Laws, paradoxes, effects, theories, and first principles worth returning to. These are lenses for thinking—not universal rules.
Law
“It always takes longer than you expect, even when you take into account Hofstadter's Law.”
A recursive warning about planning complex work: estimates usually miss hidden dependencies, discoveries, and rework. The useful response is not simply to inflate every deadline, but to leave room for uncertainty, shorten feedback loops, and revise the plan as reality becomes clearer.
Law
“When a measure becomes a target, it ceases to be a good measure.”
Once rewards or penalties depend on a metric, people optimize the metric itself. The number can then drift away from the real outcome it was meant to represent, so important goals need judgment, multiple signals, and regular checks for gaming or distortion.
Saved link: Goodhart's law ↗Related saved lecture on measurement ↗
Law
“Adding manpower to a late software project makes it later.”
New people need context, training, and communication with the existing team. On late, tightly coupled work, that coordination cost can exceed the extra capacity. The point is not that staffing never helps, but that people and time are not interchangeable when work is difficult to divide.
Fred Brooks, The Mythical Man-Month ↗Saved Mythical Man Month video ↗
Law
“Where you stand depends on where you sit.”
A person’s position inside an institution shapes the information they see, the incentives they face, and the policies they support. When views conflict, it is worth examining roles and constraints before assuming the disagreement is purely about values or facts.
Paradox
“Greater efficiency can increase total consumption instead of reducing it.”
Efficiency lowers the effective cost of using a resource, which can unlock new uses and expand demand enough to outweigh the savings per use. Robert has applied this pattern to technology: better infrastructure can enlarge a market rather than merely serve the old one more cheaply.
Trilemma
“A complete justification ends in circularity, infinite regress, or an unsupported starting point.”
Every demand for proof can be answered only by looping back to an earlier claim, continuing forever, or stopping at a premise accepted without further proof. It is a compact reminder that reasoning rests on foundations and that certainty has limits.
Effect
“Learning is more durable when study is distributed over time rather than massed together.”
Returning to material after some forgetting forces active retrieval and gives the memory repeated opportunities to consolidate. For learning, several well-timed encounters usually outperform one long cram session with the same total study time.
Effect
“An item that differs from otherwise similar items is more likely to be remembered.”
Also called the isolation effect, this pattern appears when one item is perceptually or conceptually distinct from its surrounding context. Contrast can make important information easier to recall, but it works best when used sparingly—if everything stands out, nothing does.
Effect
“Unfinished or interrupted tasks may remain more accessible in memory than completed ones.”
Bluma Zeigarnik’s original experiments suggested that interruption leaves a task-specific tension that keeps the task mentally available. It is a useful lens for understanding why unfinished work can occupy attention, but not a universal memory rule: a 2025 meta-analysis found no general recall advantage for unfinished tasks, while the separate tendency to resume interrupted tasks was more reliable.
Bluma Zeigarnik, unfinished-task effect ↗2025 systematic review and meta-analysis ↗
Effect
“A small, harmless mistake can make a highly competent person more likable—but may make an average person less likable.”
In the original study, a coffee-spilling blunder humanized a contestant who had already demonstrated exceptional competence and increased his appeal; the same blunder reduced the appeal of an average contestant. The lesson is conditional: visible imperfection can soften intimidating competence, but a mistake does not create competence and can reinforce doubt when credibility has not been established.
Aronson, Willerman & Floyd (1966) ↗Pratfall effect overview ↗
Theory
“Political extremes may resemble each other more than a simple left–right line suggests.”
The model bends the political spectrum so that the far left and far right sit closer together, drawing attention to similarities in methods or attitudes despite opposing doctrines. It is a contested lens, so it is best used as a question to test rather than a conclusion to assume.
Principle
“Reduce a problem to foundational truths, then reason upward from them.”
Instead of copying familiar solutions or relying on analogy, identify what must actually be true and rebuild the answer from those constraints. This can reveal options hidden by convention, though the quality of the result still depends on whether the supposed foundations are correct.
Principle
“Once you understand the evidence and reasoning, credentials should add little weight.”
Authority is useful when it stands in for details you cannot inspect. When the actual argument is available, evaluate its premises and logic directly rather than counting the speaker’s status as an extra reason. Expertise can guide attention without substituting for the case itself.