DoorDash Drivers Now Earn Extra by Training AI

DoorDash is expanding opportunities for its couriers to earn income beyond deliveries, now including tasks designed to train artificial intelligence and robotic systems. The company has launched a standalone “Tasks” app, allowing its 8 million U.S. Gig workers to get paid for completing assignments that contribute to the development of AI, as announced on Thursday DoorDash Newsroom.

What are DoorDash Tasks?

These “Tasks” involve activities like recording videos of everyday actions or speaking in different languages. According to DoorDash, this data is crucial for helping AI and robotics better understand the physical world TechCrunch. Pay is determined upfront based on the effort and complexity of each activity.

Examples of Tasks

The types of tasks available are varied. Examples include:

• Filming everyday chores like folding clothes, washing dishes, or making a bed Engadget.
• More complex tasks like pruning and repotting plants, which offer higher compensation.
• Recording unscripted conversations in other languages, such as Spanish TechCrunch.
• Taking photos of restaurant dishes to showcase menus.
• Photographing hotel entrances to aid delivery drivers.
• Assisting autonomous vehicles.

How the Data is Used

The footage and data collected through the Tasks app will be used to evaluate both DoorDash’s own AI models and those developed by its partners in sectors like retail, insurance, hospitality, and technology Bloomberg. For instance, video of someone washing dishes can help a humanoid robot learn how to perform the same task.

Beyond the Standalone App

In addition to the new standalone “Tasks” app, digital tasks will as well be integrated into the existing Dasher app for couriers to complete between deliveries DoorDash Newsroom.

A Growing Trend

DoorDash is not alone in this approach. Uber has also piloted a similar program, allowing drivers to earn extra income by completing tasks to train AI models TechCrunch. Other companies, like Instawork and Sunday Robotics, are also employing strategies to collect physical data for AI training purposes.

“It’s simple: you can’t deliver to a door you can’t find or get someone milk if you don’t know what’s on the shelf,” said Ethan Beatty, General Manager, DoorDash Tasks, in a statement DoorDash Newsroom.