Never make predictions - particularly about the future.

There have always been plenty of people around willing to predict the future. But, it's a risky business. The fearless forecasts of the past often turn out to be the chuckles of the future.

In 1943, Thomas Watson said: "I think there is a world market for maybe five computers." At the time, Mr. Watson was Chairman of a large company that we'll shield from further embarrassment by just giving its initials -- IBM.

In 1962, the Decca Recording Co., turned down a young group of musicians saying: "We don't like their sound, and guitar music is on the way out." The Beatles went elsewhere.

In 1895, Lord Kelvin, President of Britain's Royal Society, predicted that "Heavier than air flying machines are impossible." For a mathematician and physicist, poor old Lord Kelvin seems to have had a habit of getting it wrong. In 1900, he declared that Wilhelm Roentgen's discovery and use of X-rays was "a hoax." A year later, Dr. Roentgen was awarded the Nobel Prize for Physics.

The end of the 19th century seems to have been a bit of a Golden Age for statements full of blunders. In 1899, Charles Duell gave it as his opinion that "Everything that can be invented has been invented." Perhaps, Mr. Duell had just seen too many wacky ideas cross his desk in his job as head of the U.S. Patent Office.

It seems we never learn from the mistakes of others. In 1996, John Horgan stuck his neck out and wrote a book called The End of Science. A senior writer for Scientific American, Mr. Horgan claims all the major scientific questions have been answered and all that remains to be done is tidy up a few loose ends.

Mr. Horgan should know better. Nobody can ever know with certainty what the future will bring. That's because the future doesn't exist before it gets here. And, when it does get here it often carries a surprise or two with it. The surprises are what makes humble people out of most forecasters.

The only way we have of taking a stab at forecasting the future is to look at the past. By tracking what has happened we can project trends into the future. Critics say this kind of predicting is merely "forecasting the past." Others call it "driving in the rear-view mirror."

It's fairly safe to predict that computers will get faster and cheaper in the next decade. That's because we know that for the past couple of decades the price of computers has steadily come down while their capabilities have increased. In the 1990s, the cost of computation has halved every three years. The power to store, miniaturize, and speed up operations has doubled every 18 months.

But, that prediction might go flying out the window at midnight on 31 December 1999. Many prophets of doom predict that computers all over the world will start going berserk. That's because many of them are programmed to recognize a year by its last two digits. So, when "00" ticks over in their internal time clocks a lot of computers could shut down because they don't recognize "00" as a valid date.

This is the sort of surprise that can sometimes make experts who predict the future look like total idiots. And, it's why we should always remember that any forecast is always a best-guess estimate and not a guaranteed certainty.