring the world intently would reveal so many subtle bugs and inconsistencies in a world with otherwise extremely straightforward laws that being in a simulation should be a prime hypothesis.
52-6F-62 5y ago
I think so. It can model itself, but it couldn't replicate itself 1:1—which maybe is the wrong assumption on my part, but is the only way you could prove with certainty that you may be living in a simulation? I mean, without input from another source, that is.

by Tenoke 5y ago

lettergram 5y ago

I’m not sure that’s the case, because we don’t know the underlying system (assuming true). More importantly, If math is designed to reflect reality and his proof effectively states you can’t prove the system is correct within the system — then effectively aren’t we saying we can’t prove we are in a simulation? Honest question - I have no idea lol

qayxc 5y ago

> it's definitely not impossible a priori and especially not because of Gödel. You forgot something there: you know the underlying system and about its existence. In the absence of this knowledge it's indeed impossible to know. Here's an example from real-life physics: there's no way of telling by means of measurement, whether you're accelerating or at rest but subjected to a gravitational field. A software example would be this: how would a program written and running in Factorio (like Pong [0]) be able to tell that it's implemented by means of video game mechanics running in an emulator executed within a VM on server hardware? If the programming environment itself features no way of interrogating its runtime environment, then there's just no way. [0] https://forums.factorio.com/viewtopic.php?t=76155 https://forums.factorio.com/viewtopic.php?t=76155

Tenoke 5y ago
If we assume that this is running in a truly huge Factorio instance then monitoring the world intently would reveal so many subtle bugs and inconsistencies in a world with otherwise extremely straightforward laws that being in a simulation should be a prime hypothesis.

52-6F-62 5y ago

I think so. It can model itself, but it couldn't replicate itself 1:1—which maybe is the wrong assumption on my part, but is the only way you could prove with certainty that you may be living in a simulation? I mean, without input from another source, that is.