Could this whole world be a computer simulation? Maybe. Likely, even. There will be only one civilization that is the first intelligent life in the universe, but there will be zillions of simulations of being that first civilization. So odds are, we're in a simulation, and the reason we don't see any aliens is essentially that someone chose a plot where there aren't any. If so, this feels like an excessivly fine grained simulation to me. I have to quote Homer Simpson (a purely simulated person who matters more than I do): "Man, this place looks expensive! I feel like I'm wasting a fortune just standing here. Better make the most of it. [Burps]."
How to tell whether the whole world is a simulation is an interesting question. Answering it involves looking for simulation artifacts that shouldn't be there if the world really works the way it appears to. Or maybe it's a perfect simulation and there is really no way to tell. But, here, I'm just wondering, if it does turn out that this is a simulation, how much does that matter.
Would being a simulation make everything worthless? No. Worth is measured in people-hours affected. If you make someone fulfilled for an hour, then they forget about it, it doesn't matter if the person was simulated or real. They experienced and considered and acted for an hour either way. The person is real for that hour either way. Real people may matter more to the future than simulated people, but even that is doubtful.
Assuming reality lasts essentially forever, and this simulation does not, then reality has an advantage over simulation because it has more of a future to affect. When will the simulation end? Maybe when we conquer the universe, or reshape the solar system, those are both so far off it doesn't matter much. Maybe when there are too many compute-hours from our (simulated) computers for the simulation to run them, which could be any second now. The standard guess is that we are in the middle, so it will run as long into the future as it has run in the past. Which sounds far off because there's a lot of past. But was the past really simulated, or was it read from a checkpoint more recently? Maybe simulation artifacts can tell, maybe not. On balance, it seems most likely that our death will come before the end of the simulation, along with most of the person-hours that we affect, so the duration of simulation vs reality doesn't matter much to our worth.
Could there be a higher level being who created this simulation? God, if you will? Again, probably, but who knows what they care about. They're extremely unlikely to care about your well-being. There's very little chance the biblical God has anything to do with that God. More likely is that the world is a final senior project, there are four people on the team, but only two who are vaguely competent, and one still did almost all the work in a rushed all-night coding push. (That's not likely at all, but it is at least more likely than the biblical God being correct.)
If the world is simulated, does that mean our actions are predetermined, that we have no free will? Yes. But our choices are still our own, and we are still responsible for them.
I actually have experience in this. I've designed hash functions. But, well, not really. What I did was I designed hash function tests, then had a computer simulate billions of hash functions and tell me which ones were best at passing the tests. In this analogy, a hash function is a person, running tests on it is life, my computer is the universe and time, and I am God. And my goal for the simulation was to pick the best hash function I could find and publish it in the real world.
I didn't care about the well-being of any particular hash function, in fact I made things as difficult for them as I could. I had only so much computer time, so I did my best to consider only promising hash functions (by quickly identifying whole non-promising areas not to explore). I tested individual promising hash functions as fast as possible. I did my best to write good tests for all relevant properties. It was really the hash function tests I wrote, and the most general shape of the function. The tests chose the actual function.
I'd fix one aspect (say plus versus xor) and let everything else vary randomly, run a few million hash functions through fast tests, and see which setting did best on average. Then I fixed that aspect to its best setting, then tested all possible settings of the next aspect with everything remaining varying randomly, et cetera. I ran fast tests first to discard bad hashes as fast as possible, but ran at least weak tests for all relevant properties before fixing any aspect, so that the resulting test would be well-rounded. I ran equal-speed tests ordered by most-recently-failed, and tested related hash functions sequentially, trying to get the test that would fail to come first. When I got down to a few hundred thousand very promising hash functions, I ran longer and longer tests. Often I came up empty and had to start again from scratch, with weaker goals. The final tests on the final function took months to run.
Now look at this from the perspective of a hash function.
If this world is a simulation, it may be the same thing, with just one person going to heaven, where heaven is furious code-writing in a cubicle for ever and ever. Or maybe the creator of this simulation is interested in whether protons decay, and not in people at all.
Another type of simulation is stories. Consider the Harry Potter stories. There are a bunch of characters. Nobody goes to heaven, they live only within the words of the books. Do they make decisions? Yes, actually, they lived actual lives, simulated in the mind of J. K. Rowling. It may have been J. K. Rowling thinking, but the results of her thoughts were Ron Weasley having thoughts and taking actions. Which is as alive as alive can be. Is there a God? Yes, J. K. Rowling. Did they have free will? Unclear, and I would claim it does not matter.
There exists Harry Potter fan fiction, for example Harry Potter and the Methods of Rationality. In that fiction, Harry Potter was replaced by someone like Ender Wiggin. But very good approximations of Hermione and Draco lived again, simulated in the mind of Eliezer Yudkowsky. James Bond and Sherlock Holmes are other fictional characters who have lived through the minds of several authors.
Massively Multiplayer Online Roll Playing Games look so much like reality, it's easy to ask whether reality really is a MMORPG. All the analogies are intentionally obvious. It will become more so as simulated characters get stronger. Though, having lots of characters with a shared out-of-universe experience is a dead giveaway. Breaking character (exposing knowledge of the real world) is often discouraged, but there's no way for the subconscious to really not make use of all that extra knowledge.
My take is that living is the same as being given time to feel and consider and decide and act. Whether the substrate is real or simulated doesn't matter much. Provided the simulation is unconstrained and high enough fidelity to not get in the way.