Context: The False Philosophy Plaguing AI
The (or at least a) problem here is that the ascendant technologies, frameworks, cultures and commercial incentives have inadvertently stumbled upon the constructive and recursively self-inflected sweet-spot through which those core assumptions themselves optimally self-propagate. If the incentives and rewards for developers, researchers and corporate or institutional entities find themselves swept up and away both in and as an effective transmission medium of competitive difference, accelerated refinement of specialisation and commercial (as strategic, economic) leverage for advantage – we can clearly observe that none of this is directed at any axiomatic or paradigm level towards generalisation or unification.
Why does it matter? The organisational context within (and ultimately – as) which any intelligence evolves or emerges is not separate from the technical and conceptual or intellectual assumptions and products of that (complex, information-processing) system. We are seeking General Intelligence in a savant-like orientation to specialisation which might be great for investment interest and the PhD factory that machine intelligence has become, but which fundamentally and foundationally leads us (all) away – not only from the objective we aspire to – but also away from the powerful solutions it may bring to what are the truly significant (and only accidentally commercial or competitive, partisan or differential) big-ticket problems our Global civilisation currently faces.
General Intelligence requires general approaches and these quite concretely lie in the necessary and often unacknowledged functional and logical blindspots both implicit to and endemic of contemporary technical, commercial and institutional incentives, concepts and vocabularies.