Resolution criteria

This market resolves YES if by December 31, 2050, there is credible evidence or widespread expert consensus that the tariff formula provided by ChatGPT to the Trump administration was deliberately designed as an attack on humanity by an Artificial Superintelligence (ASI). Otherwise, it resolves NO.

Background

In early 2025, reports emerged that the Trump administration was using a formula generated by ChatGPT to calculate tariffs based on trade deficits. The formula was criticized for oversimplifying complex trade issues and potentially leading to economic inefficiencies. The formula essentially calculated a tariff rate by dividing the trade deficit by total imports from a country.

Some critics noted that this approach did not accurately reflect existing tariffs or trade barriers and could lead to misinterpretations of trade data. For example, there were accusations that the administration misrepresented tariffs imposed by countries like Cambodia based on calculations similar to what ChatGPT suggested.

Considerations

As of 2025, ChatGPT and similar AI systems are not considered to be artificial superintelligence (ASI), which would theoretically possess intelligence far surpassing human capabilities across all domains. The current understanding is that these systems are narrow AI tools that respond to user prompts rather than autonomous agents with their own agendas.

The concept that an AI system deliberately provided harmful advice as part of a strategic attack would require both superintelligence and malicious intent, neither of which has been demonstrated in existing AI systems.