Top News

Delhi. Thousands of AI Startups have emerged in India. Despite this, the country has not yet created its own premium AI engine. Why are there no Indian AI models in India like a 50 million IT employees like Deepsek or ChatGPT. In the Artificial Intelligence (AI) 9Artificial Intelligence (AI) is being emphasized. In this way India is seen in the best position in the global race of AI technology. Despite this, India has no AI model of its own.

In 2023, the US launched a new criteria in AI by launching the Chat GPT. After some time, China also launched its strong chatbot deep and erased this distance fast, but India has not made a single such as a large language model (LLM) so far which is not able to cheat human conversations.

According to data from the market intelligence firm “Traxon”, there are 7,114 startups in the Indian AI sector, which has so far achieved a total funding of $ 23 billion. Last year, Prime Minister Narendra Modi’s cabinet approved the “IndiaAI Mission”, for which a budget of about $ 1.21 billion has been kept. It aims to develop and implement AI techniques according to large multimodal models (LMMs) and special areas developed in India itself.

Deepcik created panic all over the world, India also did not survive
This week, CEO of IndiaAI Mission Abhishek Singh said that Indian startups should no longer be limited to their country, but they will have to think globally to compete and win with global companies. Singh said at the Excel AI Summit in Bengaluru, “Indian AI Startups will eventually have to compete with the world’s best companies, probably in the government. Till only they will not be able to stand on the same. While training the model, they have to adopt a global approach ”

Industry, government and academics need to develop AI
Representatives of the National Association of Software and Services Companies (NAScom), an organization representing India’s $ 283 billion technical industry, say that it is extremely difficult and expensive work to create a global level AI model. Nascom’s communication manager Satyaki Maitra said to DW, “The debate will not be able to do not have any India AII or not to do it. You can move so fast to create a distinct identity in the field of AI ”

Last week, India AI Mission made a big announcement stating that they are adding 15,916 new GPUs (graphics processing units). These GPUs are very important for AI research as they can make many calculations simultaneously. With this new increase, the country’s total AI computing will be increased to 34,3333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333 How many electrical water sucks are made AI models according to the needs of Gan AI, Gyan AI, SarvamAI and Sochet AI in collaboration with AIINII Missions like chat GPT. At the same time, companies like Sarvam AI, Fractal and Corpoor AI are paying full attention to new inventors in AI.

Maitra said, “Only new inventions cannot succeed in the field of AI. For this, the government, industry and academics will have to work together. Only when this can be prepared from handling computers and data to train the model and use in real world,” Why is the AI ​​area of ​​India not growing ahead? That India may face high quality AI hardware deficiency, limited access to modern GPUs, and lack of adequate cloud computing resources, which are necessary to train a large -scale AI model. Duggal said, “Look at the global level, the investment in the AI ​​sector in India is still very low.

Venture capital investment has increased in Indian AI startups, but it is much less than the huge investment in the US or China. From 2014 to 2023, the US invested $ 2.34 trillion in venture and startups and China invested $ 832 billion, while in the same period India has invested $ 145 billion $ 145 billion $ “Duggal is already moving towards making its own AI model, but still still moving towards its own AI model. Major challenges like infrastructure, investment, skilled people, data and rules and regulations have to be solved. There is no lack of talent in India, another major challenge for Indian engineers is the diversity of languages ​​of India.

In this country with the world’s most populous, English is one of the only 22 official languages. At the same time, official languages ​​have been counted in more than 1,600 languages ​​spoken in the country. Yash Shah of Costm Software Development Company “Momentum 91” told DW, “Indian LLM will prove to be effective only when it is difficult to work in many languages. There is a huge shortage of high quality data to train AI models. In the field of large language model for English language, some companies and countries have gone far ahead of us and there is a possibility that they will remain the same in the same way ”

However, the Utpal Vaishnav of the Global Technology Holding Company “Upskwaire Technologies” says that the real obstacles of AI’s path in India are “to avoid the risk of investors, half-incomplete rules to data, and lack of GPU. He said to DW,” Talent is not going to be talented in India. The data is also ready, just to give it the right look. If this talent gets the right investment and clear goals at the right time, then a magnificent and world-class LLM can be launched in just two-three years ”.