Don't make plans for 30 years ahead!

Don't plan too far ahead

One for the Black Friday sales: Before you make an investment in technology, make plans to mitigate your equipment's inevitable march to obsolescence.

Working with media – digital audio and digital video – puts us in a privileged position in comparison to the rest of the world. We're used to working with cutting edge technology. We don't just use it passively. Typically, we push it as hard as it will go, and then we still want more.

So we know what we expect from technology, and it knows what to expect from us, which is more of the same, but faster, smaller and cheaper.

We also know about paradigm change. We understand what it's like for our fundamental concepts to change almost overnight. The big change, from analogue to digital, has been going on for ages, but there are smaller transitions – like from HD to 4K, or from standard dynamic range to HDR, happening all the time and it's not slowing down.

Even casual consumers who stop to think about it for even a minute (and while plenty don't, that still leaves quite a few who do) know that today's iPhone is many, many times more powerful than the first one that was announced only five years ago, The iPad Pro has a screen with more resolution than the big monitors we have on our desks. It's multi-touch sensitive and has a pretty powerful processor in it. And it's portable. And it has GPS, Gyroscopes and all manner of deeply clever stuff.

When you connect all these devices together, you get a networked world that has almost no limits in any direction. There's no clear trend here except faster and more powerful - even if it is only a result of connecting things together cleverly.

So it comes as a surprise to see that, for example, the UK government is considering renewing a nuclear missile system in a plan that will take thirty years to deliver.

Now, this isn't a political publication, and this is not about the arguments for or against nuclear deterrents. It's about how anyone, who knows anything about anything, can embark on a technology project that's intended to deliver in three decades, when in that time, we might be living on the moon or may even have uploaded our entire selves to Google.

Well, maybe that's being a bit facetious. But the point is that big changes are going to happen – even bigger than the changes we think are big at the moment.

Ever-quickening pace

Technology is accelerating faster than ever. There's never been more evidence to show that technology is growing in power exponentially. This is specifically not a linear  process. It's a process of multiplication, not addition.

Some say that this is no longer the case because Moore's law is showing signs of slowing down. Even if that is true, and it's by no means certain, there is plenty of technological innovation going on outside the niche area of semiconductor fabrication to keep the slope of the curve increasingly upwards.

Just to put this into perspective, imagine that you're in charge of the biggest television and film production project in the world. You're going to build a headquarters that's going to be huge and expensive. It's so big, that it's going to take thirty years to build.

Where do you start?

Certainly not by guessing what kind of technology we're going to have in thirty years time. I mean, where would you begin? Sure, we had SD for several decades. HD (in the living room) is around ten years old and 4K has just arrived. 8K is already on the horizon and some people (mainly me, to be honest!) are talking about future video systems that won't use pixels at all, which probably won't use screens either, because they will talk to the graphics primitives in our brain.

Here's what you should do instead. Take a look at how we do things today. See what the trends are. Plan to be able to accommodate them. Don't crystallise your design with components that you're going to have to rip out in a few years. If you do have to rip anything out, make sure you make it easy. Don't bury them under expensive marble flooring.

That's the key. Plan for what you can plan for, and plan for what you can't plan for.

Managing innovations

One trend that's likely to continue is that we get more performance for less money, in less space. You can't be certain of this, but it's a well established vector. We will continue to need this extra performance because new formats and standards will demand more processing power, more storage etc.

It's not just nuclear subs. It's big organisations, too. The UK's health service has always struggled to get to grips with the latest computer tech. There's no shortage of excuses, the most frequent of which is "confidentiality." We've seen all sorts of IT initiatives planned for the NHS (the National Health Service) and most of them fail, because by the time they're supposed to be delivered or turned on, technology has moved on and implementing them would be a backwards step. So it costs even more to have them brought up to date. You can see how costs can escalate.

In some places, if you have an X-ray taken of your spine, it can take ten days for it to reach your doctor. There's no reason for this: email would do nicely (except confidentiality). All of this in an age when Facebook, a free service, accepts 800 million picture uploads per day. Confidentially must have a nuanced meaning in this context, when most patient's records are stored in a brown envelope on bookshelves at their doctor's surgery, in plain text, visible to anyone who works there.

What's the bottom line here?

Anyone that thinks technology is improving linearly can only make short term decisions, or wrong ones. If you make detailed plans for delivery in anything more than five years, it's a recipe for either massive waste or complete disaster.

You can't go on experiencing 1000 fold increases in technological capability every ten years without some massive changes in infrastructure and society.

We're in a better position than most to understand this. We need to make sure the people in charge understand it, too.

[Mushroom cloud courtesy of www.shutterstock.com]