The 6 Levels of Adaptive Learning Defined
What is Science Fiction and What is Reality?
What is Adaptive Learning? Is it the new “thing”? A quick Google search yields 51,200,000 results so clearly, it’s been around for a while. Is it science fiction? I have a colleague who imagines a Matrix-like world where knowledge is pumped into our brain as we need it – but without the big cables of course. One thing everyone agrees on is that advanced technology and AI is our future. But how do we get there?
When I want to see the future, I go to CES.
Each year around 180,000 professionals from all over the world make a pilgrimage to Las Vegas to visit CES, the annual Consumer Electronics Show. For 50 years, CES has been the global stage for innovation and since the first CES in New York City in June of 1967, thousands of technological milestones have been rolled out including:
- Videocassette Recorder (VCR) in 1970
- Compact Disc Player in 1981
- Tables, Netbooks and Android Devices in 2010
I don’t think anything rolled out at CES to date, including the smart phone, has promised as big an impact on people’s lives as Autonomous Vehicles (AV). Is it just car companies talking about AV? Not at all – Intel spent 15.3 billion dollars to purchase 84 percent of Mobileeye in 2017 to acquire camera and sensor technology needed for AV. In fact, Wired estimates that AV will add 7 trillion dollars to the global economy. Impacts include:
- 90% reduction in the 1.2m deaths annually from auto accidents (org)
- Insurance costs reduced by 80% (KPMG)
- 8 minutes of additional productivity per worker per day (based on US Census Bureau’s average commute time)
- 40% more time available for Police Officers to work on “real crimes”
- No more parking fines (your car will find and pay for the space automatically)
- Your car IS your pick-up and delivery vehicle for everything in your life (clothes, groceries, kids)
And perhaps the biggest benefit of all….
- No more waiting in line at the DMV to update your driver’s license. Priceless.
Despite all the hype, the promise of true autonomy on the roadways for the general public is likely 30+ years away. So why talk about it? Because it is driving the roadmap now! Features such as cruise control, lane drift awareness and collision avoidance systems are the baby steps to AV. In order to show how far down the path to AV we are, the Society of Automotive Engineers (SAE) created the concept of “autonomy levels” for vehicles that range from Zero to Five (six levels). Today’s most innovative, cutting-edge vehicles are still at Level One. Level 5 means that a steering wheel is optional.
Much of the current focus is on the ability to collect and disseminate data. In fact, “Unlocking the Power of Data” was Intel’s theme at CES last year. From a practical standpoint, that means that existing cars will need to be capable of monitoring their environment in a myriad of ways beyond what they do now. Sensors in the visible spectrum – located all over. Local sensors include radar, lasers, vibration, moisture, temperature – all relevant if the data is well-used. Connected data includes maps, traffic, and weather. In fact, Google’s driverless car prototype reportedly generates .75 Gigabytes of data every second.
Data and sensors are where AV meets AL (adaptive learning).
For Adaptive Learning to move from science fiction to reality, we have to start with the capture and dissemination of a massive amount of learning data. It may not be as much data as an autonomous vehicle needs but relative to the industry, it should not be underestimated.
And the possibilities are very exciting: just imagine being able to 1) benefit from autonomously crowd-sourced best practices for sales, customer service, medicine or even engineering, 2) assess an individual’s capabilities in real time or on the job using AI, and then 3) serve up the most relevant material in the right format with none of the fluff.
Unfortunately, the industry standard for success usually equates to “course completion” and currently learning and development captures and utilizes very little data. In other words, every employee who passes their on-boarding class should have the same ability to succeed on the job. Here’s an example: Google Drive (aka Waymo) sent their autonomous vehicle test car from Domino’s to pick up my pizza and delivered it to my house in 30 minutes or less. I got my pizza! In the L&D industry today, this would be considered “completion” and the training was a success.
But what happened in the process? Did a pedestrian get hit in the crosswalk while the light was malfunctioning? And what will happen next time or when a different vehicle is selected?
Clearly, we’re in the infancy of Adaptive Learning. But we think we can suggest a roadmap:
Level 0 – No Adaptive Learning. The learner is in charge and driven to seek out additional education that supplements initial experiences based on need or desire.
Level 1 – Passive Adaptive Learning is still learner-driven but provides for relevant content available for learners as in Massive Open Online Courses (MOOC).
Level 2 – Manual Adaptive Learning uses analytics to understand knowledge levels and learning styles, coaching, stretching and extending knowledge for individuals. Driven by an instructor or manager who guides the learner to additional content based on data analysis.
Level 3 – Active Adaptive Learning uses analytics to determine learner knowledge and automatically pushes relevant content, pathways and or mediums to fill the void.
Level 4 – Automated Adaptive Learning uses Artificial Intelligence to assess skill levels and gaps and automatically pushes the most relevant courses or experiences based on crowdsourced data of best practices to optimize performance.
Level 5 – Fully Automated Adaptive Learning uses AI and Machine Learning to assess skill levels and gaps to automatically push the most relevant courses or experiences based on crowd-sourced data of best practices to optimize performance, in real time.
Most companies are at Level Zero or Level One and few have the analytics to get to Level 2. That means that even if they wanted to react accordingly, they wouldn’t have the data available. (Although we would argue we’ve provided more data to our customers than any other L&D provider). And just like with Autonomous Vehicles, there is a huge leap from Level 3 to Level 4 with limited Active Adaptivity being a possibility now while truly responsive, automated adaptivity is still science fiction.
Establishing levels for Adaptive Learning is an important milestone for the industry. In the next blog post in this series, Jordan Fladell, CEO of mLevel will describe how this can play out in the real world and what you can do now to optimize learning through adaptive strategies.
If you want to see how mLevel is unlocking the power of learner data, click here to talk to one of our learning consultants.