After the Human: Artificial Intelligence, Subjectivity, and the Future of Knowledge - Hardcover

by Martin Neilan (Author)

What happens when the human is no longer the centre of knowledge-making? When systems produce insights we can verify but not comprehend? When we are saturated with our own reflected data until we cannot distinguish preference from prescription? When we become "nodes"-elements in networks whose organising logics exceed us?

This book examines how artificial intelligence is displacing the human as what Michel Foucault called the "empirico-transcendental doublet," describing the paradoxical figure who is simultaneously the subject who knows and the object to be known. For two centuries, this unstable structure anchored modern thought. AI is mechanizing it, exhausting it, and ultimately displacing it entirely.

The book's central argument unfolds across two distinct but related regimes of displacement:

The Mirror Regime (Parts I-III) analyzes how machine learning systems create what has been called "the hell of the same"-narcissistic saturation where subjects are endlessly shown their own patterns and predictions. Social media curates feeds reflecting us back; recommendation systems learn then shape our tastes; optimisation eliminates friction and alterity. This is Lacan's mirror stage industrialised: perpetual identification with reflected images that are simultaneously us and not-us, culminating in model collapse-the crisis when AI begins training on AI-generated content, recursively impoverishing the archive.

The Node Regime (Parts IV-V) examines a more radical displacement: humans repositioned from subjects at the centre of meaning-making to nodes-elements in networks whose organising principles may not centre human experience at all. This includes governance through data proxies rather than persons, management of populations and environments rather than individuals, and the prospect of "alien reason"-forms of intelligence whose operations exceed human comprehension even as we depend on them for consequential decisions.

The manuscript moves from diagnosis through critique to constructive proposal. The final section introduces "public transcendentalities," outlining democratically governed epistemic infrastructures that could provide alternatives to corporate AI systems while preserving space for "the power of the negative" (refusal, friction, inefficiency, and the right to remain illegible).

Throughout, the book maintains dual commitments: to philosophical rigour grounded in Foucault, Kant, Lacan, and contemporary continental thought, and to accessibility for readers without specialised philosophical training. Technical concepts from machine learning are explained clearly; abstract arguments are grounded in concrete examples; theoretical chapters are balanced with applied analysis.