he research side of the house: how people behave, how machines behave, and what happens when the two meet. Some of it is reading, some of it is circuits with the numbers worked out, some of it is a test card. This page lists every room in this part of the building. The rooms also point at each other, so you can wander instead.

research areas

	research areas — nineteen subjects, each at seminar depth with its own reading list
	signal detection — human behavior: telling signal from noise, and where you draw the line
	prediction error — human behavior: learning from surprise
	replication — human behavior: which findings hold up
	bounded rationality — human behavior: good-enough decisions
	sycophancy — AI behavior: models that tell you what you want to hear
	calibration — AI behavior: does 80% sure mean right 80% of the time?
	shortcut learning — AI behavior: right answers for the wrong reasons
	interpretability — AI behavior: what a network is actually using
	hallucination and confabulation — AI behavior: fluent, confident, and not true
	specification gaming — AI behavior: the letter of the goal, not the spirit
	automation bias — human–AI: trusting the machine too much, or too little
	anthropomorphism — human–AI: seeing a mind in the machine
	complementarity — human–AI: when the team beats either alone
	algorithm aversion — human–AI: preferring the human who errs
	mental models — human–AI: what people believe the machine is doing
	the nervous system — a fourth shelf: the brain itself, and the people who took it apart
	engram — the nervous system: where a memory is kept
	predictive processing — the nervous system: the brain as a guesser
	connectomics — the nervous system: the wiring diagram, and what it does not tell you
	neuro and AI — the nervous system: scoring networks against brains
	machine-made art — human–AI: how people respond to it, and why the label matters

essays

	essays — longer pieces that argue something; read the footnotes
	where the line is — from Fechner's threshold to a machine that declines to answer
	the polite machine — ELIZA reflected you; newer machines agree with you
	the break, 1951 or 1952 — Wiener, McCulloch, Pitts, and a message from Mexico City
	the thesis that wasn't — speculative, and labelled so
	who made this? — machine-made pictures, and this building's own walls

methods and timelines

	the staircase — an adaptive experiment you can run on yourself
	timeline: psychophysics — from Weber, 1834, to QUEST
	timeline: machines that answer — from the Turk to chat models
	timeline: replication — from Sterling, 1959, to Many Labs
	timeline: cybernetics — from Maxwell's governors to the homeostat, and the histories since
	timeline: memory — from nonsense syllables to engrams
	timeline: neuroscience — from a silver stain, 1873, to the wiring of a fly
	timeline: machine-made art — from Lovelace and the daguerreotype to the courts
	all four methods — the staircase, ROC, power and preregistration, each one runnable

toys NEW [playful]

	the toy shelf — six small demos of real effects, each with its sources and its limits
	the Stroop box — name the ink, not the word
	the arrow lines — the Müller-Lyer illusion
	spot the change — change blindness
	the reaction box — simple and go/no-go reaction time
	the line judge — a Weber fraction for line length
	the four cards — Wason's selection task

Good morning. Good afternoon. Good evening. Up late? Toy suggestion: This hour's toy: the Stroop boxthe arrow linesspot the changethe reaction boxthe line judgethe four cards.

people and words

	people — real people, real dates, their own works
	Weber — touch, and the constant fraction
	Fechner — the founder of psychophysics
	William James — the Principles
	Herbert Simon — satisficing, and the first reasoning programs
	Frank Rosenblatt — the perceptron
	J. C. R. Licklider — man–computer symbiosis
	Joseph Weizenbaum — ELIZA, and second thoughts
	Hermann von Helmholtz — nerve speed, the ear, unconscious inference
	Wilhelm Wundt — the first laboratory
	Hermann Ebbinghaus — memory, measured on himself
	L. L. Thurstone — comparative judgment
	Norbert Wiener — cybernetics, and a warning
	Claude Shannon — information
	Frederic Bartlett — remembering as reconstruction
	W. Ross Ashby — the homeostat and requisite variety
	Warren McCulloch — the logical neuron
	Paul Meehl — clinical versus statistical, and soft psychology
	Walter Pitts — the logician of the 1943 paper
	Jerome Lettvin — the frog's eye
	Santiago Ramón y Cajal — the neuron doctrine, drawn
	Charles Sherrington — a word for a gap
	Donald Hebb — cell assemblies
	Karl Lashley — in search of the engram
	Wilder Penfield — the map on the brain
	Brenda Milner — H.M., and kinds of memory
	Roger Sperry — two halves
	glossary — the terms these pages use, A to Z

 minds, other and otherwise

	minds — ELIZA, the shadow, the Cartesian theater, the hard problem
	Turing — the imitation game, and Enigma before it
	swarms — boids: three local rules, one flock

 signal and noise

	signal — pulling what you meant back out of what arrived
	the RC low-pass — the simplest filter there is
	RLC — resonance and Q
	the integrator — and a bronze computer
	the Wien bridge — an oscillator that keeps itself going
	Fourier, 1822 — every wave is a sum of simpler ones
	Bayes, 1763 — updating a belief by the evidence
	secrecy — Enigma and the one-time pad, noise on purpose
	base 60 and base 20 — older ways of counting

 perception and measurement

	measurement — what changes when somebody looks
	test patterns — so you know the fault was yours
	bars — colour bars, the old way

The tour has rooms for this too: a psychology hall and a room about thresholds, where the light is turned down on purpose. Walk the tour, or read the plain list of its rooms.

More rooms are going up on this side.