Herbert Simon — On AI

AI Practice Framework

The Berkeley researchers' prescription for the AI-augmented workplace — structured pauses, sequenced workflows, protected human-only time, behavioral training alongside technical training — the operational counterpart to Maslach's fix-the-…

Ascending Friction

The Orange Pill's thesis that AI does not eliminate difficulty but relocates it to a higher cognitive floor — the engineer who no longer struggles with syntax struggles instead with architecture.

Attention as Binding Constraint (Simon)

Simon's 1971 diagnosis that information consumes attention, and that in an information-rich world, attention becomes the binding constraint — the prescient framework the AI age has vindicated at civilizational scale.

Bounded Rationality

Simon's 1955 thesis that human decision-makers operate under binding constraints of information, computation, and time — producing satisficing rather than optimization, and demolishing the foundation of classical economics.

Choice Architecture (Simon Reading)

Simon's foundational insight — half a century before the term existed — that the structure of a decision environment shapes decisions more reliably than individual preferences do, and that AI constitutes the most consequential choice archit…

Chunking (Miller)

Miller's term for the cognitive operation that transcends the seven-item limit without exceeding it — the packaging of multiple items into a single retrievable unit, transforming nine unfamiliar letters into three familiar acronyms and, at …

Decomposition and Its Logic

The systematic breaking of complex problems into sub-problems manageable by bounded minds — the organizing principle of every hierarchy, division of labor, and modular system ever built, and the first casualty of AI-driven cross-domain capa…

Deliberate Practice

Ericsson's empirically established mechanism for building expertise — effortful, targeted engagement at the boundary of capability, guided by specific feedback and sustained over thousands of hours.

Designing for the Bounded Builder

The Simonian design imperative for the AI age: build organizational, educational, and personal architectures that respect the bounds of the evaluator rather than maximizing the outputs of the generator.

Division of Labour

Smith's foundational principle that specialization produces the greatest improvement in the productive powers of labour — the pin factory's logic, now being inverted by AI tools that dissolve the boundaries between specialized operations.

Flow State

Mihaly Csikszentmihalyi's name for the condition of optimal human engagement — and, in Wiener's framework, the subjective signature of a well-regulated negative feedback system.

Heuristic Search

Simon and Newell's 1972 framework for how bounded minds navigate problem spaces too large for exhaustive search — using rules of thumb to direct attention toward promising alternatives, the cognitive engine of every expert performance and t…

Near-Decomposability

Simon's 1962 principle that complex systems are buildable by bounded minds because their components interact strongly within subsystems and weakly between them — and that AI is dissolving this property faster than new architectures can repl…

Problem Space

The formal structure of a problem — initial state, goal state, operators, constraints — that Simon and Newell argued was the proper unit of analysis for understanding how bounded minds solve problems, and whose AI-era expansion demands corr…

Satisficing

Simon's 1956 neologism for the decision procedure bounded agents actually use — searching sequentially through alternatives and accepting the first that clears a threshold of acceptability, rather than optimizing across all options.

Tacit Knowledge (Polanyi-Collins Reading)

Michael Polanyi's 1966 insight that we know more than we can tell — refined by Collins into a taxonomy of three species that has become the decisive framework for understanding what AI systems can and cannot absorb from human practice.

The Ant on the Beach

Simon's metaphor from The Sciences of the Artificial — an ant's complex path across a beach reflects the beach's complexity, not the ant's intelligence — and the most uncomfortable diagnostic available for the relationship between an AI-a…

The Attention Economy

The economic system in which human attention is harvested, packaged, and sold to advertisers — the infrastructure that drives the algorithmic pathologies Gore calls artificial insanity.

The Burnout Society

Byung-Chul Han's 2010 diagnosis of the achievement-driven self-exploitation that has replaced disciplinary control as the dominant mode of power — and, in cybernetic terms, a social system operating in positive feedback.

The Fluency Trap (Brown Reading)

The cultural habit of treating fluent AI output as competent AI output — an extension of the equation between eloquence and expertise that centuries of human interaction built.

The Fourth Bound

Simon's 1971 identification of attention as the constraint that persists when information, computation, and time are all relaxed — the bound that AI cannot dissolve because it is a property of consciousness itself.

The Heuristics and Biases Program

The research tradition Tversky and Kahneman founded in the 1970s to map the systematic departures of human judgment from rational ideals — the intellectual framework this entire book applies to the AI transition.

The Judgment Bottleneck

The structural condition of AI-augmented work in which the binding constraint has shifted from generating alternatives to evaluating them — where machine-speed production meets human-speed assessment and the quality of output depends entire…

The Satisficing Threshold

The floating criterion of 'good enough' that bounded agents carry into any decision — adjusting upward as the cost of generating the next alternative falls, and rising in the AI age faster than the evaluative capacity that would police it.

Vector Pods

Edo Segal's term for the small judgment-focused groups emerging in AI-augmented organizations — decomposing work not by functional domain but by cognitive operation, restoring near-decomposability at the level of direction rather than imple…

Work (5)

Human Problem Solving

Simon and Newell's 1972 magnum opus on how bounded minds navigate problem spaces through heuristic search — the founding document of cognitive science and the framework through which AI-augmented problem-solving becomes legible.

The Architecture of Complexity

Simon's 1962 paper introducing near-decomposability as the universal structural principle of complex systems buildable by bounded minds — and the founding document of modern systems thinking about organizational design.

The Berkeley Study

Ye and Ranganathan's 2026 Harvard Business Review ethnography of AI in an organization — the empirical documentation of task seepage and work intensification that prospect theory predicts.

The Orange Pill

Edo Segal's 2026 book on the Claude Code moment and the AI transition — the empirical ground and narrative framework on which the Festinger volume builds its diagnostic reading.

The Sciences of the Artificial

Simon's 1969 landmark establishing design as a rigorous form of knowledge — distinct from natural science and focused on the interface between a system's inner environment and the outer world it must serve.

Person (5)

Allen Newell

American cognitive scientist (1927–1992), Simon's closest collaborator for four decades, co-founder of artificial intelligence, and architect of the information-processing theory of cognition that reshaped psychology and computer science al…

Byung-Chul Han

Korean-German philosopher (b. 1959) whose diagnoses of smoothness, transparency, and achievement society provide the critical idiom within which Groys's AI analysis operates — and against which Groys's emphasis on institutional frame offers…

Edo Segal

Serial entrepreneur and technologist whose The Orange Pill (2026) provides the phenomenological account — the confession over the Atlantic — that Pang's framework diagnoses and treats.

Herbert Simon

American polymath (1916–2001) — Nobel laureate in economics, Turing Award winner in computer science, co-founder of artificial intelligence — whose concept of bounded rationality reshaped economics, organizational theory, and the design of …