INFORMATION. Technology. In the Internet age, the two are inextricably
linked, yet we usually deal with them in entirely separate domains. Technology
gives us the means to solve problems, while information is an abstraction beyond
the embodiment of any particular idea.
In academia and industry, the study of information is separated from the
study of devices that manipulate it. People must generally choose between
working on hardware or software, between being a physical scientist or a
computer scientist, between developing content or the channels that carry
it.
In fact, information and technology are necessarily connected through the
behaviour of physical systems. Most everyone is aware of how the transistor
launched the modern era of information technology. Less well known, but perhaps
more significant, has been the study of the physics of information itself. The
rallying cry of the great physicist Rolf Landauer was 鈥渋nformation is physical鈥�,
meaning that because information has to take on a physical form it is subject to
the laws of physics. A small but growing number of physicists has investigated
the consequences of that observation, leading to wonders like quantum computers
and even teleporters.
Advertisement
Most people don鈥檛 realise that the very success of information technology is
forcing it up against profound physical limits. Engineers familiar with today鈥檚
technology may not appreciate how basic the challenges are that they face.
Physicists who understand the science may not grasp its relevance to practical
applications.
Take the Global Positioning System receivers that are showing up in cars and
phones. To explain how they work, you need to understand the quantum mechanical
description of how atoms interact with electromagnetic fields. This is because
GPS satellites keep time using atomic clocks, which are accurate to one part in
a thousand billion (or so). Such exquisite accuracy is needed to keep the GPS
network correct to the nanosecond鈥攖he time it takes a radio wave to travel
30 centimetres.
But there鈥檚 a bigger problem. The theory of special relativity tells us that
time slows down as things move faster, and the theory of general relativity says
that time slows down as gravity increases. In one orbit of a satellite, the
former effect changes time by about 4 microseconds and the latter by about 20
microseconds. GPS satellites handle this by broadcasting corrections based on
their altitude and velocity. We see that something as apparently mundane as GPS
depends on an intimate understanding of systems from atomic to astronomical
scales.
I realised there was a problem in not being able to make these kinds of
connections when a few years ago an undergrad and a professor from MIT asked me
in the same week how the rate at which bits can be sent down a telephone line
relates to the range of available frequencies鈥攖he bandwidth鈥攐f a
modem that is connected to the line.
Any good engineer will tell you that the answer is Claude Shannon鈥檚 channel
capacity: the bandwidth times the logarithm of 1 plus the ratio of the energy in
the signal to the energy in the noise. Not so many understand how Shannon鈥檚
limit can be surpassed by taking advantage of idiosyncrasies in the noise that
are not captured in the idealised theory. Still fewer physicists know anything
about any of this鈥攅ven though they understand the origins of noise. And
neither side has traditionally known about human perception, whose vagaries can
be exploited by not sending data that would not be missed. The apparently
innocuous inquiry about phones cuts across many intellectual and organisational
boundaries.
My response to the double question was to start teaching a course on the
physics of information technology. The success of this ambitious agenda suggests
to me that the greatest consequence of improving information technology may be
to organise intellectual inquiry around grand challenges rather than traditional
disciplines.
If this turns out to be so, then a title like 鈥淭he Physics of Information
Technology鈥� may eventually become triply redundant. The truth is that none of
those words can be properly understood without all of them.