I Tried to Disprove Simulation Theory. Here's Why I Couldn't.

ENTRY_LOG //

• Reading Time: ~7 minutes of unhinged reality-checking.

Are We Living in a Simulation? I Went Down the Rabbit Hole So You Don't Have To (Okay, You Still Kind of Have To)

Okay so here's how this started for me. I was lying in bed, scrolling, half asleep, and I saw some clip of a guy explaining that we're "probably in a simulation" because of some experiment with light and slits. And my brain did the thing where it just... wouldn't let it go. Not in a "ooh spooky" way. In a "wait, is that actually true or is this just a vibe people repeat" way.

So I went and read the actual paper. Then I read the rebuttals to the paper. Then I read the rebuttals to the rebuttals. Then I went down a three-week hole involving a Swedish philosopher, a German physicist who thinks the whole thing is basically religion with extra steps, Elon Musk's old tweets, and something called "it from bit" that I'm still not 100% sure I fully understand.

This is everything I found. I'm not going to tell you what to believe, mostly because I genuinely don't know what I believe, and I've decided that's actually the honest place to land on this topic. If you walk away from this with more questions than you came in with, good. That means it worked.

Let's get into it.

Wait, who actually started this?

Most people think this whole "simulation" thing is a Matrix-era internet meme that escaped into respectable conversation. It's not, really. The version everyone argues about today comes from one specific guy: Nick Bostrom, a philosopher who was at Oxford for years and writes about, among other cheerful topics, existential risk and how humanity might accidentally wipe itself out.

In 2003 he published a paper called "Are You Living in a Computer Simulation?" And here's the thing that genuinely surprised me, because I assumed going in that the paper says "yes, probably." It doesn't say that. At all. What it actually does is sneakier and, honestly, more interesting.

Bostrom doesn't argue we're in a simulation. He argues that one of three things has to be true, and you don't get to just pick "none of the above."

Here they are, in my own words because his original phrasing is drier than this deserves:

• Door 1: Basically no civilization like ours ever survives long enough, or gets advanced enough, to be capable of running detailed simulations of conscious minds. We blow ourselves up, or stagnate, or something else gets us first — every time, everywhere.

• Door 2: Civilizations do get that advanced, but for some reason — ethics, boredom, a rule we can't even guess at — they almost universally decide not to run these simulations once they could.

• Door 3: We're in one. Right now. You, me, this blog post.

The logic connecting these three isn't vibes, it's basically a numbers game. If options 1 and 2 are both false — meaning advanced civilizations are common AND they do run these simulations once they're capable — then because a single advanced civilization could run not one simulation but billions of them, simulated minds would vastly outnumber "real" original ones. And if that's the case, and you're just some random conscious mind picked out of the pile, the odds say you're one of the billions, not one of the rare originals.

If you want to actually read it, read here simulation hypothesis lays out the trilemma cleanly, and it links to Bostrom's original paper too.

"Okay but which of the three is true?"

I asked myself this for longer than I'd like to admit, and the answer is: nobody knows, including Bostrom. He's been pretty open that he doesn't think there's strong evidence pointing at any one of the three over the others — which, if you think about it, is kind of a wild thing for someone to say about their own famous argument. He's not selling certainty. He built a trap and is honestly telling you he doesn't know which door you're behind either.

This is the point where I want to flag something, because it tripped me up at first and I think it trips a lot of people up: there's a difference between "we're probably in a simulation" (a strong claim a lot of pop-science content implies Bostrom made) and "one of these three weird things is true, and one of them happens to be that we're in a simulation" (what he actually argued). The second one is much harder to dismiss, honestly, because dismissing it means you have to actively defend one of the other two options, and neither of those is exactly comforting either.

There's also a sneaky technical thing buried in here that's worth knowing about even if you never bring it up at a dinner party: the move from "most minds like ours would be simulated" to "therefore I'm probably simulated" relies on an assumption — that if you don't know anything that makes you special, you should assume you're a totally average, randomly picked example from the group you belong to. Sounds reasonable. Except philosophers have spent two decades arguing about whether that assumption is even valid, because depending on how you define "the group you belong to," the math changes completely. More on that mess in a second.

The "wait, which group am I even being compared to" problem

This is the part of the argument that made me sit up, because it's the kind of thing that sounds like a technicality until you realize it's actually load-bearing for the entire trilemma.

Say there really are way more simulated minds than "real" ones out there. Fine. But "out there" compared to what, exactly? If we're talking about literally any conscious experience anywhere in existence, that's one massive, vague pool. If we're talking specifically about minds having the exact experience you're having right now — including the specific experience of reading a blog post about whether you're in a simulation — that's a much smaller, weirder, way more self-referential category, and the math behaves differently depending on which one you pick.

It's a little like being told "most lottery winners report being happy," and then realizing the survey only went out to people who already knew they'd won. The framing quietly assumes the thing it's trying to prove.

And there's a deeper version of this complaint too: the whole trilemma assumes things go in a tidy line — one human species, one shared future, either we make it to "posthuman" status or we don't. But why would it be that clean? Maybe some branches of humanity become wildly advanced and others don't. Maybe some advanced descendants of ours love running simulations and others think it's deeply unethical and refuse. History has never once gone in a straight line for any civilization, so why would the argument require ours to?

Which, honestly, leads into the strangest rabbit hole in this entire topic, and it's the one that actually kept me up the latest.

If simulations can run simulations... where does it stop?

Here's the thought that got me. If Bostrom's logic is right — that advanced civilizations would eventually run tons of detailed ancestor simulations — then there's no reason that logic stops working inside a simulation. A simulated civilization detailed enough to be genuinely conscious would, presumably, eventually invent their own computers. Their own Bostrom. Their own simulation argument. And if they get advanced enough, their own simulations.

Which means even if you're fully convinced of the whole argument, you still have no idea what layer you're on. You're not necessarily one level down from "the real world." You could be eleven levels down. Forty. There might not even be a clean concept of "the real world" at the bottom, because the structure of "simulations simulating simulations" might not require — or even allow — a bottom floor at all.

I want to be honest: I don't think human brains are built to comfortably hold "there might be no bottom" as a stable thought. It's the same feeling as staring at one of those Russian nesting dolls, except every doll you open has another doll inside that's somehow also the first doll, and you genuinely can't tell anymore which one you started with. My brain just sort of slides off this idea every time I try to hold it for more than a few seconds. Maybe yours will do better. Let me know if it does.

The physicists who think this whole thing is kind of insulting, actually

Now for the other side, because I went looking for the people who think this entire conversation is a waste of everyone's time, and I found a really good one.

Sabine Hossenfelder is a German theoretical physicist, and she does not hold back. She's called the simulation hypothesis pseudoscience. Not "unconvincing." Pseudoscience. Her argument, stripped down, goes something like this: a real scientific theory has to make a prediction you could actually go check. If you can't even imagine an experiment that would tell you the theory is wrong, it's not really doing science's job — it's doing storytelling's job, dressed up in science words.

There's a quote from physicist Sean Carroll that I think nails the whole objection in one line: a theory that doesn't change what you'd expect to observe isn't really a scientific theory. If the simulation hypothesis predicts exactly the universe we'd already expect to see either way, what is it actually telling us? Hossenfelder's blunter version of the same point: saying "the programmer did it" doesn't explain anything, it just relocates the mystery somewhere you can't check it — which, as she points out, isn't that different from how "God did it" worked as an explanation a few centuries ago.

Cosmologist George Ellis is in the same camp, and he said something that made me laugh out loud reading it: that the simulation hypothesis is "totally impracticable from a technical viewpoint" and that "late-night pub discussion is not a viable theory." Which, fair. I started this whole blog post because of a video I watched half-asleep at midnight, so. Point taken, George.

But here's where it gets more nuanced than "physicist dunks on philosophy," because there's an actual specific technical objection underneath the mockery. Our universe has properties — physicists call it Lorentz invariance, basically the idea that physics works the same no matter how fast you're moving or which direction you're facing — that are smooth and continuous all the way down. Computers, fundamentally, work in discrete chunks: 0s and 1s, pixels, grid points. If reality were secretly built out of a grid underneath everything, you'd expect that smoothness to break down somewhere, even just slightly, at the tiniest scales. We've never found that break. Not in particle accelerators, not anywhere.

I'll be honest, though — not everyone agrees this objection is fatal even within physics. Computer scientist Scott Aaronson has written about this exact disagreement: he agrees the simulation hypothesis is kind of lazy as an idea, because it doesn't actually engage with the hard stuff we've learned about physics, but he doesn't think it has some uniquely fatal problem just because of the smoothness issue — because some genuinely mainstream ideas in quantum gravity already suggest that any finite region of space might only contain a finite amount of information underneath everything, which is uncomfortably close to "the universe might secretly be made of discrete chunks of information" even among physicists who have zero interest in simulation theory as a concept. So even the "this is impossible" camp has internal arguments about exactly why it's impossible, which I found kind of comforting, honestly. Nobody fully agrees on anything here.

The physicists who think the opposite — that this might actually be checkable

Now flip it, because there's a whole other lineage of very serious physicists who don't think this is unfalsifiable nonsense at all.

It traces back to John Archibald Wheeler, a genuinely legendary physicist — he's the guy who coined the term "black hole," worked with Bohr and Einstein, that level of legendary — who in 1989 put forward an idea called "it from bit." His pitch, and I had to read this a few times before it clicked, is that maybe information is more fundamental than matter. Not "matter creates information." The other way around. Every particle, every field, even space and time themselves, might ultimately come from something more like a yes/no answer to a question than from "stuff." He called the universe "participatory" — meaning the asking of questions, the act of measuring, is somehow baked into what makes anything exist in the first place.

That's not Wheeler doing sci-fi. That's one of the most respected physicists of the twentieth century saying, in complete seriousness, that "bits" might come before "things."

A computer scientist named Edward Fredkin took this even further into actual computer-science territory, proposing the universe runs like a giant cellular automaton — simple rules applied over and over to a grid, building up insane complexity from almost nothing, kind of like how Conway's Game of Life makes complicated moving patterns out of four dead-simple rules. He worked with Richard Feynman, who later did foundational work on quantum computing partly because of these conversations. And physicist Seth Lloyd has a more modern version of the same idea — that the universe might literally be running as a quantum computer.

If you take this lineage seriously, "looking for glitches in the Matrix" stops sounding like a joke and starts sounding like an actual research proposal. There have been genuine, published attempts to look for a signature of an underlying computational "grid" in nature — specifically by checking whether ultra-high-energy cosmic rays travel at very slightly different angles than they should, because a grid would have a preferred direction baked in. Nobody's found that signature. But "haven't found it yet" and "this is impossible to test, ever, in principle" are two very different sentences, and a lot of the most interesting parts of this whole debate live exactly in that gap.

And then there's the guy who's currently the most talked-about (and most argued-about) figure in this whole space: physicist Melvin Vopson. He's published research on something he calls the "second law of infodynamics" — basically proposing that information itself behaves like a physical thing with its own thermodynamic-style rules, and that it tends to compress or minimize over time, almost like the universe is quietly running a background file-compression process. He's framed this as possible indirect evidence for the simulation hypothesis. Popular science outlets ran headlines like "scientist has evidence we live in a simulation," which, predictably, mainstream physicists were a lot more skeptical about — treating it as one interesting, unproven framework, not confirmed evidence of anything.

What I noticed, reading through all of this, is a pattern: every single "physics might actually test this" thread eventually narrows down to "physics is testing whether information is fundamental" — which is a completely real, ongoing, respected research question with zero requirement that anyone "built" anything on purpose. It only becomes "we're in a simulation" the moment you add one extra assumption nobody's actually proven: that "fundamentally informational" also means "deliberately constructed by someone, for a reason." That one extra step is the entire hypothesis, basically. And nobody's bridged it with hard evidence. Maybe nobody ever will, even if it happens to be true — which is its own deeply unsettling little thought I had at like 1 a.m. one night during all this research.

Okay, the video game thing. Let's actually talk about it.

This is the part you specifically asked about, so let's do it properly, because the "reality is basically a video game" argument gets thrown around constantly and almost nobody breaks down what it's actually claiming versus what's just a meme.

There are two real versions of this argument, and they're different enough that I think they deserve separate treatment.

Version one: the pixel thing. There's a real, agreed-upon smallest meaningful length in physics, called the Planck length — it's something like 1.6 × 10⁻³⁵ meters, so small it makes an atom look like a planet. There's also a smallest meaningful chunk of time, the Planck time. Below those scales, physics as we know it just stops giving confident answers — partly because trying to look that closely would require so much concentrated energy that you'd accidentally create a black hole before you got your measurement, and the black hole swallows the information you wanted. Some physicists have actually pointed out that this is structurally identical to what happens when you zoom into a screen: you don't see infinitely fine detail, you see pixels and a refresh rate, a hard floor built into the display itself. If reality has a similar hard floor — and it does, the Planck scale is real — that's at least consistent with the idea that something underneath is running on a kind of fixed "resolution." It doesn't prove it. But it's not nothing either, and that's the honest, non-meme version of the pixel argument.

Version two: the lazy-loading thing, and this one I find genuinely more compelling as an argument, weirdly enough, because it's borrowed directly from how real modern games are actually built rather than from vague Planck-scale poetry. Huge open-world games — researchers studying this specifically point to something like No Man's Sky, which has something like eighteen quintillion procedurally generated planets — obviously don't pre-build their entire universe and store every inch of it somewhere. That'd be insane, computationally. Instead they use procedural generation: detail only gets computed the moment a player is actually there to look at it, at roughly the resolution the player can actually perceive. Walk toward a mountain in a well-built game and you'll notice it sharpens up as you get closer — not because it didn't "exist" before, but because the game didn't bother rendering the fine detail until your presence made that computation necessary.

Some researchers have actually proposed this as a serious, formal model for how an efficient simulation of conscious beings would work, if you were designing one: render detail only when and where it's being closely observed, and scale the complexity to the number of observers and what they're actually paying attention to, rather than rendering the entire universe in full detail all the time everywhere, which would be a wildly wasteful way to build anything. It's not "the universe is like a video game" as a vague comparison — it's a specific, almost engineering-style claim: if you had to build a maximally efficient simulation of conscious minds, this lazy-loading-on-observation approach is the obvious architecture. And the weirdest part is that this idea connects directly — and not coincidentally — to the single most popular piece of "proof" people cite for all of this. Which brings us to the part of this post I was honestly most nervous to write, because it's the one I had the most wrong before researching it properly.

The double-slit thing. Okay. Let's actually fix this.

If you've spent any time at all in this corner of the internet, you've heard some version of: "particles act differently when you observe them — doesn't that mean the universe waits to render itself until something's looking?" This is, by a mile, the most repeated piece of "evidence" for the simulation theory that I came across in this entire research process. It's also — and I want to be straightforward about this because I believed the popular version myself before digging in — based on a real experiment, wrapped around a popular misunderstanding that's so widespread it's basically become folk wisdom.

Here's the actual experiment, the real one. You fire tiny particles — electrons, photons, whatever — at a barrier with two narrow slits cut into it, with a detector screen behind to catch them. If you don't track which slit each particle goes through, over time the particles build up a pattern on the screen that looks exactly like interference between two overlapping waves — stripes of high density and low density, alternating. Which is bizarre, because you fired individual, separate particles, not waves. Each one somehow behaved like it spread through both slits at once and interfered with itself. Richard Feynman once said this result is "impossible, absolutely impossible to explain in any classical way" and that it contains "the heart of quantum mechanics." This isn't fringe science. It's one of the most reproduced experiments in all of physics, going back over two centuries to the version Thomas Young first ran with light in 1801.

Now the part everyone's heard: if you put a detector at the slits to track which one each particle actually goes through, the interference pattern disappears completely. The particles suddenly behave like normal little classical particles, just going through one slit or the other, no weird wave behavior at all. Measuring which path it took seems to destroy the wave-like behavior. That part is also completely real and not in dispute.

Here's the correction, and it's the single most important thing I learned in this whole research process: "observer" in this experiment has never meant a person, or a mind, or consciousness, ever. It means any physical interaction that records, even in principle, which path the particle took — a detector, a sensor, a stray particle bouncing off it, anything. A camera that nobody ever even looks at the footage of still destroys the interference pattern, just as completely as a physicist staring at a live readout. The thing that matters is whether the path information physically exists somewhere, recorded in some other system — not whether a conscious mind is aware of it. Physicists call the underlying mechanism decoherence, and it's a real, well-studied process that explains why this happens without needing anyone's consciousness involved at all.

I want to be honest: this correction actually deflated me a little when I first found it, because the "consciousness renders reality" version is so much more dramatic and quotable. But the corrected version is, if you actually sit with it, almost stranger — it means reality at the smallest scales doesn't behave like a fixed, fully-detailed thing sitting there waiting to be looked at. It behaves like something that's genuinely undetermined until it interacts with something else. That's real, completely unresolved physics — there's a whole open question called the measurement problem about what's actually happening underneath all this, and a small number of serious physicists, including the genuinely legendary Roger Penrose, have wondered out loud whether measurement leaves some kind of physical "imprint" on the structure of reality itself. That's a real, live, unresolved question in physics. It's also a much narrower and weirder claim than "the universe renders graphics for you personally," and conflating the two is exactly how a genuinely fascinating hundred-year-old physics mystery gets flattened into a much less accurate internet meme.

There's also a small set of studies — mostly tied to a researcher named Dean Radin — that claimed human intention, specifically, not measurement devices, can shift the interference pattern in lab settings. I looked into these too, because I wanted to be fair to every version of the argument. They've drawn serious methodological criticism from other scientists, including specific statistical problems with how the comparisons were structured, and haven't been independently replicated to a standard that's convinced the wider physics community. I mention this mainly because this is the exact thread that a lot of viral content quietly blends together with the legitimate, decoherence-based phenomenon, and untangling the two turned out to matter a lot more than I expected going in.

So: does the double-slit experiment prove we're in a simulation? No. Does it prove reality is weirder and less "fixed in advance" than common sense wants it to be? Completely, unambiguously yes, and that's been true since long before anyone had a computer to compare it to.

Can a simulation even have someone "home" inside it?

This part isn't about physics at all, and I think it might actually be the strongest objection in the whole debate, stronger even than the physics stuff, because it doesn't ask "is this testable" — it asks "does this even make sense as an idea."

The whole simulation hypothesis needs something philosophers call substrate independence to be true: the idea that consciousness is about the pattern of information processing happening, not about the specific physical material doing the processing. Like how a song is the same song whether it's vinyl, an MP3, or a live performance — the pattern is what matters, not what's carrying it. If that's true, then a sufficiently detailed simulation of a brain wouldn't just model a mind from the outside. It would actually produce a real, felt, first-person mind on the inside. One exactly as real to itself as yours is to you.

Interestingly, even Hossenfelder — who thinks the whole simulation idea is basically nonsense from a physics standpoint — doesn't think this particular part is the problem. She's said, more or less, that consciousness might just be a property of any system that processes enough information the right way, and it might not matter if that's neurons or transistors or transistors that think they're neurons.

But this is exactly where philosopher David Chalmers's famous "hard problem of consciousness" shows up and complicates everything. The hard problem isn't about how brains process information — that's the comparatively easier question. It's about why any of that processing comes with actual felt experience attached to it at all. Why does a pattern of signals feel like something — feel like the color red, the specific ache of missing someone, the particular experience of being you reading this exact sentence — instead of just happening mechanically with nobody home, the way a thermostat reacts to temperature without (presumably) experiencing anything about it.

If consciousness really is just about the pattern, silicon could host a mind exactly as real as a brain does. But if it's somehow tied to something more specific about biological brains that we don't understand yet, a simulation could behave perfectly, indistinguishably, while having genuinely nobody inside experiencing any of it — what philosophers sometimes call a philosophical zombie. Same outputs, same behavior, lights off inside.

This actually matters a ton for the whole trilemma, because the entire argument is about counting minds — you being one mind among billions of simulated ones. If simulated minds don't actually experience anything, there's no population of experiencing observers to be statistically lost inside. The third option in Bostrom's trilemma doesn't become less likely. It becomes meaningless, because "simulated observer" turns out to be an empty category — a perfect behavioral copy with no actual experience inside it.

And nobody knows which of these is true. Not Chalmers, not Hossenfelder, not anyone. The hard problem of consciousness remains, by genuinely wide agreement, completely unsolved. Which means the entire simulation hypothesis is built on top of a question that might be the single hardest open problem in all of science and philosophy combined. That's not a small asterisk. That's basically the whole foundation, still unverified, sitting underneath everything else we've talked about so far.

Why do genuinely smart, famous people keep taking this seriously?

I went looking specifically for who talks about this publicly, because I figured it'd mostly be internet weirdos, and honestly the list surprised me.

Elon Musk has said there's a "billion to one chance" that what we're experiencing is base reality, directly citing Bostrom's paper as his reasoning. His argument is basically Bostrom's, compressed into a video-game observation: in under fifty years we went from blocky Pong graphics to near-photorealistic open worlds. Project that trend forward even a few thousand years, he's argued, and "indistinguishable from reality" stops sounding speculative and starts sounding like an engineering inevitability.

Harvard astrophysicist Avi Loeb has a related but genuinely different idea — not that we're in a computer simulation exactly, but that our entire universe might have been deliberately created by some more advanced civilization, possibly through quantum tunneling that produces what he calls a "baby universe," which he's argued could help explain some otherwise strange features of our cosmos. What I liked about his framing is that it flips the whole emotional weight of the topic: it's not really about us being fake or lesser. It's about creation itself maybe being something competent civilizations eventually learn to do — which would make humanity, someday, the ones doing the simulating, not the ones being simulated.

Astrophysicist Neil deGrasse Tyson has called the odds "pretty high" in public talks while being careful to stress it's still completely unproven speculation, not settled science. And philosopher David Chalmers wrote an entire serious book on this, called Reality+, defending something he calls "simulation realism" — which is maybe the single most important idea in this whole post, and it gets its own section, because it quietly defuses the part most people assume is the scariest implication of all of this.

Wait — if it's true, would it actually even matter?

This genuinely changed how I think about the whole topic, so stay with me here.

Chalmers's argument, the short version: even if we are in a simulation, that wouldn't make anything that matters to you less real. Your grief was still grief. The people you love are still real relationships, made of real shared experience, regardless of what's running underneath the hood. What makes something "real," in his view, isn't the material it's built from — it's the pattern of experiences and relationships happening within it. Reality is about what's happening, not what it's happening on top of.

There's a comparison that helped this click for me: nobody thinks a relationship that started on a dating app is "fake" just because an algorithm running on some server you've never seen was the thing that introduced two people. The mechanism underneath doesn't decide whether what happened afterward was real. Maybe the same logic just keeps going, all the way down.

But — and this is the part that keeps the unease alive even after Chalmers's reassurance — there's a real difference between "the material underneath doesn't matter" and "the existence of someone, or something, running the whole thing doesn't matter." If there genuinely is a "base level" running this one, whoever or whatever is there could, in principle, change the rules, pause everything, or end it. That's not just a fun philosophical detail. That's a kind of power imbalance that makes every other power imbalance in human history look small. Chalmers's comfort about meaning doesn't automatically extend to comfort about safety, or about whether you're being watched, or recorded, or used for something you'll never get to find out about.

So which feeling is the right one — the calm "it wouldn't really change much," or the colder feeling underneath that, the one about not knowing what's watching or why? I think the honest answer is both, depending on which hour of the night you're asking yourself. I don't think that tension is supposed to go away. I think it's just what thinking about this seriously actually feels like.

The glitches people swear they've noticed (and why I don't think they count)

I have to address this because it's genuinely the most common version of this whole topic that regular people bring up, and I want to be honest that it doesn't hold up, even though I get the appeal completely.

Déjà vu has a real, boring, well-studied neurological explanation — it's basically your brain's memory and recognition circuitry briefly misfiring and tagging a brand-new moment with a false sense of "I've experienced this before." It's been reproduced in labs and tied to specific, identifiable patterns of brain activity. No simulation needed. Neuroscience already explained this one, years ago.

The Mandela Effect — when tons of people misremember the same detail the exact same way — is a memory and social-contagion thing, not a glitch in the universe. Human memory doesn't record like a camera. It rebuilds itself every time you recall something, and a similar-sounding word or visual cue can make huge numbers of people misremember in the exact same direction completely independently of each other. This has been studied directly. It's interesting. It's not evidence of a patch note in reality.

Here's my actual issue with "glitch-spotting" as a category, and it's the same issue actual physicists have with the broader hypothesis: an explanation flexible enough to account for literally anything strange that happens to anyone isn't really explaining anything. If it can never be wrong, it can never really tell you it's right either. That's the unfalsifiability problem again, just showing up at the pop-culture level instead of the physics level — and on this specific point, I think the skeptics are just straightforwardly correct.

None of this means the serious version of the hypothesis is wrong. It means the "glitch" version is a much weaker, different claim wearing the same name, and mixing the two up is probably the single biggest reason this entire topic gets dismissed by smart people before they've even read what Bostrom actually argued.

What I actually think about all this now

Philosopher Eric Schwitzgebel has a phrase for the posture I've landed on myself, even though I found it after I'd already started feeling this way: "1% skepticism." The idea is that the simulation hypothesis belongs in the same drawer as other classic skeptical scenarios — the evil demon, the brain in a vat — where yes, it's logically possible the world is radically different from how it appears, but there's no special reason to fixate on this one particular possibility over any other unfalsifiable one. You give it some small, nonzero amount of belief, and then you get on with living, with thinking, with doing the things that are actually checkable, while leaving a small permanent asterisk hovering over all of it.

There's something I actually find kind of comforting about that asterisk, once you stop fighting it. It's the same posture good scientists hold toward basically all their best theories — not blind faith, not dismissal, just genuinely open, permanently willing to be wrong, permanently aware of exactly where the evidence actually stops and the speculation starts.

And maybe that's the real point of spending three weeks on a topic that sounds, on its face, like a dorm-room conversation at 2 a.m. It doesn't hand you an answer. It hands you a much better set of questions than the ones you walked in with. What actually makes something real — the material it's built from, or what's happening inside it? Is consciousness something only biological brains can do, or something that could run on literally anything detailed enough? Does it even matter whether you can ever gather evidence for an idea, or only whether the idea holds together logically? And if it really is layers all the way down, with no privileged "real" floor anyone could ever actually locate — does that thought scare you, or does something about it, weirdly, feel like it lets you off the hook a little?

Bostrom never gave anyone a verdict. What he built was closer to a mirror, angled just right, so that assumptions about reality we usually never even notice we're making suddenly become visible for a second. Whether we're heading for extinction before we ever build the kind of machine that could simulate us, or heading toward becoming the bored, simulation-running gods ourselves one day, or are already the digital descendants of some civilization that's long gone — the trilemma basically guarantees that whatever's actually true is stranger than anything we're comfortable sitting with. No matter which of the three doors you end up standing behind.

I don't think that's something you're supposed to solve. I think it's something you just carry around with you, quietly, for the rest of your life — a low hum underneath ordinary days that gets a little louder some nights, usually right around the moment you're falling asleep, usually right when you start wondering whether the falling-asleep feeling itself is something you're actually experiencing, or something being rendered for you, one frame at a time, by something you are never, under any circumstances, going to get to meet.

Anyway. That's everything I found. Let me know what you think, because I genuinely don't know what I think yet, and that might be the most honest sentence in this entire post.

WHICH DOOR ARE YOU STANDING BEHIND?

select_option --door [1, 2, 3]

1 Humanity is doomed to wipe itself out before hitting post-human scale.

2 Future civilizations have the tech but ban running conscious entities.

3 You are a running execution instance on an optimized, lazy-loaded server.

my own skepticism (0% = Normal, 100% = Neo)

• Before the midnight video: [█░░░░░░░░░] 10% (Yeah, simulation is a fun sci-fi meme)

• After reading Bostrom's Trilemma: [███████░░░] 70% (Wait, math-wise... door number three is looking tight)

• After Sabine Hossenfelder's critique: [██░░░░░░░░] 20% (Okay, it's just physics storytelling)

• Where I live now: [█░░░░░░░░░] 10% (1% technical skepticism, 9% existential comfort)

Further reading, if you want to go down the same hole I did — click any name for their Wikipedia page: Nick Bostrom, Sabine Hossenfelder, Sean Carroll, George Ellis, Scott Aaronson, John Archibald Wheeler, Edward Fredkin, Seth Lloyd, Melvin Vopson, Roger Penrose, David Chalmers, Elon Musk, Avi Loeb, Neil deGrasse Tyson, and Eric Schwitzgebel.