With increasing attention on more than 5 million IT workers and Artificial Intelligence (AI) in education, it should be in ideal position for India that is taking shape as a global caste in AI technology. But the US set the AI standard with a chat in 2023, and China quickly shut down the difference with its powerful chatbot Dipsek, India has so far produced a uniform big language model (LLM) that can mimic human communication.
And this is not for the lacquer of ambition. For data from market intelligence firm TracxN, the Indian AI sector includes 7: 114 startups, collectively increased $ 23 billion (€ 20.15 billion) in equity funding so far. Last year, Prime Minister Narendra Modi’s cabinet approves India Mission initiative With a budget of approximately $ 1.21 billion, “aiming to start the development and disintegration of indigenous large multimodal models (LMM) and domain-specific basic models in important areas.”
This week, CEO of Indiai Mission Abbishak Singh said that Indian startups need to think beyond the turf of their home for competition and suk pedal veterans.
“He will eventually have to compete with the best in the world,” Singh said at the Excel AI Summit in Bengaluru. “The initial level of support may come from the government, but it will not maintain them in a long time.”
Singh said, “When they are doing training models, they have to take into account a global vision.”
AI development needs industry, government and educationist
Representative of the National Association of Software and Service Companies (NASSCOM), The Voice of India $ 283 billion tech industry says that the creation of a globally recognized AI model is a complex, resource-commentary process.
“The argument is not whether India can catch, but can we move rapidly and define the AI identity on our own,” Nascom’s senior manager Communications Sanvada Satyaki Maitra told DW.
Last week, the Indiaai Mission announced the linking of 15.916 graphics processing units (GPU), which are required for AI research due to their ability to calculate parallels. The latest boost will bring the total national AI computing capacity to 34.333 GPU through public private partnership.
The discovery of startups as Gaan AI, Gannan AI, Sarvamai and Sochet AI, supported by India Mission, is constructing founding models in line with India, while firm is focusing on firm AI innovation like Sarvam AI, Fractal and Kovar AI.
“However, AI success cannot be achieved through isolated innovation,” Maitra said. “It requires harmonious cooperation between the government, industry and academics to build a full value chain, from calculation and data regime to model training and deployment of real world.”
Is Indian AI back?
The country’s leading cyber security expert Pawan Duggal told DW that India is likely to face the lack of high-end AI hardware, limited access to advanced GPU and lack of encephe. Cloud computing resources, which are required to train large -scale AI models.
“There is a lack of an important investment compared to global peers. While Indian AI startups have created it by enterprise capital investment, it remains a fraction of what is in the US or China,” Duggal said.
“The US invested $ 2.34 trillion and China $ 832 billion in 2014 and startups from 2014 to 2023, while India invested $ 145 billion in the same period,” He said.
Duggal believes that India is already moving towards creating its AI model, but it is not yet solving important challenges including infrastructure, funding, talent, data and regulation.
‘There are many minds in India’
Another issue in front of Indian engineers is the diversity of languages in India, one of the 22 official languages in the most populous country in the world. In addition, official languages form only a small group of more than 1.600 languages spoken within its limits.
A major custom software development company said, “The only use of a ‘Indian’ LLM is that if it works in our different languages, which is now difficult, then given that there is a lacquer data for LLM to train most Indian languages,” said a major custom software development company.
Shah said, “For an LLM in English, there are other companies and countries that are far ahead of us and will continue in this way.”
However, Util Vaishnav, a global technology holding company, is the real obstacles of the Upescare Technologies, that the real obstacles are the real “Risk-Shased Investors, Patch Data Rules and Tight GPU supply”.
“There are many minds in India. The GPUs are on the way and our multilingual data size is awaited.
Edited by: Darko Jenjeevic
