How an Uber Eats worker created UberCheats, a web app that computed the actual distance a courier traveled, after suspecting algorithmic error and underpayment
A tech-savvy Uber Eats worker was sick of the algorithms that controlled his day. So he decided to fight back X: @mrbrownsir , @hoofnagle , @ftmag , @michaelezra , @madhumita29 , @karenojeffrey , @leylaboulton , @robertshrimsley , @lilaphasia , @walsh_e_patrick , @walsh_e_patrick , @rjbarfield1 , and @madhumita29 X: Matt Brown / @mrbrownsir : “You train the algorithm, and the algorithm trains you.” Lots going on here in a great story about tech, labor and policy The delivery rider who took on his faceless boss https://www.ft.com/... Chris Hoofnagle / @hoofnagle : Profile of determined computer scientist and Uber delivery worker who created an application to audit Uber's compensation https://www.ft.com/... @ftmag : “AI work systems are built to keep drivers apart, incentivising them to compete. As work turns into a contest, holding onto your livelihood becomes the prize.” @madhumita29 on the delivery rider who took on the UberEats algorithm https://www.ft.com/... Michael Ezra / @michaelezra : This is a remarkable story about Uber Eats and a claimed rate that drivers are cheated on their pay as the AI system underestimes mileage. Excellent work by @ArminSamii and @madhumita29 for @ftmag https://www.ft.com/... Madhumita Murgia / @madhumita29 : Go pick up an @FTMag today, for an opus by @antoniacundy, and this extract by me, on one man who took on a capricious algorithm Karen Jeffrey / @karenojeffrey : We hear a lot about the hazards of algorithmic “bosses”, but the impact on workers is rarely documented this vividly https://giftarticle.ft.com/... Fantastic @FT article by @madhumita29 about the experience of an @UberEats driver #DavidVsGoliath @leylaboulton : I love the @FT's humanity: today @johnthornhillft contemplates a future where ‘superfluous’ humans (a throwback to 19th c. Russia's 'superfluous man') might rise up against robots and @madhumita29 writes about one man's rebellion against the algorithm-boss https://www.ft.com/... Robert Shrimsley / @robertshrimsley : Really excellent piece about Uber and a man who gave a damn. But slightly terrifying too. @madhumita29 The delivery rider who took on his faceless boss https://www.ft.com/... via @ft @lilaphasia : A growing awareness of the plight of workers whose livelihoods are controlled by AI has made its way into the courts... #GigEconomy https://www.ft.com/... “The broader trend of people rising up against gig work exploitation has given a collective voice to [this] group.” [image] Patrick Walsh / @walsh_e_patrick : “As Murgia shows in CODE DEPENDENT, her wide-ranging and mildly terrifying book”... “CODE DEPENDENT is a penetrating look at how we're allowing artificial intelligence to infiltrate all parts of society”... The Telegraph reviews Madhumita Murgia's CODE DEPENDENT, out on Thursday [image] Patrick Walsh / @walsh_e_patrick : How a Tech-Savvy Delivery Rider Took on the Algorithms that Control How Gig Workers are Paid. Financial Times' Weekend Magazine today with a fascinating story from Madhumita Murgia's brilliant new book, CODE DEPENDENT - out on Thursday from Picador UK. [image] Richard Barfield / @rjbarfield1 : The delivery rider who took on his faceless boss https://www.ft.com/... via @ft Interesting article about an algorithm that has commercial & other real-world impacts - particularly given UK courts' normal assumption that ‘the computer is right’, e.g. in the Post Office scandal Madhumita Murgia / @madhumita29 : While it can seem “mildly terrifying” to read about the scope and scale of our society's automation, I hope this is a call for us all to participate in shaping the future we want, living alongside AI.
Context & Ripple Effects
Uber’s app revamp had strengthened its position in attracting drivers, while prior coverage of delivery work documented fluctuating pay and intense time pressure for couriers. UberCheats turns an individual worker’s suspicion about mileage calculations into a reproducible way to inspect the platform’s pay logic.
The episode also fits a broader worker-rights challenge to automated management, including campaigns pressing Uber over algorithmic worker rights in the UK. Its significance is not that it proves an error, but that it gives couriers a tool to test a consequential platform calculation.
First-order effects
- Couriers who can access and analyze their delivery records gain an independent mileage check against Uber Eats’ compensation calculations.
- Uber Eats faces a more concrete basis for individual pay disputes if workers identify differences between app-calculated and traveled distance.
Second-order effects
- A shareable auditing tool can make isolated concerns easier to compare across workers, increasing pressure for clearer calculation methods or data access.
- Other delivery platforms may face similar scrutiny of route, time, and pay calculations as workers apply comparable data-analysis methods.
Third-order effects
- If worker-built auditing becomes common, algorithmic management will be less of a one-way information system: platforms may need to defend compensation logic to workers, advocates, and regulators.
- The longer-term effect depends on whether audits uncover consistent discrepancies and whether workers can obtain usable underlying data; without either, these tools remain primarily individual accountability aids.
The trend: Worker-led data tools are emerging as a counterweight to opaque algorithmic management in platform labor.