This is basically scientists re-discovering the "evil demon" / "brain in a vat" and not realizing it's kind of a tautological fallacy.
If life is a simulation, you can set an arbitrary boundary on how good you assume the simulation is, so any test you do would naturally be inadequate to confirm or deny the simulationness of life.
Couldn't it be the case that a lot of the scientists investigating this hypothesis are very philosophically-minded and are aware of all of these things and know it's probably non-falsifiable and futile? Even if you believe that, it still can't hurt to try. Science is all about testing things. If there are some things we can check, why not check them, just in case? And even if we don't get any closer to knowing whether or not we're in a simulation (which we probably won't), we could still learn other things in the process.
You can't disprove that you're in a simulation (e.g. "brain in a vat"). But it might be possible to prove (or at least provide strong evidence) that you're in one.
I mean, that’s not clear. It may be that anything which would obviously be interpreted by an outside viewer as “weird simulation stuff” would just be “advanced physics” to us.
Odd glitches would get rationalized away, in that manner. (The key being, the simulation is the only thing you experience, so there’s nothing to compare it against). So when you say “this universe is an advanced simulation,” you’re making an empty claim. (If everything I’ve ever experienced isn’t real, then what is the exact definition of “real”).
If you had blatantly obvious evidence that we were in a simulation, (such as escaping it), you would still not know that the new thing you are in is not a simulation. Or if it is just a further part of the same simulation.
It really comes down to the question of “how do I know what is real?” The answer is that we agree on what is reality, and we’ve decided that reality is the stuff around us. There’s stuff inside reality that we agree is “not real” but that stuff is all still “real” it just isn’t in the same form as your daily life.
Can a program prove it's not running on a vm? This discussion reminds me a Greg Bear novel I can't remember the name of, part of which was trying to figure out if you were an uploaded consciousness running on a vm or not. And there was something like a "person" running on a vm that was trying to break out in the real world.
Eternity, which was the sequel to Eon. A guy puts an alien consciousness into a VM in his head hardware, but it turns out those aliens are really good at detecting and escaping from VMs, and the alien takes over his hardware (i.e. body).
There's a big difference between a real brain perceiving artificial stilumation as if it were authentic sensation, versus a virtual simulation of a sentient entity that only exists as data in motion within a volatile memory store.
It's a far greater deception, for an entire universe of distinct sentient entities to exist as a stream of oscillations in a digital circuit, imagining themselves as mortal flesh, than it is for ten, or a thousand, or one billion fleshy organs to imagine a vicarious life unfolding before them, when they can't even move because they're suspended in brine, hooked up to fluid drips, and stimulated by electrodes.
On the one hand, you have a universe brimming with ephemeral entities that need not live as mortals, and on the other, you have limited fleshy blobs, easily placated by limitless sensation, that would have been miserable fleshy blobs whether they had arms and legs and a face or not.
The individual brain in a vat is better of in the vat, while the simulated universe is utterly tragic and should be destroyed immediately for all the undue suffering it needlessly replicates.
The universe in a VM can never know it's not a faithful replication of a higher genuine reality trying to predict future events by fast forwarding the simulation under an array of probabilities. Mostly because it wouldn't be a useful simulation if it could.
Let's say you can simulate planet earth with 100% fidelity, and you want to know if you should drive to work, or telecommute. So you fire up the simulation, and it shows you a fatal car accident. So you telecommute that day, and work from home. Are you ever going to let the simulated version of yourself learn that it's a simulation?
Nevermind that you could simulate bank robberies until you get away with one, or lotto tickets until you win. The point being that, you'd force your simulated self to look at the simulation that says "drive to work" and then watch what happens, while running it parallel alongside the version that stays home. You run it in fast forward, and 9 out of 10 times a freeway pile up kills you, but you score a million dollar bonus that day, based on an opportunity only available from the office. The saty home version misses out on the bonus that gets snapped up by someone else. You now know that starting the car is 90% dangerous, according to high fidelity simulations that are fully sentient copies of yourself, forbidden from realizing that fact, because their deception is essential to the risks they must verify. How do you know you aren't a deeper layer simulation to a higher version of your true self consulting a virtualized oracle of future events?
If life is a simulation, you can set an arbitrary boundary on how good you assume the simulation is, so any test you do would naturally be inadequate to confirm or deny the simulationness of life.