To revolutionize artificial intelligence (AI) for India’s billion-plus population, tech giants Microsoft and Google have set their sights on Karya, a startup founded by the ambitious 27-year-old computer engineer, Manu Chopra. Let’s explore the strategic collaboration between these industry giants and Karya, shedding light on the pivotal role the Indian startup plays in reshaping the AI landscape.
Empowering Local Languages
Manu Chopra’s Karya has embarked on an ambitious mission to collect text, voice, and image data in India’s myriad local languages. The crucial task of tagging and labeling this data is undertaken by speakers of these languages, forming the foundation for AI chatbots and virtual assistants to deliver contextually relevant results across diverse linguistic landscapes.
Addressing Industry Failures
Chopra, in conversation with Bloomberg, highlights the disparity in pay for the laborious task of data tagging, exposing a failure within the tech industry. Despite the normalization of AI tools, underpaid workers contribute significantly to global AI datasets, especially in local languages.
Microsoft and Google’s Strategic Goals
Microsoft leverages Karya’s resources to source local speech data for its AI products, aiming to reduce gender biases within language models. Simultaneously, Google partners with Karya to gather speech data from 85 Indian districts, with plans for expansion to encompass every district and dialect. The end goal is to develop a generative AI model capable of handling 125 Indian languages.
The Significance of Indian Languages
With existing generative AI models predominantly focused on the English language, India’s diverse linguistic landscape presents a goldmine for Silicon Valley companies. The potential user base of over a billion individuals propels India into a strategic position for these tech giants.
Challenges and Karya’s Innovative Approach
Non-English datasets often suffer from low quality and limited availability, exacerbating the challenges faced by large language models. Karya, headquartered in Bengaluru, addresses these issues by targeting rural workers for tagging and labeling local language content. The app operates offline, providing voice support for those with limited literacy, fostering inclusivity.
Impact of Karya
Supported by grants, Karya has witnessed over 32,000 crowdsourced workers logging into the app, contributing to a staggering 40 million paid digital tasks. This collaborative effort between tech giants and the Indian startup signifies a paradigm shift in making AI more accessible and relevant to India’s linguistic diversity.
In this evolving landscape, the alliance between Karya, Microsoft, and Google stands as a testament to the transformative power of AI when harnessed with a commitment to inclusivity and innovation.