Amazon's touchless technology, which allows customers to pick up what they need from a shelf and “just walk home” without going to the checkout, really relies on human workers in India to verify purchases. I wonder if there is one?
The Seattle-based retailer announced Tuesday that it would replace Just Walk Out technology with smart carts in more than half of its 40 Amazon Fresh grocery stores, without actually saying anything.
Just Walk Out technology sends a receipt after shoppers leave the store, while Amazon Dash Cart displays the amount charged for each item in real-time on the screen, allowing shoppers to bypass the checkout. You can also. Amazon says the changes being made at its Amazon Fresh grocery store are in response to customer feedback, but it will continue to offer Just Walk Out products at more than 130 third-party partners, including airports, university stores and cafes. He said he would continue to use technology.
The company claims that sensors, cameras and other tools at these locations help shoppers track what they buy. However, several media outlets reported that there may be more to it than just that, with hundreds of Indian workers playing a key role.
How does Just Walk Out know what I'm about to buy?
AWS, a separate division of Amazon, says on its website that customers using Just Walk Out technology can use Amazon One (where customers can register their palm and connect it to their payment method), a credit/debit card, or a credit card. He said he could enter the store. Purchase items in the app and go. Customers are automatically charged for their purchases.
“Sensors, cameras, and deep learning tools sense what consumers are removing from the shelves,” the website says.
An Amazon spokesperson further explained: “Just Walk Out technology is enabled by artificial intelligence, such as computer vision, and deep learning techniques, including generative AI, to accurately determine who gets what in any retail environment. Decide: Amazon has built a synthetic dataset that mimics millions of realistic shopping scenarios in any environment, including changes in store format, lighting conditions, and even shopper crowds. to ensure accuracy.”
However, some media outlets reported that Indian workers may also be heavily involved.
According to CNBC, like many artificial intelligence systems, Amazon's system relies on human moderators and data labelers, who review Just Walk Out transactions and label the footage to make it It will support training of AI models to make it work. According to CNBC, The Information reported last year that the team is made up of more than 1,000 employees, primarily based in India. The Information reported that an Amazon spokesperson confirmed at the time that the company is hiring human moderators, but declined to say how many people it employs for the role.
Business Insider cited further reporting from the Information on Tuesday, saying Just Walk Out still relies heavily on humans, according to an anonymous person who said the Information was working on the technology.
According to Business Insider's The Information, about 700 of every 1,000 Just Walk Out sales needed to be reviewed by Amazon's team in India in 2022. Internally, Amazon required manual checks for only 50 out of 1,000 sales, according to the report.
What does Amazon say?
An Amazon spokesperson disputed media reports in a statement Thursday.
““The misconception that Just Walk Out technology relies on human reviewers monitoring shoppers live from India is misleading and inaccurate,” an Amazon spokesperson said in an emailed statement to USA TODAY. “As with many AI systems, the underlying machine learning model is continuously learning,” he said. It is improved by generating synthetic data and annotating the real video data.
Smart technology:Why Amazon is eliminating Just Walk Out checkouts at grocery stores
“Our employees verify a small portion of shopping visits by reviewing recorded video clips to ensure that our systems are operating to a high standard of accuracy. This is because we continually improve both our algorithms and use human input to modify them.”
Betty Lin-Fisher is a consumer reporter for USA TODAY. Contact her at blinfisher@USATODAY.com or follow her on X, Facebook or her Instagram @blinfisher.. Sign up here for The Daily Money newsletter, free for consumer news on Fridays.
