DoorDash launched Tasks: gig workers record themselves to train AI

DoorDash — the company most people know as a food delivery platform — quietly launched a new product called Tasks. It's an app where ordinary people record videos of their everyday life for a small reward. Doing laundry, frying eggs, taking a walk in the park — all of this can now be monetized.

A Wired journalist tried it firsthand and reached an unsettling conclusion: Tasks is not a side gig, but a window into a bleak future of the gig economy. The scheme works simply to the point of absurdity. A user opens the app, sees a list of tasks, and chooses one that suits them.

A task might sound like this: "Record a 3-minute video of you cooking food in your kitchen." You hit record, complete the task, upload the file. A reward is credited for the video — typically a few cents or dollars, depending on complexity.

The money accumulates and is withdrawn through standard payment systems. Who buys these videos? Large technology companies that need training data for AI systems.

This involves computer vision, motion recognition, understanding of everyday scenes. To teach a neural network to understand how a person fries eggs, you need to show it thousands of recordings of real people frying eggs under different conditions, in different kitchens, with different lighting. DoorDash acts as an intermediary: it aggregates performers and sells the result to corporate clients.

Essentially, it's Amazon Mechanical Turk, but for visual data in the era of generative AI. The problem lies in economics. If previously DoorDash gig workers could earn $15–25 per hour from food delivery before deducting transportation costs, Tasks offers significantly less.

A journalist who spent several hours in the app earned amounts that are hard to call even supplemental income. Meanwhile, the barrier to entry is minimal — all you need is a smartphone with a camera — which inevitably creates competition among a huge number of performers and pushes rates down. This is a classic trap of platform economics: the easier it is to enter, the cheaper labor becomes.

The nature of the work itself is also interesting. Unlike delivery or taxi services, Tasks doesn't require physical movement — tasks are completed at home. This lowers the barrier for people who cannot or do not want to work outside their homes.

But it also means competitors are not limited by geography: a performer from Manila or Lagos can complete the same task as a user from New York for far less money. Globalization of the training data market in action. There's also the question of privacy.

When you record a video at home, you inadvertently share data about your interior, habits, and family members who happen to appear on camera. Where these recordings go, who processes them, how long they're stored — Tasks, according to the journalist's account, doesn't make this sufficiently transparent to performers. The market for training data for AI is valued at billions of dollars and continues to grow as major laboratories compete for dataset quality and diversity.

Niches like Tasks have been filled before: Scale AI, Remotasks, Appen, CloudFactory have long worked with gig performers for data annotation. DoorDash simply came to this market with a recognizable brand and millions of users in its database. This demonstrates where platform economics is heading — from physical delivery of goods to delivery of human behavior as raw material for AI.

The same mechanisms, the same precarity, a new type of task. As long as technology companies will pay for training data, platforms will find ways to collect it — cheaply, at scale, and with the participation of people who need supplemental income. Tasks is not the distant future.

It is the present, already available in the App Store.