At the final stage of the Asian tour, the head of the Openai, Sam Altman, said that India is able to become one of the leaders in the AI race, especially in creating small “reasoning” models (SLM). Altman expressed interest in the dynamic ecosystem of local developers, which can become a key element of the Openai expansion, Digitimes reports. According to Altman, India has become the second largest market for the company, because over the past year the number of users has tripled. During the visit, Altman met with representatives of the country’s government.
Although the costs of developing new models are still high, Altman admitted that progress in the field of AI can significantly reduce dependence on expensive equipment, and the return on investments in AI will grow exponentially. This will lead to an annual tenfold decrease in the cost of “units of intelligence” – a conditional measurement of computing efficiency of AI. According to the businessman, the world has reached the stage of incredible progress in the field of “distillation” of models.
Image Source: Shubhangee Vyas/Unsplash.com
Although the training of even small models remains quite expensive, it is small models with the possibility of “reasoning” that will literally lead to an explosion of creativity. And India should be on the front edge of progress. Altman especially distinguished the potential of projects in healthcare and education, where AI is able to become the driving force of transformations. Now Indian companies have paid attention to open models, including Deepseek.
It is worth noting that during a visit to India in June 2023, Altman described the country’s chances of creating AI-models of the ChatGPT level as “completely hopeless”. Now he is pleasantly surprised by the achievements of India in this area. The head of Openai separately explained that the previous statements were difficult to competition with IT giants when creating economically effective models. Altman separately emphasized that the cost of the API Openai fell significantly, and hinted that in the future the emergence of the Open Source of initiatives is also possible.
India is actively promoting AI projects as part of the Indiai initiative, supported by investments in the volume of ₹ 103.7 billion rupees ($ 1.2 billion). The authorities intend to create the local large language model (LLM) for ten months. Representatives of the Ministry of Communications and Information Technologies of India said that the creation of the basic structure has already been completed, now the efforts of the developers are aimed at creating options for models that meet the unique language and cultural requirements of the country.