It governs a lot of your digital life these days, but the story of where it first materialized is likely deeper than you know.
You’re surrounded by it.
You can’t send an e-mail without touching it.
You can’t watch your favorite streaming series without welcoming it into your home.
You don’t know where you’re going, but it’s tracking your every move and telling you to turn left.
It is “the cloud,” and whether you realize it or not, it has probably taken over your digital life.
But what is “the cloud” actually?
The cloud is a system of millions of hard drives, computer servers, signal routers and fiber-optic cables.
These elements are, in a way, like the water droplets, ice crystals and aerosols that make up a true cloud.
They are nebulous. They are constantly shifting. But they are in close connection with one another over large distances of space and time.
The cloud’s true purpose is to float unseen all around us, silently, creating ever present connectivity.
And just like a real cloud, it can bring benefits—and danger. The cloud is making services more affordable and accessible to people all over the world. It helps businesses update products for their customers and enables remote work across industries. But it also opens us up to behavior tracking at each and every Web site we visit. It can put our virtual privacy in the hands of tech giants. And should the cloud ever fail, our increasing, mostly unknowing dependency on it will become painfully clear.
To understand why, let’s take a step back to the early days of the cold war with the Soviet Union.
After the Soviets beat the Americans to space in 1957, the U.S. Department of Defense decided it needed to work harder on research and development. The following year, it formed the Advanced Research Projects Agency. That agency would create the rough draft of the Internet a decade later. It was called ARPANET, and it connected four university computers via telephone lines.
ARPANET was the offspring of a novel and controversial view of computers championed by engineer and psychologist J.C.R. Licklider.
In 1962 Licklider, known as “Lick,” was put in charge of the agency’s Information Processing Techniques Office. And with his budget, he advanced a view of these machines that was very different from that of his peers.
Lick didn’t believe in building new computers for every project. Instead he wanted to consolidate resources into a network of “thinking centers” that individuals could access as needed. This vision, which became the basis for ARPANET and later the Internet, is the premise of cloud computing.
When you access data on the Internet, you’re actually requesting files from a server. The files are broken into tiny packets of information, which may travel together or take completely different routes back to your device, where they’re reassembled. The particulars may get a bit complicated, but all that really matters is that you’re connected to the network.
One of the first visual representations of a network as a cloud comes from 1971.
The previous year, telecom giant AT&T introduced Pittsburgh to the Picturephone, a kind of early videoconferencing system. It was meant to run over digital systems and phone lines. In 1971 Irwin Dorros, then at the company’s Bell Telephone Laboratories, published a schematic of the system.
It showed a couple of cloudlike shapes that would tie together the hardware needed—even if Dorros didn’t know which computers or phone lines would be working together at any one time.
By the early 1990s, engineers had gotten used to referring to the Internet this way. But the idea of “cloud computing” really went mainstream in the 2000s. In March 2006 Amazon launched its first cloud-based service as part of its Amazon Web Services, or AWS.
Initially, Amazon just planned to build a platform to help other companies set up online shops. But it quickly started to recognize that many of the tools and databases it was building could be useful outside of e-commerce.
Amazon began renting out server space and database tools that allowed companies to launch and maintain applications much more cheaply than starting from scratch. For instance, Zillow, the online real estate Web site, uses AWS to store 100 terabytes of house pictures and data rather than maintaining the necessary servers itself.
Using an outside server can be safer than managing your own, because it’s much less likely to overload, and many of these servers will back up your files for you. Most cloud computing services also come with optimization tools to handle traffic spikes and lulls, and many have data centers around the world, making sites load faster for international users.
The cloud can also give you access to more computational power than you can easily get on your own, letting you effectively use a supercomputer from your smartphone.
Cloud services generally fall into three categories: software, platforms and infrastructure.
Cloud-based software are just applications that run on the Internet so users don’t have to download anything. This category is especially popular for widely used programs such as the instant messaging platform Slack or the file sharing app, Dropbox.
Next, cloud platforms, such as Google’s App Engine, are digital environments that developers build their software to run on.
Last, cloud infrastructure provides server space that a customer manages remotely.
Consolidating the digital world onto a few powerful servers is extremely efficient. And the cloud brings us unprecedented connectivity. It’s the basis for the “Internet of Things,” in which embedded sensors connect physical objects such as farm vehicles or building thermostats to the Internet. Once connected to the cloud, all of these things can act autonomously—doing work for us without human intervention.
But while the cloud can make our jobs more efficient and our lives more flexible, we pay for those privileges with our data and security. With every action we take online, we hand personal information to companies trying to maximize how much they can profit on us. Many Web sites and apps regularly track our digital movements and sell those data to marketers. Cloud service providers can also collect data from applications built on their servers. This is often done to monitor efficiency, but some worry the data could be abused.
As we connect more of our daily lives to the cloud, we become reliant on a network that controls everything from who we meet on dating apps to whether our credit cards work.
We also lose sight of how fragile the system is. The cloud only exists because of physical parts, such as paper-thin fiber-optic wires that are easily damaged and degrade over time. And when software issues from a company you’ve never heard of can knock huge chunks of the Internet offline, it’s hard not to wonder if we’ve made a bad bargain.