Infrastructure, superstructure and not enough sockets

It can be hard to reverse bad decisions. It can be hard to recover from having failed to anticipate the future. It can be hard not having enough power sockets in hotel rooms.

There is an old (but sadly discredited) story that the design of Roman chariots constrained the design of the space shuttle. It would be a good story if it were more true, but it’s still quite a good story even if it’s not true at all. It is undoubtedly the case that the decisions we make today will constrain the choices available to be made in the future, just as it is true that decisions made in the past constrain the choices we have today.

Sarah Baskerville didn’t like her hotel room in Dundee. I understand and have shared her frustration that power sockets are few in number and badly placed in far too many hotel rooms. But the problem isn’t – or isn’t just – cheap hotel rooms, it is that fixed wiring tends to be around for much longer than the pattern of usage it is designed to support. The most extreme case in my house is a room with eight wall sockets which between them currently power over thirty devices, but this too is an instance of a more general rule. To adapt (I think) Brad DeLong, you need more power sockets even after you allow for the fact that you need more power sockets. The clear corollary of that rule is that you will always run out.

The wiring of a hotel room doesn’t need to have been specified all that long ago to feel grossly inadequate now (and often, in my experience, as much about location as absolute number, making short term fixes even harder).

We are used to the flexibility of software and increasingly of devices – of the superstructure of modern living. We want that flexibility to extend to infrastructure, but on the whole it can’t, or can’t without considerable expense, which amounts to much the same thing.

That isn’t a new or gadget specific problem. We can’t have air conditioning on the underground because the way the tunnels were specified over a hundred years ago means there is no room for it. Cyclists dice with death on the streets of London because of a consistent design philosophy which gives precedence to cars and in which billions have been invested. We use copper cable to carry data not because it’s a good way of doing it, but because it’s the installed way of doing it – and it takes superhuman efforts to overcome the inertia of the installed base.

We just have to deal with the problems inflicted on us by those who went before us. Is there anything we can do to avoid inflicting similar problems on those who follow us? Probably not, as that’s the same as asking whether we can predict the future but not be caught out by unforeseen discontinuities – it could happen, but it almost certainly won’t. As Charlie Stross puts it

The near-future is comprised of three parts: 90% of it is just like the present, 9% is new but foreseeable developments and innovations, and 1% is utterly bizarre and unexpected.

If hotel room designers had paid more attention to the 9%, it might be easier to plug in all the gadgets. But in a hotel room generation’s time (however long that may be), when there are wall to wall power points, perhaps we will be in a 9% zone where we won’t need any of them any more, or even in a 1% zone where there some means of storing and transmitting power which makes the whole infrastructure utterly redundant.

So what can we do, beyond reminding ourselves yet again of the wisdom of Yogi Berra (or perhaps somebody else) that it’s hard to make predictions, especially about the future? We can attempt to insulate different layers of architecture from each other, to minimise the extent that the faster moving are constrained by the slower moving. We can increase the speed of obsolescence and renewal, trading modernity for hard cash. We can do more to ensure that design thinking reflects the 9% as well as the 90%, though again at a cost, since some of the 9% will never go mainstream. We can choose to give our money to hoteliers with more sockets and better wifi than their competitors.

But in the end, perhaps all we can do is try more explicitly to counter our tendency to get the balance wrong in how we look at the very near and slightly further futures. As Roy Amara, who played a big part in putting thinking about the future onto a more rigorous basis, put it:

We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.