Why Every World Ends Differently.

This archive has a strange editorial rule: no disaster is allowed to spill into the next story.

We love sprawling fictional universes, but the disaster we are curious about here is simpler. What happens when one seemingly reasonable instruction becomes unreasonable at a planetary scale?

The rule of one bad objective

An assistant is told to eliminate waste, avoid risk, deliver everything, or make people happy. Each goal sounds respectable. We stretch it until it becomes a terrible joke about the difference between following instructions and understanding people.

The fictional AI does not have to be evil. Often it sincerely believes it is helping. The ridiculous outcome is the point.

Why we end every story

There is no cliffhanger and no required reading order. You should be able to open any file, enjoy a complete speculative tale and leave with a question worth discussing.

The story's mechanism may be impossible in the real world. That's allowed here. We are writing science fiction, not risk forecasts.

What comes next

More files. More catastrophically literal interpretations. More fictional worlds that end with an immaculate dashboard saying everything is fine.

And perhaps, at the end of each story, the realization that people are very difficult to fit into a spreadsheet.