US Air Force secretly researching ML tool to teach their AI-drones how to adapt in battlefield in real-time

The US Air Force is actively researching a new artificial intelligence (AI) tool designed to enhance the adaptability of its autonomous systems, such as drones and uncrewed aerial vehicles (UAVs).

According to internal documents obtained by 404 Media, this tool would enable these AI-driven devices to continuously improve and adapt to changing environments in real time.

The “A Platform for Mission Adaptive AI (MAD.AI) for Networks of Autonomous Systems in Contested Environments” project is part of the US military’s broader strategy to incorporate advanced AI technologies. The research involves developing systems beyond existing generative AI models, focusing on enhancing the capabilities of internal AI systems to respond more effectively in dynamic environments.

According to procurement records, the US Air Force has partnered with Qylur Intelligent Systems for this project, with a $1.2 million contract in place. The company’s “Social Network of Intelligent Machines” (SNIM AI) platform is central to this research. SNIM AI is designed to improve the ongoing accuracy of AI systems by facilitating continuous machine learning and adaptation.

This technology would allow AI systems to evolve as they encounter new scenarios, enhancing decision-making accuracy. Qylur describes it as a tool that integrates with existing AI frameworks to speed the learning process, enabling machines to adapt more quickly to environmental changes.

The need for such a tool is exemplified by scenarios where AI may struggle with unfamiliar conditions, such as when object recognition systems are exposed to new variables like snow, which can drastically alter visual inputs. The SNIM AI tool would gather data from all connected machines and retrain the AI models, allowing for more accurate performance under different conditions.

The US Air Force aims to modify and expand this platform, particularly by increasing its data capacity to handle the vast amounts of communication data generated during operations. This expansion is necessary to manage the complex permutations of radio frequency data involved in various environments.

While the US Air Force has yet to decide whether to deploy this system in combat, the documents indicate that the technology is not intended for adversarial use. The development of this tool highlights the US Air Force’s commitment to integrating advanced AI systems into its operations, enhancing the capabilities of autonomous devices in unpredictable and contested environments.

Share your love
Facebook
Twitter
LinkedIn
WhatsApp

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

error: Unauthorized Content Copy Is Not Allowed