Fifty years ago this week, Tomorrow’s World started a 38-year run of weekly TV programmes designed to showcase what Prime Minister Harold Wilson had characterised as the “white hot technological revolution”. The name of the game for the programme makers was, in general, show-and-tell. During its lifetime the show featured more than 7,000 gizmos, ideas and discoveries, from new tin openers to computers to the latest in heart-surgery techniques. And more importantly (the clue’s in the show title), what these innovations would do to the future.
Some they got right, including: breathalysers, digital watches, CDs, ATMs, camcorders and barcode readers. Some didn’t do so well: Concorde, hover-everything, Mars by 2000.
All things considered, though, the programme did OK, given that the perennial difficulty with prediction lies in the insidious Murphy’s Law. If anything can avoid turning out as you predicted, that’s what will happen. In the main, this is due to the way innovation works. Usually because ideas or objects come together in a new way, and when they do, 1+1=3. Example: in the 19th century, the perfume spray and petroleum came together and the result was the carburettor.
On other occasions the ripple effect of an innovation goes beyond the foreseeable. The first typewriter triggered a requirement for carbon paper. But it also brought women out of the kitchen into the office. And boosted the divorce rate.
Accidents happen, too. In the mid-19th century, a London chemist noodling around with coal tar (looking to make artificial quinine) fails to do so. However, the sludge he comes up with turns out to be the first artificial dye. Then somebody spills some into a bacteria culture dish, where it stains only one bug. Bingo: chemotherapy.
You get the point. And the variables involved in change can be massive. No wonder Tomorrow’s World was fighting a losing battle. Best they could do was take an object and predict how it would develop. There was no way to see how it would interact with all the other things in the future. The numbers were just too big. Back then we didn’t even have calculators.
But today (as we used to say back in Tomorrow’s World), things are about to change.
Key in this, is Big Data. That’s “big” as in: “everything”. The amount of data we’re generating today is enormous and, so far, most of it has had to be trashed for lack of space to save it. But recent computer advances mean we can now use it, rather than lose it – starting with the previously “superfluous” data exhaust that we, and all the organisations we use, spew out each time we do anything. Exhaust such as: bank records, insurance, health care, travel, passport, school, family, shopping, car, holidays, income, credit cards, location, phone, friends, tastes, photos, lifestyle. And everything we ever do or say on the internet.
And five years from now there’ll also be 200 billion tiny sensors (they already exist and you can buy them on a roll of sticky paper: tear off, stick on, and monitor) gathering and transmitting data on objects. On what’s known as the “Internet of Things” – everything from underwear, to oil-well performance, to whether your fridge needs stocking-up.
Over the next 50 years, as we collect, store, analyse and manage all this data with the help of trillions of powerful “smart dust” computers (each the size of a speck of dust and embedded in everything; one of these already exists) then we’ll do really clever things with Big Data.
First of all, algorithms: search procedures that you use to find what you’re looking for in the ocean of online data. It’s an “ocean” that includes as many as 30 million comments, 20 million tweets, 3.8 million blogs, 7.1 million photos, and one billion Wikipedia searches. Every hour!
But in terms of prediction, the best trick with the search algorithm is not to ask it to find something specific, but rather, ask it to find any patterns in the data. That way you don’t limit the search to stuff you expect to find, or that you already know something about. So what you get are the “unknown unknowns”. What you didn’t know you didn’t know. Like how the perfume spray was going to generate the carburettor.
If you were to do this trick with the Big Data you’d gathered, say, from every lab, research institution, manufacturer and university in the country, the patterns found by the algorithms could reveal when many of the ideas and research, happening at thousands of separate locations, were likely to come into contact with each other, initiate 1+1=3 innovation, and then cause change. You could, of course, then run another search for the ripple-effect patterns you could expect to see happening.
So now that you have a near-perfect prediction of innovations “in the works”, so to speak, you then do the same pattern-search in social media of our tweets, Facebook postings, blogs, and the rest. And what you’d hope to find is the expression of massive, real-time public opinion, reflecting users’ lives, their desires, likes, hates, intentions and all that. For the first time, the true vox populi.
Bringing those two sets of patterns together might show what the entire public (not just some small, “statistically meaningful” group) thought of any upcoming innovation and whether or not they wanted it. So now you’d know what’s coming and whether or not it will be accepted. Running the exercise backwards could also provide potential innovators with a public wish list.
With Big Data predictive analytics you could not only predict innovation with extreme accuracy, but also learn what to do to make it happen. Effectively, you could design the future. You may not like this idea. But don’t worry… just remember Murphy’s Law.
James Burke talks Tomorrow’s World on Radio 4 tonight (Saturday 4th July) at 8.00pm