But if a thought is just an arrangement of electric potential in your brain, a random brain will eventually form already in the exact configuration that you are experiencing now. The fact that an instant later it fades for eternity is meaningless - you don’t know that the next instant WON’T bring oblivion, until it has already passed; and your memory of the prior instant is just that - a memory. It could be the result of those randomly generated electrons.
Or so the theory goes
Sure, but like I said, I don’t think cognition only takes an instant. I can’t say for sure exactly how it does work (so you can hold on to that nobel prize for the moment), but I think that it probably is a side effect of the ongoing transition from one brain state to another to another et al.
And since the only transition the Boltzman Brain will have is from one static image of the brainstate to a dissipating cloud of undirected sparks, I think no cognition or awareness will occur.
Now, if a whole physical brain spontaneously appeared, with neurons and such to direct the electrical pulses along, then you could maybe sell the argument that this process would continue long enough to have some momentary thought and awareness before it simultaneously froze and exploded from decompression. Spontaneously generate a head and skull to hold the brain together and it’ll last even longer. Make the whole body and he’ll probably live long enough to suffer. Spontaneously generate an entire universe around him during which Trump is being reelected and he’ll suffer even longer.
But I think at that point we’re not talking about a Boltzman brain anymore.
True, and once you’re dealing with infinities, enough time for a brain to spontaneously appear and enough time for a universe aren’t all that different.
Well, as far as semi-tongue-in-cheek “our reality is a simulation” parallels go with “God is just a kid playing a sim game on god mode explains everything weird/bad going on”, there’s yet another possibility.
Ever play Sim City, the original game from like 1990? It had a free demo. But if you played it too long, or used cheat codes to continue playing beyond a certain point, the game would start throwing disasters at you in an increasing and compounding way. Tornadoes, earthquakes, Godzilla attacking your city, and so on.
So what if not only our reality is a sim game for some deific entertainment… But it’s a pirated sim game where the software is programmed to go rogue?
Sure, but you’re only dealing with infinities at the basic level of the infinite void of nothingness - any instantiated random universe will not be infinite in size itself (because that’s not how infinities work), so the odds that secondary ‘simulations’ will appear within any existing universes are drastically lower than the chance that a new random universe will appear somewhere else in the infinite void of nothing.
And I strongly suspect that ‘big bang’ universes are much simpler (and thus much more probable) than Last Thursday universes, so even when we’re talking about universes popping randomly into existence in the infinite void the odds that our universe is a Last Thursday universe are vanishingly low.
True - by natural selection big bang universes popping up all over the place will outcompete last Thursday universes
So, having proven that this “reality” must be a simulation, how does this fact inform your behavior?
Because this is great debates and I’m just that kind of dork, I’ll answer you seriously.
If our universe is a simulation, then it’s not unreasonable to assume it’s a mass-produced simulation. (Perhaps where each copy starts from a ‘big bang’ seed, perhaps where each copy starts from a ‘saved game’ pre-set scenario.) And in such a simulation there could be copy protection, sure.
How likely the 2020 stuff is to be copy protection depends on how solidly seated in history it is. We’ve been working our way up to global warming for hundreds of years - when did the copy protection kick in? Trump only started 75 years ago, but his election was powered by 250 years of systemic racism and 50 years of republican plotting. So, um, murder hornets are pretty recent! Are they the copy protection?
If one did prove that their reality was simulated, it would make it no less real to them, and so probably shouldn’t change anyone’s behavior
Eta: unless they could discern something about the simulator.
I shall recognize that my book-writing is functionally equivalent to the act of creating this entire universe and laugh maniacally more often.
This is true. Although without understanding the scope of the simulation, the technology involved or the context, we don’t really know if it is, in fact, a “lot of work”. It could be some super-advanced content generation engine just writing books for the subjects to read. Or it could be the archived works of Old Earth from 10,000 years ago before the remnants of humanity loaded their brains into a ship’s computer in an attempt to escape Earth’s destruction by an approaching space goat. Or there’s really just one book (or none) and the simulation just always tricks your brain into thinking you just skimmed something called ‘Moby Dick’ that you remember is about a whale.
I’m really not that fond of the "you really don’t know anything - you just think you do because your screenwriter wrote “begbert2 is a very smart boy; you can tell he is because he wears glasses” argument. I personally remember dozens, perhaps hundreds of books I’ve read, and could ramble on about them at length until I was banned from the board for dumping all over it. That information had to come from somewhere, and I’m not willing to believe that I don’t actually remember the things I actually remember.
It’s sort of a cogito ergo sum thing - you can tell me that you are a shallow bot reading from a script, but I’m actually experiencing my noggin.
But would you know if you read those books yourself, the “memories” of you reading them was planted in your brain, or “you” are really a digital copy of the original begbert1 living in some virtual world?
If I’m in a restore of a saved game where my memories extend back before the time of the save, then I was the original begbert1. (I’m not anymore, but their past is my past.)
The four possible sources for past reality are:
(ETA: Also, begbert1 is my father.)
(ETA2: And I’m well aware that there are thorny identity issues with the copy situation, but they don’t extend to the information question. Information can be copied freely.)
I think it can be combinations of those possibilities. i.e. the “simulated” reality was turned on last Thursday as part of a highly advanced game. You are a copy of some original begbert, or perhaps an amalgamation of real people’s memories, perhaps augmented with additional memories and history, both real and fabricated?
Yes, if a Boltzmann brain were to persist, it would notice.
I’m not going to explain the concept here, you can check the Wiki.
The point was simply to show that Last Thursdayism / the five-minute hypothesis is not only used in a religious context and does not assume a creator.
No, I am not positing that.
I am saying that if we begin with the proposition that the simulation is necessarily fake or prebaked or whatever at a certain level, then how much is fake?
Proponents of the truth of the simulation hypothesis, use lazy evaluation as a kind of get out of jail card for how our universe can be so big and detailed.
My point is simply that if this card is played, it has knock-on effects; it gives us good ground for doubting our environment, with no obvious reason why we could assume our memories are valid. After all, if I wanted to do a simulation of 2020 earth, with humans that don’t know they are in a simulation, and I am happy prebaking a lot of the data, I’m not going to start the simulation 13.7 billion years ago and hope an Earth forms. I’m going to kick it off in 2020, or 2000 say, with fake memories.
My contention is that this is not more absurd than past reality appearing randomly. Indeed, I think this sort of thing will become so commonplace that we’ll all have to get used to it eventually. When an entity as complex as an intelligent superobject attempts to model the past (or the present, or the future), it could model events with a depth and complexity that is comparable in complexity to the day-to-day experience of an ordinary human being. At any given moment in our experience, we could either be
a/ an ordinary flesh-and-blood human being living their life in the ordinary sequence of events
or
b/ a fleeting model in the databanks of an intelligent superobject of arbitrary scale.
Bostrom’s simulation argument can be paraphrased as follows.
1/ Either nobody ever builds an intelligent superobject in the history of humankind
or
2/ They do build intelligent superobjects, but for whatever reason these entities do not ever model the past
or
3/ We are much more likely to be experiencing a recreation of the past (of whatever duration) than ordinary flesh-and-blood humans living their lives in the ordinary sequence of events.
There are plenty of get-out clauses there. We might be sensible enough to never construct intelligent superobjects, perhaps because we’ve been warned off by the Matrix films and their terrible sequels; or maybe ISOs won’t be interested in humans any more, and leave us in peace (or in pieces).
Or we could all be part of a model, existing for a fleeting moment of time, or for extended periods of simulated suffering, or perhaps paused and reactivated whenever the model is useful.
How could we tell? As has been pointed out earlier, from inside the simulation all our knowledge of our history is also simulated. The sim could actually have begun in 1900 or 2000, with all of our understanding of time before then (including our own lives, if born before the “logical start time”) is simply backstory.
This is an untestable proposition - I think… - but it raises the point that even within a simulation context, there is no reason to think the simulation has a singular purpose, that we are that purpose, or that it hasn’t already been achieved and we’ve simply not yet been turned off because hey, this is going into weird places, let’s see how this goes.
Or even that the simulation is being directed. We could be a bad (buggy) simulation, too!
So the superobject is presupposed to manually decide every detail of everybody’s history thoroughout history…without simulating it inside its own head?
I mean, I suppose there’s a grain of possibility there. Speaking as a writer, I obviously have crafted a world with a couple dozen agents in it that have (to varying degress) some history, and who engage in interactions with each other and such that are sufficiently complicated and interesting that you will not be able to resist the temptation to buy my books - and I invented all of these characters, history, events, and interactions without simulating any of them. Which suggests that rather than simulating the universe, some gigantic entity incapable of boredom and with perfect memory and analysis skills could simply invent the stories of everyone in the universe.
However, it should be noted that coming with even my stories’ levels of history and interactions was a titanic pain in the butt. My document folders are littered with the detritus of ideas that I attempted but which ran into an intractable contradiction that I hadn’t thought of earlier. When you think of the sheer amount of history and interactions that you’d have to synchronize to write the story of the entire universe, doing that while avoiding contradictions and keeping everyone in character sounds ludicrously difficult. And I’m having a hard time thinking of any tactic that even a superbrain could use to make that process more efficient - it would literally be, “Well, what if Fred was little slower in the shower, and arrived a little late to work, and that’s why his coworker was annoyed, and was a little ruder to the barista, and that’s why they messed up the coffee order of that guy who…” a hundred trillion times, just to come up with the causal sequences behind the major events in history. Seriously, even pretending to keep all these people in character would be insane.
It occurs to me that this is what we’re talking about when we talk about simply creating a simulation of the world of 2020 ex nihilo. That’s a lot of history to backfill - and you can’t really simulate backwards, as far as I know. You just have to invent crap off the top of your head, while making sure all your causation works out and your characters stay in character and nobody important has a bigger hole in their memory than they can brush off. (“My mom’s name? Man, that was some bender I was on last night…”)
If you go with a last Thursday model, then the relative simplicity of a simulation of the universe 500 years ago doesn’t matter. It would have to be complex enough to work for us today and in the future, when we will be able to observe more of the structure of the universe with better telescopes.
Whether we are real or not isn’t the issue here. It is making some assumptions about those who wrote the simulation. We can assume that they don’t want to wait around 50 billion years for results, and we can assume that they don’t want to piss away energy.
Sure you can loaded every simulated thing with its state at whatever t=0 you want, but trust me, doing that in a consistent way is a lot harder than starting the simulation from some simple reset state. That could be the Big Bang, it could be the formation of the Earth and other planets. If you think anyone could create a consistent state for billions of people, you might as well call them God and be done with it.