Back in 2018, Amazon opened its first cashier-less grocery store. The technology -- which it later branded as "Just Walk Out" -- was intended to allow customers to do exactly that. All you had to do was put items into your cart and walk out of the store. And then, through the magic of sophisticated computer vision, machine learning, and lots of sensors, you would be billed and sent a receipt.

However, this month the company announced that it will be moving away from this technology, and instead focusing on its Dash Carts (more on this shortly). It turns out that the technology wasn't nearly automated enough.

Last year, The Information reported that "Just Walk Out" was relying on at least 1,000 off-site workers in India to constantly review video footage and figure out who had bought what. This is why it apparently took so long to receive a bill sometimes; humans far away were working to figure out if that was a persimmon in your hand, or a tomato.

I'm not an expert on this space, but I'm guessing it is not (yet) feasible to do what Uniqlo and other retailers now do with their supply chains and checkouts. So this was the workaround. Whatever the case, Amazon has now said that it will be focusing on its Dash Carts, which are kind of like roaming checkout counters. They come with screens and scales for weighing things.

Obviously the ideal solution is to not have to do or scan anything. But being able to avoid check-out lines still feels like meaningful progress. I just wonder if these smart carts will encourage or discourage spending. Because now everyone will have a live receipt in front of them. That might discourage spending unless you can offset it with rewards and/or other incentives.

This morning the first Amazon Go store opened to the public in downtown Seattle. It’s more convenience store than grocery store, but the big deal is that there are no cashiers and no lines.

You enter the store through a gate and with your phone and Amazon’s app. As you walk around the store and pick up items they get automatically added to your online cart on Amazon. So everything goes right into the offline bag you’ll be leaving the store with. Place an item back on the shelf and it is instantly removed from your “cart.” Walk out of the store and you’re automatically charged.

It’s not yet clear how exactly the technology works, but Amazon says that all of this is accomplished through sophisticated computer vision (cameras), machine learning, and lots of sensors. 

What’s really remarkable is that it doesn’t rely on every product having a special chip or sensor attached to it. I would think that was one of the biggest hurdles to overcome in order to remove the pain point of grocery store lines.

Now that this is up and running, I can only imagine the customer behavior data that they must be collecting. Heat maps of every shelf showing conversion rates for every imaginable customer segment. (Are tall people more likely to buy products displayed higher up?) Correlating people’s food purchases to their broader Amazon shopping habits. And the list, I’m sure, goes on. 

There is even speculation that Amazon will begin licensing this technology to other retailers, similar to what it does with Amazon Web Services. That seems like a reasonable assumption given the data play we just talked about. Assuming the tech works, it’ll get copied. So they may as well embed themselves.

In case you were wondering, the Bureau of Labor Statistics pegs the number of cashiers in the US at about 3,555,500 (2016 number). And this number is projected to remain more or less flat until 2026.

That doesn’t feel right to me.

Below is a keynote talk by Benedict Evans about what’s going on in tech today and what may happen in the next ten years. It covers: the growth of mobile; S-curves; Google / Apple / Facebook / Amazon (who knew Amazon had so many employees?); machine learning; autonomous vehicles/impact to cities; mixed reality; crypto-currencies; and so on.

For those of you interested in crypto-currencies – and that appears to be everyone these days – it’s interesting to hear how Evans describes their current position at the beginning of the curve: “The tech works, but what’s the use case?” This is not to say the potential isn’t huge. It is. Automated trust. Distributed and programmable money. But the future is still unclear.

If you can’t see the video below, click here.

[youtube https://www.youtube.com/watch?v=cVYDkPidXrU&w=560&h=315]