Cities famously tend to suffer from congestion, and the congestion tends to be more extreme the bigger a city gets. Every morning, millions of people squeeze into buses and trains, or climb into cars and in turn squeeze into crowded arterial roads to get from their homes to their place of work. Huge amounts of energy are wasted this way, as well as human hours that could have been spent in leisure, social activity or productive work. The cost of computing bumps up GDP, but it is a net loss to the economy, because most people who do it don't want to do it for its own sake — it is a cost to them. Very few cities seem to be able to control that cost.
The reason is this: it's a product of the natural way that cities tend to grow. Cities usually start as confluences on trade routes — in harbours or at crossroads or near some key resource, such as a goldmine. Traders naturally want to locate as close to the confluence as possible, in order to catch as much passing trade as they can. Thus, they crowd together, and where they congregate, property values rise until those living in the area have to move to its ouskirts. People building transport systems are also attracted to the points of congregation, so they build train stations that stop in the town centre, and roads or railways that radiate outwards from it, accelerating the process of concentration. The level of concentration in a thriving city, if uncontrolled, will always be the maximum that is technically possible. Naturally, therefore, someone had to invent the skyscraper, and concentration exploded further. At the same time, people want peace and quiet, and pleasant open spaces to enjoy in their leisure. This motivates them to locate their homes further and further away from the urban centre as transport systems get better and better, and as the city itself grows. There's a kind of vicious circle: the transport system increases in capacity and speed, so the people spread out more, and the average journey time to work stays the same, or even grows. The only limit to journey time is what people can afford, and that's not determined by what people would like to afford, but by competition for jobs. Therefore, if planning policies don't control the tendency, and the city continues to grow, commuting journey times will grow until either there's no time left outside work and commuting for anything besides sleep, or until the transport system is fully congested and cannot expand further.
When a city is planned from scratch, these problems can be avoided entirely. The solution is to enact zoning policies that compel trade to locate in lots of small centres, or (better still) in long, stretched out ribbons instead of one great big blob. Also, a prohibition on the building of skyscrapers is advisable. If some buildings are much taller than others, the likely outcome is that there will be some areas with much higher density than others, and this will create the kinds of traffic flow that lead to congestion. If a strict policy of even density and ribbon-shaped districts is followed, journey times will tend to be short, and congestion can be zero, no matter how big the city gets.
To illustrate with an example: imagine a city whose sprawl covers 5,000 sq. km. (the size of Atlanta), with the commercial centre being a 1,000 sq. km. blob in the middle. Most people wishing to reach the centre will have a journey of more than 15 km., so they'll have to take some kind of transport, placing high demands on infrastructure. Additionally, those near the centre will have to use the transport system any time they want to visit the countryside. Supposing the city were stretched out into a ribbon just two kilometres wide, with the commercial centre being stretched along that middle. In that case, nearly everyone could live within a ten or fifteen minute walk of the city centre, at the same time as living within walking distance of the countryside. The main cause of congestion is thereby completely eliminated, and for many, the reason to own a car also disappears.
If we can replicate that in the planning of Desertopolis, we can cut down the transport energy requirements to below the level we previously estimated we would need, perhaps a very great deal below.
