Information has content, and it has a physical representation. These two levels of description are usually considered separately as hardware and software, but many of the most serious challenges and significant opportunities in information technologies lie right at this interface. This is what the MIT Media Lab's Physics & Media Group studies.

Billions of dollars have been spent developing and fabricating ever-faster CPUs, which are then put in a dumb box. Increasingly, the box is the problem, not the CPU: it may be possible to use ballistic carrier transport to build a GHz processor, but most computers right now can't even tell when you get up and leave. Now consider the great mature technologies that are so satisfying, such as a Stradivarius violin or a Gutenburg Bible. It's not possible to separate their interface from their application: they are beautifully integrated systems. Their performance significantly exceeds that of comparable modern technology. The effective rate at which a Stradivarius solves its equations of motion to map a player's gestures into sounds is gigaflops, far beyond what a typical electronic keyboard can do. The pages of a nicely printed book have much better contrast and acceptable viewing angle than the best active-matrix LCD panels, and they don't need batteries. Right now it is very sensible to prefer to read a book on paper than on a computer screen; only when new technology can match the performance of old technology will meaningful generalizations be possible.

The Physics and Media Group is at the MIT Media Lab, rather than in a more conventional Physics department, because it is so important to ask these kinds of questions in an environment that can help pose problems that need solving, and recognize unexpected solutions. The research in the group is done at three levels. The bottom layer studies the basic materials and mechanisms for sensing, computing, and communications. This includes questions about the physics of computation (such as a project to build a computer based on atomic spins), ways to remotely sense the characteristics of materials to determine their identity, state, and location, and the mathematical methods for smart systems to describe their physical environment. These ideas are then reduced to practice in useful devices such as shoe computers and musical instruments. Finally, we work closely with creative, industrial, and academic collaborators in a range of application domains.

Along with the research, associated courses have been developed to cover the the required physics and mathematics background for this kind of work. The members of the group include MIT undergraduates in the UROP program, and Media Lab graduate students as well as some graduate students from other parts of campus.