During the 20th century there were huge advances in our understanding of information, and the methods and electronic technologies necessary for its processing. The central tasks of information processing are the storage, communication and transformation of data, and the cornerstone of their implementation since the 1940's is the digital computer, which was first built of electro-mechanical relays, then thermionic valves, and now silicon, and instantly overtook the human computors in speed and accuracy of addition and subtraction. The detailed methods for performing these tasks are of course provided by us — intelligent humans — and are not yet intrinsic to the information processors themselves. We have yet to incorporate this creative biological intelligence into technology. Indeed, the feature that distinguishes biological intelligence from more simple information processing is the biological agent's ability to extract and exploit appropriate knowledge from the world, and to apply that knowledge for its economic advantage — as in feeding itself, defending itself, breeding (sometimes with itself), and communicating its needs and intentions, mostly with others. The important consideration here is that "knowledge" is not simply stored data. It is a flexible, meaningful, purposeful, context-sensitive organisation of immediately accessible data — a cross between an encyclopedia and a DIY manual, located in the brain.
Progress toward autonomous intelligent systems capable of real-world interaction has been slow simply because this intrinsically goal-oriented form of data organisation and processing that characterises brains remains poorly understood in engineering terms. It is humbling that the intelligent co-ordination of flying, navigation, foraging, and communication by a honey bee (which possesses a rather modest brain of only one million neurons) far exceeds the intelligent performance of the most advanced autonomous artificial systems we can build. In the present age of the global proliferation in the use of silicon-based technologies that are increasingly sophisticated and personalised, it is worth re-emphasising that every significant expression of knowledge on this planet has its source in biological nervous systems, not in technology.
Few branches of neuroscience, however, are actually striving to understand, let alone resolve this fundamental disparity. The major advances in autonomous intelligence now come from developments in machine learning algorithms whose relation to brain architectures and processes are remote. However, Neuromorphic Engineering does have as its goal the development of artificial neural systems whose processing architecture and principles are based on those of biological nervous systems. They are usually composed of hybrid analog and digital electronic circuits that emulate biological sensors and their processing neuronal networks. Initially, the big challenges for neuromorphic engineering were simply to understand how to exploit the properties of silicon devices to build circuits in a neuromorphic style; and to combine sensors and neuron-like computational elements into simple reactive systems, integrating them in such way that they expressed their processing and behaviour in a "neurally inspired" way. These goals have now been reached, and the focus is shifting towards the conceptual and technical problems of how to induce more sophisticated behaviours in neuromorphic systems. The ultimate aim is to achieve a neuromorphic "cognition" that resembles that of animals in being capable of creating, storing and manipulating knowledge of its external and internal world, and of using this knowledge for economically advantageous behaviour. The induction of effective brain-like cognitive behaviours in autonomous artificial agents offers enormous economic, technological and social benefits, and so the abstraction of engineering principles from the biological brain is a major challenge for the 21st Century. Harnessing the principles of biological intelligence can be expected to have a major impact on the technology market as autonomous intelligence becomes incorporated into equipment, vehicles, buildings, utilities and clothing, which some see as the relentless march to the mergence of a truly "prosthetic man".