In contrast to today's silicon computers our brains achieve their marvelous abilities using components that act a thousand times slower than the clock speeds of modern computers. Instead, our biological „super computers“ use close to a hundred billion neurons in parallel. These neurons communicate via the exchange of discrete voltage signals called action potentials. Trying to understand the neural code, i.e. how these signals encode information about our perceptions, thoughts, and actions has remained a major challenge. Progress has been limited by two roadblocks. First, the responses of individual neurons seem to be very unreliable, making it hard to map a neuron to a precise function. Second, experimental techniques have limited us to sampling the activity of only small numbers of neurons at the same time. Recent experimental and theoretical advances now promise to finally overcome these barriers.