Uber Turns Its App into an AI Training Playground, Taking on Industry Leaders

Uber is making a bold move that could reshape how artificial intelligence (AI) is developed. The ride-hailing giant has launched a pilot program in the United States, giving drivers and couriers the opportunity to earn extra money by completing microtasks that help train AI models. With this initiative, Uber is stepping into territory dominated by companies like Scale AI and Amazon’s Mechanical Turk.

Harnessing a Global Workforce for AI Training

Uber’s program allows its drivers and couriers to complete tasks such as:

• Recording voice samples
• Taking photos of objects or locations
• Submitting documents in multiple languages

For example, a driver might take a photo of a Spanish-language menu and earn up to one dollar per submission.

By integrating these tasks into its app, Uber is transforming from a ride-hailing service into a dynamic AI training platform. CEO Dara Khosrowshahi highlighted that this initiative aligns with Uber’s broader goal of becoming the go-to platform for flexible work, providing drivers with more ways to earn.

Taking on Established Data-Labeling Platforms

Data-labeling is crucial for training advanced AI systems. Companies like Scale AI and Amazon’s Mechanical Turk have long dominated this space, often outsourcing tasks to low-cost labor markets abroad. Uber’s approach is different: by embedding AI training tasks directly into its app, it offers a more localized and potentially efficient alternative.

Uber first tested this model in India, where drivers performed tasks like image classification and text analysis during downtime. The strong engagement from that pilot paved the way for the U.S. rollout.

Enhancing the Driver and Courier Experience

Uber is not stopping at AI training. The company has rolled out several updates aimed at improving the everyday experience for drivers and couriers:

• Redesigned trip offer cards for clearer navigation
• Heatmaps showing high-demand areas
• Guidance for handling multi-order deliveries

Safety features have also been expanded. For instance, the Women Rider Preferences tool now allows female drivers to receive trip requests exclusively from female passengers in more cities.

Other improvements include:

• Greater transparency on account deactivations, giving drivers a chance to respond before action is taken
• A Delayed Ride Guarantee, ensuring drivers are compensated for trips delayed due to traffic or customer actions

A Strategic Move into AI Solutions

Uber isn’t new to AI data-labeling. The company has previously used “human-in-the-loop” systems, combining automated AI with human judgment to improve model accuracy. To strengthen its capabilities, Uber recently acquired Segments.ai, a Belgian startup specializing in data annotation tools.

Uber’s AI Solutions division now offers:

• Image and video annotation
• Text labeling
• 3D point cloud processing
• Sentiment detection

With support for over 100 languages, Uber aims to provide comprehensive AI model training for a variety of global applications.

What This Means for the Future of AI

Uber’s initiative raises important questions about AI development and the role of human labor. By integrating training tasks into its app, Uber not only creates new earning opportunities but also challenges traditional data-labeling platforms to innovate.

As AI evolves, the demand for high-quality, labeled data will continue to grow. Uber’s model—leveraging its large workforce—offers a scalable and potentially more sustainable solution. The program’s success will depend on:

• Driver participation
• The quality of AI models trained with the collected data

Uber’s approach could significantly reshape the AI landscape, proving that ride-hailing platforms can evolve into AI powerhouses while benefiting their workforce.