The aim is to contribute to the realisation of hybrid deductive intelligence and its applications, and ultimately to universal fulfilment.
The whole of mathematics, particularly the theoretical development of mathematics, reaching its most abstract and abstruse parts, though most likely addressing applicable mathematics first, is regarded as a core competence for the level of intelligence sought. Intelligence of a this order is beyond the present state-of-the-art for completely automated reasoning. And so, in this project we envisage, as in the field of Interactive Theorem Proving, that less trivial results will be realised (at least in the early stages) by man and machine working together, to which arrangement I refer by the term hybrid intelligence. Hybrid intelligence of course, we already have, that’s how mathematics is done, the challenge is enable the machines to contribute at higher levels, and to integrate the competence into more general problem solving capabilities.
Many believe that the pace of change is growing and that in relatively brief timescales radical advances in artificial intelligence will be secured. Because these projects are conceived of as contributing to a technological domain which is expected to be radically transformed, its future relevance depends upon how well it is aligned, not so much with todays information technology, but with the very different environment into which we will soon be thrust. For this reason I seek, in describing the project, to place it in the context of that future as I see it, and to make a connection between the details of the proposed architecture and the character of that future.