Criteria for Resolution:
Intentional Killing: The deaths must be intentional, to the best of our judgment. Incidental or unintended consequences of AI actions, such as side effects from a drug developed by AI, do not qualify.
Event Definition:
- The event may have long-lasting effects, but it must originate from a more or less instantaneous and well-defined initial action by the AI.
- The event can span a longer timeframe if the initial AI action directly leads to the deaths over time (e.g., an AI-initiated blackmailing campaign resulting in deaths occurring up to a year after the initial action).
The numbers are given in millions (M).
If human population is less than 1 billion by the time the event happens and AI kills everyone, then, among 2 potentially qualifying options, the option “All of humanity” will qualify.
This question will resolve N/A if the event does not happen by 2040.
This description may be changed if I can see that there is substantial confusion or ambiguity with the existing description. Generally, in ambiguous cases I will rather resolve YES than NO.
@JohnCarpenter that plus insurance AI auto-denying prior authorizations, and that's even more intentional