AFTERWORD: THE ROAD AHEAD 259
the same way as we see and hear. They then process this
information in real time in order to move through space.
In fact, in the course of generating this new technology the
engineers at DeepMap were reminded of the extraordinary
power of the human brain. From speed limits to navigating
rules of intersections, merging, and taking a car through
curves, humans rely on their senses and their prior experience
to process an enormous amount of information while driving.
Automated vehicles will require similarly sophisticated
systems to replicate this operation. And, for this reason,
the measurements involved in this mapping database must
be far more precise than any road map in history, down to
a centimeter.
All of this presents a rather colossal challenge for DeepMap.
Yet it is also an oddly old-fashioned task of creating wayfinding
aids to get from one point to another. This same quest drove
so many mapmaking efforts over the past five centuries, in the
service of discovery, conquest, settlement, or general mobility.
To teach cars to drive themselves, we are in a way replicating a
very basic cartographic operation. But the creation of platforms
for autonomous vehicles also reminds us that maps are far more
than instruments of navigation. Maps make this technological
innovation possible. In turn, the technology demands an
entirely different kind of map. This reciprocal dynamic has been
at work throughout American history: across five centuries,
maps have both reflected and mediated change.