The US Navy has recently announced its collaboration with General Dynamics Information Technology (GDIT) to address the pressure caused by inadequate manpower resources. They are introducing the company’s technology to develop a dedicated artificial intelligence assistant system.
US Navy and General Dynamics IT Collaborate
The US Navy, in partnership with General Dynamics Information Technology (GDIT), has launched a new artificial intelligence assistant system called Amelia, which utilizes the same generative AI technology as ChatGPT.
GDIT won the bid for this AI assistant system project from the US Navy in 2021. They integrated over 90 different network service platforms used by the Navy into a single platform. The development of Amelia, which cost approximately $136 million, aims to provide robust consultation services to all Navy personnel, Marines, and the general public.
Amelia has recently been adopted by several US banks, allowing customers to fulfill the majority of their financial service needs without relying on customer service representatives. The Navy and GDIT are currently testing Amelia in various scenarios that it may encounter. Initially, Amelia will handle data integration and query demands from over a million Navy personnel, Marines, civilians, and the general public, starting with basic permissions.
During the testing period, Amelia has demonstrated excellent processing efficiency, with over 90% of queries resolved through initial phone calls or text messages. The system can also detect the emotional state of users based on their tone. If it determines that a query cannot be resolved, it automatically transfers the call to a technician for assistance.
Read More
Currently, the Navy and GDIT are working on expanding Amelia’s training and database scale. They anticipate that in the future, the system will have higher security clearances and be able to handle more complex command requests, similar to the AI assistant “Cortana” in the classic game “Halo.”
Amelia is expected to go live between July and August this year. During this time, its performance will be demonstrated under various shared query conditions to assess how it compares to ChatGPT.
