Do you remember when Six Sigma was hot? Nearly 20 years ago, Motorola University revitalized Six Sigma with modern information technology tools to drive $2 billion worth of improvements to their bottom line. “Digital Six Sigma” required training thousands of Six Sigma experts – black and green belts to design great products and improve existing processes. Training for black belts included four weeks of in-class coursework spread over four months, and in the off weeks they had to apply the learning to an important project sponsored by a senior leader.

For the green belts, Motorola University used all the latest e-learning and blended learning and won an industry award, but still struggled to help 5,000 people apply the new skills onto their jobs in over 50 countries around the world. This is because each learner is unique, with different skills, challenges and needs, and it was not affordable to help each one transfer their skills to the job in a systematic and personalized way.

While it is not a cure-all, new work in artificial intelligence (AI) has substantially addressed this classic training transfer problem by supporting learners in-between sessions with an artificially intelligent coach. When combined with a real-life, expert coach or teacher, there are two major types of AI that are now being used to help learners practice real-world skills in their jobs as a new form of electronic performance support: expert systems and deep learning “flight simulators.”

  1. Expert Systems

One new form of AI is a new type of calibrated coaching. After completing a short computer-adaptive assessment, each learner can pick a goal and schedule short, tweet-sized suggestions for how and what they should practice. On the mornings and days they prefer, learners get a push notification they can use to suggest fresh ways they can sharpen their skills throughout their work day. Each e-coaching statement is written by an expert and calibrated like an assessment instrument, so that it can target that unique proficiency level of each learner. In this way, e-coaching is always in that person’s “Goldilocks Zone” – neither too hard, nor too easy, but just right for their current skill level.

Further, learners can schedule a second push notification on their smartphone or tablet to remind them to journal about the lessons they’ve learned from trying to apply the artificially intelligent coaching to their jobs. Most learners prefer to capture these private journal entries using the built-in voice-to-text like Siri on their smartphone, another form of AI that makes reflecting on deliberate practice more natural, especially when the learner is tired at the end of a long day.

This mobile reminder serves two key purposes. First, it is developmental for learners to reflect on what they learned and refine their mental models. Second, the journal gives the teacher or coach a private, confidential window into how that specific learner is doing, and whether he or she can benefit from praise, nudging or other forms of support. In the past, teachers and coaches could only do this through methods that felt like nagging by phone, SMS or email spam to learners.

  1. Deep Learning “Flight Simulators”

A second new form of AI leverages computer science advances with deep learning to give people a sort of pocket flight simulator to allow people to practice new behaviors, get immediate assessment and qualitative feedback, and improve before having to perform real job tasks. In this way, the AI provides a safe place to practice difficult, dangerous or embarrassing skills in a private setting where only the coach or teacher can see the result through their web portal.

The first such flight simulator has just won the 2018 Society for Industrial-Organizational Psychology Bray/Howard Assessment Grant, for a prototype designed to measure and improve persuasion. Using the science from eminent professor emeritus, Robert Cialdini, of Arizona State University, the “Instant Coach” has a chatbot interface to help people improve their persuasive appeals before they need to do them in real life. The Instant Coach flight simulator has four different modes. One mode allows the user to simply speak into their phone and get immediate feedback on how effectively they’ve used Cialdini’s Principles of Persuasion, and gives concrete suggestions for improvement. Another group of modes are more structured to help a person who doesn’t know where to start, learn how to construct an ethical approach to influencing in a specific situation and measures how much progress they have made.

Because AI is developing rapidly, it is very likely that AI will continue to augment the context and relationship components that are important in training and coaching, to better mass-personalize people’s ability to transfer training into their real jobs and lives.

Stealing Jobs?

Some coaches and teachers are fearful that this advanced AI will someday take their jobs. But a more realistic perspective is that expert teachers and coaches have an irreplaceable role in understanding the context around the learner, connecting with the learner as a person, and providing socio-emotional support that no AI can do, and will not be able to do anytime soon. We feel obligated to each other, and only an expert teacher showing unconditional positive regard will be able to appreciate the struggles of a learner in applying job skills, and empathize, providing the most holistic scaffolding for skill development possible. That’s not the strength of AI, and we’re not likely to feel close to machines the way we appreciate our mentors, teachers and coaches.

But that’s not the purpose or benefit of AI. The best use of AI is to augment that relationship, where that AI is an extension of our relationships that we could not have without it. For example, by having each learner journal about the lessons they learned, mistakes they made, or pleasant surprises they found, the coach can remotely praise, nudge or suggest areas for improvement. Similarly, with “Instant Coaches,” that have been engineered to coach in a specific area, a very narrow focus. A coach can complement that narrow focus by looking at the pattern of learner progress over time, and adjust the next lesson or coaching session tailored to what that unique individual had mastered or struggled to improve upon, based on all the activity that has occurred since they last met.


AI is now practical to help address the chronic issue trainers face, in helping support them when they’re not in the classroom, MOOC or e-learning course. It is not a panacea, however. AI will not be able to understand context or social relationships like a real teacher or coach anytime soon and care needs to be taken to make sure AI is unbiased. Regardless, AI can be a tremendous help to close the gap between the class and the workplace.