In the heart of Raghurajpur, Odisha, Swarnalata and many others have found an extraordinary opportunity through Karya, a non-profit initiative. While most master Pattachitra painters in her community earn just $12 a month, the organization empowers rural Indians to create high-quality datasets for tech giants like Microsoft and Google. These datasets are the backbone of Large Language Models offering them an opportunity to earn $60 a week.
Rural individuals, like Swarnalata, spread across 100 districts in 22 states, earn a substantial $5 per hour by performing “simple dataset generation tasks” for Karya. Since its launch in 2021, the organization has disbursed payments to over 30,000 rural Indians, who have collectively completed a remarkable 40 million paid digital tasks.
Karya’s distinctive approach as a “data cooperative” stands out by granting workers royalties for their data sales, not only offering supplemental income but also the promise of passive earnings for rural India. The driving force behind this initiative is Manu Chopra, a Stanford University engineering graduate deeply connected to the cause of poverty due to his upbringing in Delhi’s basti. His unwavering commitment to addressing this issue makes Karya a unique and impactful endeavor.
Manu Chopra’s path toward ‘tech-for-good’ began as a computer science engineering student at Stanford, which eventually led him to join Microsoft Research in 2017. His mission was clear and it was to combat extreme poverty by providing underprivileged individuals access to dignified digital employment. After extensive travels across India, he observed that many rural Indians possessed smartphones and bank accounts, yet remained among the world’s poorest. This insight led to a transformative idea – what if they could earn a livelihood by leveraging their existing skills, particularly their linguistic abilities?
