How Can AI Help Ensure Effective Education for Everyone?
Technology evolution is playing an increasingly important role in transforming education. Technology is making education more personalized by aligning the learning to individual students’ strengths and weaknesses. It has also played a critical role in responding to the disruption brought by Covid-19. Emerging technologies like Artificial Intelligence are now poised to deliver an even more democratized educational experience.
The Education Scalability Challenge
Historically, the great thinkers were taught one-to-one by their learned teachers. This model had constraints in terms of educating the masses. Even today, personal teaching is expensive and requires significant efforts from both the teacher and the student. Classroom teaching emerged as a poor but cheaper and easier means to educate the masses.
The availability of the internet addressed the scalability constraints of education. Students all over the world use Massive Open Online Courses (MOOCs) for their various learning needs. What lacked was the flexibility of teaching strategies to help students approach education to best suit their needs. With Deep Neural Networks, AI can offer many insights to the teachers effectively. These insights allow teachers to adjust their teaching approach and plan for the best learning outcomes for students.
The Academia Is Driving the AI Innovation
With the advent of capabilities, AI will be an integral part of the future educational landscape. It will be instrumental in assessing student capabilities, planning learning programs, and even predicting student performance. The predictions can lead to further adjustments to teaching and counseling when appropriate.
As in the past, universities are at the forefront of AI innovation impacting the future of education. For example, Stanford University is attempting to develop a mobile solution to facilitate inquiry-based learning and generate real-time analytics. Their Stanford Mobile Inquiry-based Learning Environment (SMILE) program will bring the power of AI solutions to portable devices.
Researchers believe educators will welcome the changes brought by AI. However, they warn that the digital divide may hinder adoption. Those with better device and infrastructure availability will benefit the most. Addressing this digital divide should be a critical concern for all.
AI and Education: A Historical Perspective
The earlier efforts to introduce technology through rule-based AI failed due to cost prohibitions. Such programs were available since the 1960s but were limited to research institutes. In the following two decades, researchers made many advances to develop programs that guided students through learning paths.
Current EdTech solutions utilise the same rule-based approach, implemented in the form of “Decision Trees.” They guide students to pre-defined learning paths based on pre-set performance criteria.
However, evolving “Deep Learning” networks and AI algorithms today can autonomously learn student performance patterns over time. Based on this learning, they can adapt the learning paths and content to optimize the teaching process.
These evolving AI teaching platforms’ growing capabilities can uncover better insights through deep and continuous student performance data processing. Better, deeper insights help students achieve even better outcomes than what they achieve through traditional classroom or even personal tutoring.
How Is the General Availability Evolving?
The efforts to roll-out practical, commercial solutions are also on the rise. Knewton, founded by a former Kaplan Inc employee and later acquired by John Wiley & Sons, was the first modern commercial application that used machine learning algorithms for such profound insights. The solution determined the right content to deliver for every individual student based on the answers they provided to Knewton.
Riiid, a startup from South Korea, developed algorithms to track student progress, understand their emotional state, and optimize engaging and motivational content. Over a million students in Korea and Japan adopted their English-learning app. Riiid has accumulated the world’s largest dataset of student-AI interactions as a result.
Riiid initially focused on the test-prep market. However, post-Covid-19, they received queries from all over the world to utilize their solution in multiple scenarios. EdTech solution providers are keen to deploy algorithms by Riiid to offer better solutions for personalized learning.
Riiid is conducting a global competition – The Riiid AIEd Challenge – to optimize the algorithms with experts and innovators worldwide. More than 1000 teams are participating in this challenge on Kaggle to develop a “knowledge tracing” algorithm using Riiid’s EdNet dataset. The top three teams will demonstrate their solutions at the Association for the Advancement of Artificial Intelligence Conference in February 2021 and share a prize of $100,000.
In the past, the ImageNet Large Scale Visual Recognition Challenge was responsible for advancements in Computer Vision. Self-Driving cars were the outcome of the US DARPA Grand Challenge. So, the Riiid challenge is one step forward in that series.
Another company that is leveraging the partnership with academic institutes is Korbit. Yoshua Bengio, one of the most prominent AI researchers who won the 2019 Turing Award (considered the Noble of computer science), established the Korbit. Korbit is carrying out this research in collaboration with Cambridge University and the University of Montreal.
Many similar start-ups are leveraging academic research in this area for commercial EdTech applications. More than $6 Billion expenditure is predicted on AI Education by 2025 (HolonIQ) through such start-ups.
China, as a leader in the sector, has invested over $1 Billion in AI Education. Yixue Education developed Squirrel AI in partnership with Carnegie Mellon University. The company established the CMU-Squirrel.AI Research Lab. It now has more than 2000 learning centers in 200 cities in China. They are many other such examples in China.
Transforming Education for A Better Future
There is broad consensus that these futuristic solutions will aid human teachers while extending personalized learning to students. They can fulfill the learning needs across the world, positively impacting millions of students.
With the adoption of non-text-based content like chatbots and Augmented and Virtual Reality techniques (AR/VR) for a more realistic experience, the focus is to make education more engaging and meaningful. These efforts are still in the nascent stages and will take decades to reach the sophistication levels that AI has the potential to make a reality.
If that potential is realized, AI can make education available to the masses through the internet. The AI solutions can usher in a new era through transformed education in the post-Covid world.