If you care enough, you can keep a vintage car going indefinitely. When it stops, you strip it down, fix any broken parts, or fabricate new ones. But the human body, alas, is not like that. It cannot be shut down for repairs. And however clever its internal maintenance, it always fails in the end.

Biology explains why we meat machines are mortal, in spite of the wilder predictions of the nano- technologists. Any animal has to allocate finite resources between normal functions, reproductive functions and maintenance. The trade-offs between them mean that if you want to reproduce at all, there is always a limit to maintenance. Healing wounds, detecting cancers, repairing DNA and cleansing the body of toxic chemicals are all worthwhile - but not as worthwhile as passing on your genes.

So the bad news from the gerontologists is that, the more we know about the biological basis of ageing, the better we understand that it is inevitable. In particular, we see why there will never be a simple path to life extension. All the hopeful recent press articles about arresting loss of DNA from the ends of chromosomes probably do not hold the key to endless renewal of our cells. But what would count as life extension? A reasonable expectation of life today is well beyond the biblical three-score and ten. The good news is that we may have already produced life extension, of a kind, without realising it.

There are two parts of the demographic story, both remarkable, but one more familiar than the other. The familiar part is the historical demography of the past century or so - that many more people now live to enjoy what would once have been regarded as old age, say to 60 and beyond. "Our populations are uniquely old. There have never been populations as old as this anywhere in the world - and they have grown old suddenly," says Peter Laslett of the Cambridge Group for the History of Population and Social Structure. For most of human history, fewer than one person in ten was over 60. In the next century, in the affluent world, more than one person in three will be.

É«ºÐÖ±²¥

Then comes the less familiar news. Not only are there more old people, but those old people are getting older. The decline in mortality has been greatest among 60, 70 and 80 year-olds, and is still steepening. James Vaupel of the Max Planck Institute for Demography has outlined the consequences: the older the age group, the faster the percentage annual growth. The number of centenarians is growing by 8 per cent a year. In the 1920s, fewer than 100 people in England and Wales passed 100 each year; now there are almost ten 100th birthday parties every day.

The most important factor, according to Vaupel, is the decline in mortality after 80. The deferral of death has been most striking among women, already the longer-lived half of the species. There are, he estimates, half a million more women over 80 alive in England and Wales today than would still be with us if the death rates of the 1950s had remained.

É«ºÐÖ±²¥

Individually, it is not clear what you have to do to get on the right side of this curve. The standard health advice applies, although the oldest living human - Frenchwoman Jeanne Calment, 122 - still takes a daily cigarette. But beyond that, says Vaupel, "If I wanted to give you advice on how to live to 90 instead of 70, I'd have a hard job knowing what to tell you". It is plain, though, that for the first time there is a sporting chance that any of us could survive for almost all the apparent maximum human lifespan. This could obviously strain already creaking health services, and Vaupel is anxious to raise an alarm here. "The idea that old-age mortality is intractable hangs on, especially among policy-makers. That belief is pernicious", he says. It means that the official forecasts of numbers of older people are too low, projected health care expenditures are too low, and research spending in the area is too low.

All proper cautions. But he also emphasises that we may not just survive, but keep our faculties largely intact. "My feeling is that it is difficult to live a long time if you are sick." The biologists and clinicians can document a falling off in performance in virtually any body system you care to mention, of course. Muscle power. Respiration. Speed of nervous transmission. Eyesight. Hearing. But the steady cell loss, the inexorable decline in physical and intellectual performance, need not equate to outright disability.

Gerontologists are now taking more of an interest in "successful" ageing - George Martin of the University of Washington in Seattle calls it "sageing". They want to overcome the sampling bias built into the clinic, where people only come to call if they have a problem. We know, for example, that neurons undergo "dendritic sprouting" - making more elaborate connections as the cells decrease in number. Compensation? Maybe. Certainly, people who keep up the habit of intellectual work retain their power to recall complex material. Good news for university faculty here, according to a study whose subjects worked at the University of California at Berkeley.

For mental and physical abilities, staying in practice helps a lot: use it or lose it is good advice for young and old. But use it for what? Almost everyone used to work until they dropped. Now, just as life expectancy rises, we are moving people out of work earlier. How will we understand the shape of a life if paid employment takes up 35-odd years out of, say, 95? What are all these people going to do?

É«ºÐÖ±²¥

We have hardly begun to address this question. We have created a cult of youth which shows no signs of abating just because youth is on the wane: we are all old people pretending to be kids. Until we overcome this cultural lag between the facts of age structure and our preferred images and values, we will not begin to sort out how the brave new world of the old might look.

There are starting points. It is widely assumed that the old extended family is vanishing - low fertility means fewer brothers, sisters or cousins. But the family is growing in two other dimensions. One is the complex elaboration of networks of step-kin, and we need to know what obligations step-children and step grand-children will feel toward their elders if they become dependent, for example. The other, more obvious, dimension, is the rapid growth in numbers of three and four-generation families. The "beanpole family" is the family of the future, according to Laslett. Adults in their fifties or even sixties will have parents still living. Their children will have grandparents alive when they too have children.

The policy questions which arise are rich and complex. Who is going to pay for whom over these long lives, for one thing? Future election campaigns will see arguments over pensions which will make the ones we had this time round look pretty tame. Research is already under way here, with Nobel laureate James Mirrlees of Trinity College Cambridge trying to improve economic modelling of pensions and social security. There are theoretical and empirical questions to pursue around intergenerational equity, as well as issues around how to predict whether people will save for retirement, go on saving after retirement, or never manage to save at all.

Even less clear is the way these new-style families will manage mutual care. One scenario has fit grandparents looking after working mothers and fathers' children, to everyone's benefit. There is some evidence this is already happening, as Lord Young of Dartington spelled out in a talk in which he called on the new Government to set up a grandparents' commission. But there are many uncertainties. Which grandparents is it likely to be? The maternal grandmother seems the most-used carer at the moment. How many "third-agers" will want to tie themselves down for a fresh bout of child-rearing? Will they be living close enough to the children? Will they be looking after their own, still older, parents instead?

É«ºÐÖ±²¥

All these remain open questions. The answers will determine whether our other traditional assumption, that long life is desirable, is valid. How many of us will emulate the legendary jazz musician Eubie Blake, who on his 98th birthday merely joked, "if I had known I was going to be around so long, I would have taken better care of myself"? And how many will simply be at a loss what to do with all those uncommitted years? As Peter Laslett put it, "we shall have to find a way of filling up the extra time".

Jon Turney is lecturer in science communication, University College London. This article is based on last week's British Academy/Royal Society meeting on ageing.

É«ºÐÖ±²¥