Uber labeling for AI

Uber uses ephemeral workers to enter the AI-powered labeling industry, Bloomberg reports. It’s a sign that the transportation and delivery company is interested in expanding its independent contractor business model to meet the demands of the rapidly evolving world of machine learning and multilingual models.

The company is fresh “Scalable Solutions” section. claims that it can connect companies with “sophisticated analysts, testers and independent data operators” through its platform. This is an extension of an internal team of members from the US and India that tests fresh features and performs other tasks, such as converting restaurant menus to options selected in Uber Eats.

Uber was already using AI and machine learning in its business, and now it’s making it available to others for a fee. The company hires ephemeral workers to label, test and localize data for other companies, including Aurora, Luma AI and Niantic.

The behind-the-scenes reality of AI model training is that many workers must perform tedious tasks, such as selecting the best human-sounding chatbot output or marking frame-by-frame obstacles, such as pedestrians, in footage of a self-driving car.

Companies that create AI models often hire workers in developing countries to perform these tasks and pay tiny amounts for each exercise. This was reported by one engineer from India Bloomberg they were tasked with comparing and assessing the accuracy of AI-generated answers to sophisticated coding problems and were paid 200 rupees per set, or about $2.37.

Uber currently enrolls people from Canada, India, Poland, Nicaragua and the United States, paying various amounts per task completed, with wages paid to workers monthly. It is also looking for people from different cultural backgrounds to facilitate improve artificial intelligence in various markets.