DAF-AI Accelerator hosts Learning Machines Course with MIT Media Lab

  • Published
  • By Master Sgt. Amber Williams
  • Department of the Air Force AI Accelerator
The Department of the Air Force – Artificial Intelligence Accelerator hosted two iterations of the Learning Machines Training course in conjunction with the Massachusetts Institute of Technology Media Lab, Nov. 28-30 and Dec. 12-14. These two sessions comprised the first round of this experimental course.

The three-day, hands-on immersive learning experience consists of highly structured coding projects, discussions on AI policy and ethics, and a capstone project that challenges Airmen to build and interact with their own machine learning models. The course was structured with skillsets such as acquisitions in mind, and for Airmen who already have an interest in coding and machine learning, though neither is a requirement.

The course material was previously created and offered by the MIT Media Lab, but the capstone project, which consisted of training drones for a simulated search and rescue mission, was added to exemplify real-life implications of the technology.

"The capstone project is a culmination of all the lessons they've had," said Maj. Megan Muniz, Chief Learning Officer at the DAF-AI Accelerator. "The capstone is their opportunity to immerse themselves with the technology."

"They learned how to train data and the fundamentals of supervised learning, reinforcement learning, and computer vision to a simulated real-world scenario," said Muniz.
"This course did a really great job of getting you immersed and hands-on with AI very quickly," said Capt. Chris Shaheen, a contracting officer at Hanscom Air Force Base and former fellow at the DAF AI Accelerator.

The task of integrating AI into the military needs the combined efforts of individuals from all career fields,” said Shaheen. “Working with such a diverse group from all backgrounds reinforces that together we can rise to the challenge.”

The DAF AI Accelerator plans to use course feedback and lessons learned from the Learning Machines Training to explore the possibility of a future course scalable to the entire Department of the Air Force. This type of training would give Airmen the skills to critically assess various applications of machine learning to augment daily operations in the future and build a foundation of trust between operators and AI capabilities.

“Shared understanding and trust are crucial across the spectrum of military operations. That is true when humans team with one another, and it is no different when humans are working with and relying on AI in operations” said Col. Garry Floyd, DAF-MIT AI Accelerator director. “This course is just one of the programs we've used at the AI Accelerator to help promote effective human-machine teaming... trusting AI starts with education... and I'm proud of our efforts in that space.”