Whether generative AI can assist the U.S. military in planning responses to potential global conflicts or performing more mundane tasks such as providing quick access to inside information. started a test to confirm
Bloomberg reported on July 6 that the US Department of Defense and an unnamed ally are testing five AI Large Language Models (LLMs) for the first time in an experiment conducted by the Department of Defense Digital and AI Office. rice field.
Information about which LLMs are being tested is being kept confidential, but AI startup Scale AI has reportedly come forward with its “Donovan” model being one of the five.
Air Force Colonel Matthew Strohmeyer told Bloomberg that the LLM’s initial tests were “very successful.” […] Very fast,” he said, adding that the Pentagon is “learning that this is possible” but is “not ready for primetime at this time.”
One test Strohmeyer described confirmed that the AI model sent an information request in 10 minutes. This is breakneck speed as requests often take days and involve multiple people.
LLMs are already given classified operational information to generate responses about real-world problems. The test will see if it can help plan for a possible escalation of an already tense military situation with China in the Indo-Pacific.
Related: AI will choose Bitcoin over centralized cryptocurrencies — Tether CTO
Although the experiment is only scheduled to run until July 26, the U.S. military has been studying AI’s potential capabilities in warfare for some time.
In May, the UK government’s Defense Science and Technology Laboratory (Dstl) hosted the US and Australia for the first joint test of target tracking and detection by AI-equipped military drones.
Dstl said the trail achieved a “world first,” including live retraining of AI models during flight and AUKUS exchanging models. AUKUS “aims to rapidly introduce these technologies into military capabilities.”
Magazine: AI Travel Bookings Are Hilarious, 3 Weird Uses of ChatGPT and Crypto Plugins
