Uber India has initiated a new program to enable its rideshare and delivery drivers to supplement their income by participating in AI data classification tasks directly through the company’s app. Megha Yethadka, who spearheads Uber AI Solutions, announced the initiative on LinkedIn, highlighting that drivers often have available time to engage in these tasks during breaks or at the end of their shifts. The tasks include reviewing images, classifying text, recording audio clips, and more, supporting various global enterprises and consumer applications.

Previously handled by freelancers outside Uber’s ecosystem, the program’s initial results are promising, and there’s a potential for global expansion. According to Prabhjeet Singh, president of Uber India and South Asia, these digital tasks are already active in 12 Indian cities and engage tens of thousands of drivers.

Uber’s CEO Dara Khosrowshahi has also acknowledged the alignment of this initiative with Uber’s core ability to distribute tasks globally, envisioning it as a novel way to employ workers who fuel AI advancements worldwide.

In addition to the AI task gig, Uber disclosed its massive 350-petabyte data lake and introduced a tool named HiveSync, inspired by Airbnb’s ReAir, to efficiently manage and replicate data for enhanced system redundancy using a single region compute model. Developed since 2016, HiveSync processes about 300 petabytes within 800,000 Hive tables and replicates eight petabytes daily, positioning itself for an upcoming open-source release as Uber migrates analytics and ML systems to Google Cloud.