DoorDash just launched a stand-alone “Tasks” app that turns its massive delivery workforce into an AI training army. The company will pay couriers to complete assignments like filming everyday activities and recording speech in different languages, according to TechCrunch AI.
The concept is straightforward: gig workers already crisscross cities all day. Why not pay them to collect real-world data while they’re at it?
“This data helps AI and robotic systems understand the physical world,” DoorDash wrote in a blog post announcing the launch.
What couriers will actually do
- Film everyday activities: one example involves recording yourself washing at least five dishes while wearing a body camera, holding each clean dish in frame before moving on
- Record speech in other languages: contributing audio data for multilingual AI training
- Photograph restaurant menus: taking real photos of dishes to help restaurants showcase their food
- Document delivery locations: snapping photos of hotel entrances so other drivers can find drop-off spots
- Close Waymo car doors: part of DoorDash’s existing partnership with the self-driving company
Pay is shown upfront and based on effort and complexity. No mystery pricing.
Who’s actually using this data?
Here’s where it gets interesting. Bloomberg reports that the footage won’t just train DoorDash’s own AI models. It’ll also feed into systems built by partners across retail, insurance, hospitality, and tech. DoorDash is essentially positioning itself as a data collection platform, not just a delivery service.
With over 8 million registered Dashers who can reach almost anywhere in the U.S., that’s a serious geographic footprint for gathering real-world training data.
“There are more than 8 million Dashers who can reach almost anywhere in the U.S. and who want to earn flexibly beyond delivery,” said Ethan Beatty, general manager of DoorDash Tasks. “That’s a powerful capability to digitize the physical world.”
Where it’s available (and where it’s not)
The Tasks app and in-app task listings are rolling out in select U.S. locations, with some notable exclusions:
- Not available in: California, New York City, Seattle, and Colorado
- Expansion plans: more task types and countries coming later
Those exclusions likely reflect stricter gig worker regulations in those states and cities. DoorDash didn’t explain the reasoning, but the pattern speaks for itself.
DoorDash isn’t alone here
Uber made a similar move late last year, announcing plans to let drivers earn extra income by uploading photos to help train AI models. The trend is clear: gig economy companies are sitting on massive, distributed workforces and realizing that delivery is just one thing these workers can do.
What stands out with DoorDash’s approach is the scale of ambition. A stand-alone app signals this isn’t a side experiment: it’s a new business line. Selling real-world data collection as a service to partners across multiple industries is a fundamentally different revenue stream than food delivery.
The big question going forward: how will workers feel about being AI data collectors, and will the pay justify the effort? DoorDash says compensation is transparent and task-based. Whether that holds up in practice will determine if Tasks becomes a meaningful earning channel or just another underpaid gig.
More details are available in the original report on TechCrunch AI.