We all share it: the survival instinct. And this means we kill. Every single day.
Each of us is kind of a big war robot, controled by a huge community of microorganisms that collaborate for their survival. Our immune system is their armee on the small scale, elliminating existential risks. Our body is optimised to nurture the whole community and to fight for the whole community's survival on a bigger scale, being only partly in control. (None of us can willingly control any of our microorganisms that make our body work. It is nearly impossible to go against the commands like hunger and thirst, need to sleep, sexual needs... and only live a rationally driven life.) Even as a vegan, you can't completely avoid to kill - be it by accident (like eating a worm in an apple or stepping on an ant) or just because plant life already contains organisms who fight for their survival as well. One could start argumenting that we can at least avoid the best way possible to hurt sentient beings. But are we equipped by nature to do so?
Physical level - when we believe we or our offspring becomes threatened - won't we be ready without hesistation to cause harm in order to protect us/our child? Doesn't this open the opportunity to any species capable of manipulating our perception to turn us into killer machines? On a more rough skill level, human beings do this already with each other. With propaganda and populism and the weapon of fear you can manipulate humans to do a lot.
Psychological level - Are we even able to live without suffering? Imagine a world where all human needs are fullfilled. Will we really be happy? Our happiness system doesn't seem to be build for eternal happiness. It is a system that shall guide us to avoid things that harm our "robot" and seek for benefitial opportunities. And it callibrates with the ups and downs we experience. It's the same with all senses: If we are in a silent room, our hearing will scale up so we suddenly hear our heartbeat. Our happiness system works similiar: I can painfully tell you that a dopamine high that's completely off the usual limits will only get back to balance with a heavy dopamine low afterwards. As a musician, I can tell you that after achieving dopamine-highs like playing first time at the big festival of your dreams and being pushed by several thousands of people in front of stage, I am not a happy smiling person for days, but instead the system shuts down and recalibrates afterwards - I am more likely to sit on my couch with a blanket, stare at my wall and try to adjust back to feel joy from drinking a cup of tea and feeling sad about an untidied kitchen - or worse even start feeling numb, empty and sad just because I am miles away from the feeling on stage or feel so much pain about whatever small thing that has gone wrong this time. (Some drugs shall have a similiar effect, but I never tried that, so I can't say anything in comparison about it.) So according to the white room, the "all needs fullfilled"-scenario means our dopamine system will adjust as well. And even if we practise to behave cautiously, a wrongly chosen word from another person will have the dimension of a drama because it is the biggest threat we experienced for days. At the best, we suffer, at the worst, things escalate in society and we develop enemies and fight wars about simple misunderstandings of daily life.
As a result, it seems impossible to avoid any conflict between humans. And it seems impossible to completely avoid the feeling of suffering for any person. So if we want to optimise the amount of days we spend in happiness, we have to go for a lot of very small ups in rising itensity (and sudden, stronger downs to recalibrate if it is not possible to climb up the ladder eternally). As this is hard to achieve in real life every single day, the best way to achieve that level of happiness is a simulation. We are designed to always optimise, to always want more. We are happy if we are in comparison a bit better than our competitors/fellows. (But of course in real life, all humans can't be better than their neighbours.) We have to be challenged to activate the system, but not so much that the goals seem (almost) impossible to reach (too ambitious goals and depression should sound familiar to some people here I guess?). And the best design would be if you achieve the biggest possible goal the day you die. We all have different interests that reward us the most, but I can't imagine a real world society where all desires can be matched. 
Bonus: If we build such a simulation, we could also include animals, as the happy catch for the cat never is the happy moment for the mouse at the same time.
So welcome to dreamland, the world that frees us the best way possible of suffering. To insure this, we can hand over control to our super intelligent machines that we so perfectly aligned to our needs in order to optimize our happiness creation constantly. Perhaps they can include some tasks they still can't handle in a good way in the game design until we reach the day they found more efficient ways to handle them and program us new sources of joy instead.
Is this the future we should strive for? It feels wrong and right at the same time to me.
(Perhaps that's why we didn't meet other species yet. They just enjoy their time in their dreamland. Edit: If there's no mechanism to create children, that would mean that those who die in the simulation are also the last of their kind. An ultimate time of happiness before the end of all life and therefore also the end of a lot of potential suffering.)