By Geoffrey Chen
1 / 12
According to the author, what is the ‘single decisive withdrawal’ that this book makes regarding the nature of knowledge?
A. The distinction between discovery and creation is reinforced to protect human expertise.
B. Human understanding is dismissed as a secondary and irrelevant cognitive error.
C. Reason is replaced by statistical probability as the only valid form of logic.
D. The knower is no longer treated as a necessary condition for knowledge to operate.
That’s right!
The central premise of the book is that AI demonstrates that knowledge can function effectively without being anchored in a present human subject.
2 / 12
In the contemporary epistemic scene, how has the temporal relationship between knowledge and understanding changed?
A. Understanding must now be reached instantaneously at the moment of discovery.
B. The two processes have become identical, merging into a single act of ‘deployable comprehension’.
C. Knowledge is delayed indefinitely until a human subject can verify the internal basis of a decision.
D. Judgments are generated and acted upon first, and only afterward do interpretation and explanation catch up.
That’s right!
The author describes an ‘inversion’ where knowledge-like outputs enter practice prior to the articulation of reasons.
3 / 12
What is the ‘Ghost’ in Hume’s philosophy that the author identifies as a precursor to modern statistical regimes?
A. The fragility of causal necessity and the reliance on habit-stabilised expectations.
That’s right!
Hume observed that we project necessity onto regularities through habit, which anticipates the shift toward reliability in AI.
B. The spiritual remnant of the subject that persists in mathematical logic.
C. The irrationality of the human mind when faced with complex data.
D. The hidden variables that would explain the world if only we had enough data.
4 / 12
How does ‘Reliability’ differ from ‘Necessity’ as a form of epistemic legitimacy?
A. Necessity is based on performance, whereas reliability is based on logical axioms.
B. Reliability is an all-or-nothing achievement that prohibits any exposure to error.
C. Reliability is diachronic, accruing force through endurance and resilience under repeated use.
That’s right!
Reliability earns authority through a track record of performance rather than an immediate proof of why something must be the case.
D. Reliability is synchronic and must be proven at the exact moment of assertion.
5 / 12
The author contrasts ‘Theorems’ and ‘Laws’ to illustrate two forms of stability. What is a key characteristic of ‘Theorem-like’ stability?
A. It is absolute and ‘brittle’, holding all at once or not at all.
That’s right!
Theorems are logical derivations where a single error in the proof or a counterexample within the system undoes the entire result.
B. It allows for productive failure that refines the theory’s scope.
C. It gains authority through repeated empirical confirmation over time.
D. It depends on the physical world ‘pushing back’ against the observer.
6 / 12
What is the ‘Statistical Turn’ mentioned in the text?
A. The requirement that all human experts learn to use advanced probability theory.
B. The replacement of rule-following and conceptual design with structure that stabilizes through iteration and performance.
That’s right!
The statistical turn moves away from deriving structure from first principles toward inferring it from large-scale regularities.
C. The discovery that the universe is fundamentally made of numbers rather than matter.
D. The political decision to prioritize data-driven policies over traditional ethics.
7 / 12
Why does the author argue that ‘hallucination’ is a misplaced objection to connectionist AI systems?
A. Because it indicates the system is beginning to develop a human-like imagination.
B. Because it is a byproduct of high-level coherence and competence in pattern-matching rather than a simple breakdown.
That’s right!
The author notes that only a highly fluent system can produce plausible-sounding but false statements, showing it is functioning exactly as optimized (for coherence).
C. Because the system is technically incapable of being wrong due to its training data.
D. Because human experts are actually more likely to hallucinate than machines.
8 / 12
How is the role of the ‘Expert’ transformed in the age of delegated knowing?
A. The expert’s role shifts from producing judgments to auditing, monitoring, and managing the implications of outputs.
That’s right!
Expertise becomes supervisory and interpretive rather than the primary generative source of knowledge.
B. Experts are replaced entirely by automated governance protocols and checklists.
C. The expert must now focus exclusively on theoretical explanation without practical application.
D. The expert becomes the sole point of origin for all valid judgments.
9 / 12
What does the author identify as the ‘Human Remainder’ after cognitive and epistemic tasks have been delegated?
A. The unique capacity for higher-level mathematical reasoning and symbolic logic.
B. Commitment and responsibility—the willingness to stand behind consequences and assume answerability.
That’s right!
The author argues that while cognition can be delegated, the normative stance of commitment cannot.
C. The ability to process large datasets without algorithmic bias.
D. The biological necessity of being the primary medium for all information retrieval.
10 / 12
In the author’s view, how does ‘Reliance’ function as a replacement for ‘Belief’ in institutional settings?
A. Reliance is only used when the system is known to be 100%accurate.
B. Reliance requires a deeper level of psychological conviction than belief.
C. Reliance is a mental state that guarantees the truth of the system’s output.
D. Reliance is a practical orientation to treat a judgment as sufficient to proceed, sustained by performance records.
That’s right!
It moves legitimacy from inner commitment to the outward track record of the system.
11 / 12
What is meant by the ‘Privatisation of Understanding’?
A. The transition of all academic research into secret, high-security facilities.
B. The legal ownership of AI models by private corporations.
C. Explanation no longer serves as a shared public standard but as local orientation for individuals or groups.
That’s right!
The author argues that because understanding no longer ‘authorises’ the decision, different people can use different stories for the same outcome.
D. The elimination of all public explanations in favour of proprietary algorithms.
12 / 12
When knowledge becomes ‘Infrastructure’, what does it primarily act as within a civilisation?
A. A background condition that shapes action and decisions without requiring continuous justification.
That’s right!
Like electricity or roads, delegated knowing becomes a default environment we ‘inhabit’ rather than a series of claims we debate.
B. A symbolic representation of the shared values of the community.
C. A rare resource that requires constant active reasoning to access.
D. A tool used exclusively by a specialized elite to control the ‘knowing subject’.