Soundbytes: Sense and Serendipity

There?s an irritating sign at Watford Junction station. It says, simply, ?Network South East Welcome to Watford Junction?. This unexceptional sign is a typical product of the times. It illustrates the common problem of information overload. The sign has seven words where one (or two at the most) would do. It risks hiding its information content behind a thicket of words which perform no service.

To be certain you have arrived at Watford Junction, all you need is a sign to confirm where you are. If the sign just said ?Watford?, that would probably be enough. There is a station at the far end of the Metropolitan Line also called Watford, but there isn?t much danger of confusing the two.

The rest of the wording is pointless. The inclusion of ?Network South East? is a trumpet-blowing exercise that contributes nothing to a passenger?s understanding of the situation. Is it thanks to Network South East that he or she has arrived there? Or should it be Railtrack, or North London Railways, or Virgin? This part of the sign reminds me of the wording often seen on municipal or contracted-out dustcarts, to the effect that this is such-and-such borough council Keeping Your Rubbish Moving, as though, far from this being their duty as recipients of your taxes, you should be grateful to them.

?Welcome to? is a harmless expression. But if the railway operators employed enough people ? as porters, say, or to enhance your feeling of security ? there would be more chance of visitors feeling genuinely welcome.

The sign might still redeem itself by using typography to emphasise ?Watford Junction?. It does not. The only variation is the use of outlining to draw attention to South East. A foreign visitor scanning the sign might still be wondering where this place ?South East? is as the train pulls out, next stop Motherwell.

The point I?m labouring is that much of the information offered up since the recent dawn of the Information Age is not information at all. Most of it is just electronic noise. It amounts to interference of the kind that made watching the Poland/England game on Channel 5 such a testing experience for many of us. Somewhere in the picture England were winning, but the blizzard in which the game was apparently played made it difficult to be sure.

The amount of general background noise is increasing. Technology encourages it. When signwriting was an art and words on a board cost money, a company would have thought twice before using seven words where two were plenty. Now everyone can be their own signwriter. You can bet that most will lack restraint. You only have to look at the progression from desktop publishing through to the internet. Everyone is a publisher.

So the business to be in now is not information technology but information management. Sales of IT reflect greatly to the credit of suppliers and resellers. But the volume of sales has created a problem. An equivalent of Gresham?s Law ? that bad money drives out good ? has begun to apply to information. Demand in the future is going to come from people who, having acquired IT, need some way of dealing with its output. They will need a means of winnowing the wheat from the chaff.

I?m not persuaded that intelligent agents will make much of a difference. Anything with ?intelligent? in the description must be automatically suspect. Software that learns to recognise patterns in your behaviour may be fine to help with your grocery shopping, but would you want to trust it with anything more important? Would you even want it to determine what you read in the newspaper? Until someone builds serendipity into an intelligent agent, I?ll continue to think of it as a counter-intelligence agent.

At the other end of the scale, suppliers are starting to make big claims for object-relational databases. They promise to extend the benefits of relational technology to new types of data. These are usually referred to as unstructured and streamed data. The standards organisations and most of the suppliers are laying down the ways and means of taking database architectures into the new area.

But to regard this as a solution to the problem is to mistake data for information. At best, such databases will bring all the data into one place. The history of IT suggests the result will be a further explosion in the volumes of so-called information.

The trouble with so much of the supposed information on offer is that it is cheap. It is apparently cheap to source and certainly cheap to distribute. Thus it accumulates. But if information is genuinely a key corporate asset, it should have a value. Knowing how to value information may be the first step towards managing it.