One of my first jobs—not even out of University—was working at the now defunct GEC Marconi. I was hacking on neural network code to do real-time compression of video down a copper wire. Hopelessly primitive stuff by today’s standards, but interesting work at the time. It never went anywhere, although sometimes—if I squint—I think I can see bits and pieces of the code I wrote back then… elsewhere.

I remember having to walk over to the next building—to the IT guys—and ask them to download something from the outside world for me. There was a single 64kpbs line into the facility and it stopped when it reached the server room.

Email was batched, there was no real time communications with the rest of the company, let along the rest of the world. We weren’t really connected to the Internet, we were connected to some network that was in turn connected to the Internet, but wasn’t really part of it, yet.

If you wanted to talk to someone you picked up a telephone handset—it had a wire coming out the back which disappeared into the wall—and then you sent sparks down the wire by turning a rotary dial to move a mechnical switch into place to connect you to the person you were dialling. Today that seems almost crazy.

“The most compelling reason for most people to buy a computer for the home will be to link it to a nationwide communications network. We’re just in the beginning stages of what will be a truly remarkable breakthrough for most people — as remarkable as the telephone.” — Steve Jobs, the Playboy Interview (1985)

Today I’m writing this essay inside a web application hosted on a server a lot closer to mean sea level than my current position—which is 38,000 feet over Minneapolis—about half way between SFO in San Francisco and JFK in New York. I’m typing the words on a slab of aluminium, plastic, glass and silicon. It’s amoungst the thinest laptops in the world and has a glowing Apple logo on the back. There are no moving parts.

Just a few years ago the slab of aluminium I’m using to type on would be inconceivable, and the fact that I’m connected to the Internet from a seat in a commercial jet liner would be impossible.

Today I take it for granted, I take for granted that yesterday I talked to my wife—despite me being in San Francisco and her more than 5,000 miles away back in England. That I could listen to my son being read a bedtime story, that I could say “Goodnight” to him and he could reply and tell me he misses me. That I could wave to him, he could see it, and wave back. All without wires.

A commonplace, a mere bagatelle, a technological trick that I could see both my wife and my son in living colour.

The very phrase “living colour” dates me, and as someone whose job it is to know about things before other people know about them—to spot trends—to predict the future, I’m none the less a relic. The first computer I ever touched was fed with punch cards and had a switch panel on the front, big metal switches, used to program the bootloader so that the operating system could be loaded, so that you could then load a program that actually did something for you.

If you took my slab of aluminium back in time and showed it to me then it would appear utterly magical. That was only thirty years ago. That’s no time at all.

Yet people seem blind to the past, and for some reason every next big thing in technology is seen as a final and last solution to all of the problems of the computing industry, and right now that thing, the solution, is called the cloud.

But the trend towards cloud architectures we’ve seen over the last few years isn’t sustainable. With tens of billions more Internet connected devices arriving over the next few years—far faster than any predicted increase in bandwidth to outside world—data is increasingly going to become a local problem, rather than a cloud problem.

That’s a good thing, not a bad thing.

You never have better context for data than at the moment of collection. Reconstructing that context later—or elsewhere—tends to be expensive, if not outright impossible. Local decision making, where the data is taken, and as close to the time as the data was taken as possible, can improve outcomes and the quality of those decisions.

Think ahead to the time when everything is smart, and everything is networked, when computing has diffused out into our environment. Imagine for a moment if there were a computer in literally every object in your house. In your office. In your neighborhood. Imagine computing being blown about by the wind. Imagine if it was literally ubiquitous.

The phrase data exhaust will be no longer a figure of speech, it’ll be a literal statement. Your data will exist in a halo of devices surrounding you, tasked to provide you with sensor and computing support as you walk along. Calculating constantly, consulting with each other, predicting, anticipating your needs. You’ll be surrounded by a web of distributed sensors, computing, and data.

The cloud is just another artefact of the stage of our technological progress, another turn of the historical wheel. There is no solution to all of the problems, because the problems, and the solutions, keep changing.

If you learn anything from history, learn that.