DoorDash now pays couriers to train AI robots

DoorDash just turned its army of delivery drivers into AI trainers. The company launched a separate app called Tasks that pays couriers to complete small digital jobs — filming themselves loading a dishwasher, folding laundry, or recording unscripted conversations in Spanish — all to help train artificial intelligence and robotics models.

Your delivery driver is now teaching robots to fold clothes

The Tasks app lists paid micro-jobs that have nothing to do with delivering food. Instead, couriers record short video clips of everyday household activities — the kind of training data that AI companies desperately need to teach robots how to move through the real world.

Think of it this way: before a robot can learn to load a dishwasher, it needs to watch thousands of humans doing it from different angles, in different kitchens, with different dish layouts. That's what DoorDash drivers are now providing.

Why DoorDash has a secret weapon other AI companies don't

Most AI companies that need training data hire through specialized platforms like Scale AI or Surge AI. But DoorDash already has millions of gig workers across the country — people who are used to picking up flexible, short-notice jobs through an app.

That's a massive advantage. Instead of recruiting and onboarding new data workers from scratch, DoorDash can push a task to thousands of drivers in specific markets instantly. The result: diverse, real-world training data collected at scale.

The bigger trend: gig workers are becoming AI's teachers

DoorDash isn't alone. Earlier this year, Waymo started paying DoorDash drivers to close robotaxi doors that passengers leave open. Uber, Lyft, and Amazon Flex workers have also been recruited for AI-related side tasks.

This reflects a fundamental shift in the gig economy. As AI models get more sophisticated, they need increasingly specific, real-world data — and the people best positioned to capture it are the millions of workers already on the ground with smartphones in their pockets.

What this means for gig workers

If you're a DoorDash driver, this is a new income stream you can tap between deliveries. No special skills required — just a smartphone and the willingness to film everyday activities. It's flexible work on top of already-flexible work.

What this means for AI development

For robotics companies, this solves one of the hardest problems in the field: getting enough diverse, high-quality video data of humans performing everyday tasks. Lab-recorded data tends to look artificial. Data from real homes, real kitchens, and real people? That's what makes robots actually useful.

The privacy question nobody's asking yet

There's an important question here that hasn't been fully answered: who owns the data? When a driver films their kitchen for $5, do they have any say in how that video is used? Can it train military robots? Can it be resold to third parties?

As AI training becomes a bigger part of gig work, expect regulators and unions to start asking these questions — loudly.

Related Content — Get Started with Easy Claude Code | Free Learning Guides | More AI News