Our Final Century by Martin Rees

In the 20th century, science gave two superpowers the ability to destroy civilisation. In the 21st, get ready for a world in which any disaffected teenager can wreak technological havoc. Scared? Martin Rees wants you to be.

The Astronomer Royal wants to concentrate minds on some medium-term threats to our survival which stem from technologies emerging now. Genetically modified organisms, the minute self-reproducing machines of nanotechnology, or super-intelligent robots: each, in its way, is a threat to humanity. Add environmental disasters and astronomical hazards, and he puts the chances of our civilisation celebrating the turn of the 22nd century at no better than 50:50.

This is eye-catching, but pretty much guesswork, because of the nature of these dangers. But that is not reassuring, because they are mostly "zero-infinity" risks where there is a very small chance of something really bad happening: a wacky high-energy physics experiment destroying the universe, an asteroid hitting Beijing, that kind of thing.

The new ingredient is that a few deranged souls may now skew the odds. Think Columbine, Dunblane, Jonestown. Think (as you will) 11 September 2001. The 747 heading for your local nuclear power plant is no freak accident. It is a suicide mission.

Aside from diverting an airliner or smuggling uranium out of a former Soviet republic, the technologies with most appeal to fanatics and fruitcakes are biological. And it is "bio-error or bio-terror" which most alarm Rees. The really bad things biotechnology might lead to – engineered viruses and ultra-virulent pathogens – are closer to reality than the other nasties he speculates about. Last year, Rees publicly staked $1,000 on a bet that an accidental or deliberate release of bioengineered pathogens will kill a million people by the year 2020. He hopes to lose the bet, but thinks he will collect.

After this compelling problem, you might expect the rest of the book to offer serious thoughts about solutions. But aside from urging us to reduce the gap between rich and poor, and prepare to tolerate the kind of surveillance formerly only imposed in police states, Rees does not come up with much cause for hope. In fact, the latter part rather loses focus. As well as a socially concerned scientist, Rees is a much-feted cosmologist, and his cosmological interests gradually take over.

True, the ensuing discussion about the prospects of identifying extra-terrestrial life puts our earthbound concerns in context. It is still an open question whether life abounds in the universe, in which case our own fate will merit little more than a cosmic footnote, or whether human consciousness is unique. If so, then the present century is crucial for the whole of creation. If we survive, we or our successors may colonise the galaxy. If not, life may have used up its best chance of getting properly established in an unfriendly universe.

Purely terrestrial loyalties will satisfy most people that the hazards Rees highlights are worth addressing. His book is worth reading for its sober assessment of the human prospect. But it raises more questions than answers, and I was left hoping that more determined efforts to address them appear – and soon.

The reviewer's new book is 'Lovelock and Gaia: signs of life' (Icon)