A new academic program developed at MIT aims to teach U.S. Air Force and Space Force personnel to understand and utilize artificial intelligence technology. In a recent peer-reviewed study, program researchers found this approach to be effective and well-received by employees from a variety of backgrounds and professional roles.
This project is funded by the Department of the Air Force and the MIT Artificial Intelligence Accelerator to contribute to AI education research, particularly on how to maximize learning outcomes at scale for people with diverse educational backgrounds. We are aiming for
MIT Open Learning experts leveraged existing MIT educational materials and resources to create curricula for three common types of military personnel: leaders, developers, and users. We've also created new, more experimental courses aimed at Air Force and Space Force leaders.
MIT scientists then led a research study to analyze the content, evaluate individual learners’ experiences and outcomes during the 18-month pilot, and ultimately recommend innovations and insights that would allow the program to be scaled up.
They evaluated how 230 Air Force and Space Force personnel engaged with the course materials using interviews and several surveys provided to both program learners and staff. They also worked with his MIT faculty to conduct a content gap analysis to identify ways to further improve the curriculum to address desired skills, knowledge, and mindsets.
Ultimately, the researchers found that military personnel responded positively to hands-on learning. They appreciated the asynchronous and time-efficient learning experience that fit into their busy schedules. Although we strongly valued the experience created through team-based learning, we were looking for content that included more specialized and soft skills. Learners also wanted to see how AI directly applies to day-to-day operations and the broader mission of the Air Force and Space Force. They were also interested in having more opportunities to engage with other people, including colleagues, instructors, and AI experts.
Based on these findings, which program researchers recently presented at the IEEE Frontiers in Education Conference, the team will expand the educational content and add to the portal the next iteration of the study, which is currently underway and will extend through 2023. Adding new technical features.
“We aim to expand learning opportunities, not just from research questions, but from understanding the science of learning around the scale and complexity of projects. But ultimately, with the Air Force and We also aim to provide real translational value to the Department of Defense. This research has real-world impact for them, and we're very excited about it,” said MIT RAISE, dean of MIT's School of Digital Learning. said principal investigator Cynthia Breazeale, director of Responsible AI for Social Empowerment and Education and head of the Media Lab's Personal Robotics Research Group.
Build your learning journey
At the start of the project, the Air Force provided the program team with a set of profiles that captured the educational backgrounds and job descriptions of six basic categories of Air Force personnel. The team then created three archetypes that they used to build “learning journeys” — a series of training programs designed to give each profile a set of AI skills.
Lead-Drive archetypes are individuals who make strategic decisions, Create-Embed archetypes are technicians who implement AI solutions, and Facilitate-Employ archetypes are end users of AI augmentation tools.
“The priority was to convince lead-drivers of the importance of this program,” said lead author Andrés Felipe Salazar Gómez, a research scientist at MIT Open Learning.
“Even within the Department of Defense, leaders were questioning whether AI training was worthwhile,” he explains. “We first had to change the mindset of our leaders to make this training available to other learners, developers, and users. At the end of the pilot, we could see that they were embracing it. Their way of thinking has changed.”
The three six- to 12-month learning journeys combine existing AI courses from MIT Horizon, MIT Lincoln Laboratory, MIT Sloan School of Management, Computer Science and Artificial Intelligence Laboratory (CSAIL), and Media Lab. It included a combination of teaching materials. , and the MITx MicroMasters program. Most educational modules were delivered entirely online, either synchronously or asynchronously.
Each learning journey included different content and format based on user needs. For example, the Create-Embed journey included his five-day in-person hands-on course with a research scientist at Lincoln Laboratory, which gave us a deep dive into the technical content of AI. The Facilitate-Employment journey, on the other hand, is a self-paced, asynchronous learning experience. Primarily utilizes MIT Horizon material designed for a more general audience.
Researchers also created two new courses for the Lead-Drive cohort. One is a synchronous online course called “The Future of Leadership: Human-AI Collaboration in the Workplace.” Developed in collaboration with Esme Learning, the program is based on leaders' requests for more training on ethics and human-centered AI design, as well as content on human-AI collaboration in the workplace. The researchers also created an experimental three-day in-person course for him called Learning Machines: Computation, Ethics, and Policy. This immersed leaders in a constructivist-style learning experience where teams collaborated on a series of hands-on activities using autonomous robots. It culminated in an escape room-style capstone contest that brought it all together.
The learning machines course has been a huge success, Breazeale said.
“At MIT, you learn through building and teamwork, and I thought, what if we let executives learn about AI this way?” she explains. “We found that our work went much deeper and we developed a stronger intuition about how these technologies work and what it takes to implement them responsibly and robustly. I think it will deeply impact how we think about executive education around these disruptive technologies in the future.”
Collect feedback and improve your content
Throughout the study, MIT researchers used surveys to reach out to learners to obtain feedback about the content, pedagogy, and technology used. We also had MIT faculty analyze each learning journey to identify educational gaps.
Overall, the researchers found that learners wanted more engagement opportunities with peers through team-based activities or with faculty and experts through the synchronous component of online courses. And while most reps found the content interesting, they wanted more examples that were directly applicable to their day-to-day work.
Now, in the second iteration of the study, researchers are using that feedback to enhance the learning process. They design knowledge checks that are part of self-paced, asynchronous courses that help learners engage with content. We're also adding new tools to support live Q&A events with AI experts and foster community building among learners.
The team is also looking to add specific Department of Defense examples throughout the educational modules and include scenario-based workshops.
“How do we upskill our 680,000 employees in diverse roles at scale, across all levels? This is an MIT-wide problem that we at MIT Open Learning have been working on since 2013. We are leveraging world-class efforts to democratize education globally,” said Maj. John Radovan, DAF-MIT AI Deputy Director. accelerator. “By leveraging our research partnership with MIT, we will be able to study how best to educate our workforce through focused pilots, so we can quickly double down on unexpected positive outcomes and improve the learning curve.” We can build on lessons learned and this is how we can accelerate positive change for Airmen and parents.”
As research progresses, the program team is focused on how this training program can be realized on a larger scale.
“The U.S. Department of Defense is the world's largest employer, and when it comes to AI, it's really important that all employees speak the same language,” said MIT Horizon Senior Director and MIT Collective Intelligence Center Executive Director. Kathleen Kennedy says. “But the challenge now is to extend this to ensure that individual learners get what they need and stay engaged. It will help you see how it can be used with large groups of types.
