This only works with half-assed early 21st century A* pathfinding. If we are in a simulation good enough to simulate general intelligence, pathfinding is a non-issue.
Well, that depends on what the level and extent of the simulation is. For example, if this is a video game, most people might be computationally simple NPCs based on legacy code.
Despite many orders of magnitude more processing speed and memory, we still can’t get companions to walk through doors properly. If engineering taught me anything, it’s that every job is done half-assedly and will stay that way as long as it possibly can.
Or we might be in half a simulation, where player AI actually comes from outside the world (like in all our games), but NPCs are handled via (flawed) AI. I’m not sure how I could easily tell the difference between the two. If you’re serious about annoying the simulators though, going off the rails is probably more interesting than finding implementation bottlenecks, sure.
(Note, just to make it explicit: I’m not really taking the whole simulation cluster of ideas serious. It’s just that common simulation hypothesis assumptions about necessary complexity and low-level simulation seem quite implausible.)
You are postulating a non-reductionist world in which you are the only or one of a small set of agents with actual minds.
Correct. Maybe not strictly non-reductionist, but for all practical purposes, yeah.
If current science and philosophy tell us anything about our simulation, it is that it is very detailed and low-level.
“Us”? Are you running a particle accelerator? All I know is that there exists somewhat-consistent literature to back up a certain vast, low-level interpretation of my actual experiences, and even then I have no idea how the (certainly plausible-seeming) ideas about physics get me to an explanation of minds or consciousness at all, so if they are all wrong or massively incomplete, I wouldn’t know.
That might be because there is a vast, detailed universe that follows relatively simple and consistent laws. (I certainly favor that interpretation.)
It might also be because someone designed a plausible physicalist setting for a game and presented it to me in a convincing way. I don’t think that’s actually very hard to do.
For example, I regularly have lucid dreams that have a similar level of detail as my waking state (even though I know I’m dreaming). It looks the same, people act realistically, I can talk to them. I don’t think my brain simulates any kind of detailed physics for that. The illusion doesn’t last very long, sure, but scaling it up still seems simpler than getting a quantum physics implementation to work properly.
I’m not saying physics is a hoax, but it would certainly be much easier to bullshit me into believing that I’m living in a detailed physicalist universe than to actually simulate even just one planet of it. So if you want to complicate the simulation, you maybe shouldn’t even assume that there actually is a quark layer or anything like that.
You are postulating a non-reductionist world in which you are the only or one of a small set of agents with actual minds.
Correct. Maybe not strictly non-reductionist, but for all practical purposes, yeah.
Of course I know this to be incorrect. but of course I would say that rite? :)
I’m not saying physics is a hoax, but it would certainly be much easier to bullshit me into believing that I’m living in a detailed physicalist universe than to actually simulate even just one planet of it. So if you want to complicate the simulation, you maybe shouldn’t even assume that there actually is a quark layer or anything like that.
Ok good point. So then if we are in a simulation, we can probably fuck with it.
I had some good ideas on this the other day, but I was thinking of it from the acasual trade angle instead of the “universe is sim” angle. One was you should entangle your mental state with some really hard to compute stuff, like some crazy multiphysics with convective fluid flows. That only works for acausal trade, because you are being simulated to find out what you would do so the simulator is constrained to be accurate. In the general case, they can just bullshit a result.
If they are doing a simulation tho, it will be because they have some question that they can’t just bullshit. The trick would be to find out what that is and do some crazy high-load physics around that.
This only works with half-assed early 21st century A* pathfinding. If we are in a simulation good enough to simulate general intelligence, pathfinding is a non-issue.
Well, that depends on what the level and extent of the simulation is. For example, if this is a video game, most people might be computationally simple NPCs based on legacy code.
Despite many orders of magnitude more processing speed and memory, we still can’t get companions to walk through doors properly. If engineering taught me anything, it’s that every job is done half-assedly and will stay that way as long as it possibly can.
Or we might be in half a simulation, where player AI actually comes from outside the world (like in all our games), but NPCs are handled via (flawed) AI. I’m not sure how I could easily tell the difference between the two. If you’re serious about annoying the simulators though, going off the rails is probably more interesting than finding implementation bottlenecks, sure.
You are postulating a non-reductionist world in which you are the only or one of a small set of agents with actual minds. I think that’s wrong.
If current science and philosophy tell us anything about our simulation, it is that it is very detailed and low-level.
(Note, just to make it explicit: I’m not really taking the whole simulation cluster of ideas serious. It’s just that common simulation hypothesis assumptions about necessary complexity and low-level simulation seem quite implausible.)
Correct. Maybe not strictly non-reductionist, but for all practical purposes, yeah.
“Us”? Are you running a particle accelerator? All I know is that there exists somewhat-consistent literature to back up a certain vast, low-level interpretation of my actual experiences, and even then I have no idea how the (certainly plausible-seeming) ideas about physics get me to an explanation of minds or consciousness at all, so if they are all wrong or massively incomplete, I wouldn’t know.
That might be because there is a vast, detailed universe that follows relatively simple and consistent laws. (I certainly favor that interpretation.)
It might also be because someone designed a plausible physicalist setting for a game and presented it to me in a convincing way. I don’t think that’s actually very hard to do.
For example, I regularly have lucid dreams that have a similar level of detail as my waking state (even though I know I’m dreaming). It looks the same, people act realistically, I can talk to them. I don’t think my brain simulates any kind of detailed physics for that. The illusion doesn’t last very long, sure, but scaling it up still seems simpler than getting a quantum physics implementation to work properly.
I’m not saying physics is a hoax, but it would certainly be much easier to bullshit me into believing that I’m living in a detailed physicalist universe than to actually simulate even just one planet of it. So if you want to complicate the simulation, you maybe shouldn’t even assume that there actually is a quark layer or anything like that.
Of course I know this to be incorrect. but of course I would say that rite? :)
Ok good point. So then if we are in a simulation, we can probably fuck with it.
I had some good ideas on this the other day, but I was thinking of it from the acasual trade angle instead of the “universe is sim” angle. One was you should entangle your mental state with some really hard to compute stuff, like some crazy multiphysics with convective fluid flows. That only works for acausal trade, because you are being simulated to find out what you would do so the simulator is constrained to be accurate. In the general case, they can just bullshit a result.
If they are doing a simulation tho, it will be because they have some question that they can’t just bullshit. The trick would be to find out what that is and do some crazy high-load physics around that.