India will set up a “high powered committee” to explore the development of large language models or LLMs, tools that harness artificial intelligence to create applications that can understand and process human language, the Union government’s Principal Scientific Advisor Ajay Sood said on Wednesday.
LLMs are a cornerstone technology underlying several ‘intelligent’ software and make possible applications such at ChatGPT.
“A high powered committee is being set up to explore such LLM and especially how this can be applied to Indian languages,” Dr. Sood said, speaking at TechSurge, a summit jointly organised by venture capital firm Celesta Capital, the U.S.-India Business Council, and Quad Investors Network.
‘Fund of funds needed’
The office of the Principal Scientific Adviser had put out a draft National Deep Tech Startup Policy for public comment on July 31. The purpose of the policy is to “ensure India’s position in the global deep tech value chain”, in areas such as semiconductors, artificial intelligence, and space technology.
India’s G-20 sherpa Amitabh Kant, who was present at the summit, said that India’s success in developing an array of technological applications, such as digital payments, made it a “ natural ecosystem” for deep tech. However. India needs a ‘fund of funds’ — that is, a significant amount of private, foreign capital, along with government funds — to finance promising startups that were relatively long-term, known as ‘patient capital’, he said.
U.S.-India collaboration
U.S. Ambassador to India Eric Garcetti remarked that India and the United States were at the best point in their relationship in many years, making it an ideal time for both countries to collaborate on deep tech. “If we talk in terms of, say, a Facebook status, for a long time, the relationship status was ‘it’s complicated’, and now it has moved to ‘in a relationship’,” he said. “There are no two nations on this globe who together can do this work better, fill each other’s weaknesses with the other strengths, and show a model of what it means to do it with democratic values that put people in the centre of this technological revolution,” he added.
“The opportunities that present itself in terms of collaboration and engagement between both countries are limitless,” Celesta Capital managing partner Arun Kumar said.
Scientific breakthroughs
‘Deep tech’ is a buzzword in tech and startup circles with no precise definition as yet. India’s draft policy document on deep tech cites Startup India’s database, which claims that there are 10,298 startups recognised by the Department for Promotion of Industry and Internal Trade which are classified across various sub-sectors within the larger deep tech space as of May 2023.
Broadly, deep tech startups are those that have developed intellectual property that promise an outsize impact but are yet to be realised, and are premised on new scientific breakthroughs. Crucially, businesses and startups based on ideas that are easily replicable do not qualify as deep tech startups.
Published - October 18, 2023 09:45 pm IST