On the 24 Nov 2023, at the ICOBI 2023 at NSBM Green University, I explored the exciting potential of Generative AI in reshaping education. We delved into how these advancements can not only enhance learning experiences but also address key challenges like the 2-Sigma problem. It's a journey of combining innovative technology with the indispensable human element to create a more inclusive, personalized, and effective education system. Let's embrace this digital transformation for the betterment of our society.
ICOBI 2023 Keynote: Generative AI in Education: Catalyzing Digital Transformation for Societal Advancement
1. Keynote: ICOBI 2023, NSBM Green University of Sri Lanka - 24 November 2023
Generative AI in Education:
Catalyzing Digital Transformation for Societal
Advancement
Roshan G Ragel, BSc Engineering, PhD
Professor | Computer Engineering
University of Peradeniya
2. Senior Lecturer (since 2007), Professor (since 2017),
Department of Computer Engineering, University of
Peradeniya
IEEE Affiliation: Professional Member (since 2005), Senior
Member (since 2014)
Research Focus: IoT, Wearable Computing, Bioinformatics, AI
& ML
Professional Roles:
○ Consultant CEO, Lanka Research and Education
Network (LEARN)
○ Member, Presidential Task Force for Education 2020
○ UGC Standing Committee on IT & Online Education,
since 2020
○ Member of the Presidential Committee on AI for
Economic Development
○ Member of the Expert Panel for UNESCO Roundtable
for AI in Education for APAC
Prof. Roshan G Ragel: Biography
4. • Benjamin Bloom’s 2-Sigma problem (1984)
• Bloom's Observation: Average tutored student outperforms 98% of the control class
The 2-Sigma Problem
5. • Central Challenge: Replicating one-to-one
tutoring benefits on a larger scale
• Resource Intensity: Bloom's recognition of
impracticality in conventional systems
Urging educators and researchers to find
scalable solutions with Digital
Transformation (Dx) and Generative AI
Challenge Posed by the 2-Sigma Problem
6.
7.
8. Primary Goals of Digital Transformation in Education
• Digital transformation aims to make education more
accessible and inclusive through technology, online
learning and assertive tools
01. Enhance accessibility
and inclusivity
• Education institutions use data for informed decision
making, optimizing resources and improving student
outcomes
02. Drive data-driven
decisions
• Digital transformation removes geographical barriers,
fostering cross-cultural learning and global networks
03. Globalize Education
• Institutions streamline expenses through efficient
resource allocation and the adoption of cost-effective
solutions like open educational resources (OER)
04. Manage Costs
9. Mastery of Material
● We should make a system where the only possible grade
is an ‘A’
○ It is not simply giving A to everyone
○ Rather, everyone should be allowed to reach to get an ‘A’
● There is no point of we recording the followings
○ Who is a fast learner
○ Who started to learn a material first (such as before starting the
course)
10. In the current system, at the time of the exam,
All
three
are
smart
14. • Benjamin Bloom’s 2-Sigma problem (1984)
• Bloom's Observation: Average tutored student outperforms 98% of the control class
The 2-Sigma Problem and Personalized Learning
18. Source: U.S. Department of Education, Office of Educational Technology, Artificial Intelligence and Future of Teaching and
Learning: Insights and Recommendations, Washington, DC, 2023. https://shorturl.at/aFJU5
Future Technology Capabilities
Familiar Technology capabilities Future Technology Capabilities
Input • Typing
• Clicking and dragging
• Touching and gesturing
• Speaking
• Drawing
• Analyzing images and video
Processing • Displaying information and tasks
• Sequencing learning activities
• Checking student work
• Assisting students and teachers
• Planning and adapting activities
• Revealing patterns in student work
Outputs • Text
• Graphics
• Multimedia
• Dashboards
• Conversations
• Annotating and highlighting
• Suggesting and recommending
• Organizing and guiding
21. The core issue - Resource Constraints
• Inadequate teacher-to-student ratios
• limited time for individualized instruction
• Insufficient financial resources
Guskey and Pigott (1988), Slavin (1987) highlighting resource challenges
Highly effective, one-to-one tutoring
"too costly for most societies to bear on a large scale" (Bloom, 1984)
GenAI: Resource Revolution and Beyond
22. • Beyond financial investment: Structural and systemic factors
• Introducing GenAI as a potential transformative solution
• Automating content generation and adapting to individual learner
needs
• Emphasizing the need for innovative approaches in educational
practices
GenAI: Resource Revolution and Beyond