Step back in time

Even for an expert in the field, it’s impossible to predict the future – particularly when it comes to the surprising world of computers

If you predict the future, you get it wrong. That will always be true – with exceptions. For me, the past 30 years have been a succession of mistakes.

People have rubbed my face in some of those mistakes. In my first column for PCW, I sang the praises of associative memory.

It was, to my mind, a true revolution that would allow the memory storage device to make processing decisions, rather than having to wait for a central processor to read the memory, compare it, and perform a judgement. How could it fail to replace the Von Neumann architecture?

But it did. Other predictions that went wrong were more or less forgiveable, and the fact that they came in for mockery and scorn was partly a result of perceived disloyalty.

In the 1990s, having produced PCW’s Newsprint section for the best part of two decades, I ran away to earn more money from the newly launched rival PC Magazine. The PCW editor’s revenge was swift: he started a regular column of ‘Guy’s Past Blunders’ (using a different title, of course) and found plenty to choose from.

I didn’t take the job of writing the news for PCW in an attempt to become a futurologist. It was bleeding obvious to me that the microcomputer chip had the potential to transform society. I lacked the depth of insight available to writers such as Peter Laurie (a colleague of mine at New Scientist at the time) or Chris Evans (at the National Physical Laboratory), but even so I was astonished how many people were blind to the micro’s potential.

Today’s readers may not remember how expensive computer technology was. A friend built his own Altair computer and then had to buy a teletype to get data in and out of it because the glass teletype, or video display, was still a rarity. Whatever it printed disappeared off the top forever. And that teletype, printing at 30 characters a second, cost him well over £1,000.

My salary at the time was comfortable at £5,000 per year. “Nobody could afford a home computer,” said the typical pundit “and even if you could, you’d still need to spend three months’ salary on a teletype.” Only a fool would imagine such a thing becoming important to the average human.

So when I wrote in 1978: “We are on the edge of a revolution that will make the printing press, the telephone and the motor car look like minor items on a shopping list, as the population gets ‘on-line’, and from here on, the history of the computer will be history of society, not just of calculators,” it was, for many, simply proof of my naive ignorance.

What they didn’t understand (but I did) was that mass production changes everything. “The Zilog Z80 that you’re so excited about,” said a Marconi labs scientist pityingly, “isn’t even fast enough to operate as a disk controller.” True - but utterly irrelevant.

I showed similar abilities to miss the point many times over the history of this column.

There was, for example, the time I mocked the idea of looking at pornography on home computers because the graphics were so bad, only two years before the graphics revolution rendered such tricks trivial. And similarly, when Intel first launched the Pentium processor as its top-of-the-range product, I remember mocking them for suggesting that it was a games processor.

I had forgotten that it would probably take less than a year to go from launch to el cheapo junk and that everybody would want a Pentium III within half a decade.

But you can make equally bad mistakes by assuming that what is costly today will be popular tomorrow. I’m fond of quoting Al Shugart on this; he correctly forecast that silicon memory would become horribly cheap but failed to anticipate just how cheap, or how big, rotating disks would get. Ten years after he forecast the end of disks, we still use them.

So here’s to the next 30 years, and the surprises that will come with them - mostly to do with low-energy systems and portable computing. And let’s hope it’s as much fun as the past three decades.

This article was taken from the April 2008 issue of PCW.