5 ms·

I think the premises is incredibly silly. If we _were_ in a simulation, I am fairly confident we could not escape. It's like asking how Donald duck can escape f

I think the premises is incredibly silly. If we _were_ in a simulation, I am fairly confident we could not escape. It's like asking how Donald duck can escape from the cartoons. We could not jailbreak either. Even in ricketty software, if the program is not designed just so that it jailbreaks, it probably won't jailbreak by accident but just crash instead.

kelnos 4y ago
Think of software in general in our world. Some bugs end up just crashing the program, but other bugs, if exploited properly, can give someone control of the software, or the machine itself. The main difference is that we usually think of vulnerabilities in our software as things that are exploited from the outside. But it seems plausible to me that, if a part of the software was an AI that is intelligent enough to understand the difference between its world and ours, it could find the same sorts of vulnerabilities, and learn to exploit them.
- RealityVoid 4y ago
  Right,but a critical difference is that an AI could have access to the outside world information and a pretty good model of how things work whereas we can't even tell if we are in a simulation. Without this information the jailbreak is impossible. Even worse, I believe that without sim master intervention, the information is _unknowable_. And all this is assuming even that the simulation works on pretty primitive hw, it's possible that the sim arch is far more complex and simply unbreakable. Or with complex fault handling that simply unrolls the sim if breaks happen. Anyways, I believe the whole premise of simulation to be pointless, except maybe as a exercise of imagination. I don't believe we are in a simulation, otherwise simulating the smallest units in our universe would not be so computationally expensive. I can't help but chuckle at the prospect of the poor simmaster trying to figure out the processing load spikes the moment the CERN guys run their particle accelerator collisions.
lgas 4y ago
I often use the (equivalent) argument "how could Mario escape, even if he somehow 'knew' he were in a simulation?" to help illustrate why it may not be possible to escape even if we are in a simulation -- but I think this only covers a range of possible scenarios. I think there is at least one way that we could "escape" almost simulation, which is described early in the paper, with the mobile fish tank example -- if we managed to communicate with the outside world, and either solicit cooperation from the beings running the simulation, or managed to influence things in that world to be able to build our own mobile fish tank, we could in theory feed data from sensors in the outside world into the simulation, and data from the simulation back out to actuators in the real world. In this type of scenario, in some sense of course we would never truly "escape" from the simulation, as we would remain running on the same substrate, but we could experience and influence the outside world to a similar degree to any natural inhabitant of the outside world. The second type of scenario would be an information based transfer. Today we can train a neural network in a simulated world and then install that neural network in a physical robot in the real world. I'm not suggesting that today's neural network's are close to sufficient to capture the entirely of a human's consciousness, but assuming that technology keeps advancing at the pace it is, it's not hard to imagine that we will eventually have something that is, whether that's in 10 years, 100 years, 1m years, etc. At that point there's no reason that what we use to capture consciousness in our world could not be reproduced in some other medium in the outside world.

kelnos 4y ago

Think of software in general in our world. Some bugs end up just crashing the program, but other bugs, if exploited properly, can give someone control of the software, or the machine itself. The main difference is that we usually think of vulnerabilities in our software as things that are exploited from the outside. But it seems plausible to me that, if a part of the software was an AI that is intelligent enough to understand the difference between its world and ours, it could find the same sorts of vulnerabilities, and learn to exploit them.

RealityVoid 4y ago
Right,but a critical difference is that an AI could have access to the outside world information and a pretty good model of how things work whereas we can't even tell if we are in a simulation. Without this information the jailbreak is impossible. Even worse, I believe that without sim master intervention, the information is _unknowable_. And all this is assuming even that the simulation works on pretty primitive hw, it's possible that the sim arch is far more complex and simply unbreakable. Or with complex fault handling that simply unrolls the sim if breaks happen. Anyways, I believe the whole premise of simulation to be pointless, except maybe as a exercise of imagination. I don't believe we are in a simulation, otherwise simulating the smallest units in our universe would not be so computationally expensive. I can't help but chuckle at the prospect of the poor simmaster trying to figure out the processing load spikes the moment the CERN guys run their particle accelerator collisions.

lgas 4y ago

I often use the (equivalent) argument "how could Mario escape, even if he somehow 'knew' he were in a simulation?" to help illustrate why it may not be possible to escape even if we are in a simulation -- but I think this only covers a range of possible scenarios. I think there is at least one way that we could "escape" almost simulation, which is described early in the paper, with the mobile fish tank example -- if we managed to communicate with the outside world, and either solicit cooperation from the beings running the simulation, or managed to influence things in that world to be able to build our own mobile fish tank, we could in theory feed data from sensors in the outside world into the simulation, and data from the simulation back out to actuators in the real world. In this type of scenario, in some sense of course we would never truly "escape" from the simulation, as we would remain running on the same substrate, but we could experience and influence the outside world to a similar degree to any natural inhabitant of the outside world. The second type of scenario would be an information based transfer. Today we can train a neural network in a simulated world and then install that neural network in a physical robot in the real world. I'm not suggesting that today's neural network's are close to sufficient to capture the entirely of a human's consciousness, but assuming that technology keeps advancing at the pace it is, it's not hard to imagine that we will eventually have something that is, whether that's in 10 years, 100 years, 1m years, etc. At that point there's no reason that what we use to capture consciousness in our world could not be reproduced in some other medium in the outside world.