Computer optimization is a powerful art form these days. Starting from a crude approximation in parameter space, an iterative program can 'hill-climb' very rapidly to a useful optimum. Daedalus now wants to apply it to the greatest art form of all, which these days is advertising. Enormous sums of money are cheerfully spent seeking the optimum advertisement that will grip potential purchasers, and the speed and lack of conventional wisdom of the computer is bound to help.

When Standard Oil changed its name to Esso, only human decision was needed. When the latter changed further to Exxon, a computer program churned out thousands of possible names, ignorant of the limitations of the letter 'x' in English, and a human observer reacted to the outcome. So, says Daedalus, what we need is a big screen, or many little screens, all watched by potential customers wired up to headphones, electroencephalograph, heartbeat and galvanic skin-response detectors, and so on, all of which are fed back to a master computer controlling the whole thing.

The initial screen image, maybe a boot or a bicycle, will be so faint that the subjects will probably take it to be an imagining of their own, if a bit self-willed. But, far faster than any focus group, the machine will take everybody through a whole range of options until its hill-climbing optimization routine has identified that sudden frisson of delight or recognition in at least one subject. The whole languages, the gigabytes of visual data, now easily stored and accessed in modern computers, make feasible the entire ambitious concept — indeed, much of the expense of the project will lie in setting up the initial database of resources. Fortunately these can be almost continuously upgraded with more words, music and even existing advertisements, as the project proceeds.

Thereafter, the machine will improve the sound and image to a powerful democratic optimum. The project will be far more expensive than the conventional activities of advertising agencies. But it will also be far faster and far more likely to produce great and powerful ads.

For his hill-climbing project, Daedalus has settled on advertising for its dominance and financial rewards. But his machine should also be able to optimize pop music, high art, architecture, and other art forms that also fall short of what one might reasonably expect.