ultrashort SF story: Superprank

The hardest thing to understand is how compressed our minds are. We're only adapted to deal with a tiny set of circumstances. Stuck inside our shrunken illusions, we barely notice these limits; let alone the vastness beyond them.
Most possible situations we could never handle at all.

The rumbling clouds were like a movie background. They slid past in waves between bits of blue sky, the bottoms dark gray with patches that were almost black in the distance. Occasional lightning flashes were like scenes from a Spielberg movie, but the rain just wouldn't start.
He had spent three hours digging a hole in his backyard for what would become a small swimming pool. Soon, he got lost in the job as the shovelled pile of earth grew higher. Next week the professionals would deliver the vinyl tub.
It seemed like a good time to take a beer from the cooler he'd placed on the terrace. He stretched his back, and sat down on the edge of the hole he had made. Leaves rustled in a gust of wind.

He had never known what a scream of absolute terror sounded like. There were many things he hadn't known. The sound seemed to come from all around him. First he looked up, then all around.
The dark mass on the horizon was moving so slow it seemed to be part of the sky. It was behind the speeding clouds, moving at its own speed.
A living mountain. While the sun was behind it, he could make out a texture of dark skin and monstrous scales. A golden flame emerged from the top and arced to the ground like a meteor. It landed somewhere over the horizon.
There was a sudden shock in the earth under him. Then the first sounds arrived, like the world being torn in half.

He realized he wasn't scared, but waiting for something. Just very curious how he would react. In fact there was no way to know how to respond to such a thing.
That was when he first knew he was in a simulation.

I HATE superhero movies

They should make a movie with a hero called Anti-Bureaucracy-Man, because the major supervillain in my life is having to do things that I don't want to do because they don't make anything better, but not doing them would make things a hell of a lot worse.

The solution to the Alignment Problem is to split superhuman AIs into subhuman (but brilliant) components

We should not try to create superhuman minds at this time.
Instead, we should try to create superhuman imagination modules, planning systems, simulators, data analyzers etc. These would be potential components of superhuman minds, but they would not be directly integrated.
There should be nothing there with a will or a sense of identity.
In the simplest terms, nothing that could theoretically feel pain. IMHO that has always been the most immediate risk of the vast "black box" training systems that AI researchers seem to fear - not civilization ending AI plots.