DoorDash is expanding its services beyond food delivery, now offering its network of 8 million U.S. Drivers a new way to earn income: training artificial intelligence. The company launched a standalone app called “Tasks” on Thursday, allowing couriers to complete assignments designed to improve AI and robotic systems, a move that reflects a growing trend of companies leveraging gig workers to bolster their AI development efforts. This new initiative, focused on AI training data, comes as the demand for real-world data to refine AI models continues to surge.
The Tasks app presents couriers with a variety of assignments, ranging from everyday chores to more specialized tasks. Couriers can earn money by filming themselves performing activities like washing dishes, folding clothes, or even having a conversation in another language – specifically, Spanish speakers are being asked to engage in “natural, unscripted conversations” with friends and family. Pay is determined by the effort and complexity of each task, DoorDash says. The company’s cofounder and chief technology officer, Andy Fang, expressed enthusiasm about the potential of the program, stating, “We reckon this will be huge for building the frontier of physical intelligence. Glance forward to seeing where this goes!” according to a social media post.
Beyond Delivery: The Rise of ‘AI Training Gigs’
DoorDash isn’t alone in tapping into its workforce for AI development. Last year, Uber piloted a similar program, allowing drivers to earn extra income by completing digital tasks like uploading photos to help train AI models. The data annotation industry, which involves labeling and categorizing data for AI training, has also experienced significant growth in recent years, with companies like Oxford Economics reporting its economic impact has boomed in recent years. However, DoorDash’s approach is notable for its focus on capturing physical data – how people actually *move* and interact with the world.
This shift towards capturing physical data is driven by the necessitate to train AI to understand and navigate the complexities of the physical world. For example, a humanoid robot learning to load a dishwasher requires data on how humans perform that task. Other companies are employing similar strategies. Instawork, a staffing app, was recently reported by the Los Angeles Times to be recruiting workers in Los Angeles to record themselves cleaning their homes while wearing head-mounted cameras. Sunday Robotics, a California-based robotics developer, even ships “skill capture gloves” to individuals across the country to collect motion data during household tasks.
Tasks Within the Dasher App and Beyond
While the standalone Tasks app represents a new avenue for earnings, DoorDash is also integrating similar opportunities into its existing Dasher app. Couriers can now earn additional income by completing tasks such as verifying a restaurant’s holiday hours, taking photos of demanding-to-find drop-off locations to assist other drivers, or even assisting autonomous vehicles. Ethan Beatty, general manager of DoorDash Tasks, explained that the company realized its existing capabilities for solving logistical challenges could be valuable to other businesses, stating, “These are the kinds of real-world problems we’ve been solving for over a decade, and we realized the same capabilities that helped us could help other businesses too.”
DoorDash anticipates that the Tasks app will initially focus on activities that directly contribute to AI and robotics training, but plans to expand the range of available tasks over time. The company emphasized that the Tasks app is currently a “small pilot” compared to the broader range of opportunities available within the regular Dasher app, where couriers can already complete a variety of assignments between deliveries.
Implications for the Gig Economy and AI Development
The emergence of these “AI training gigs” raises questions about the future of operate and the evolving relationship between gig workers and technology companies. While offering a new income stream for drivers, it also highlights the increasing reliance on human labor to fuel the development of artificial intelligence. The potential for this type of work to develop into a significant part of the gig economy remains to be seen, but it underscores the growing demand for diverse datasets to train increasingly sophisticated AI models. DoorDash has already seen over 2 million tasks completed by Dashers since 2024, according to a company statement, demonstrating early adoption of the concept.
DoorDash plans to continue expanding the Tasks program, offering more opportunities for couriers to contribute to AI development while earning additional income. The company will be closely monitoring the program’s performance and gathering feedback from Dashers to refine the offerings and ensure a positive experience. Updates on the program’s progress and new task opportunities will be available through the DoorDash website and app.
As AI continues to advance, the need for high-quality training data will only grow, potentially creating further opportunities for gig workers to participate in the development of this transformative technology. The next step for DoorDash will be to assess the initial results of the Tasks app and determine how to scale the program effectively while maintaining quality and fairness for its drivers.
Share your thoughts on DoorDash’s new AI training program in the comments below.
