= 27d567bc2d48d9f40e9ccc20ea370b15)
sfb comments on Harry Potter and the Methods of Rationality discussion thread, part 5 - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (648)
Assume they are in a simulation - why would it have an event buffer created able to compute time travel at all, and why pick 6 human hours (equivalent) as the magic number constant for it?
Presumably simulating a human brain is harder than simulating the same mass/volume/atom count of solid metal, so as the population increases has the time-turner interval shrunk correspondingly? Seems unlikely it would settle on such a round number if it was changing with population. (Or is that why magic is getting weaker - as the simulation computer fills up?)
Why not? I think it'd be kind of cool to simulate a universe with magic and I'd feel altogether clever if I could implement time travel in it. :)
OK, simulating time travel for the hell of it sounds good, I still question the buffer limit idea:
"Simulated time travel now working, but what should we set the Horological Constant value to?"
"Dunno. Three years? A thousand years? Just enough time to undo saying something rude? MAXINT seconds? AVAILABLEMEM? User configurable?"
"Oh whatever, it's time to go home, I'll just put six simulated hours and be done with it. Shall I start locking up?"
That would be a disappointing reason for the 6 hour limit, and an unconvincing way to make the universe work so that the plot works. I hope the 6 hour limit is either not real, or something more interesting.
My first hypothesis would be simple processing complexity. Time travel is complicated. It is the kind of problem that grows ridiculously with time and space. The programmer has been able to invent an algorithm that simplifies it for low dimensionality but even with that algorithm higher order time travel would still take too much time on the given hardware.
Second: the programmer initially programmed the buffer in to allow for short term time processing. Things like the soda that prompts you to drink, spell dodging time hop magic, etc. It didn't even occur to him that the wizards would find a way to exploit the mechanism for long term use.
Another possible reason to have a time-buffer in a world simulator (which, btw, I don't believe the HP:MORverse is) is that the simulation doesn't actually do everything in real time.
Rather, you may have situations where process A and process B are defined as taking the same number of simulated time-slices, but process B takes more actual time to simulate for whatever reason, and so the simulation of process A is halted until process B catches up. (This presumes that it's not possible to reallocate simulated processes across simulating resource threads with arbitrarily fine granularity.)
Which means that at any given real-world moment, some parts of the simulation are at timeslice T, some parts are at timeslice T+1, and so forth. The six-hour limit might simply reflect the typical spread, and simulated time-travel might be a hacking of the system that is bound by that spread, rather than an explicitly simulated capability with an explicitly simulated upper bound.
Something like this is true of the only reality-simulating system we know of, namely our own brains. For example, color phi is a kind of simulated time travel where, in response to a perceived event E1 at time T, your brain constructs an illusory event E2, which you experience as occurring before T. This works because different parts of your brain construct your experience of time T at different rates, and tag those parts as occurring at T; the experience of simultaneity is constructed by your brain.
If the six-hours is to avoid too much time-skew, it's a hack and one would expect better from simulation-builders.
There are plenty of ways to efficiently calculate differing time-space regions, using lazy evaluation or equivalents thereof. For example, the famous Hashlife algorithm for Conway's Game of Life does exactly that - different regions can be billions or trillions of generations apart thanks to memoization. Lazy evaluation proper allows weird techniques like the reverse state monad or circular programming (aka time-traveling).
It may be that the limit is not due to the physics itself but because of the intervention of an early wizard. Unbounded time travel is one of the most powerful abilities imaginable. The first witch to exploit this and take ultimate power would obviously want to prevent others from overthrowing her. It would be in her best interest to find a way to put constraints on the powers of others. Longer than six hour anomalies may well trigger destructive countermeasures.
It occurs that magic is basically the ability to hack the simulation; the wizards who developed time travel didn't know what was actually safe, what was unsafe, what would crash the simulation, and what the simulation could actually even do - so they picked the weakest version of time travel to implement (both branches consistent) and slapped a bunch of arbitrary limits on it so it (hopefully) couldn't break anything major.
Also, Merlin in T.H. White's The Once and Future King lives his life backwards in time, from old age to youth - is that canon!Merlin, and does that property carry over to the Merlin to whom the characters refer in HP:MOR?
Merlin is not in the canon except as a curse-word of sorts (By Merlin's beard!).
And as the figure referenced by the Order of Merlin, which is awarded to people who perform exceptional deeds, and as the first person to get his face on a chocolate frog card
If the simulation is filling up that might make time travel less powerful but I don't see why the processing/resource constraints should make other forms of magic also need to be restricted.
(fixed simulation resources / increasing population) = less magic per person
(fixed simulation resources for magic / increasing population using magic simultaneously) = less spare magic at any given moment
Were the sort of ideas I was thinking. But, if it's a simulation it needn't compute in realtime, so it is a weak suggestion.