A confession: I have been too complacent about technological fixes for the twin problems of climate change and finite oil and gas reserves. Without looking very closely at the numbers, I figured that if politicians would finally get their act together, and if we avoided some of the more unlucky possibilities (such as the release of methane ice from the oceans), cheap, clean energy would be within our grasp, given suitable research incentives and some technological brilliance.

Looking at progress in computer chips, I dreamt about how cheap photovoltaic solar panels might become over the next 50 years. Solar wallpaper, solar paint – who needs fossil fuels? Most climate-change scenarios look at a 100-year time scale. Surely, in that time, we should have figured out a way to take greenhouse gases out of the atmosphere again.

I still wouldn't rule out such techno-fantasies, but having read a remarkable book by David J.C. MacKay, a Cambridge physicist (you can download it at www.withouthotair.com) I am far more pessimistic about the potential of technology to help us out. In Sustainable Energy – Without the Hot Air, Professor MacKay makes this point very simply by sidestepping the economics altogether. Technological progress and economic growth loosen the corset of cost-benefit analysis, but not the laws of physics. No matter how cheap and efficient solar collectors become, there is only so much solar power available per square metre of land. Hydroelectric energy is constrained by the quantity of rainfall and the height of reservoirs above sea level. The most perfectly designed windmill is limited by the energy of the wind. It would barely be possible to make the numbers add up even if renewable energy generators were free.