Latham's fantasia requires a certain patience, however. Downloading it from his homepage on the World Wide Web takes close to an hour with a standard PC; growing tired of the ever-mutating dazzle, however bracingly strange it first appears, takes rather less. And your eyes hurt.
This is the state of computer art exemplified: innovative, increasingly sophisticated and available, and yet held back by a suspicion that it may be nothing more than a kind of visual Muzak - one man's masterpiece is another man's screen-saver. As with other products of the communications boom, the potential claimed for digital art oscillates wildly between, for its evangelists, a revolution in artistic production and consumption, and, for the sceptics, a few technical advances of limited significance.
What is certain is that the new computer technologies are changing art. Software programmes for drawing, image manipulation and animation, like Adobe Photoshop and Freehand, sell in their thousands from the same high- street shelves as spreadsheet packages. Work produced with these tools dots the Internet, islands of colour and movement, amateur or otherwise, in its ocean of type. Some pieces are collected in virtual galleries, to be examined, bought and sold free of the intimidations of Cork Street.
The wider consumption of art is altering too. Stacks of CD-Roms are a commonplace in gallery shops whenever a major exhibition arrives. More grandly, Bill Gates of Microsoft is piling up the rights to some of the world's larger art collections into an electronic library of his own. So far, his private company, Corbis, has bought licences to sell on reproductions - for CD-Roms and other multimedia products - from the Hermitage Museum in St Petersburg, the National Gallery, and the vast photographic inventory of the Bettman archive. Corbis owns well over a million images and adds another 25,000 every week.
Then again, the appetite for electronically stored art is often exaggerated, usually by the companies paying for the storing. Of the nearly 409,000 people who visited the Tate's Cezanne show last spring, only 985 bought the accompanying CD-Rom. At pounds 39.95 - with a computer and CD-Rom drive costing well over pounds 1,000 needed to use it - the multimedia art disc remains a boutique purchase.
Public galleries may offer a better hope of genuinely mass access. Last month, in its first review of museum usage since the Thirties, the Government announced plans to direct up to pounds 500m from the National Lottery into making their contents electronically available. The National Gallery has already done so. Since 1991, a slim first-floor room in its Sainsbury Wing called the Micro Gallery has drawn one visitor in every 60 to a bank of colour computer monitors. Free to use, each gives entry to an intricately cross- referenced archive of all the parent gallery's pictures, linked up by crisp context and criticism. And the Micro Gallery works: students take notes, French tourists plan and print out careful gallery tours, young children flick between Canalettos. In contrast to CD-Roms, the Micro Gallery aims to be an appetiser rather than a substitute for physical artworks, and tries to maximise its appeal by pretending not to be computer-based at all. Users touch the screens; they have no keyboards.
A similar ambivalence about technology is beginning to colour art made with computers. Karen Noor is a photographer who manipulates her pictures on the screen, and teaches too at Goldsmiths' art college. "I see some student work done on computers, and it's just effects," she says. "Form without content." In the preface to Multimedia Graphics, a recently published book of contemporary computer art, the designer Neville Brody, a champion of digital design, worried about "the lack of adventure and risk being taken ... Most CD-Roms are little more than books read using VCR controls."
Some of these limitations are technical. All computer-animated art must, for example, make a trade-off between speed and detail. And the visual quality of a screen, lit from behind, is quite different to that of a solid artwork, reflecting light. But more constricting is an eternal truth about any creative form: only a small minority of artists are much good. "If they didn't succeed in traditional media, they won't be rescued by computers," says R Mike King, who teaches computer art and animation at London Guildhall University. His best students are often the most adventurous with pencils too.
TIMETABLE OF REVOLUTION
The impact of new technology on art is unpredictable: screen-printing gave us Pop Art, holograms a few quickly-kitsch window displays. The imagination and enthusiasm, or otherwise, of artists and their public is at least as important as the technical advances themselves. This is the enthusiastic version of future events.
1999: large thin screens begin replacing paintings on the walls of the rich and fashion-conscious. The images they show come from CD-Roms and databases of old artworks or, more daringly, computer animation. Restless collectors change what they see via on-screen controls.
2000: Turner Prize given to a computer artist.
2005: art enters virtual space. Consumers put on headsets and lose themselves in apparently three-dimensional works. Half an hour wandering a Dali landscape, converted to digital by an army of programmers, replaces a long wait at the Tate.
2010: Artificial Intelligences, or computers with limited independent thought, start producing art works all by themselves. Less need to buy up young talent from Goldsmiths' (now employed reworking old paintings for the new media - see 2005.)
2020: creation by humans becomes largely confined to artists funded by computer companies. Bill Gates's Corbis corporation runs out of visual images to buy licences for. Silicon Valley swamped by villa building and comparisons to Renaissance Tuscany.