About 14 000 people will contract HIV today. And tomorrow and the day after that, and every day for the foreseeable future. That's 5 million by the end of the year, most of whom will be dead within a decade.

Figures like these bring home the devastating impact of AIDS and the urgent need for a cheap, effective vaccine. At a stroke, a vaccine could stop the tide of infection and stem the need for more, costly treatments. It could even help people who already have the virus stay healthy.

Back in 1990, drugs companies and researchers confidently predicted we'd have a vaccine against HIV-I within 10 years. These were rash statements. The virus has turned out to be more cunning and stealthy than anyone expected, and our knowledge of how vaccines bolster the immune system hasn't been good enough. A dozen years on, we still have no clear-cut candidate for a vaccine.

So you might expect the announcement of two large-scale trials of AIDS vaccines to be applauded. Yet they have been criticized as a monumental waste of money. The trials will test almost identical vaccines, neither of which is expected to offer great protection against the virus; What's more, both are funded by the US government, one through the National Institutes of Health and the other through the Department of Defense.

The NIH and the DoD have a long history of rivalry in AIDS research, but in this case it seems sensible for the NIH to back down. Although the NIH is under pressure "to be seen to be doing something", duplicating work of questionable value is itself questionable. Better to join forces with the military for this trial and spend the money saved—which amounts to about $60 million—elsewhere.

There are, after all, reasons for optimism. A new wave of vaccines from industry and academia has nearly completed safety tests. It makes sense to carry out limited trials of all these newcomers, to identify which ones offer the best protection, before committing tens of millions of dollars to larger trials.

Such a strategy would need the agreement of drugs companies, government agencies and medical charities—something that's not as Utopian as it sounds, The NIH has already signed a deal to test a new AIDS vaccine made by the pharmaceuticals giant Merck. And the International AIDS Vaccine Initiative, a not-for-profit funding organization based in New York, has pioneered new ways to divide up intellectual property rights for successful vaccines.

What's needed is cooperation and coordination, not competition. The important thing is to find the fastest route to an effective vaccine. Every day we forget that, another 14 000 people pay the price.