In the shadow of Amazon’s offices in downtown Seattle, people enter a tiny grocery store, take whatever they want, and then walk out. And nobody runs after them screaming.
This is what it’s like to shop at Amazon Go, the online retail giant’s vision for the future of brick-and-mortar stores. There are no checkout clerks, or even checkout stands. Instead, a smartphone app, hundreds of regular and infrared cameras on the ceiling (black on black, so they blend in), computer-vision algorithms, and machine learning work together to figure out what you’re picking up and charge you for it on a credit card connected to your Amazon account.
For now, it’s still largely an experiment. Over the past year, employees at Amazon’s headquarters acted as the guinea pigs, ducking in to grab plastic-wrapped sandwiches, chips, and yogurt. Meanwhile, the company—historically focused on selling all kinds of things online but increasingly interested in the world of offline retail with moves like its 2017 purchase of Whole Foods—has been munching on the data those interactions produced and refining its approach to AI-focused shopping.
But Amazon is about to move to the next phase of its plan: opening the store up to the public. Starting Monday, anyone who wants to can head over to the store in downtown Seattle and shop after checking in with an Amazon Go app. Eventually, the company hopes to open up more of these stores, said Dilip Kumar, vice president of technology for Amazon Go, as he showed me around the one on Seventh Avenue on a recent morning.
“People can come in and regardless of how crowded or less crowded it is, you control the amount of time that you’re actually spending at the store. You’re no longer subject to the vagaries of how long it takes you to shop,” Kumar said.
Amazon is not the only company working on checkout-free shopping—Stockholm-based Wheelys has tested an autonomous store in China, while a Silicon Valley startup called Standard Cognition is working on its own version of cashier-free checkout. But it is by far the most prominent company to try it. And its clout as a retailer on and off the Web, plus its ability to build something as complicated as a checkout-free store from start to finish with its own tools and businesses (beyond Whole Foods, its Amazon Web Services is available to host all the data this kind of service requires), makes it the most likely to succeed in the near future.
Actually shopping at Amazon Go feels a little weird. You open an entry gate by scanning a QR code in the Amazon Go app, and then you can just walk in and put your phone away (for some reason, this was really hard for me; I kept feeling I needed the phone out for it all to work). You grab what you want and place it in a bag, or just hold on to it while wandering the well-stocked aisles.
Kumar won’t say how good the company’s technology is at tallying your purchases beyond “very, very accurate,” but the store does stock a lot of items that look very similar, like Diet Dr Pepper and Caffeine Free Diet Dr Pepper. And the prices seem competitive.
With the exception of a back corner where a guy was stationed to check customers’ IDs before they grabbed a bottle of pinot grigio or a six-pack of beer, nobody really seemed to be paying any attention to us. There’s no need, in theory: when you take an item off the shelf, Amazon’s cameras and AI should work to determine what it is and charge you for it immediately, and if you put it back on the shelf, the charge will be removed.
When we were ready to leave, we just, like, left. The whole experience was seamless and quick, and yet it was unnerving to have so little contact with other humans. The lack of cashiers made me wonder what will happen to people who currently hold those jobs if Amazon Go, or something else like it, spreads—will they take on in-store concierge roles, or be replaced entirely by AI? Even though I’m often lost in the world of my smartphone when I’m in stores, the social interaction can be nice.
Kumar admitted it does take a little while for people to get used to simply walking out, so maybe it’s like shopping on Amazon.com: at first, it’s strange to buy things that way, but over time it becomes practically indispensable, for better or worse.
Troll farms reached 140 million Americans a month on Facebook before 2020 election, internal report shows
“This is not normal. This is not healthy.”
The Facebook whistleblower says its algorithms are dangerous. Here’s why.
Frances Haugen’s testimony at the Senate hearing today raised serious questions about how Facebook’s algorithms work—and echoes many findings from our previous investigation.
She risked everything to expose Facebook. Now she’s telling her story.
Sophie Zhang, a former data scientist at Facebook, revealed that it enables global political manipulation and has done little to stop it.
Covid conspiracy theories are driving people to anti-Semitism online
Old and overtly anti-Semitic fantasies are gaining new adherents, and far-right activists have been working to convert anti-lockdown beliefs to anti-Semitism too.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.