The government is worried that one third of people between the age of 50 and the state retiring age of 65 are not working. Worse still, a recent report stated that the age at which executives in large companies are commonly made redundant has dropped from 46 to 42.

I was lucky. In 1981, I was made redundant at the ripe old age of 51, by a computer company for whom I had worked 28 years. My wife feared that I would never work again. However, living by my wits, I am still gainfully employed in the computer industry, and celebrated my 70th birthday last month.

This is somewhat miraculous, and makes me, I suppose, having seen off all my contemporaries and most people up to 20 years younger than me, an expert on the art of survival in the IT industry.

It is therefore my duty to offer some advice to people approaching 40. My first tip is: start networking now. Get to know people in your field in other companies. Attend meetings of your local branch of the British Computer Society or other professional body. People in your own company will be no use at all. If you get the chop, they will not want to know you.

I have given this advice before, and been roundly berated by programmers, deep into complex projects. They think I am trying to distract them from their serious job, which will be inevitably recognised by their management as being a big contributor to the company's success. OK, maybe the management is grateful, but the next project needs Java or XML expertise. Terribly sorry, but you don't have that, so we are hiring a new team of teenage scribblers. You are out.

I have seen this before. There were the people who would not accept Cobol or Fortran because "machine code programming gives you more control". There were those who saw PCs as Mickey Mouse, because they were not as massive as mainframes, and could be used by human beings, not just PhDs in white coats. There were those who saw no point in colour screens or graphics. More recently, I have seen people reject the internet as impossibly insecure, and some who cannot see the importance of full-screen video.

I have come to the deeply worrying conclusion that the majority of computer people are like that, particularly those who work in the IT departments of large companies. They learn whatever is the flavour of the year when they start, become very arrogant of those with older skills, and five or 10 years later, reject anything that follows. They seem to resent the computer's increasing ease of use, because it decreases their priestly power. These people, and I estimate them to be 80% of the profession, deserve to be made redundant at 45, or to play only menial roles in the computer industry, the industry of perpetual innovation.

But I am concerned about the remaining 20% of perpetual innovators, who are in danger of getting thrown out with the dross, just because they are over 42.

For starters, get out of large corporations if you work in one. As a journalist, I meet a lot of corporate "suits" and few of them are over 50. Do not ask so-called "human resource" consultants for help. They are even more ageist than those who hire them.

Small start-ups are different. I interviewed Michael Dell when he was 22, but he had a 50-year-old minder alongside him. I find this a pattern. Whiz-kids need the wisdom of a greybeard somewhere close.

This is particularly important now. The dot.com entrepreneurs seem to think that a sexy website will make them a fortune. They are wrong. They need a greybeard who can do the plumbing behind the website, to give a smooth flow of data down the supply chain. And somebody, also probably of an advanced age, who will make sure that the goods get delivered.

The trick is to find a whizkid before he becomes over-confident and thinks he can rule the world without your help. So, you have to suss out the next decade's technology and its pioneers, and latch on to them. Most people tag along with today's technology until it runs out of steam in five years time. Being ahead of the game is risky, but, if you choose correctly, can add 10 years to your career.

You do not have to hitch your wagon to someone else's star. The people I have met in the past 20 years who have impressed me the most are aged visionaries who saw e-commerce coming decades ago, and have laboured with little reward to get the standards and legal infrastructure into place.

One of these was a guy no one will have heard of: Bernard Wheeble. He produced, through bodies like the International Chamber of Commerce and the United Nations, most of the ground rules for the legal acceptance of electronic signatures and electronic bills of lading, which are now incorporated in e-commerce legislation across the world.

Wheeble was one of the most electric speakers on the international e-commerce conference circuit into his 90s. I used to be pinned back in my seat by the crisp logic of his arguments.

I asked him once why he kept at it. He answered that he had to move fast to get the legal infrastructure in place before the e-commerce revolution took off. He died two years ago, but, by the skin of his teeth, had just had time to do what he had planned.

If you have a vision, age is no barrier. The only barrier is death.