There’s a paradox in all disruptive new technologies. The powerful lure of new ideas can make them seem inevitable. Yet it often takes very long for these ideas to have an effect. Large sums have been sunk in the existing infrastructure, behaviours change slowly and smarter new ideas may come along. That makes it hard for the layperson to distinguish between the true visionaries of the Information Age and the hype merchants who are simply riding the latest tech bandwagon.

So it is with “cloud computing”, the latest faddish phrase to emit from the computer world. This is one of those simple, all-embracing ideas that seems to promise much. Think of it as the mass production phase of computing: rather than information being stored and manipulated on billions of isolated and inefficient PCs, it would be drawn into a small number of vast computing “clouds”. Each one would be a distributed, centrally managed mass of machines operating in harmony. Given their economies of scale, these clouds would process information at a cost-per-instruction many times lower (and eat up far less energy) than other systems. If individual machines crashed, the cloud as a whole would reform and continue unaffected. The sum of human knowledge would no longer be locked away on isolated hard drives, but accessible from anywhere, and widely shared.