The Conditions of Intelligence

The Conditions of Intelligence traces artificial intelligence from its technical foundations to its philosophical, cultural, and political consequences. Beginning with the optimisation paradigm that drives modern AI, and its theoretical limits, the book asks what intelligence actually is, and what current systems demonstrably are not.

The argument moves through four registers: the genuine capabilities and documented failure modes of large language models; the philosophical obstacles to alignment, among them value pluralism, Arrow's impossibility, and the hard problem of consciousness; the practical techniques that address part of those obstacles but not all; and the broader conditions, historical, cultural, geopolitical, in which AI development is taking place.

Drawing on thinkers from Isaiah Berlin and Hannah Arendt to Cajal, Chollet, and Quine, and examining traditions from Shinto ethics to Ubuntu philosophy, the book insists that AI is not only a technical problem. It is a question about what kind of intelligence we are building, for whom, under whose governance, and at what cost to labour, to the epistemic commons, and to the physical world.

The answer is not yet known. The question is urgent.