If you had to guess, how long do you think it would take for the U.S. Air Force to build and deploy a course curriculum? Whatever that guess was, maybe refine it with this new information: you’re building a course to teach already highly trained military personnel best practices for handling State-of-the art equipment (deliberate capital ‘S’, here). The answer, as we’ve learned, is 18 months of devoted effort for an experienced and professional team.
Now, what if you only had 3 months? That was the reality for the U.S. Air Force when they first found Growth Engine. Why the 6x time crunch? Well, let’s just say that things move quickly when you are dealing with U.S. Air Force equipment in many senses of the expression. R&D is constantly coming up with new and better solutions, strategy and tactics are constantly evolving, and there is a steady churn of highly motivated people advancing through the ranks of this high-speed machine. If the best of the best needs 18 months to prepare training for the best of the best on the latest and greatest, that becomes a bottleneck. In fact, a team of three Subject Matter Experts (SMEs) were tied up on this project, and the pace was not going to cut it.
So, the U.S. Air Force found Growth Engine, a new and advanced Natural Language Processing (NLP) solution, and a new vision was formed: Rather than a team of humans going through all the pertinent information and meticulously designing a training curriculum from scratch manually, what if that could happen programmatically? As in, what if a computer program could test and evaluate human understanding of a broad and technically dense collection of material? This is not a question anymore. From the perspective of Growth Engine, it never was a question, because that’s what Growth Engine’s Adaptive Training Platform does—it automates the creation of training content. And it does so with a speed that even the U.S. Air Force recognizes.
Before beginning to work with Growth Engine, the U.S. Air Force required about 62 days to complete about 100 relevant questions used to assess trainee understanding. To simplify matters, let’s treat these questions like they belong in a pop quiz from our schoolhouse days. Ok, so U.S. Air Force personnel are going to take a series of quizzes to ensure they understand the proper use of equipment before they can take an exam certifying their mastery of the equipment, then that equipment is placed under their control. With the complexity of the equipment that the U.S. Air Force deals with, and with the stakes involved in the use of that equipment, 100 questions does not cover much of the training regimen. Unfortunately, of these 100 questions, only 60 passed the intensive peer review process to find a place in the final draft curriculum. A tiny fraction of the required content to begin with had taken much too long, and then it had just shrunk again by another 40%.
Growth Engine’s Adaptive Training Platform produced thousands of questions right out of the gate. It produced more, even, than were necessary. Does this mean it created extra work in the sense that the body of “quiz questions” needed to be pared down? No, because the generated questions can be refined in a batch process by the SMEs and review board. This happens not in months, not in days, not in minutes, but in seconds. Essentially, the problem of question creation was not so much solved as eradicated completely.
What does this mean for training? From now on, it is possible for people to begin their training as soon as any new equipment, process, or protocol becomes approved and available. It is exciting to think what this will do for the pace of innovation within the U.S. Air Force!
What about the SMEs who were doing the curriculum creation work, manually? Are they out of a job? Far from it. These women and men may have been the most qualified personnel to create curricula, but that does not mean it was their only important work. Like most organizations, SMEs are rare and extremely valuable. Whether deep diving into an operations issue to resolve it in a timely fashion, engaging in quality assurance work, or optimization oversight, there is no shortage of mission-critical projects that can use their skill and experience.