When the Human Genome Project published the first sequence of the genetic code in 2001, the cost of applying its insights was vast. But now genotyping services, which examine a third of the six billion DNA letters, can be bought for £700 and a whole individual genome can be mapped for £70,000. It is likely that, a decade from now, every baby born will have its genetic code mapped at birth. One reading, taken like the test for cystic fibrosis from pricking the heel of a new-born baby, will produce the single unit of heredity responsible for how we develop from conception, how we grow and mature, how we live, and how we die.

Any policy document on health can be given a patina of profundity by saying that the health service should seek to prevent more in order to have to cure less. Genome technology puts that ambition in sight. It will allow far more accurate prediction of, and screening for, conditions such as diabetes, heart disease and many cancers, even before symptoms are visible. Drug prescription can be made more personally appropriate.

Viewed from a distance, the greatest policy success of the 20th century was public health. In the first decade of the 20th century life expectancy for men was only 49 years. By the final decade of the century it had risen to 73. What the advances in medical knowledge did during the 20th century for mortality, our understanding of genetics promises to do for morbidity. It is not just that we will live longer but the lives we live will be less burdened by avoidable illness.

The barrier to progress is no longer technological. It is that, as a nation, we have to yet to arrive at settled ethical positions. The principal concern is about access to information. Though the individual will have an interest in his doctor reading the information, he may not - indeed if the news is bad, he will not - wish the same access for his insurer or his employer. We might not wish a school to include such information on our children in their admissions deliberations. Will we permit a genetic predisposition to be entered as a plea in mitigation in court?

Even when the question of access to information is settled, there are very many thorny questions about the uses to which genetic information can legitimately be put. The most pressing ethical questions always involve children, where we act necessarily on behalf of another. For example, should a parent have the right to have their children tested for diseases that will emerge in adult life? What if a parent refuses to consent to a test that is clearly in their child's best interest? What if a parent decides to pursue a genetic “enhancement” that involves significant risks for a child, or that may limit a child's life prospects?That ethical difficulty is redoubled when we are talking about the possibility of a child yet to be born, and the use of genetic information in decisions about reproduction.

We should not be naive about the accuracy of genetic testing. Between the sequence and the conclusions for human life there is all manner of uncertainty and interpretation. The terminology of the insurance industry reminds us that only death - or life as it prefers to call it - is assured. That will not change as a result of genetic mapping. But our scientific learning is running ahead of our ethical wisdom, and this is now an argument that needs the fresh air of public debate.