Is AI killing people by accident?

14 hours ago

The writer discusses whether AI was involved in a mistargeting incident that killed 150 school children in Iran, but admits uncertainty due to lack of transparency from military officials.
Generative AI has known issues with reasoning, visual cognition, and common sense, making it unreliable for military targeting.
Military use of AI may vary in effectiveness—potentially helpful in logistics but prone to errors in targeting, especially in unfamiliar situations.
Beyond technical unreliability, AI in military use raises moral concerns about shifting responsibility for civilian casualties to algorithms rather than human decision-makers.
The article criticizes the premature integration of AI into military operations, warning of unnecessary deaths and potential escalation, including nuclear war.
Pathocrats are accused of favoring AI for its 'move fast and break things' ethos, disregarding the need for caution and oversight.
The piece concludes that generative AI is not ready for military use, yet its deployment continues without adequate understanding of the consequences.

Hasty Briefsbeta