in

#TechBossThanksSamAltmanForDaringUsToDream #DreamingBig

Why this tech boss thanked Sam Altman after building Indian LLM under $5 million: ‘For daring us to dream’

CP Gurnani, former CEO of Tech Mahindra and founder of AIonOS, developed a large-language model (LLM) for just $5 million and thanked OpenAI boss Sam Altman for the challenge. The Indian LLM can communicate in about 40 different local languages and dialects. Gurnani shared that Altman initially challenged India’s ability to create an LLM, but Tech Mahindra’s chief innovation officer quickly developed a plan. Despite Altman’s skepticism, the Indian LLM was successfully created within five months for less than $5 million.

Altman’s comments about the competitiveness of training foundation models sparked controversy, leading him to clarify that startups should focus on creating new and innovative solutions. Gurnani emphasized India’s potential in the AI race, predicting that the country will become a major player in the industry in the future.

Tech Mahindra launched Project Indus, the first phase of its LLM, using the “GenAI in a box” framework for easy deployment in enterprises. Nikhil Malhotra, global head of the makers lab at Tech Mahindra, highlighted the efforts behind Project Indus, which involved collecting data from Hindi-speaking populations to build the Indus model.

Overall, Gurnani’s success in developing the Indian LLM under budget and Altman’s challenge to dream big showcase the potential for innovation and growth in India’s AI industry.

Source link

Source link: https://www.hindustantimes.com/business/why-this-tech-boss-thanked-sam-altman-after-building-indian-llm-under-5-million-for-daring-us-to-dream-101719994229154-amp.html

What do you think?

Leave a Reply

GIPHY App Key not set. Please check settings

Sousa Brothers

Meta unveils faster 3D Gen AI tool for textured models. #technology

Photos of Australian kids have been found in a massive AI training data set. What can we do?

Discovering Australian children’s photos in AI data, what now? #privacy