The US government is exploring the power of artificial intelligence (AI) to expedite tasks that can take humans days to complete, according to a report by Bloomberg. Air Force Colonel Matthew Strohmeyer disclosed that a large language model (LLM) AI completed a task in just 10 minutes, which he described as “highly successful” and “very fast.” The Pentagon is reportedly testing these models with secret-level data, potentially leading to AI-assisted decision-making in the Defense Department.
Key Points:
- The US defense complex is testing artificial intelligence to accelerate tasks that typically take humans hours or even days.
- An AI large language model (LLM) successfully completed a task in just 10 minutes, as confirmed by Air Force Colonel Matthew Strohmeyer.
- The Pentagon is feeding AI models with classified data to see how they perform in real-time conditions.
- The Defense Department’s director of emerging capabilities policy office, Michael Horowitz, emphasized the desire to use AI “ethically” in supporting US defense capabilities.
- The State Department issued guidelines in February that included ensuring human control and oversight of AI tasks and called for accountability in the military use of AI within a responsible human chain of command and control.
