The SHAKTI-man of India, Veezhinathan Kamakoti, director of IIT Madras, has had a busy few years. The genius behind India’s first microprocessor has been making an attempt actually onerous to construct semiconductor capabilities in India, and put a cease on low-cost imports from China. “This is something which we have achieved over the last 10 years and I have completely given up every other thing to work for SHAKTI,” he advised AIM in an unique interview.

Kamakoti began constructing the SHAKTI microprocessor in 2014, an open processor constructed on high of the open RISC-V structure, which is funded by the Ministry of Electronics and Information Technology. He stated that there’s a want for constructing {hardware} that helps constructing AI fashions inside the nation, which led him to take the cost with the SHAKTI microprocessor.

Humble Kamakoti has been emphasising for the previous couple of years to extend the funding into semiconductor startups, which continues to be a really dangerous space for a lot of VCs. 

“I am only Saraswati, not Lakshmi for these startups,” Kamakoti stated that there are 5 firms popping out of the IIT Madras lab: Incore, Mindgrove Technologies, Chakra Electronics, Vyoma Systems, and SecurWeave, and all of them are constructing on high of SHAKTI core, however he’s solely the mentor for these firms, “It is very interesting to see that we have an indigenous processor ecosystem in India,” he added. 

Building belief for the chip

Ever since finishing his first course in AI in 1992 from IIT Madras, Kamakoti has been discovering methods to make use of AI for social good and fixing complicated issues inside India. This led him to ultimately begin specializing in {hardware} and led to engaged on SHAKTI, which can also be specializing in edge AI use instances. 

Kamakoti stated that it is vitally onerous to construct belief within the semiconductor enterprise as a result of traders have to have full belief when investing in these startups. “Though the government is providing full support for improving the semiconductor industry in India, investors are still hasty and cautious when it comes to investing, as it is very costly, and cheaper alternatives are always available. We are now closing the gap,” he added.

“Today, Shakti C class is very stable and we have made multiple chips with it. We are ready to give to startups as well, which is not a joke,” stated Kamakoti highlighting that they’re making it open for everybody for democratising processor primarily based SoC (system of chips) design within the nation.

“Unlike a software company which has a cloud and is developing a software solution with demos, hardware startups take a lot more trust and investment, which the government is promoting, but we need a lot more VCs to take interest as well,” Kamakoti emphasised. 

BharatGPT could be the Copilot for India

Kamakoti was the chairperson of the AI Task Force, which was constituted in 2016 by the Ministry of Commerce and Industry. The AI Task Force report highlighted how AI can be utilized as an financial transformer for the nation. Apart from this, at present he’s additionally a vital member of the BharatGPT initiative, which was began by IIT Bombay. 

Highlighting specialised {hardware} for each subject together with IoT, community processing, and computing, Kamakoti stated that it is usually important for LLMs to be constructed and catered for particular domains. He believes that that is the place BharatGPT goes to play a vital position. “The initiative requires validation from experts from every domain, and that is what BharatGPT stands for,” he defined. 

Giving examples of coaching small fashions for the authorized subject or medical subject and the way they’ll help professionals within the subject, Kamakoti stated the open supply strategy of BharatGPT will assist it change into the copilot for various domains. “It is not a search engine, but a technology that is doing a lot of NLP, and giving you contextual output,” which he believes would assist individuals strategy each subject with much more confidence. 

Reflecting on how there are 1000’s, if not hundreds of thousands of paperwork in Indic languages, it’s the want of the hour to extract data from them, together with historic information and science. “For example, there is some proof that Newton’s laws were present in our ancient texts at least a 1000 years ago,” and this could profit the entire of the world. 

“LLMs are like faithful kids”

Talking about generalised fashions like ChatGPT and Bard, Kamakoti stated that LLMs are like trustworthy children, “if you give them relevant data, they will do exactly what you want them to.” He stated that there’s a dire want for knowledge scientists and to concentrate on extracting and gathering info for tokenising these Indic fashions. 

“This would obviously require a lot of money and compute,” highlighting his love for farming, Kamakoti stated that whereas we’re additionally speaking sustainability with AI fashions, they’re additionally very essential for serving to varied sectors similar to agriculture. “There is a lot of knowledge hidden in rural India, which needs to be brought out,” he added, highlighting that each subject, even inside the similar area, requires totally different quantities of knowledge. 

Focusing on the current funding of INR 110 crore from Sunil Wadhwani to IIT Madras for constructing a college for AI and knowledge science, Kamakoti recalled {that a} reporter requested him if the identical cash may as a substitute be spent on constructing 10-12 faculties in India. To which, he replied, “I can use AI to enhance the capabilities of the teacher. I don’t want to replace the teacher, but augment them, in the end creating a very good copilot,” Kamakoti concluded.