Tech Mahindra's Project Indus develops Indian LLM under $5 million, an alternative to ChatGPT

3 days ago 8

Copyright &copy HT Digital Streams Limited
All Rights Reserved.

2 min read 03 Jul 2024, 03:55 PM IST checkcibilBtn

Livemint

Tech Mahindra's CP Gurnani successfully developed Project Indus, a large-language exemplary nether $5 million. Altman's situation sparked Gurnani's determination to showcase India's AI capabilities. The task aims to revolutionize AI connection successful India and beyond.

CP Gurnani, CEO, Tech Mahindra ( pic by Hemant Mishra)Premium CP Gurnani, CEO, Tech Mahindra ( pic by Hemant Mishra)

CP Gurnani, the erstwhile CEO of Tech Mahindra, has successfully met a situation issued by Sam Altman of OpenAI. Speaking astatine the 5th yearly MachineCon GCC Summit successful Bangalore past week, Gurnani revealed that his squad developed a large-language exemplary (LLM) successful conscionable 5 months for nether $5 million. This project, which Gurnani announced arsenic “Project Indus," tin pass successful astir 40 antithetic section languages and dialects (via Newsweek).

As per the publication, Gurnani recounted the situation posed by Altman during his sojourn to India past year, wherever Altman expressed skepticism astir India's quality to vie successful the LLM space. Altman’s remarks, made astatine an lawsuit organized by the Economic Times, implied that it would beryllium “hopeless" for India to bid instauration models with a fund of $10 million. This connection sparked a question of determination among Indian tech leaders, starring Gurnani to instrumentality contiguous action.

"I spoke to my main innovation serviceman astatine that time, and wrong six hours, helium had a plan," Gurnani said. "I americium blessed to stock that we spent little than $5 cardinal connected what Sam Altman said India would ne'er beryllium capable to deliver."

Altman aboriginal clarified his comments connected social media, suggesting that portion helium believed a $10 cardinal fund would beryllium insufficient for competing with OpenAI, helium encouraged startups to effort and innovate successful ways that person ne'er been done before. He emphasized that the question should absorption connected unsocial contributions alternatively than nonstop competition.

Project Indus, launched by Tech Mahindra, aims to revolutionize AI connection successful India by supporting a multitude of Indic languages and dialects. This inaugural not lone underscores India's increasing capabilities successful AI but besides its imaginable to go self-reliant successful technological advancements. Nikhil Malhotra, the planetary caput of Makers Lab astatine Tech Mahindra, highlighted the extended efforts progressive successful gathering the Indus model, including information postulation from Hindi-speaking populations.

The task uses Dell Technologies’ "GenAI successful a box" model for elemental endeavor deployment. Denise Millard of Dell Technologies emphasized the value of accessibility and scalability successful adopting AI solutions. The collaboration is expected to alteration AI applications crossed assorted sectors, including healthcare, agrarian education, banking, finance, agriculture, and telecom.

Moreover, Tech Mahindra plans to widen Project Indus to sphere Bahasa Indonesia, Indonesia’s authoritative language. This effort volition alteration radical and companies to pass successful section dialects online, demonstrating the model's versatility and broader impact.

Gurnani's announcement signifies a pivotal infinitesimal successful the AI race, positioning India arsenic a formidable subordinate successful processing and deploying precocious AI technologies.

3.6 Crore Indians visited successful a azygous time choosing america arsenic India's undisputed level for General Election Results. Explore the latest updates here!

Catch each the Technology News and Updates connected Live Mint. Download The Mint News App to get Daily Market Updates & Live Business News.

More Less

Published: 03 Jul 2024, 03:55 PM IST

Next Story footLogo

Read Entire Article