Centre plans to develop AI model tailored for India's use, Rs 2,000 cr allocated as initial investment

This ambitious plan will begin with an investment of Rs 2,000 crore, aimed to be launched after the ongoing 2024 Lok Sabha elections, according to the report.

With this initiative, the government plans to position India among nations like the United States and China, which have already started to work on similar programs.

The proposed endeavour may be led by the IndiaAI Innovation Centre, said the report.

This centre is likely to be set up under the Rs 10,000 crore IndiaAI Mission, the report said.

“The government will likely tap eminent higher education institutes and prominent researchers working on AI in the private sector to work on a foundational model,” a senior official said, the report added.

The initiative is envisioned as either a large action model (LAM) or a large multimodal model (LMM) with the aim to cater to the needs of a diverse range of applications and services, the official added.

Foundational models, basically pre-trained generative transformers, are the cornerstone for the development of other AI models and work by using existing data to generate new responses based on user inputs.

Addressing the specific requirements and preferences of India, the official stated, “This [foundational model] will aim to provide output in more than one native language, borrowing from all the work that has been done so far on projects such as Bhashini.”

The IT ministry launched Bhashini, an AI -based language translation platform and model in 2022.

As such, both private entities and governmental bodies develop foundational models.

According to data from the Stanford Center for Research on Foundation Models, more than 330 such models have been developed by private companies and governments across the world as of April 2024.

AI models developed by Microsoft-backed OpenAI, Google, and Amazon enjoy a competitive edge though they face difficulties such as hallucination, bias, and comprehension deficits.

The Indian government is looking to make use of publicly available data, digitised literature, and anonymised non-personal data obtained from a range of stakeholders to train its model.