Skip to navigationSkip to content


Our home for bold arguments and big thinkers.

Reuters/Jason Redmond
Amazon Go in action.

In our new situational economy, the challenge is to stay human

Quentin Hardy
By Quentin Hardy

Head of editorial, Google Cloud

The combination of sensors, smartphones, and cloud computing is enchanting the world. On average, there is now almost one connected device for every acre of the world’s inhabitable land. One upshot to our newfound connectivity: Nowadays we read the world, and the world reads us back. That may become the heart of a new kind of commerce and services, where behavior itself is the means of payment.

That is what will happen in the cashier-free stores of Amazon Go. You enter a store with phone authentication, and once you’re logged in, cameras and sensors note what you take off the shelf. Walk out of the store with a bottle of ketchup, and a card is debited. The financial trigger is where you are and what you hold.

It happens with Face++, owned by Chinese company Megvii Inc., which looks at some 83 facial features to authenticate people or send money. Face++ operates in the province of Zhejiang, but the technology is used widely in China, and other countries combine recognition of facial features and location for getting a cab, surveillance, and even dating.

It may also lead to the creation of signs and advertisements that change themselves depending on how you read them. For customers of Google, my employer, deploying software may eventually be a matter of simply describing what you need it to do. Which mean time clocks and piecework may give way to new kinds of productivity measurement for workers.

Turning “situational awareness” into products and commerce is something like the way a package leaves a warehouse in online shopping. In that case, the transfer from loading dock to truck is logged, initiating a perky email telling the buyer that her stuff is on its way, and charging her credit card.

That is a point in a relatively stable supply chain. But in the new technology, frictionless commerce is and will be created through combinations of location awareness, mobile apps, computer vision, and sensors, all managed remotely in globe-spanning cloud computing systems of companies like Amazon Web Services, Microsoft Azure, and Google Cloud. Hook stuff together, and every transaction can spin up an ad hoc virtual chain.

“It’s the fluidity of a system without any borders,” said Bernardo Huberman, a lecturer at Stanford University. “Right now in places like Denmark you can pay for everything with your phone, but the phone will go away too.”

Mr. Huberman, who was a senior researcher at Hewlett-Packard’s Labs before it was shut down last October, built a prototype airport sign that was equipped with inexpensive eye trackers. The sign had basic instructions in several languages, and when it figured out where a traveler’s eyes concentrated, it would translate itself into that person’s language.

“You could do this in stores, too,” he said. “Notice someone stops to look at pictures of belts, and they’ll see belt displays.”

Things are increasingly invisible and frictionless at the business end, too. Software makers used to install computer servers, then add databases and connective middleware before they created a service. Now they sign into clouds, and name the virtual servers and associated software they want. That isn’t just less work: It’s a different way of thinking about things.

Instead of designing a product or service in terms of its siloed attributes—a maps program, a payments program—coders will think about the needs of a particular point, and pull in that capability through programming interfaces. Describe what is needed, and increasingly there’s an asset to do it. That makes it lots easier to create virtualized commerce based on nothing more than our changing attributes, like where we are, what we hold, and what we look at.

That process will soon have even less friction, said Urs Hölzle, Google’s senior vice president of technical infrastructure. As employee number eight at the company, he has built one of the world’s largest computer networks. He thinks it will soon become even easier to describe and build these invisible systems.

“Five years from now, you won’t think about what type of machine you want, you’ll just say what you want to do and it will get configured automatically,” he said. “People think of the cloud as cheap computing, but it’s more about automating things.”

Of course, automating things and eliminating friction is something computer technology has done for decades. Software, perhaps the greatest wealth creator of our time, is itself an invisible product, a series of statements about how to organize something.

But just like with geology or evolution, incremental changes eventually make differences. A new land takes shape, with new creatures.

What new creatures will we be, both as makers and consumers, in this world? Clearly, when we sign up for these services, we are allowing third parties to have greater awareness of us. That kind of third party exists today, whether in home-security cameras, social-media networks, location permissions, or customer-loyalty cards. Large and complex invisible systems may present their own issues, with more parties sharing data.

“Let’s not kid ourselves—there are lots of issues of privacy here,” said Huberman. “But the fact is, most of this stuff is already here. It is just getting packaged more densely. Systems will have to become flexible enough to handle all of your contexts.”

If Hölzle is correct about automating how software is written, and the main job of designing things well will be naming the attributes well, then privacy and appropriate personalization will have to be part of the naming. Regulators, consumers, and companies have all discussed this sort of thing for years, but there will be new urgency.

A bigger shift will be how we think about our own work, which is a big part of identity and meaning for most. For much work, hourly pay, redolent of punching time cards early in the first Industrial Revolution, will likely be supplanted by other measurements.

The tech that enables an Uber or an Amazon Go will hit workers, too. Imagine a chiropractor getting an automatic bank deposit when a customer leaves the table, or a developer paid the moment one customer (or 1,000, or 1 million) interacts with her code. Throw in fitness trackers or internal sensors, and re-imagine payments to healthcare workers.

“Compensation is ultimately an exchange of value,” said Josh Reeves, the founder and chief executive of Gusto, a cloud-based provider of payroll services. “There will be ways to make better work shifts, because the system sees days when dropping off the kids is a hassle, or better commissions from figuring out what and where a person is good at selling.”

Reeves worries about the costs of defining the system wrong. “If everything is just about compensation, though, it’s going to get depressing,” he added. “It’s worrisome that in Uber-type jobs there’s no way to move up. Purely incentivized behaviors about dollars lead people to game a system. Things like satisfaction, pride, and fulfillment have to be part of how we describe people.”