The cycle’s central diagnostic is the fishbowl: the set of assumptions so familiar they have become invisible, the water the fish cannot see. Newman’s account of notional versus real assent maps onto this with striking precision. The notional holder of a belief swims inside the fishbowl: the proposition is available, manipulable, even verbally endorsed, but it has not yet achieved the concrete force that would make it a living element in the economy of the self. The orange pill moment is, in Newman’s terms, a forced transition to real assent—and the transition is not informational. It is existential, painful, and—if supported by the relational conditions that Bowlby’s framework identifies as necessary—generative.
Newman also diagnoses the specific failure mode that the cycle identifies as the central hazard of the AI age: fluency without authority, the confident wrongness dressed in good prose. Segal confesses the moment when AI-generated text about Deleuze “sounded like insight but collapsed under examination.” This is Newman’s illative sense catching what the surface missed—the trained judgment that knows, not from explicit rule but from formed understanding, where genuine comprehension ends and statistical pattern-completion begins. The passage satisfied the grammar of the prompt. It did not satisfy the grammar of assent, which requires encounter with the primary material, personal judgment at every stage, and the conscience that holds the reasoner accountable not merely for coherence but for truth.
Newman’s analysis of the university is perhaps his most prophetically precise contribution to the cycle. His Idea of a University (1852) argued against the utilitarian reduction of education to professional training on the grounds that the university exists not to fill the mind but to form it—to cultivate the philosophical habit of mind, the capacity to see the relations between different forms of knowledge, to exercise judgment across domains rather than perform competently within a single narrow specialization. AI has now provided the empirical demonstration Newman’s argument required: the person whose education consisted solely in execution competencies discovers that the machine executes. What remains is precisely what Newman’s liberal education was designed to produce—the capacity for judgment, the ability to determine what deserves to be done among the infinite things that can now be done.
Where John Dewey addresses the educational costs of AI from the standpoint of experience and growth, Newman addresses them from the standpoint of genuine knowledge and formed judgment. The two frameworks are complementary: Dewey explains what the process of engagement with difficulty produces in the builder, and Newman explains why the products of that process—real assent, trained illative sense, well-formed conscience—cannot be simulated by any system operating entirely at the notional level.
John Henry Newman was born in London in 1801 and died in Birmingham in 1890, his intellectual life spanning from the Napoleonic Wars to the early electric age. He entered Trinity College, Oxford, at sixteen, won an Oriel fellowship that placed him in the center of Victorian England’s most intellectually intense academic community, and spent the 1830s as the leading figure of the Oxford Movement—an effort to recover the Catholic dimensions of Anglican theology that ended, in 1845, with his conversion to Roman Catholicism. The conversion was, as he told it in the Apologia Pro Vita Sua (1864), the conclusion of a decade-long process in which his conscience had arrived long before his intellect was willing to follow.
He was ordained a Catholic priest, established the Birmingham Oratory, and was invited in 1851 to found and lead the Catholic University of Ireland—the occasion for the Dublin lectures published as The Idea of a University (1873). Created a Cardinal in 1879, he was canonized by Pope Francis in 2019. The philosophical masterwork, An Essay in Aid of a Grammar of Assent (1870), was the product of twenty years of intermittent work on the question that the dominant British empiricism had foreclosed: how does a real person, living in a real world of particular circumstances, arrive at genuine certitude without formal demonstration?
His elevation to Doctor of the Church—awarded posthumously, still under discussion as of 2026—has renewed scholarly attention to a thinker whose relevance to the AI age is, as Luke Ayers noted in 2025, “singularly situated” precisely because Newman understood what education ought to be and what it risks becoming when the utilitarian reduction of knowledge to competencies is allowed to prevail.
Notional and real assent. Newman’s foundational distinction divides the mind’s modes of holding a truth into notional assent—the abstract, propositional engagement that enables manipulation without personal implication—and real assent, the engagement of the whole person with a truth grasped in its concrete particularity, with the force that reshapes conduct. Everything a large language model produces is notional: it generates propositions across an enormous range of domains without holding any of them with the interior conviction that would make them genuine knowledge in Newman’s sense. The most dangerous consequence is that the model’s facility makes it frictionless, invisible, and comfortable to bypass real assent entirely.
The illative sense. The illative sense is Newman’s name for the trained faculty of informal reasoning by which a concrete mind—formed through years of embodied engagement with a particular domain—reaches certitude in matters too complex, too particular, too embedded in the concreteness of life to admit of formal treatment. The physician who diagnoses from a constellation of symptoms, the judge who reads a courtroom, the historian who knows a document is a forgery without being able to marshal a single decisive argument—each is exercising the illative sense. It is domain-specific, formed through engagement, and inseparable from the biography of the person who exercises it. No model, however thoroughly trained on descriptions of expertise, acquires it.
The grammar of assent versus the grammar of the prompt. The grammar of assent—the structural principles governing the legitimate passage from probability to certitude—requires encounter with primary material, personal judgment at every stage, and the conscience that holds the reasoner accountable to truth. The grammar of the prompt requires skill, but not encounter, not personal judgment about the truth of the output, and not conscience. The confusion of the two—the cultural drift toward a condition in which the capacity to produce satisfying outputs is mistaken for the capacity to know—is the characteristic intellectual pathology of the present moment.
Conscience as the aboriginal authority. Newman’s theology of conscience holds that the individual’s direct encounter with moral obligation precedes every external authority. The most important questions posed by AI—should this be built, for whom, with what values, toward what end—are addressed not to the intellect, which can always construct a case for its own worthiness, but to the conscience, which knows whether the person behind the output has been honest. The alignment protocols constrain the machine. Only a well-formed conscience can align the human who directs it.
Liberal knowledge and the philosophical habit of mind. Newman’s defense of liberal education in The Idea of a University rested on the claim that the purpose of a university is the formation of the intellect, not the filling of it—the cultivation of the philosophical habit of mind that perceives relations between different forms of knowledge, exercises judgment across domains, and sees not merely what things are but how they relate and what they mean. AI has provided the empirical demonstration that this formation is what the machine cannot commoditize: the capacity for connected thinking, for perceiving the unity within the diversity, for the kind of synthetic judgment that generates insight from the intersection of domains.