India launched Bodhan AI , a homegrown artificial intelligence system, last week. This initiative aims to create a digital public infrastructure for education, comparable to foundational models like ChatGPT.
Bodhan AI: A Digital Public Infrastructure for Education
The Ministry of Education and IIT Madras unveiled Bodhan AI as a Section 8 non-profit company. Director of IIT Madras, V Kamakoti, stated the company’s goal is to ensure equitable, accessible, and quality education for all students.
Bodhan AI is envisioned as a Digital Public Infrastructure (DPI). This means it will serve as a foundational technology layer that others can build upon. Similar to how the Unified Payments Interface (UPI) underpins various payment apps, Bodhan AI will provide the core AI capabilities for educational tools.
Education Minister Dharmendra Pradhan referenced the National Payments Corporation of India (NPCI) and its UPI system as an analogy. UPI enables digital payments across different apps. Bodhan AI aims to do the same for educational technology, making advanced AI accessible for learning.
How Bodhan AI Will Function
Students and teachers will not directly access Bodhan AI. Instead, they will use applications developed by schools, states, and edtech startups. These applications will leverage Bodhan AI’s infrastructure to offer services.
Potential applications include analyzing student responses, providing personalized feedback, identifying learning gaps, and offering insights to teachers. The system could also assist with mentoring and career guidance.
IIT Madras’ Centre of Excellence in AI for Education is developing Bodhan AI. The infrastructure is expected to be ready within six months. The plan is to pilot the system with four states initially.
Role of Edtech Startups and States
Bodhan AI will create a foundational AI code, acting as the core infrastructure. Edtech startups will then build applications on top of this foundation. These applications will be the end-user interface, similar to payment apps like Google Pay or PhonePe.
An example provided involves an edtech app designed to analyze a student’s reading aloud. A teacher would record the student’s voice. The app would send the audio to Bodhan AI for processing, analyzing pronunciation and fluency. Bodhan AI handles the complex AI processing, providing feedback through the app.
The system will support multiple languages and process various data formats, including text, optical character recognition (OCR) data, and attendance records.
Addressing Accessibility Challenges
To ensure reach in remote areas with limited resources, Bodhan AI is developing “Frugal AI.” This approach focuses on creating AI systems that require less money, hardware, and infrastructure while still delivering effective results.
Development and Future Plans
Bodhan AI will provide a large language model (LLM) designed to be as effective as models like ChatGPT. This LLM will serve as a base for further specialization in areas like science, social sciences, and medicine.
The system is intended to be accessible to the public, functioning as digital public infrastructure. This ensures it is not limited to a select group of institutions.
The Bodhan AI Conclave brought together various stakeholders, including representatives from states like Maharashtra, Andhra Pradesh, Odisha, Karnataka, and Goa. These states are expected to be early adopters.
Safeguards will be integrated into the infrastructure to ensure its use is strictly for educational purposes. Audit mechanisms will monitor platform usage. Applications will undergo checks similar to app store reviews.