"It is a common enough desire. As humans, we want to know what is lurking outside our perimeter, beyond our flickering circle of firelight. We have built lenses and Geiger counters and mass spectrometers and solar probes and listening stations on remote Antarctic islands. We have drenched the world in information in the hope that the unknown will finally and definitively go away. But information is not the same as knowledge. To extract one from the other, you must, as the word suggests, inform. You must transmit. Perfect information is sometimes defined as a signal transmitted from a sender to a receiver without loss, without the introduction of the smallest uncertainty or confusion.
In the real world, however, there is always noise.
Since 1965 the Russian Academy of Sciences has published a journal called Problems of Information Transmission. It is, insofar as it is possible for a scientific publication (even a Russian one) to convey an emotional tone, a melancholy read. Threaded through recondite papers on Markov Chains and Hamming Spaces and binary Goppa codes and multivariate Poisson flow is a vocabulary of imperfection, of error correction and density estimation, of signals with unknown appearance and disappearance times, of indefinite knowledge and losses due to entropy. Sparse vectors are glimpsed through a haze of Gaussian white noise. Certainty backslides into probability.
Information transmission, it emerges, is about doing the best you can. “
( Hari Kunzru, Transmission, Dutton Adult, 2004; pp. 375,6.)