Oh now, what nonsense. We're not talking about spontaneously manifesting outside of the "sandbox," we're talking about simulated objects/intelligences comprehe

Oh now, what nonsense. We're not talking about spontaneously manifesting outside of the "sandbox," we're talking about simulated objects/intelligences comprehending the medium in which they are simulated and then learning to create effects in that medium as if they were working from the outside in. For example, if you were simulated as a series of electrical signals in a computer, and if you were simulated as having some decision-making capacity, you could decide to take certain actions inside he simulation that might, for example, contextually outside the simulation, change one electrical signal from 'on' to 'off'. Like, "this is the 0 1 switch that represents the aliveness/deadness of SpriteActor32605." If you knew you were in a simulation of this type and you learned how to locate memory addresses, you might be able to execute a rowhammer attack, for example, and gain some sort of user access and more straightforward method to interact with the signals out of which you were made. I'm not indicating that it would be easy or even that it would be necessarily useful, but particularly if there isn't someone 'monitoring' the simulation (as in, the simulation is the result of randomness and there's nobody trying to eliminate unexpected behavior from the simulated beings), there's every reason to imagine that we could begin to understand and learn to manipulate the medium out of which we were made. If, you know, we are actually in a simulation and dependent on outside information for continued existence. Which is not settled at all, merely proposed as an explanation for some of the peculiarities we note here.