Is it better to get to the aiport an hour or two sooner than your flight leaves, or one minute too late?
Is it better to be prepared for exponential general-purpose molecular manufacturing five years or ten years before it is developed, or six months too late?
Getting to the airport earlier than necessary means you have to sit around and wait. Getting to the gate even 30 seconds too late means you have to wait for the next flight...if there is one.
Preparing responsible policy to effectively deal with the implications of advanced nanotechnology a few years before it arrives means we have more time to improve the policy. Waiting even one month too long and allowing runaway effects to get started could mean disaster.
How much do we have to do to prepare?
First, all of the risks have to be well understood. Then we need to work out a series of plans for dealing with each risk -- a task made much harder by the fact that measures to reduce one risk may increase another. Then all this information has to be presented, convincingly, to the people who make the policy. There are a lot of them, in many different organizations. Then systems have to be designed to administer the policy. Then the systems have to be implemented.
Each of these steps will take time. And this isn't a complete list. Technological safety measures will have to be carefully invented and developed. Public opinion, and then public support, will be necessary at several stages. Nations must learn to cooperate in ways that have not yet been tried.
How long will all this take to complete? A conservative estimate would be twenty years. A rush program might be able to get it done in ten years.
Because our analyses show that molecular manufacturing almost certainly will be developed within the next twenty years, and perhaps in less than ten, we'd better get started right away. We can't afford to miss this flight.