Replacing the traditional classroom with AI augmented Education.

Let's stop trying to figure out how to insert AI into the traditional classroom and let's instead use a new model entirely.

AI augmented education represents a significant change in the education and training fields over traditional classroom and e-learning models.?It creates an educational environment in which each learner can achieve the?Mastery of a topic without the binding constraints of the traditional classroom. Traditional education approaches rely on structured, time released, subject siloed content, physical or virtual class participation and are based on a?one-teacher-to-many-students model. Alternatively, AI augmented education is a many-teachers-to-one-learner model providing a true?one-to-one?highly personalized educational approach. By the use of AI teaching assistants, the ability to support the learner with multiple teachers and AI tutors, AI augmented education has the potential to?eliminate grades, grade levels?and the?time constraints?imposed on learners for the completion of course materials.

This empowers the learner to work towards?Mastery?at their own pace.?

AI augmented learning effectively deconstructs and completely replaces the traditional classroom environments of today, whether brick and mortar or e-learning. AI augmented learning creates a dynamically interactive relationship between the teacher and the learner that fundamentally changes how education can be approached and will result in more effective knowledge transfer. AI augmented learning can also eradicate feelings of failure within learners as they strive towards Mastery?and success.

How we got here

AI augmented learning has its roots in the personalization technologies and database algorithms which have emerged within the last few decades. These systems over time begin to understand and responsively anticipate the interests and behaviors of the participant. In AI augmented learning, these personalization systems are designed to develop an understanding of the individual students’ learning styles, educational needs, personal interests, talents and abilities. In effect, AI becomes the learner’s personal tutor, and over time, begins to know the students learning strengths and weaknesses at a fundamental level. AI teaching assistants and AI tutors then provide increasingly personalized educational materials tailored specifically to the learner’s knowledge levels and learning profile.?

There are now very powerful AI systems available in the education sector - the issue is how they are being implemented.

Changing the model

To date, there has been no educational model designed to use AI to completely replace the current system of?one-to-many?with a?one-to-one?approach. Instead, AI, when installed, has been added onto the traditional education environment by?simply “bolting” the AI onto the existing classroom as a mere accessory. This is much akin to bolting a jet engine onto a horse and buggy - it can be done but it does not make any sense to do so. The failure to utilize AI to its fullest?one-to-one advantage is rooted in the current educational model’s reluctance and even inability to change. Taking this “bolt on” approach does not allow the AI to be fully centric to the student.?

In order for the learner to be centric to the educational system and fully utilize the power of AI augmented education, it is necessary to construct an entirely different approach where the human educator and a dedicated AI Tutor and AI Teachers Assistant provide increasingly individualized educational content and support to the learner.?

The instructor’s role, therefore, changes significantly when AI is used for ongoing?one-to-one?instructional interaction. Their role shifts from the “sage on the stage” to that of a supportive educator who is responsible for the success of their student in achieving?Mastery?of the subject - eliminating the arbitrary measurement of a grade as a measure. The educator is challenged to provide inspiration, excitement and enthusiasm (what I believe that teachers do best) as opposed to teaching towards a test. The learner benefits from individualized support and is also free to learn at their own pace. Education begins to resemble a research or library model and less like the education factory model in use today.

Bloom’s 2 Sigma Problem

It is of particular interest that AI also can provide a solution to Bloom’s 2 Sigma Problem (Bloom, 1984). In 1984, Benjamin Bloom studied an educational approach which significantly improved the effectiveness of the learning process. His approach delivered improvements by a factor of two standard deviations, or 2 Sigma, in student performance as depicted in the Achievement Score graphic below (Nebra, 2018). Average students using Bloom’s approach systematically outperformed students who were using traditional, instructor led classrooms.?Bloom’s approach was composed of two key elements,?Mastery Learning?and?One-on-one tutoring.

Mastery Learning. Mastery of the topic is required before the student can move onto the next advanced stage, there are no grades and no pre-supposed timelines, as the student can attain Mastery at their own individual pace.?

One-on-one tutoring.?The student is assigned a tutor who, in addition to providing course guidance, provides an atmosphere of encouragement and inspiration.

In spite of its obvious success, Bloom's approach was considered too costly in human and other resources to implement. In 1984 it was identified as an unsolved educational deployment problem and was therefore named the "2-Sigma Problem".

Today, armed with advanced AI, we can begin to eliminate these resource constraints. It is possible to envision an education model that provides each learner with an AI tutor and each teacher, or teachers, with an AI teaching assistant. This approach completely replaces the traditional and e-learning environments with a?one-to-one?system focused solely on the needs of the individual in order for them to achieve?Mastery?of topic.?

Using 21st century technologies and methods, The 2-Sigma Problem can now be solved.

The?Learner Centric Adaptive Learning Model – The Gordon Model

With AI it is feasible to enhance the?the initial AI learning environment to completely re-engineer how education is provided. These enhancements extend?the initial AI learning environment?with a number of collaboration and support mechanisms designed to assist the learner in achieving?Mastery within a field of study as well as their ability to become a productive and collaborative member of society.

These components include:

·??????A Content Library / Mastery Engine?that provides a common content and evaluation repository that is curated by both human and AI librarians.

·??????Subject Matter Educational Experts?(SMEE’s) that are composed of both human and AI experts who provide on-demand subject matter support focused on the needs of each learner. In other words, just in time education.

·??????A series of?External Interface?components for the health and maintenance of the educational environment.

·??????A Knowledge Mapping module: one of the most powerful features of this model?in that it is designed designed to provide a real time view into an individuals' abilities, capabilities, interests and learning abilities. This data becomes the road map for future individual development as well as a transportable data file for potential employers. It has the potential to eliminate the idea of a course transcript in that it represents the learners educational status in real time.

·???????A?Collaboration Universe?module to interconnect the learner with a globally connected group of learners pursuing the same field of research or study. This Universe is moderated by both human teachers and AI.

·??????A?21st Century Skill component ?dedicated to issues which are?designed to facilitate personal growth, civility as well as critical thinking elements such as media, information and AI literacy.

This extended and enhanced environment would provide dedicated focus to the learner to support them in their pursuit of?Mastery.

There are no grades, no timelines, no silos - there is only success.

Summary

We live in a world in which we expect, indeed, demand one-to-one?experiences in all that we do. The classical classroom education model was designed for a previous age, a mass-produced age of?one-to-many. This current education system has remained virtually unchanged for well over a century. With the advent of advanced artificial intelligence we can begin to implement a new education model - a model designed for the 21st century. This model attends to each individual’s needs and then molds itself as an extension of each learners personal interests and directly supports their continued intellectual growth. Indeed, AI fundamentally will require lifelong learning - to the point that it becomes learning for life.

An?AI education learning model that is focused on learner success?implies significant disruption to the 100+ year old education system. Large components of the existing education model would be eliminated, others created and teacher roles would be re-defined - and I believe education would be significantly improved and more fun. We will need more teachers, not fewer, and the current segmentation of students by arbitrary categories such as age, would be obsolete. Classrooms will begin to disappear and perhaps morph into a laboratory / library / research environment. Subject matter success is defined by progress towards?Mastery, not by the amount of time a learner sits in a room or by the ambiguous measure of a grade.

?“AI would have allowed me to follow my passion at my own pace. This system would have eliminated the feeling of me being a failure” –?student quote, research survey.

An AI education model would allow the learner to fully understand a topic instead of being forced to “move onto the next chapter” simply to keep the class on schedule. Additionally, it would allow the learner to chase their own curiosity. Teachers, then become even more important - and we will need a lot more of them. Human teachers will be responsible for infusing excitement, identifying insights and providing inspiration for each learner, as opposed to attending to administrivia and constantly testing.

Instead of failure being an option, success would be the?only?option.

I believe that learners, and educators, may again find excitement in being simply being curious. An insatiable curiosity that results in reigniting a love of learning.

Undoubtably, there will be resistance to re-engineering education and there will be valid concerns about how to get from the “as-is” to the “to-be” state, but the time has come to support each individual in their educational success. We must not allow stubborn adherence to traditional methods to deter us from moving into a new educational model. It is time to leave the 19th?and 20th?centuries behind and build an education system on the foundations of 21st?Century technologies.

We have the tools. The time is now.

Besides, this will be great fun as we learn to love learning again.

References:

Bloom, B. S. (1984). The 2-sigma problem: The search for methods of group instruction?as effective as one-to-one tutoring.?Educational Researcher,?13(6), 4.

Gordon, K. W. (2020).?A conceptual design for an adaptive learning technology implementation?model?https://search.proquest.com/docview/2435208951/previewPDF

Nebra, M. (2018, October 03). Solving The Education Conundrum Of Our Time: Bloom's 2??Sigma Problem - Class Central. Retrieved August 19, 2020, from https://www.classcentral.com/report/blooms-2-sigma-problem/

About: Dr. Kim William Gordon holds his PhD in Educational Leadership and is owner of EDTECH RESEARCH LABS, an AI augmented Adaptive Learning research lab.

Current independent research project: "Revisiting Blooms 2-Sigma Problem"

Dr. Gordon is available for Conceptual Design support, lectures and presentations.

Kim William Gordon, PhD?/ 001-314-570-8177 / [email protected] / https://www.dhirubhai.net/in/kimwilliamgordon/