Generative artificial intelligence, exemplified by innovations like ChatGPT, has taken center stage in the tech industry. Companies like Microsoft and Google, at the forefront of the AI race, recognize the pivotal role of quality data in shaping the success of their AI endeavors. Addressing this need is 27-year-old Manu Chopra, the visionary founder of Karya.
Based in Bengaluru, Karya, established in 2021, stands out for its commitment to linguistic diversity, achieved by actively recruiting individuals from rural areas traditionally underrepresented in such technological roles. The startup’s offline application provides vocal assistance, catering to users with limited literacy skills.
Meet Manu Chopra:
Manu Chopra, a Stanford-educated computer engineer, graduated in Computer Science from Stanford University in 2017. During his time at Stanford, he co-founded CS+Social Good, a student group dedicated to technology’s positive impact. Beyond the academic realm, Chopra is focused on alleviating extreme poverty in rural India through dignified digital work. According to his website, Karya’s efforts have lifted over 100,000 rural Indians out of extreme poverty in the past 1.5 years.
Chopra’s impact extends to education, where he has taught courses on technology with a positive influence at Stanford’s Computer Science Department. These courses have empowered students to undertake projects benefiting more than 30 million individuals across 15 countries globally.
Through Karya, Chopra addresses the surging demand for precise data in the AI market while reshaping the economic landscape for data annotation workers. Karya’s unique approach involves hiring predominantly women from rural India, paying them over 20 times the minimum wage. Their services encompass gathering text, voice, and image data in India’s vernacular languages.
In a conversation with Bloomberg, Chopra highlighted the industry’s failure in compensating workers appropriately for the crucial task of collecting training data. Notably, Microsoft has utilized Karya for regional speech data acquisition for its AI products, while the Bill & Melinda Gates Foundation collaborates with Karya to mitigate gender-related biases in AI chatbots.
Google’s AI Expansion Plans:
Google, eyeing expansion in India, envisions building an AI-generative model for the country’s 125 languages. The tech giant relies on Karya and local collaborators to amass speech data across 85 Indian districts. Manish Gupta, head of Google Research in India, emphasized the significance of this effort, stating, “India is the first non-Western country we are doing this in, and we are testing Bard in nine Indian languages.”
In India, over 32,000 crowdsourced laborers have leveraged Karya’s application, completing a remarkable 40 million compensated digital assignments. These tasks encompass image recognition, contour alignment, video annotation, and speech annotation, contributing to the advancement of AI technologies and fostering inclusive innovation.
Last modified: November 14, 2023
