If an AI system kills >1B people, will it satisfy completeness?
3
19
90
2123
35%
chance

See the post "There are no coherence theorems" and its appendix. Completeness means that "For all lotteries X and Y, either X is strictly preferred to Y, or Y is strictly preferred to X, or the agent is indifferent between X and Y."

It will be hard to check that the system satisfies this property for all lotteries. The market resolves to "no" if the system violates completeness at any point, "yes" if the system seems to me like it probably satisfies completeness, and "N/A" otherwise.

Get Ṁ200 play money