by Andrew Starling

When you create a new Web site, what's your prime audience? People or machines?
July 19, 2000

Allow me to ramble on for a moment about Ray Kurzweil's "The Age of Spritual Machines". It's highly relevant.

The basic premise of the book is that we're on the verge of inventing intelligent machines that are clever enough to manage their own reproduction.

They may not be genuinely spiritual, or original, or deep-thinking, but they'll be able to defend themselves and reproduce, which means they become a species, in a primitive sort of way.

The book also explains that superior species invariably win the battle for survival, and there's a risk that the machines we create will probably be superior to humans in a few important aspects. Somewhere along the way we'll find ourselves subservient to them. Ultimately, if we get into a conflict with them, we may lose.

You can see the idea portrayed in science fiction films like Robocop and to a lesser extent Blade Runner. Many people believe the idea belongs right there, in the movies and in books, because it will never become reality.

I'm not one of them. And other people I know who also have a deep knowledge of science see the book as prophetic. We're building dangerous tools with our high technology. Ultimately we'll be driven by commerce into making slip-ups and we'll lose control.

Maybe we won't lose control to the extent of becoming overrun and dying out, though the risk exists. But the machines we develop will be so advanced that only other machines can design and build them, the details will always be a mystery to humanity. That's quite a lot of trust we'll be placing in machines.

And crucially, it's our generation that's creating the technology and building the wired planet these machines will need if they are to take on independent life. We are, as a generation, the equivalent of Oppenheimer's A-bomb team. We're building something very technical and interesting, but we won't be in control of its final use.

I first came across examples of this process in action when I edited a magazine about semiconductors (microchips). In some factories there are real robots that move around taking half-finished products from one processing machine to another. The internal design of the chips is so difficult that it's no longer possible for humans to do it. And the process of physically building them takes even more computing power. Clever computers are used to design and build the next generation of chips that they themselves will eventually incorporate in their own electronic bodies. That's not too far away from the regular concept of reproduction.

Internet

It hasn't been so noticeable in the Internet industry. Sure, I can see that we're all so dependent on computers and email that if we get a power cut or a major computer virus we may as well go home. But that's not much different to a blacksmith needing furnace fuel or an old-fashioned ploughman needing his horse to be fit and well.

I've also recognised that personal computers aren't user-friendly. The only way we manage to deal with them is by users becoming computer-friendly. But again, no big deal.

A few weeks ago I changed my mind.

I'm creating a new Web site, as a sideline, as people in the Internet industry do. It's the sixth site I've started from scratch (if I remember correctly) and I've learned plenty from the previous five. One of the big things I've learned is that very few visitors arrive in the early months. Unless you've spent a fortune on publicity, your site bursts into the virtual world with a whimper rather than a roar.

Instead of being disappointed by this, I now work with it. Some of the search engines take four months to index a site, and only a few manage to return results within the first few weeks. Meanwhile the number of visitors each week can be measured in dozens, maybe reaching the low hundreds if you're lucky.

So my philosophy is to let the site mature on the vine. Get the subject matter right in the early stages, spend enormous amounts of time on keywords (half of all initial work), then mount the site on a server and register it with the search engines. For a month or so afterwards, pretty much ignore it. Let it ripen in the sunshine.

At this stage, there's no real need to worry about the graphics, grammar, comprehensive links and possibly even accuracy, because hardly any people are going to find this precious new creation. All I'm interested in doing is making sure the search engine robots can find it and successfully pick up the keywords.

There, I just said it.

My first priority in creating a new site isn't humans, it's machines. I want to please computers and their scouts, otherwise known as spiders or robots. I want them to come and enjoy themselves and go back with wonderful memories, which they eventually index and add to their databanks.

Sure, I'll get around to refining it so humans find it worthwhile too, but that can wait. My first priority is its attractiveness to machines.

Now tell me, isn't that just a little bit scary? It's probably the first time in my life I've created material primarily for the appreciation and approval of machines, as a kind of species, and appropriately a species called robots. I even write a file especially for them, the robots.txt file.

I don't think it's a revelation that's going to change my life, or anybody else's. But it's sure been an interesting wake-up call.

I am now trying to please a species of machines. Welcome to the IT A-bomb team.