Time travel could lead to semi-paradoxical situations occur. Like someone going back in time to save computer files. Instead of being "saved" or "not saved", it's sometimes they are saved, and sometimes they are lost. If we are in a simulation, the more we travel back/forward in time, the more rendering has to take place. And the rich owners of the systems wouldn't want to hpay $1 million ($100k in todays dollars, for example) for rendering minimal useful data. Compute in the future will probably dwarf present computers, but still isn't infinite. So, you might think that a simulation can't hold 8 billion people. But have you ever interacted with even half of the 8 billion people? You probably have interacted with under 10,000 people. And seen less than 1 million at any one time. So a well designed simulation will only load the extra people when needed to, and stick with a few thousand, or hundred. Since so many people, including me, spend a lot of time on the internet, it will be even easier to simulate or model it. Maybe sleeping helps make it easier to model us people in whatever simulation we are in. Glitches my be rare, and even if we are able to break out, we'll need to understand how to create a new simulation which we control, and enter it ourselves. For me, I'd add friends to the simulation, create a few buildings, like houses, bunkers and common areas, and have a lot of fun interacting. I would also program a website where you could send shell commands, text, media and mouse movements to the simulation. Whether we are in a simulation or not, if we are able to run a simulation ourselves (like creating an AI generated chat log, and using our imagination to build the environments around the chatters, etc), then I think there's a 50/50 chance we are in a simulation ourselves. Just because we can't yet simulate consciousness, doesn't mean we can't put ourselves into a simulation. There might only be 2 billion simulated people across 2,000 or more simulations, since whoever designed the simulation might only want to study a specific aspect of humanity, and doesn't need to be scaled up to a full population.