b72e103de770c3b18a3e009fa49c21e06b0f6f04
new file mode 100644
index 0000000..44a47cc
@@ -0,0 +1,45 @@
+---
+visibility: public-edit
+---
+
+# applications of math
+
+many people say math is useless. "fine, addition and multiplication are useful, but I learned those early. why is calculus useful?"
+
+this wiki is my answer. math applies to everything — not just in the obvious "you need it for engineering" way, but in a deeper way: advanced math gives you ways of thinking that restructure how you see the world. the biggest use of advanced math is the ability to organize and structure things.
+
+the wiki has three layers, from the most immediately practical to the most profoundly structural.
+
+---
+
+## layer 1: the immediate and basic
+
+the math you use every day without calling it math. these seem trivial, but they're everywhere — and getting them wrong has real consequences.
+
+- [[counting-and-measurement|counting and measurement]] — all of science, commerce, and engineering starts with measuring things. precision vs accuracy. units as a type system. measurement theory.
+- [[arithmetic-everywhere|arithmetic everywhere]] — addition, subtraction, multiplication, division. so embedded in daily life that we forget they're math. compound interest. dosage calculations. cooking ratios.
+- [[ordering-and-comparison|ordering and comparison]] — sorting, ranking, prioritizing. total and partial orders. arrow's impossibility theorem. elo ratings. pareto optimality.
+- [[probability-in-daily-life|probability in daily life]] — "should I bring an umbrella?" bayesian reasoning. base rate neglect. the monty hall problem. expected value. risk assessment.
+- [[patterns-and-estimation|patterns and estimation]] — fermi estimation. order-of-magnitude reasoning. mental math. pattern recognition and its failures.
+
+## layer 2: the STEM foundation
+
+the classic "math is the language of science" argument. true and important, but well-known — so these pages are concise.
+
+- [[physics|physics]] — the original applied math. newton's laws as differential equations. quantum mechanics as linear algebra. the unreasonable effectiveness of mathematics.
+- [[computer-science|computer science]] — boolean algebra → circuits → computers. algorithms and complexity. cryptography. information theory. machine learning.
+- [[engineering-and-modeling|engineering and modeling]] — math modeling competitions. monte carlo simulation. optimization. spectral methods. "all models are wrong but some are useful."
+- [[biology-and-medicine|biology and medicine]] — population dynamics. epidemiology. genetics. EEG signal processing. medical statistics.
+
+## layer 3: the profound and structural
+
+this is the most interesting layer — and the one most people miss. abstract math doesn't just solve problems. it gives you **ways of thinking about fundamental patterns**. each branch provides a lens.
+
+- [[the-organizational-lens|the organizational lens]] — my core thesis. advanced math's biggest use is as an organizational tool for seeing structure in everyday things.
+- [[calculus-as-thinking|calculus as thinking]] — not computing derivatives. thinking about change, accumulation, limits, and continuity as universal patterns.
+- [[linear-algebra-as-thinking|linear algebra as thinking]] — every idea as a vector. every process as a matrix. eigenvalues as stability. semantic space and word embeddings.
+- [[set-theory-as-thinking|set theory as thinking]] — unions, intersections, MECE partitions. binary classification as set membership. the discipline of precise categorization.
+- [[multivariable-calculus-as-thinking|multivariable calculus as thinking]] — gradient (which direction?), divergence (source or sink?), curl (circulation?). gradient descent and optimization.
+- [[abstraction-as-power|abstraction as power]] — why abstraction is useful. the abstraction ladder. category theory. the trade-off between generality and specificity.
+- [[topology-as-thinking|topology as thinking]] — invariants. what stays the same under deformation. connectedness. fundamental groups. topological data analysis.
+- [[symmetry-and-groups|symmetry and groups]] — group theory. noether's theorem. the rubik's cube. music theory. why symmetric solutions tend to be elegant.
\ No newline at end of file
new file mode 100644
index 0000000..2b080b5
@@ -0,0 +1,7 @@
+---
+visibility: public-edit
+---
+
+# log
+
+append-only record. format: `## [YYYY-MM-DD] type | title`
\ No newline at end of file
new file mode 100644
index 0000000..ef83ef9
@@ -0,0 +1,94 @@
+---
+visibility: public-edit
+---
+
+# schema
+
+this wiki uses the LLM wiki pattern. the LLM incrementally builds and maintains a persistent, interlinked knowledge base. knowledge is compiled once and kept current, not re-derived on every query.
+
+## three layers
+
+**`raw/`** — immutable source documents. articles, transcripts, notes, data. the LLM reads from raw but never modifies it. this is the source of truth.
+
+**`wiki/`** — LLM-maintained compiled pages. summaries, entity pages, concept pages, comparisons. the LLM owns this layer entirely — creates pages, updates them when new sources arrive, maintains cross-references, keeps everything consistent. you read it; the LLM writes it.
+
+**`schema.md`** (this file) — tells the LLM how the wiki is structured, what conventions to follow, and what workflows to use. you and the LLM co-evolve this over time.
+
+## page format
+
+every wiki page has two layers:
+- **compiled truth** (above the `---` divider): current best understanding, rewritten freely when evidence changes
+- **timeline** (below the `---` divider): append-only evidence trail with dates, never edited after writing
+
+the compiled truth is a living synthesis. when new evidence arrives, rewrite it. the timeline is an immutable log — only append, never edit.
+
+## frontmatter
+
+```yaml
+---
+title: Page Title
+type: topic | person | project | idea | source-summary
+visibility: public
+tags: [area1, area2]
+sources: [raw/meeting-2026-04-10.md, raw/article-link.md]
+---
+```
+
+## wikilinks
+
+**inline wikilinks are the most important thing you do.** they are what makes this a knowledge graph instead of a folder of notes.
+
+- weave links into prose naturally: `[[mark-khrapko|Mark]] mentored him on [[hiring|hiring philosophy]]`
+- link on first mention only per section
+- use display text: `[[hiring-philosophy|hiring philosophy]]` not `[[hiring-philosophy]]`
+- cross-category links are the most valuable (person↔topic, project↔idea, experience↔reflection)
+- every page should have at least 2 outbound links
+- no 'see also' sections — if it's not worth mentioning in prose, it's not a connection
+- conservative: better to miss a link than create a wrong one
+
+## operations
+
+### ingest
+
+when processing a new source:
+1. read source → identify entities (people, topics, projects)
+2. for each entity: check if wiki page exists (read index.md first)
+3. if exists: update compiled truth, append timeline entry with date and source
+4. if new: create page with compiled truth + first timeline entry
+5. **add inline wikilinks to related pages as you write** — this is the critical step
+6. update index.md with the new page
+7. append to log.md: `## [YYYY-MM-DD] ingest | Source Title`
+
+a single source might touch 10-15 wiki pages. that's normal.
+
+### query
+
+when answering questions:
+1. read index.md to find relevant pages
+2. read those pages and synthesize an answer
+3. if the answer is valuable, file it back into the wiki as a new page — explorations should compound in the knowledge base, not disappear into chat history
+4. append to log.md: `## [YYYY-MM-DD] query | Question summary`
+
+### lint
+
+periodically health-check the wiki. look for:
+- contradictions between pages
+- stale claims that newer sources have superseded
+- orphan pages with no inbound links
+- important concepts mentioned but lacking their own page
+- missing cross-references
+- data gaps that could be filled with new sources
+
+append to log.md: `## [YYYY-MM-DD] lint | Summary of findings`
+
+## index.md
+
+the index is a catalog of every page in the wiki. each entry has a link, a one-line summary, and optionally metadata like date or source count. organized by category. the LLM reads the index first when answering queries to find relevant pages. update it on every ingest.
+
+## log.md
+
+append-only chronological record. every entry starts with `## [YYYY-MM-DD] type | title` so it's parseable with simple tools. types: `ingest`, `query`, `lint`, `update`.
+
+## source tracking
+
+every fact should be traceable. use `sources:` frontmatter to list which raw files contributed to a page. timeline entries cite their source: `**2026-04-10** (from meeting-notes.md): key insight here`
\ No newline at end of file
new file mode 100644
index 0000000..45560c9
@@ -0,0 +1,57 @@
+---
+visibility: public-edit
+---
+
+# arithmetic everywhere
+
+addition. subtraction. multiplication. division. you learned these before you were ten. they're so embedded in daily life that doing them doesn't feel like "doing math" — it feels like common sense. but that's exactly the point.
+
+## the invisible math
+
+you wake up. you check the time (subtraction: "I have 45 minutes before I need to leave"). you pour cereal (estimation, ratios). you check your bank account (addition, subtraction). you drive to school (speed × time = distance, even if you don't think of it that way). you split lunch with a friend (division). you tip the waiter (multiplication by 0.2, or whatever you tip).
+
+by noon you've done dozens of arithmetic operations without once thinking "I'm doing math." this is arithmetic's greatest success: it's so useful that it became invisible.
+
+## the operations are deeper than they seem
+
+each arithmetic operation captures a fundamental relationship:
+
+**addition** — combining. merging. accumulating. any time two things come together to form a whole, that's addition. revenues from two product lines. ingredients in a recipe. hours worked across a week. the concept of "putting things together" is so basic that it's hard to imagine thinking without it.
+
+**subtraction** — difference. comparison. removal. "how much more?" "how much is left?" "what changed?" subtraction is how we detect change, measure progress, and find gaps.
+
+**multiplication** — scaling. repetition. area. any time you have "some number of groups, each of some size" — that's multiplication. it's also how we handle rates: price × quantity, speed × time, probability × outcomes. dimensional analysis is just multiplication with units attached.
+
+**division** — sharing. averaging. ratio. per-unit thinking. "miles per gallon," "dollars per hour," "points per game" — all division. it's how we normalize, compare, and think about rates.
+
+## where it gets interesting
+
+### dosage calculations
+
+medical dosing is arithmetic that kills people when it goes wrong. a drug might be prescribed at 5mg per kg of body weight, administered in 3 doses per day, diluted in a solution of 10mg/mL. that's multiplication, division, and unit conversion — elementary school math — but errors in this chain cause roughly 7,000 deaths per year in the US.
+
+### cooking ratios
+
+a vinaigrette is 3 parts oil to 1 part vinegar. a bread dough is roughly 5:3 flour to water by weight. once you think in ratios instead of recipes, you can cook anything in any quantity without looking anything up. this is the power of multiplicative thinking — it scales.
+
+### compound interest
+
+einstein (probably) never called it the eighth wonder of the world, but compound interest does demonstrate something genuinely profound about multiplication: repeated multiplication (exponentiation) grows shockingly fast. $1000 at 7% annual return becomes $7,612 in 30 years. you've added $0 — multiplication did all the work.
+
+this same principle underlies population growth, viral spread, and nuclear chain reactions. it's why [[probability-in-daily-life|probability in daily life]] matters so much for financial decisions.
+
+### mental shortcuts
+
+people who are "good with numbers" usually aren't doing harder math — they're doing the same arithmetic with better shortcuts:
+- to tip 20%, find 10% (move the decimal) and double it
+- to multiply by 15, multiply by 10 and add half
+- to check if a number is divisible by 3, add its digits
+- to estimate 18 × 22, compute 20² - 2² = 400 - 4 = 396 (difference of squares — this is algebra sneaking into arithmetic)
+
+## the philosophical point
+
+arithmetic works. that's actually strange. why should the abstract rules governing numbers — which are themselves abstract objects — map so perfectly onto physical reality? why does 3 apples + 4 apples always give 7 apples, never 8?
+
+this is a baby version of wigner's "unreasonable effectiveness of mathematics" question that becomes much more dramatic at the [[physics|physics]] level. but it starts here, with the fact that addition works on apples, dollars, people, photons, and ideas — despite these things having nothing else in common.
+
+and these same four operations generalize far beyond numbers. [[linear-algebra-as-thinking|linear algebra]] extends arithmetic to vectors and matrices — you can add vectors, scale them by numbers, multiply matrices. the operations feel familiar because they *are* arithmetic, lifted into higher dimensions.
\ No newline at end of file
new file mode 100644
index 0000000..455846a
@@ -0,0 +1,47 @@
+---
+visibility: public-edit
+---
+
+# counting and measurement
+
+the most fundamental math is counting. one, two, three. it sounds so trivial that calling it "math" feels generous. but everything starts here.
+
+every science begins with measurement. before galileo could discover that heavy and light objects fall at the same rate, someone had to figure out how to measure time. before we could understand disease, someone had to count who got sick and who didn't. before we could build bridges, someone had to measure distances and loads.
+
+## measurement theory
+
+measurement is deeper than it looks. there's a whole branch of math — measurement theory — that asks: what does it even mean to measure something?
+
+consider temperature. we say it's 70°F outside. but what is that number? it's not like counting apples. you can't line up 70 "temperature units" on a table. temperature is an abstraction — we've assigned numbers to sensations using a scale that's, in some sense, arbitrary. fahrenheit and celsius give different numbers for the same reality.
+
+this distinction matters. you can say "20 apples is twice as many as 10 apples." but can you say "40°F is twice as hot as 20°F"? no — that's meaningless. (0°F isn't "no temperature.") the type of measurement determines what math you can do with it.
+
+measurement theorists classify scales:
+- **nominal**: just labels. jersey numbers. you can't average them.
+- **ordinal**: rankings. 1st, 2nd, 3rd. you know the order but not the gaps.
+- **interval**: temperature. equal differences are meaningful, but ratios aren't.
+- **ratio**: weight, length, time. has a true zero. ratios are meaningful.
+
+these scale types are really a classification problem — [[set-theory-as-thinking|set theory]] in disguise. each scale type defines a category with specific rules about what operations are valid.
+
+every time someone averages ordinal data ("our average customer rating is 3.7 stars"), they're making a measurement theory error that we've all collectively agreed to ignore.
+
+## precision vs accuracy
+
+precision is how many decimal places you report. accuracy is whether you're close to the truth. they're independent.
+
+a broken clock is precise (to the minute) but inaccurate. saying "there are about 8 billion people on earth" is imprecise but accurate. the danger is confusing the two — a number with lots of decimal places *feels* more trustworthy, but it might just be precisely wrong.
+
+in my math modeling work, this comes up constantly. you can run a simulation that outputs 15 significant digits, but if your input assumptions are only good to 2 digits, those extra 13 are fiction. garbage in, precise garbage out.
+
+## units
+
+units are math's type system. they prevent you from adding nonsensical things. you can't add 5 meters to 3 seconds — the units don't match. this is exactly like type-checking in programming: the compiler (or your brain) catches the error before you get a nonsensical result.
+
+the mars climate orbiter crashed because one team used metric and another used imperial. a $327 million spacecraft destroyed by a unit conversion error. that's how fundamental counting and measurement really are — get them wrong and everything built on top collapses.
+
+## the deep point
+
+counting and measurement aren't just "basic math" — they're the interface between abstract numbers and physical reality. every measurement is a tiny philosophical act: you're claiming that some aspect of the messy, continuous, complicated world can be captured by a number. that claim is always an approximation, always a choice, and understanding the nature of that approximation is where the real math begins.
+
+this connects directly to [[calculus-as-thinking|calculus as thinking]] — the question of "how precisely can we measure change?" is what led newton and leibniz to invent calculus. and [[patterns-and-estimation|estimation]] is the practical skill of knowing when high precision matters and when it doesn't. measurement also connects to [[ordering-and-comparison|ordering]] — once you can measure things, you can rank them, but the type of measurement determines what kind of ordering is valid.
\ No newline at end of file
new file mode 100644
index 0000000..f059e88
@@ -0,0 +1,64 @@
+---
+visibility: public-edit
+---
+
+# ordering and comparison
+
+every time you make a to-do list and decide what to do first, you're performing a mathematical operation. every leaderboard, every ranking, every "which restaurant should we go to?" — that's math. specifically, it's the math of orders.
+
+## total orders and partial orders
+
+when you sort a list of numbers from smallest to largest, that's a **total order**: every pair of elements can be compared. 3 < 7, always, no ambiguity. numbers are easy.
+
+but most real-world ordering isn't total. it's **partial**. consider prioritizing tasks:
+- "study for exam" is more important than "do laundry"
+- "finish essay" is more important than "do laundry"
+- but is "study for exam" more important than "finish essay"? depends on deadlines, weight, your energy level
+
+you can't always compare. some things are just... incomparable. and that's fine — partial orders are a legitimate mathematical structure, not a failure of decision-making. partial orders are sets with a specific relation — which makes them a [[set-theory-as-thinking|set theory]] concept at heart.
+
+## the hidden math in sorting
+
+sorting algorithms are one of the most studied topics in computer science (see [[computer-science|computer science]]). but the key mathematical insight is about *comparisons*: to sort n items, you need at least n log(n) comparisons. this is a proven lower bound — no cleverness can beat it.
+
+this has a practical consequence: if you're trying to rank 100 job applicants by doing pairwise comparisons, you need at minimum about 665 comparisons. ranking is inherently expensive. that's why we use heuristics, filters, and thresholds instead of trying to create a perfect total ordering of everything.
+
+## comparison and transitivity
+
+the most important property of a "well-behaved" ordering is **transitivity**: if A > B and B > C, then A > C.
+
+numbers obey this. but human preferences often don't. you might prefer pizza to sushi, sushi to burgers, and burgers to pizza. that's a cycle — a violation of transitivity. it means your preferences can't be represented as a simple ranking, and someone who knows your preferences can manipulate you by controlling the order of choices presented.
+
+arrow's impossibility theorem proves that no voting system can always convert a group's preferences into a transitive ranking while satisfying a few basic fairness criteria. democracy is mathematically hard because ordering is mathematically subtle.
+
+## metrics and distances
+
+comparison requires a notion of "how much better." this leads to **metrics** — mathematical distance functions. a metric needs to satisfy:
+1. d(x, y) ≥ 0 (distances are non-negative)
+2. d(x, y) = 0 iff x = y (only identical things have zero distance)
+3. d(x, y) = d(y, x) (symmetry)
+4. d(x, z) ≤ d(x, y) + d(y, z) (triangle inequality)
+
+these seem obvious, but many real-world "distances" violate them. travel time isn't symmetric (one-way streets). perceived similarity isn't symmetric (people say "north korea is like china" more than "china is like north korea"). whenever you're comparing things, it's worth asking: does my comparison method actually behave like a real metric?
+
+metrics are deeply connected to [[topology-as-thinking|topology]] — every metric defines a topology, and the topological properties (connectedness, compactness) determine what kinds of ordering and comparison are possible in a space.
+
+this connects to [[linear-algebra-as-thinking|linear algebra as thinking]] — when you embed things as vectors, the distance between them (cosine similarity, euclidean distance) gives you a metric that enables meaningful comparison in high-dimensional spaces.
+
+## ordering in practice
+
+### elo ratings
+
+chess ratings, competitive gaming, and matchmaking systems all use elo or elo-like systems. the core idea: after each match, update both players' ratings based on the *expected* vs *actual* outcome. if a 2000-rated player beats a 1500-rated player, not much changes. if the 1500 beats the 2000, big update.
+
+elo is mathematically elegant because it converts a partial order (we only observe some matchups) into a total order (a single number per player). it's also a beautiful example of [[probability-in-daily-life|bayesian updating]] — each game is new evidence that shifts our estimate.
+
+### pareto optimality
+
+when comparing on multiple dimensions, a solution is **pareto optimal** if you can't improve one dimension without worsening another. a car that's faster AND cheaper AND safer than another car dominates it. but usually there are trade-offs, and you end up with a *set* of pareto-optimal choices — the pareto frontier.
+
+this is the mathematical way of saying "it depends on what you value." ordering breaks down when there's no single axis of comparison, which is most of the time.
+
+## the deep point
+
+ordering feels like common sense, not math. but the math of orders reveals why ranking is so hard: transitivity failures, incomparable options, multiple dimensions, and the fundamental impossibility results of social choice theory. understanding this doesn't make decisions easier, but it does make you stop expecting a "correct" ranking where none exists. and comparison always rests on [[counting-and-measurement|measurement]] — what you can order depends on what you can measure.
\ No newline at end of file
new file mode 100644
index 0000000..827d841
@@ -0,0 +1,64 @@
+---
+visibility: public-edit
+---
+
+# patterns and estimation
+
+"about how many piano tuners are in chicago?"
+
+this is a fermi estimation problem. you're not supposed to know the answer. you're supposed to *reason* your way to a reasonable estimate using math you already know.
+
+chicago has about 2.7 million people. maybe 1 in 20 households has a piano — call it 5%. average household size is about 2.5, so roughly 1 million households, meaning ~50,000 pianos. a piano should be tuned once or twice a year — say 50,000-100,000 tunings per year. a tuner can do maybe 4 per day, works 250 days per year, so about 1,000 tunings per year per tuner. that gives 50-100 piano tuners in chicago.
+
+the actual answer is somewhere around 100-200. we're within a factor of 2, which is excellent for a problem we "knew nothing about."
+
+## why estimation matters
+
+estimation is the skill of getting approximately right answers with minimal information. it's arguably more useful than precise calculation because in most real situations, you don't have precise inputs.
+
+in my math modeling competition work (HiMCM, MCM/ICM), the first step for every problem is estimation. before building a fancy model, you sanity-check: what's the right order of magnitude? if your model predicts that a city needs 50,000 ambulances, something is wrong. if it predicts 50, that might be reasonable. [[engineering-and-modeling|engineering and modeling]] always starts with this kind of gut-check.
+
+## order-of-magnitude reasoning
+
+the most powerful estimation tool is thinking in powers of 10:
+- 10⁰ = 1 (a person)
+- 10¹ = 10 (a classroom)
+- 10² = 100 (a lecture hall)
+- 10³ = 1,000 (a small school)
+- 10⁴ = 10,000 (a stadium section)
+- 10⁵ = 100,000 (a large stadium)
+- 10⁶ = 1,000,000 (a city)
+- 10⁹ = 1,000,000,000 (a country, roughly)
+- 10¹⁰ = 10 billion (the world, roughly)
+
+once you have these reference points, you can locate almost any quantity. "how many restaurants in new york city?" well, 8 million people, maybe 1 restaurant per 100 people... so ~80,000. (actual answer: about 27,000 — we're in the right order of magnitude, and the discrepancy tells us something interesting about the restaurant-to-person ratio.)
+
+the key insight: being wrong by a factor of 3 is fine. being wrong by a factor of 1,000 means you're confused about something fundamental. order-of-magnitude reasoning catches the catastrophic errors.
+
+## pattern recognition
+
+humans are pattern-recognition machines. we see faces in clouds, hear words in noise, find trends in random data. this is both our greatest mathematical strength and our greatest mathematical weakness.
+
+**when it works**: noticing that sales spike every december. recognizing that a function is growing exponentially. seeing that two seemingly different problems have the same structure. pattern recognition is the engine of mathematical intuition.
+
+**when it fails**: seeing patterns in random noise. the "hot hand" debate in basketball raged for decades partly because humans are so eager to see streaks in random sequences. financial "technical analysis" finds patterns in stock charts that often aren't really there. conspiracy theories are pattern recognition run amok.
+
+the antidote is statistical thinking: asking "how likely would this pattern be by chance?" if you flip a coin 100 times, you'll almost certainly get a streak of 6 or 7 heads somewhere. that's not a pattern — it's expected randomness. this connects directly to [[probability-in-daily-life|probability]] — distinguishing signal from noise is fundamentally a probabilistic question.
+
+## mental math tricks
+
+some useful estimation heuristics:
+
+**the rule of 72**: to find how long it takes money to double at x% interest, divide 72 by x. at 6% interest, money doubles in ~12 years. this works because ln(2) ≈ 0.693, and 72 has many factors making division easy.
+
+**dimensional analysis**: if your answer has the wrong units, it's wrong. this catches a surprising number of errors. "speed = distance × time" — nope, the units don't work out. must be distance / time.
+
+**anchor and adjust**: start with something you know and adjust. "how tall is that building?" well, each floor is about 3 meters, I count 15 floors, so ~45 meters. you're using [[arithmetic-everywhere|arithmetic]] and visual estimation together.
+
+**break it down**: any complex estimation becomes tractable when you decompose it into simpler estimates. this is the core of fermi estimation — you might be wrong on individual factors, but errors tend to cancel when you multiply several rough estimates together.
+
+## the connection to abstraction
+
+estimation is where [[counting-and-measurement|layer 1]] thinking meets [[the-organizational-lens|layer 3]] thinking. the act of estimating forces you to build a mental model: what are the relevant quantities? how do they relate? what's the structure of the problem?
+
+this is exactly what harrison means by "the organizational lens" — you're not just computing, you're *structuring your understanding* of a situation. a good fermi estimate reveals the key parameters of a system, which parameters matter most (sensitivity analysis), and which you can safely ignore. that's mathematical thinking at its most practical. pattern recognition itself is what [[abstraction-as-power|abstraction]] formalizes — noticing that two different situations share the same structure is both the essence of estimation and the essence of abstract mathematics.
\ No newline at end of file
new file mode 100644
index 0000000..14f615b
@@ -0,0 +1,70 @@
+---
+visibility: public-edit
+---
+
+# probability in daily life
+
+"should I bring an umbrella?" that's a probability question. "is this email a scam?" probability. "should I take the highway or surface streets?" probability. you're a probabilistic reasoning engine whether you know it or not — the question is whether you're a good one.
+
+## bayesian reasoning without knowing it
+
+you get a text from an unknown number: "hey, it's me, I got a new phone." who is it?
+
+your brain immediately does bayesian inference:
+- **prior**: who would text you? friends, family. probably not your dentist.
+- **likelihood**: who recently complained about their phone? who writes "hey, it's me" vs "hi, this is [name]"?
+- **posterior**: combining these, you make a guess.
+
+this is bayes' theorem: P(hypothesis | evidence) ∝ P(evidence | hypothesis) × P(hypothesis). you update your beliefs based on new information, weighted by how likely that information would be under each hypothesis.
+
+bayes' theorem isn't just a formula — it's a *thinking tool*. it tells you that evidence that's equally likely under all hypotheses is worthless. a medical test that comes back positive for everyone doesn't tell you anything. the value of evidence is its ability to *distinguish* between possibilities.
+
+## base rate neglect
+
+a disease affects 1 in 10,000 people. a test for it is 99% accurate. you test positive. what's the probability you have the disease?
+
+most people say 99%. the real answer: about 1%.
+
+why? because 99% accuracy means 1% false positive rate. if you test 10,000 people, 1 has the disease (true positive) and 100 get false positives. so you're 1 out of 101 positive results — roughly 1%.
+
+base rate neglect — ignoring the prior probability — is one of the most common reasoning errors humans make. it affects medical diagnosis, criminal justice ("the DNA matches!"), and hiring ("they passed the coding test!"). the math is elementary. the mistake is universal.
+
+## the monty hall problem
+
+you're on a game show. three doors. one has a car, two have goats. you pick door 1. the host (who knows what's behind each door) opens door 3, revealing a goat. should you switch to door 2?
+
+yes. switching gives you a 2/3 chance of winning.
+
+this problem is famous because it's genuinely counterintuitive. the key insight: the host's action gives you information. when you first picked, you had a 1/3 chance of being right. that means there's a 2/3 chance the car is behind one of the other doors. the host eliminated one of those doors for you, concentrating the 2/3 probability onto the remaining door.
+
+the deeper lesson: new information should change your beliefs. refusing to update ("I'll stick with my choice") is not rationality — it's stubbornness. this connects directly to bayesian reasoning: your posterior should change when you get new evidence.
+
+## expected value
+
+should you buy a $2 lottery ticket with a 1-in-10-million chance at $5 million?
+
+expected value: (1/10,000,000) × $5,000,000 = $0.50. you're paying $2 for $0.50 of expected value. mathematically, it's a bad deal.
+
+but expected value isn't everything. the *utility* of $5 million isn't 2,500× the utility of $2,000. money has diminishing marginal utility — the first million matters way more than the fifth. this is why insurance makes mathematical sense: you pay more than the expected loss because the *disutility* of a catastrophic loss is much worse than the disutility of small premiums.
+
+this same logic explains why you should wear a seatbelt (low-probability event, catastrophic downside), diversify investments (reduce variance even at the cost of expected return), and why casinos always win (they play the expected value game over thousands of bets; you only play once).
+
+## risk assessment
+
+humans are terrible at assessing risk intuitively:
+- we overestimate dramatic risks (plane crashes, shark attacks, terrorism)
+- we underestimate mundane risks (car accidents, heart disease, falls)
+- we treat very small probabilities as zero (until it happens to us)
+- we treat very small probabilities as large when the outcome is vivid (nuclear meltdown)
+
+the fix isn't to "stop being biased" — it's to do the math. [[patterns-and-estimation|estimation]] skills help here: if you can get the order of magnitude right, you're already ahead of most people's intuitions.
+
+## independence and conditional probability
+
+two dice rolls are independent — the first doesn't affect the second. but many real-world events that *feel* independent aren't. your probability of getting a job depends on the economy, which depends on interest rates, which depend on inflation, which depends on supply chains, which depend on geopolitics.
+
+the gambler's fallacy — "I've flipped 5 heads in a row, so tails is due" — is an error about independence. the coin doesn't remember its history. but in many real situations, history *does* matter: a basketball player who's made their last 5 shots might genuinely have a higher probability of making the next one (hot hand is real, it turns out, though weaker than people think).
+
+understanding when events are truly independent vs when they're correlated is one of the most practically important probabilistic skills. [[engineering-and-modeling|engineering and modeling]] relies heavily on getting these dependencies right — in monte carlo simulation, assuming independence when events are correlated gives wildly wrong results.
+
+when probability moves from discrete events (coin flips, dice) to continuous distributions (heights, temperatures, stock prices), you need [[calculus-as-thinking|calculus]] — continuous probability is built on integration, and concepts like probability density functions only make sense through the lens of limits and infinitesimals. the [[counting-and-measurement|measurement]] of probability itself raises deep questions about what we're actually quantifying.
\ No newline at end of file
new file mode 100644
index 0000000..44a47cc
@@ -0,0 +1,45 @@
+---
+visibility: public-edit
+---
+
+# applications of math
+
+many people say math is useless. "fine, addition and multiplication are useful, but I learned those early. why is calculus useful?"
+
+this wiki is my answer. math applies to everything — not just in the obvious "you need it for engineering" way, but in a deeper way: advanced math gives you ways of thinking that restructure how you see the world. the biggest use of advanced math is the ability to organize and structure things.
+
+the wiki has three layers, from the most immediately practical to the most profoundly structural.
+
+---
+
+## layer 1: the immediate and basic
+
+the math you use every day without calling it math. these seem trivial, but they're everywhere — and getting them wrong has real consequences.
+
+- [[counting-and-measurement|counting and measurement]] — all of science, commerce, and engineering starts with measuring things. precision vs accuracy. units as a type system. measurement theory.
+- [[arithmetic-everywhere|arithmetic everywhere]] — addition, subtraction, multiplication, division. so embedded in daily life that we forget they're math. compound interest. dosage calculations. cooking ratios.
+- [[ordering-and-comparison|ordering and comparison]] — sorting, ranking, prioritizing. total and partial orders. arrow's impossibility theorem. elo ratings. pareto optimality.
+- [[probability-in-daily-life|probability in daily life]] — "should I bring an umbrella?" bayesian reasoning. base rate neglect. the monty hall problem. expected value. risk assessment.
+- [[patterns-and-estimation|patterns and estimation]] — fermi estimation. order-of-magnitude reasoning. mental math. pattern recognition and its failures.
+
+## layer 2: the STEM foundation
+
+the classic "math is the language of science" argument. true and important, but well-known — so these pages are concise.
+
+- [[physics|physics]] — the original applied math. newton's laws as differential equations. quantum mechanics as linear algebra. the unreasonable effectiveness of mathematics.
+- [[computer-science|computer science]] — boolean algebra → circuits → computers. algorithms and complexity. cryptography. information theory. machine learning.
+- [[engineering-and-modeling|engineering and modeling]] — math modeling competitions. monte carlo simulation. optimization. spectral methods. "all models are wrong but some are useful."
+- [[biology-and-medicine|biology and medicine]] — population dynamics. epidemiology. genetics. EEG signal processing. medical statistics.
+
+## layer 3: the profound and structural
+
+this is the most interesting layer — and the one most people miss. abstract math doesn't just solve problems. it gives you **ways of thinking about fundamental patterns**. each branch provides a lens.
+
+- [[the-organizational-lens|the organizational lens]] — my core thesis. advanced math's biggest use is as an organizational tool for seeing structure in everyday things.
+- [[calculus-as-thinking|calculus as thinking]] — not computing derivatives. thinking about change, accumulation, limits, and continuity as universal patterns.
+- [[linear-algebra-as-thinking|linear algebra as thinking]] — every idea as a vector. every process as a matrix. eigenvalues as stability. semantic space and word embeddings.
+- [[set-theory-as-thinking|set theory as thinking]] — unions, intersections, MECE partitions. binary classification as set membership. the discipline of precise categorization.
+- [[multivariable-calculus-as-thinking|multivariable calculus as thinking]] — gradient (which direction?), divergence (source or sink?), curl (circulation?). gradient descent and optimization.
+- [[abstraction-as-power|abstraction as power]] — why abstraction is useful. the abstraction ladder. category theory. the trade-off between generality and specificity.
+- [[topology-as-thinking|topology as thinking]] — invariants. what stays the same under deformation. connectedness. fundamental groups. topological data analysis.
+- [[symmetry-and-groups|symmetry and groups]] — group theory. noether's theorem. the rubik's cube. music theory. why symmetric solutions tend to be elegant.
\ No newline at end of file
new file mode 100644
index 0000000..d17848b
@@ -0,0 +1,60 @@
+---
+visibility: public-edit
+---
+
+# biology and medicine
+
+biology resisted mathematization longer than physics. living systems are messy, stochastic, and deeply complex. but math is steadily becoming biology's most powerful tool — and in medicine, mathematical errors are measured in lives.
+
+## population dynamics
+
+the lotka-volterra equations model predator-prey interactions:
+- prey grow exponentially when predators are scarce
+- predators grow when prey are abundant
+- the system oscillates: more prey → more predators → fewer prey → fewer predators → more prey → ...
+
+this is a pair of coupled differential equations, and the oscillating solution explains real population cycles (like the famous lynx-hare cycle in canadian fur trapping records). the math predicts the qualitative behavior — boom and bust — without knowing anything about the specific animals. [[calculus-as-thinking|calculus]] is the language here: growth, decay, rates of change, equilibrium points — all the core concepts show up in biological modeling.
+
+more sophisticated models handle competition, mutualism, migration, and age structure. conservation biology uses these to predict extinction risk and design nature reserves.
+
+## epidemiology
+
+the SIR model (susceptible → infected → recovered) is the foundation of epidemic modeling. it's three differential equations:
+- dS/dt = -βSI (susceptible people get infected at a rate proportional to contact with infected)
+- dI/dt = βSI - γI (infected people either infect others or recover)
+- dR/dt = γI (recovered people are immune)
+
+the basic reproduction number R₀ — how many people one infected person infects on average — determines whether an epidemic grows or dies out. R₀ > 1 means epidemic; R₀ < 1 means it fizzles.
+
+during COVID, everyone suddenly cared about these models. "flatten the curve" was a mathematical statement: reduce β (through masking, distancing) to keep the infection peak below hospital capacity. the math was simple; getting people to act on it was the hard part.
+
+## genetics and bioinformatics
+
+DNA is a string over a 4-letter alphabet (A, T, C, G). comparing DNA sequences is a string-matching problem. finding genes is a pattern-recognition problem. building evolutionary trees is a graph theory problem.
+
+the human genome project was as much a computational/mathematical achievement as a biological one. sequence alignment algorithms (like BLAST) use dynamic programming — a mathematical technique — to compare your DNA sequence against billions of known sequences in seconds.
+
+CRISPR guide RNA design, protein structure prediction (AlphaFold), and drug-target interaction modeling are all mathematical problems at their core.
+
+## EEG and brain signal processing
+
+in my research on anesthetics and brain monitoring, the raw data is EEG signals — electrical voltage measurements from electrodes on the scalp. the raw signal is a mess: brain activity plus muscle artifacts, eye blinks, electrical noise.
+
+the math pipeline:
+1. **fourier analysis** decomposes the signal into frequency bands (delta 0.5-4 Hz, theta 4-8 Hz, alpha 8-13 Hz, beta 13-30 Hz, gamma 30+ Hz)
+2. **filtering** removes artifacts and noise
+3. **classification** — in our case, a CNN trained on spectrogram images to detect depth of anesthesia
+
+the same [[engineering-and-modeling|signal processing techniques]] used in engineering — fourier transforms, wavelets, spectral methods — are critical here.
+
+the goal: can we tell from brain signals alone how deeply anesthetized a patient is? too light and they might wake up during surgery. too deep and you risk complications. the math turns a subjective clinical judgment into an objective measurement — exactly the [[counting-and-measurement|counting and measurement]] problem, but for consciousness.
+
+## medical statistics
+
+clinical trials are [[probability-in-daily-life|probability]] in its highest-stakes application. does this drug work, or did we get lucky with our sample? p-values, confidence intervals, randomization, blinding — the entire machinery of evidence-based medicine is statistical.
+
+and the errors are consequential. p-hacking (running many statistical tests until one comes out significant) has contributed to a replication crisis across biomedical research. base rate neglect in diagnostic testing (see [[probability-in-daily-life|probability]]) leads to unnecessary procedures and missed diagnoses. getting the math right literally saves lives.
+
+## the deep point
+
+biology is where math meets the messiest, most complex systems we know. the models are always dramatically simplified — a cell is not a differential equation, and a brain is not a neural network. but the simplifications reveal structure that would be invisible otherwise. the SIR model doesn't capture every detail of epidemic spread, but it explains *why* epidemics have the shape they do. that explanatory power — seeing the pattern through the noise — is what math brings to biology.
\ No newline at end of file
new file mode 100644
index 0000000..052845e
@@ -0,0 +1,55 @@
+---
+visibility: public-edit
+---
+
+# computer science
+
+the digital world is built on math. not approximately, not metaphorically — literally. every computer, every app, every website is a tower of mathematical abstractions from boolean logic at the bottom to algorithms at the top.
+
+## boolean algebra → circuits → computers
+
+in 1937, claude shannon's master's thesis showed that boolean algebra (true/false, AND/OR/NOT) could be implemented with electrical switches. this single insight is the foundation of all digital computing. every transistor in your phone is computing a boolean function. a modern chip has billions of them, all doing math.
+
+the instruction your CPU is executing right now — reading a value from memory, comparing two numbers, jumping to a different part of the program — is a sequence of boolean operations on binary numbers. [[arithmetic-everywhere|arithmetic]] at the hardware level. and boolean algebra is really [[set-theory-as-thinking|set theory]] in disguise — AND is intersection, OR is union, NOT is complement. the entire digital world rests on set operations.
+
+## algorithms and complexity
+
+an algorithm is a precise mathematical procedure for solving a problem. the key question isn't just "can we solve it?" but "how fast?"
+
+complexity theory classifies problems by how their difficulty scales:
+- **O(1)**: constant time. looking up an element in a hash table. doesn't matter how big the table is.
+- **O(log n)**: binary search. searching a sorted list of a million items takes ~20 steps.
+- **O(n)**: linear. reading every element once.
+- **O(n log n)**: sorting. the best you can do for comparison-based sorting (connects to [[ordering-and-comparison|ordering]]).
+- **O(n²)**: quadratic. naive pairwise comparison. starts hurting around n = 10,000.
+- **O(2ⁿ)**: exponential. brute-force search. at n = 100, there are more possibilities than atoms in the universe.
+
+the P vs NP problem — are there problems that are easy to verify but hard to solve? — is one of the millennium prize problems. a million dollars if you solve it. most computer scientists believe P ≠ NP, which means there are inherently hard problems that no clever algorithm can make easy.
+
+## cryptography
+
+the security of the internet rests on a mathematical asymmetry: multiplying two large prime numbers is easy, but factoring the result back into primes is (believed to be) hard.
+
+RSA encryption works because of this. your browser uses it right now. the math is number theory — a branch of "pure" math that was considered the most useless branch of mathematics for centuries. G.H. Hardy, the famous number theorist, wrote proudly that his work had no practical applications. he was wrong — it now secures trillions of dollars in transactions.
+
+this is a beautiful example of [[abstraction-as-power|abstraction as power]]: the most abstract math becoming the most practical.
+
+## information theory
+
+shannon (again) founded information theory in 1948. the key idea: information can be measured mathematically, in bits. a fair coin flip is 1 bit of information. a die roll is log₂(6) ≈ 2.58 bits.
+
+shannon's theorems set fundamental limits on communication: there's a maximum rate at which you can transmit information through a noisy channel, and no encoding scheme can beat it. every wifi standard, every cell phone protocol, every streaming video codec is engineered to approach these mathematical limits.
+
+the connection to [[probability-in-daily-life|probability]]: information theory is deeply connected to probability. the "surprise" of an event is -log₂(probability). unlikely events carry more information than likely ones. "the sun rose today" is low-information; "a meteor hit the earth" is high-information. entropy — the average surprise — measures uncertainty.
+
+## machine learning and linear algebra
+
+modern AI is mostly [[linear-algebra-as-thinking|linear algebra]] and [[calculus-as-thinking|calculus]]. a neural network is a sequence of matrix multiplications and nonlinear functions. training is gradient descent — [[multivariable-calculus-as-thinking|multivariable calculus]] applied to a loss function with millions of parameters.
+
+the semantic space example from my essay is a direct application: word embeddings represent meanings as vectors, and the geometry of the vector space captures semantic relationships. "woman" + "king" - "man" ≈ "queen" works because the vector arithmetic in embedding space mirrors conceptual relationships.
+
+[[topology-as-thinking|topological data analysis]] is an emerging tool — using persistent homology to find the "shape" of high-dimensional datasets that linear methods miss.
+
+## the deep point
+
+computer science demonstrates that math isn't just about numbers — it's about computation, information, and the fundamental limits of what can be known, computed, and communicated. some problems are provably unsolvable (the halting problem). some are solvable but take longer than the age of the universe. understanding these limits is itself a mathematical achievement.
\ No newline at end of file
new file mode 100644
index 0000000..de02306
@@ -0,0 +1,55 @@
+---
+visibility: public-edit
+---
+
+# engineering and modeling
+
+"all models are wrong, but some are useful." — george box
+
+this is the most important sentence in applied math. a model is a mathematical simplification of reality. it's wrong by definition — reality is messier than any equation. but if it captures the right aspects of reality, it's useful. engineering is the art of building models that are wrong in the right ways.
+
+## math modeling competitions
+
+i've competed in HiMCM and MCM/ICM (the high school and college math modeling competitions). these are 4-5 day events where you get an open-ended problem — something like "model the spread of an invasive species" or "optimize emergency evacuation routes" — and have to build a mathematical model from scratch.
+
+the process teaches you something textbooks don't: choosing what to model and what to ignore. a model that includes everything is just reality, and you can't solve reality. the skill is identifying the 3-4 variables that matter most and writing down the simplest equations that capture their relationships.
+
+[[patterns-and-estimation|estimation]] is the first step — sanity-checking that your model gives reasonable outputs before you trust it with anything important. and every model output needs to be evaluated against [[counting-and-measurement|measurement]] — precision vs accuracy, units, and whether the numbers even make physical sense.
+
+## monte carlo simulation
+
+some problems are too complex for exact solutions. monte carlo methods handle them by running thousands of random simulations and looking at the statistical results.
+
+want to estimate π? throw random darts at a unit square. count how many land inside the inscribed circle. the ratio approaches π/4. want to price a financial option? simulate thousands of possible stock price paths and average the payoffs. want to assess flood risk? generate thousands of possible weather scenarios and see how often the levee fails.
+
+monte carlo is powerful because it converts analytical difficulty into computational cost. you don't need to solve the equation — you just need to simulate it enough times. this connects to [[probability-in-daily-life|probability]]: monte carlo is basically "run the experiment a million times in silico."
+
+## optimization
+
+most engineering problems are optimization problems: minimize cost, maximize efficiency, find the best design subject to constraints.
+
+linear programming alone (optimizing a linear function subject to linear constraints) underlies scheduling, supply chain management, and resource allocation across every major industry. the simplex method, which solves these, is one of the most practically important algorithms ever invented.
+
+nonlinear optimization is harder — you might get stuck in local minima. [[multivariable-calculus-as-thinking|gradient descent]] is the workhorse method: follow the slope downhill. it's how neural networks are trained, how engineers design aircraft wings, and how economists find market equilibria.
+
+## spectral methods
+
+the fourier transform decomposes any signal into a sum of sine waves at different frequencies. this is the mathematical backbone of:
+- audio processing (your music is stored as frequency data, not raw waveforms)
+- image compression (JPEG uses a related transform, the DCT)
+- signal processing (filtering noise from an EEG or radio signal)
+- structural engineering (resonance analysis — at what frequency does a bridge vibrate dangerously?)
+
+in my EEG research for the anesthetics paper, signal processing was critical. brain signals are noisy — you need to extract meaningful patterns from messy electrical data. fourier analysis and wavelet transforms let you decompose the signal into frequency bands (delta, theta, alpha, beta, gamma) that correspond to different brain states. the CNN then learns patterns in these transformed signals.
+
+## differential equations in engineering
+
+heat transfer, fluid dynamics, structural mechanics, circuit analysis, chemical reactions — all governed by differential equations. the navier-stokes equations (fluid flow) are so important and so hard that solving them is another millennium prize problem.
+
+engineers rarely solve these equations analytically. they discretize them — break continuous space into a mesh of points — and solve the resulting system numerically. finite element analysis, finite difference methods, and computational fluid dynamics are all "convert a continuous math problem into a very large but solvable discrete math problem."
+
+this is where [[calculus-as-thinking|calculus]] meets [[linear-algebra-as-thinking|linear algebra]]: discretizing a differential equation turns it into a matrix equation. the continuous meets the discrete, and the computer handles the rest.
+
+## the deep point
+
+engineering is where mathematical abstraction crashes back into physical reality. the model has to work. the bridge has to stand. the prediction has to be close enough. this creates a productive tension between mathematical elegance and practical necessity — and it's why the modeling mindset ("what can I ignore? what matters most? how wrong is acceptable?") is one of the most transferable skills math teaches. modeling is [[abstraction-as-power|abstraction]] applied under constraints — you need the right level of simplification, and the penalty for getting it wrong is measured in collapsed bridges, not lost marks.
\ No newline at end of file
new file mode 100644
index 0000000..7a3e794
@@ -0,0 +1,47 @@
+---
+visibility: public-edit
+---
+
+# physics
+
+physics is math's oldest and most dramatic application. the same equations that describe a ball rolling down a hill also describe the orbit of mercury around the sun, the vibration of a guitar string, and the probability of a particle tunneling through a barrier.
+
+## the unreasonable effectiveness of mathematics
+
+in 1960, physicist eugene wigner wrote a famous essay asking why math works so well for describing physics. he called it "the unreasonable effectiveness of mathematics in the natural sciences." it's a genuine mystery.
+
+consider: mathematicians developed non-euclidean geometry in the 1800s as a pure intellectual exercise, asking "what if parallel lines could meet?" decades later, einstein needed exactly that math for general relativity. the universe, it turned out, actually uses non-euclidean geometry. the math was ready and waiting.
+
+this keeps happening. complex numbers were invented to solve polynomial equations. they turned out to be essential for quantum mechanics. group theory was developed as abstract algebra. it turned out to describe the fundamental symmetries of particle physics — [[symmetry-and-groups|symmetry and groups]] explains how noether's theorem connects every symmetry to a conservation law, one of the deepest results in all of physics. mathematicians keep building tools that physicists later discover the universe was using all along.
+
+## newton's laws are differential equations
+
+F = ma looks simple. but a = d²x/dt² — acceleration is the second derivative of position with respect to time. so newton's second law is really a differential equation:
+
+F = m(d²x/dt²)
+
+this single equation, given a force law, tells you the entire future trajectory of a particle. you throw a ball: gravity provides F = -mg, and solving the differential equation gives you the parabolic trajectory. every physics problem in mechanics is "set up the differential equation, then solve it."
+
+[[calculus-as-thinking|calculus as thinking]] is born directly from this: position, velocity, and acceleration are related by derivatives. physics gave calculus its first and most natural application.
+
+## maxwell's equations
+
+four equations describe all of electricity and magnetism. just four. and when maxwell wrote them down, he noticed they predicted electromagnetic waves traveling at the speed of light. "light is an electromagnetic wave" was a mathematical prediction before it was an experimental fact.
+
+this is the pattern: write down the math, follow where it leads, discover new physics. the math doesn't just describe what we already know — it predicts things we haven't seen yet.
+
+## quantum mechanics is linear algebra
+
+the state of a quantum system is a vector in a hilbert space. observables are matrices (operators). measurement is projection. the probability of an outcome is the squared magnitude of a component. entanglement is a tensor product.
+
+if you know [[linear-algebra-as-thinking|linear algebra]], you already have the mathematical framework for quantum mechanics. a quantum state is literally a vector. the schrödinger equation is a linear differential equation. superposition is vector addition. the entire weirdness of quantum mechanics — superposition, entanglement, measurement — is encoded in the mathematics of vectors and matrices.
+
+## general relativity is differential geometry
+
+gravity isn't a force — it's the curvature of spacetime. mass tells spacetime how to curve; curvature tells matter how to move. the math behind this is riemannian geometry: metrics, tensors, curvature, geodesics.
+
+einstein's field equations relate the curvature of spacetime (a geometric quantity) to the distribution of matter and energy (a physical quantity). solving these equations predicted gravitational lensing, black holes, and gravitational waves — all confirmed experimentally, decades after the math predicted them. this is where [[multivariable-calculus-as-thinking|multivariable calculus]] reaches its most extreme form — the field equations are systems of coupled nonlinear partial differential equations in curved spacetime.
+
+## the deep point
+
+physics doesn't just *use* math — it suggests that math *is* the language of reality. the laws of physics aren't written in english or chinese; they're written in differential equations, symmetry groups, and geometric structures. whether this means the universe "is" mathematical (the Mathematical Universe Hypothesis) or whether math is just an unreasonably good model, nobody knows. but the track record is staggering: every time physicists have followed the math, even when it led to absurd-sounding predictions, the predictions turned out to be right.
\ No newline at end of file
new file mode 100644
index 0000000..fa1fe84
@@ -0,0 +1,63 @@
+---
+visibility: public-edit
+---
+
+# abstraction as power
+
+a triangle has three sides. the number three has nothing to do with triangles — it's an abstraction. but once you have the concept of "three," you can apply it to apples, people, dimensions, and anything else. you've traded specificity for generality, and gained enormous power in the exchange.
+
+mathematics is, at its core, the discipline of productive abstraction.
+
+## the abstraction ladder
+
+consider how abstraction builds on itself:
+
+1. **counting objects**: 3 apples, 5 oranges (concrete)
+2. **numbers**: 3, 5 — stripped of what they're counting (first abstraction)
+3. **arithmetic**: 3 + 5 = 8 — rules for combining numbers (abstraction of operations)
+4. **algebra**: a + b = c — letters replacing specific numbers (abstraction of arithmetic)
+5. **functions**: f(x) = x² — abstraction of relationships
+6. **spaces**: vector spaces, topological spaces — abstraction of geometry
+7. **category theory**: the mathematics of mathematical structures — abstraction of abstraction
+
+each level loses some detail and gains generality. at level 1, you know you're talking about apples. at level 7, you might be talking about anything — but the structural insights apply to everything at once.
+
+## the trade-off
+
+abstraction has a cost: the more abstract you go, the harder it is to connect back to concrete reality. this is why abstract math feels "useless" — the connection to specific applications is indirect.
+
+but the trade-off is worth it when the same structure appears in many different domains. [[linear-algebra-as-thinking|linear algebra]] works for physics, data science, economics, and natural language processing because the abstraction (vector spaces) captures structure shared by all these domains. if you'd stayed concrete — "column of numbers" — you'd never see the connections.
+
+the skill is knowing when to abstract and when to stay concrete. abstract too early and you're doing math for math's sake, disconnected from reality. abstract too late and you're solving the same problem over and over without realizing it's the same problem.
+
+## numbers and shapes
+
+one of the oldest and most beautiful connections in math: numbers and geometry are secretly the same thing.
+
+- 1 corresponds to a point
+- 2 corresponds to a line segment
+- 3 corresponds to a triangle
+- 4 corresponds to a tetrahedron
+
+the number line puts numbers in geometric space. the coordinate plane connects algebra and geometry (every equation is a shape; every shape is an equation). this connection — algebra ↔ geometry, the discrete ↔ the continuous — runs through all of mathematics.
+
+descartes' invention of the coordinate plane was an abstraction breakthrough: it showed that geometry and algebra, which had been separate fields for 2,000 years, were the same thing viewed from different angles. one abstraction unified two worlds.
+
+## category theory: the mathematics of mathematics
+
+category theory asks: what do all mathematical structures have in common? it studies objects (which could be sets, groups, spaces, anything) and arrows between them (which could be functions, transformations, continuous maps, anything that preserves structure).
+
+this sounds absurdly abstract, and it is. but it's also powerful:
+- it reveals when two seemingly different mathematical constructions are "really the same" (natural isomorphism)
+- it identifies universal patterns that recur across all branches of math (products, coproducts, limits, colimits)
+- it provides a language for talking about mathematical structure itself
+
+in [[computer-science|computer science]], category theory has found practical applications: functional programming languages like Haskell use categorical concepts (monads, functors) as programming abstractions. a monad in Haskell is the same mathematical object as a monad in category theory — the abstraction crosses from pure math to software engineering.
+
+## the deep point
+
+abstraction is not escape from reality — it's compression of reality. a good abstraction captures the essential structure and discards the irrelevant details. the power of mathematics is that its abstractions are *remarkably good* at capturing the structures that matter.
+
+this is the core of the [[the-organizational-lens|organizational lens]]: each branch of math provides an abstraction layer, a way of seeing structure. [[set-theory-as-thinking|sets]] abstract classification. [[calculus-as-thinking|calculus]] abstracts change. [[linear-algebra-as-thinking|linear algebra]] abstracts direction and transformation. [[topology-as-thinking|topology]] abstracts connectivity. [[symmetry-and-groups|groups]] abstract symmetry.
+
+the question isn't "is this abstraction useful?" — it's "is this the right level of abstraction for the problem at hand?" too concrete and you're lost in details. too abstract and you're lost in generality. the sweet spot — where the abstraction reveals structure without obscuring specifics — is where mathematics does its best work.
\ No newline at end of file
new file mode 100644
index 0000000..19a471f
@@ -0,0 +1,59 @@
+---
+visibility: public-edit
+---
+
+# calculus as thinking
+
+calculus is not about computing derivatives and integrals. those are the exercises, not the ideas. the ideas are about change, accumulation, limits, and continuity — and they're thinking tools that apply to everything.
+
+## the derivative: how things change
+
+the derivative answers: "how fast is this changing right now?"
+
+not "how much did it change over the last hour" — that's an average. the derivative is the instantaneous rate. the speedometer reading, not the trip average. the slope of the curve at a point, not the slope of the line connecting two points.
+
+this distinction matters everywhere:
+- a company's revenue grew 20% last year (average). but is it accelerating or decelerating *right now*? that's the derivative.
+- a patient's fever is 102°F (a measurement). but is it rising or falling? that's the derivative.
+- unemployment is 4% (a level). but is the rate of change positive or negative? and is the rate of change itself changing? that's the second derivative.
+
+the second derivative is the rate of change of the rate of change. when people say "things are getting worse more slowly," they're describing a negative first derivative with a positive second derivative. calculus gives you the vocabulary to be precise about change.
+
+## the integral: accumulation
+
+the integral answers: "what does all this add up to?"
+
+if the derivative tells you velocity at each moment, the integral tells you total distance traveled. if the derivative tells you income rate, the integral tells you total earnings. if the derivative tells you flow rate, the integral tells you total volume.
+
+the fundamental theorem of calculus says these two operations — differentiation and integration — are inverses. knowing how something changes tells you where it ends up. knowing where it ends up tells you (with some ambiguity) how it changed. this duality between rates and totals is one of the deepest ideas in all of mathematics.
+
+in practice, integration is "adding up infinitely many infinitely small things." the area under a curve is a sum of infinitely thin rectangles. this idea — decomposing something continuous into infinitesimal pieces, handling each piece, then adding up — is a thinking pattern that extends far beyond math. any time you're analyzing a continuous process by breaking it into tiny steps, you're thinking in integrals.
+
+## limits: approaching without arriving
+
+the limit asks: "what happens as we get infinitely close?"
+
+you can never divide by zero. but you can ask what happens as the denominator approaches zero. you can never reach infinity. but you can ask what happens as n gets larger and larger.
+
+limit thinking is about the trend, not the destination. it's about asymptotic behavior: what does this system look like in the long run? does it converge (settle down) or diverge (blow up)? does it oscillate? does it approach something it never reaches?
+
+this is directly useful for thinking about:
+- diminishing returns: each additional hour of study helps less than the last. the learning approaches a limit.
+- convergence in iterative processes: will this negotiation converge to an agreement, or will it diverge?
+- asymptotic analysis in [[computer-science|computer science]]: how does this algorithm behave as the input gets very large?
+
+## continuity: small changes → small effects
+
+a continuous function is one where small changes in input produce small changes in output. no jumps, no teleportation.
+
+most real-world systems are approximately continuous, and we rely on this constantly. if you turn the steering wheel a tiny bit, the car turns a tiny bit (not 90 degrees). if you add a grain of salt to a soup, the taste changes imperceptibly (it doesn't become inedible). continuity is the mathematical version of "the world is predictable."
+
+when continuity breaks — phase transitions, tipping points, market crashes — we're surprised precisely because we were assuming continuity. the water was getting warmer and warmer (continuous) and then suddenly it's boiling (discontinuous). recognizing where continuity assumptions fail is a crucial thinking skill.
+
+## the connection to other layers
+
+[[counting-and-measurement|measurement]] gives you a snapshot. calculus tells you the story — how things are changing, where they're heading, what they'll add up to. it's the mathematical upgrade from static to dynamic thinking.
+
+[[physics|physics]] is where calculus was born: Newton invented it to describe motion. but the thinking patterns — rates, accumulation, limits, continuity — are universal. any time you're reasoning about change, you're doing calculus, whether or not you write down an equation. in [[biology-and-medicine|biology]], calculus describes population growth, enzyme kinetics, and the spread of disease — every differential equation in the SIR model is calculus applied to living systems.
+
+when the change happens in multiple dimensions simultaneously, you need [[multivariable-calculus-as-thinking|multivariable calculus]] — gradients, divergence, curl — which extends these ideas into the full complexity of real-world systems. and when calculus meets [[linear-algebra-as-thinking|linear algebra]] — as it does in finite element analysis, neural network training, and dynamical systems — the two frameworks reinforce each other: linear algebra provides the structure, calculus provides the motion.
\ No newline at end of file
new file mode 100644
index 0000000..af81d24
@@ -0,0 +1,73 @@
+---
+visibility: public-edit
+---
+
+# linear algebra as thinking
+
+linear algebra is the mathematics of directions, transformations, and projections. every idea can be a vector. every process can be a matrix. and the deep structures — eigenvalues, rank, nullspace — reveal things about systems that are invisible from any other angle.
+
+## everything is a vector
+
+a vector is just a list of numbers — but that framing undersells it. a vector is a *point in a space*, and the space can represent anything:
+
+- a color is a vector in RGB space: (255, 128, 0) = orange
+- a location is a vector in geographic space: (40.7, -74.0) = new york
+- a person's test scores are a vector: (SAT: 1520, GPA: 3.9, AP count: 12)
+- a word's meaning is a vector in semantic space: "king" = (0.2, -0.5, 0.8, ...)
+
+once something is a vector, you can do vector operations on it. you can compute distances (how similar are these two things?), averages (what's the "center" of this group?), and projections (how much of this is in that direction?). at its core, this is [[arithmetic-everywhere|arithmetic]] lifted into higher dimensions — addition, subtraction, scaling — but now operating on objects far richer than single numbers.
+
+the semantic space example from my essay: word embeddings represent meanings as vectors in 300-dimensional space. the dot product of two word vectors measures their semantic similarity. and vector arithmetic captures analogies:
+
+vector("king") - vector("man") + vector("woman") ≈ vector("queen")
+
+the direction from "man" to "king" (roughly: royalty) is the same as the direction from "woman" to "queen." this isn't a coincidence or a trick — it reflects genuine structure in how language encodes meaning. linear algebra made that structure visible and computable.
+
+## every process is a matrix
+
+a matrix is a linear transformation — it takes vectors in one space and maps them to vectors in another. but the conceptual reach is much broader:
+
+- google's PageRank algorithm represents the web as a matrix (pages × links) and finds the dominant eigenvector
+- a markov chain represents state transitions as a matrix: what's the probability of going from state A to state B?
+- a rotation, reflection, or scaling in any dimension is a matrix multiplication
+- a system of equations is a matrix equation: Ax = b
+
+the power is that all of these — web ranking, probability, geometry, equation-solving — use the same mathematical machinery. once you know how to work with matrices, you can work with all of them.
+
+## eigenvalues: what stays stable under transformation
+
+this is the deepest idea in linear algebra. an eigenvector of a matrix is a direction that doesn't change under the transformation — it just gets scaled. the eigenvalue tells you by how much.
+
+why does this matter? because eigenvectors reveal the *essential behavior* of a system:
+
+- in PageRank, the dominant eigenvector is the steady-state importance of each page
+- in principal component analysis (PCA), eigenvectors are the directions of maximum variance in data — the "most important" dimensions
+- in dynamical systems, eigenvalues tell you whether the system is stable (|λ| < 1), unstable (|λ| > 1), or oscillating (complex λ)
+- in quantum mechanics, eigenvalues of an operator are the possible measurement outcomes
+
+eigenvectors answer: "what is the intrinsic structure of this transformation?" strip away coordinates, strip away representation — what's left is the eigenstructure.
+
+## projection: seeing components
+
+projecting a vector onto a direction extracts "how much of this is in that direction." in 3D, projecting onto the x-axis gives you the x-component.
+
+but projection is conceptually richer than that:
+- in statistics, linear regression is projection: you're projecting data onto the best-fit line
+- in signal processing, fourier analysis is projection: you're projecting a signal onto each frequency component (see [[engineering-and-modeling|engineering]])
+- in machine learning, dimensionality reduction is projection: compressing high-dimensional data into a lower-dimensional space while preserving the most important structure
+
+any time you're decomposing something into components, you're projecting.
+
+## rank and dimensionality
+
+the rank of a matrix tells you how many independent dimensions the transformation actually uses. a 1000×1000 matrix might have rank 3 — meaning all that apparent complexity lives in just 3 dimensions.
+
+this is profound for data analysis. real-world data is almost always lower-dimensional than it appears. a dataset with 100 features might effectively live in a 5-dimensional subspace. finding that subspace (via PCA, SVD, or other methods) is one of the most powerful techniques in data science and [[computer-science|computer science]] more broadly.
+
+the concept generalizes: when something seems complex, ask "what's the effective dimensionality?" how many independent factors actually drive this system? that's the rank. this move — from surface complexity to underlying simplicity — is a form of [[abstraction-as-power|abstraction]]: ignoring the noise to find the signal.
+
+## the deep point
+
+linear algebra is the mathematics of structure in spaces. once you start seeing things as vectors in spaces — meanings, states, data points, possibilities — you gain access to a powerful toolkit: distances, transformations, decompositions, and the deep structural insights that eigenvalues and rank provide.
+
+[[the-organizational-lens|the organizational lens]] applied here asks: "what are the dimensions? what are the transformations? what stays stable? what's the effective dimensionality?" these questions apply to far more than matrices.
\ No newline at end of file
new file mode 100644
index 0000000..a425a83
@@ -0,0 +1,70 @@
+---
+visibility: public-edit
+---
+
+# multivariable calculus as thinking
+
+single-variable [[calculus-as-thinking|calculus]] asks: how does one thing change with respect to another? multivariable calculus asks: how does one thing change with respect to *many* others simultaneously? since almost nothing in the real world depends on just one variable, this is where calculus becomes truly powerful.
+
+## the gradient: which direction is uphill?
+
+the gradient of a function at a point is a vector that points in the direction of steepest increase. its magnitude tells you how steep.
+
+this is one of the most useful concepts in all of mathematics:
+
+- **gradient descent** in machine learning: to minimize a loss function with millions of parameters, compute the gradient (which direction increases the loss fastest) and step in the opposite direction. this is how every neural network is trained. the math is: go downhill.
+- **optimization** in any domain: if you have a function measuring "goodness" that depends on multiple knobs, the gradient tells you which knob to turn and by how much. this applies to engineering design, portfolio allocation, experimental parameter tuning — anything with a quantifiable objective.
+- **sensitivity analysis**: the partial derivatives (components of the gradient) tell you which variables matter most. if ∂f/∂x is huge and ∂f/∂y is tiny, then f is very sensitive to x and barely affected by y. this answers the crucial question: "what should I focus on?"
+
+the gradient is the mathematical answer to "where should I put my effort?" — aim where the slope is steepest.
+
+## divergence: is stuff flowing in or out?
+
+divergence measures whether a point in a vector field is a source (stuff flows out), a sink (stuff flows in), or neither.
+
+in [[physics|physics]], divergence is everywhere:
+- a positive charge creates a divergence in the electric field (field lines flow outward)
+- a drain in a bathtub is a sink for the water velocity field
+- an incompressible fluid has zero divergence everywhere (what flows in must flow out)
+
+metaphorically:
+- a company with positive cash flow divergence is generating more money than it's spending (a source)
+- a team losing members faster than it's hiring has negative "people divergence" (a sink)
+- a stable system has zero divergence — inputs balance outputs
+
+## curl: is there rotation?
+
+curl measures the tendency of a vector field to circulate around a point.
+
+physically:
+- a whirlpool has curl (the water rotates)
+- a tornado has curl (the air circulates)
+- a current-carrying wire creates curl in the magnetic field
+
+metaphorically:
+- circular dependencies in a project (A depends on B, B depends on C, C depends on A) are a kind of curl
+- feedback loops in systems — positive feedback spirals, negative feedback stabilizes — have a rotational quality
+- any cyclical process (boom-bust cycles, predator-prey oscillations from [[biology-and-medicine|biology]]) has the flavor of curl
+
+## surface and line integrals: adding up along paths and surfaces
+
+a line integral adds up a quantity along a path. how much work does a force do as an object moves along a curve? that's a line integral. how much does a stock price change over a trading day? if you model the instantaneous changes, the total is a line integral.
+
+a surface integral adds up a quantity over a surface. how much fluid flows through a net? how much heat radiates from a surface? these are surface integrals.
+
+the deep theorems of multivariable calculus — green's theorem, stokes' theorem, the divergence theorem — connect these: they say that what happens on the boundary determines what happens in the interior (and vice versa). this is a profound structural insight: you can understand the inside by studying the outside.
+
+## the connection to optimization and ML
+
+modern machine learning is, mathematically, multivariable calculus applied to very high-dimensional spaces. a large language model might have 175 billion parameters. the loss function lives in 175-billion-dimensional space. gradient descent navigates this space by computing (approximately) the gradient and following it downhill.
+
+the challenges are all multivariable calculus challenges:
+- **saddle points**: not every place where the gradient is zero is a minimum. some are saddle points (minimum in some directions, maximum in others). in high dimensions, saddle points vastly outnumber true minima.
+- **conditioning**: if the loss surface is much steeper in some directions than others (bad conditioning), gradient descent oscillates instead of converging. this is why techniques like Adam and learning rate scheduling matter.
+- **local vs global**: gradient descent finds local minima, not necessarily the global minimum. in high dimensions, though, local minima tend to be nearly as good as the global minimum — a fact that's still not fully understood theoretically.
+
+## the deep point
+
+single-variable calculus is about change in one dimension. multivariable calculus is about change in the real world — where everything depends on everything else, where you need to find the best direction in a high-dimensional space, and where the relationships between boundary and interior reveal deep structural truths. much of [[engineering-and-modeling|engineering and modeling]] comes down to setting up and solving multivariable calculus problems — from optimizing airfoil shapes to simulating fluid dynamics.
+
+the thinking tools — gradient (which direction?), divergence (source or sink?), curl (is there circulation?) — are metaphors that apply far beyond physics. any time you're navigating a complex system with multiple interacting variables, you're in the domain of multivariable calculus.
\ No newline at end of file
new file mode 100644
index 0000000..332bbc9
@@ -0,0 +1,82 @@
+---
+visibility: public-edit
+---
+
+# set theory as thinking
+
+a set is a collection of things. that's it. but from this absurdly simple starting point, you get a way of thinking that clarifies almost everything.
+
+## the basic operations
+
+**union** (A ∪ B): everything in A or B or both. combining categories. "people who like pizza OR people who like sushi" = "people who like at least one of pizza or sushi."
+
+**intersection** (A ∩ B): everything in both A and B. finding overlap. "people who like pizza AND sushi" = "people who like both."
+
+**difference** (A \ B): everything in A that's not in B. filtering. "people who like pizza but NOT sushi."
+
+**complement** (Aᶜ): everything not in A. the opposite. "people who don't like pizza."
+
+these four operations let you manipulate categories with precision. every database query (SQL's WHERE, JOIN, UNION, EXCEPT) is set operations on rows — a fact that makes [[computer-science|computer science]] fundamentally dependent on set-theoretic thinking. every search engine query (AND, OR, NOT) is set operations on documents.
+
+## MECE: mutually exclusive, collectively exhaustive
+
+this is the consulting framework, but it's really just a set theory concept: a **partition** of a set.
+
+a partition divides a set into subsets that are:
+- **mutually exclusive**: no element belongs to two subsets (A ∩ B = ∅)
+- **collectively exhaustive**: every element belongs to some subset (A ∪ B ∪ C ∪ ... = the whole set)
+
+why does McKinsey care about this? because sloppy categorization causes sloppy thinking. if your market segments overlap, you'll double-count customers. if they don't cover everyone, you'll miss a segment. MECE is the antidote.
+
+examples of MECE partitions:
+- people by age bracket: 0-17, 18-34, 35-54, 55+ (no overlap, no gaps)
+- revenue by product line (if every dollar comes from exactly one product)
+- time by day of the week
+
+examples of NON-MECE breakdowns:
+- "students, athletes, and musicians" — a student-athlete-musician is in all three (not mutually exclusive)
+- "domestic and international customers" — what about customers with dual presence? (ambiguous)
+
+the discipline of asking "is this MECE?" catches a huge number of analytical errors.
+
+## venn diagrams
+
+venn diagrams are the visual language of set theory. two or three overlapping circles can clarify relationships that paragraphs of text obscure.
+
+but venn diagrams have limits. with 4+ sets, they become unwieldy. and they encourage you to think about small, discrete categories when sometimes the right framework is continuous (a spectrum, not a partition). knowing when to use set-theoretic thinking and when to use continuous thinking (more like [[calculus-as-thinking|calculus]] or [[linear-algebra-as-thinking|linear algebra]]) is itself a useful skill.
+
+## binary classification as set membership
+
+the most basic question in set theory: "is x in the set S?" yes or no.
+
+this is exactly the binary classification problem in machine learning: is this email spam or not? is this transaction fraudulent or not? is this patient sick or not? the entire field of classification is about drawing boundaries in some feature space that define the set of "positive" examples.
+
+connections to [[probability-in-daily-life|probability]]: in practice, you rarely get a hard yes/no. you get a probability — "this email is 87% likely to be spam." the boundary between "in the set" and "out of the set" becomes fuzzy. fuzzy set theory formalizes this.
+
+## set operations on ideas
+
+the most powerful application of set-theoretic thinking is to ideas themselves:
+
+- what's the **intersection** of "things I'm good at" and "things the world needs"? (that's roughly your career sweet spot)
+- what's the **union** of all the skills needed for this project? who on the team covers what?
+- what's the **difference** between what we planned and what we delivered? (the gap)
+- is our feature list **MECE** with respect to user needs? are there gaps? are there overlaps?
+
+any time you're categorizing, comparing categories, or checking for gaps and overlaps, you're doing set theory. the formal language just makes it precise.
+
+## russell's paradox and the limits of naive sets
+
+consider the set of all sets that don't contain themselves. does it contain itself?
+
+- if it does, then by definition (it contains sets that *don't* contain themselves) it shouldn't
+- if it doesn't, then by definition it should
+
+this paradox (bertrand russell, 1901) destroyed naive set theory and forced mathematicians to build more careful foundations (ZFC axioms). the lesson: even the simplest-seeming concepts — "a collection of things" — can harbor contradictions if you're not careful about definitions.
+
+this is why definitions matter. not pedantically, but fundamentally. sloppy definitions lead to paradoxes, and paradoxes lead to wrong conclusions. set theory teaches you to be precise about what you mean by a category — and that precision is valuable far beyond mathematics.
+
+## the deep point
+
+set theory is the mathematics of classification. the impulse to classify connects to [[ordering-and-comparison|ordering and comparison]] — you can't sort things into sets without first deciding what makes them similar or different. taken further, the study of how sets relate to each other through notions of "nearness" and continuity leads to [[topology-as-thinking|topology]], which generalizes set theory into the study of shape and space.
+
+in a world drowning in information, the ability to categorize precisely — to define boundaries, check for completeness, identify overlaps, and reason about relationships between categories — is one of the most practical mathematical skills there is. this kind of precise categorical reasoning is itself a form of [[abstraction-as-power|abstraction]]: stripping away particulars to focus on structure. you don't need to know the axiom of choice to benefit from thinking in sets.
\ No newline at end of file
new file mode 100644
index 0000000..aa68700
@@ -0,0 +1,75 @@
+---
+visibility: public-edit
+---
+
+# symmetry and groups
+
+symmetry is the most fundamental organizing principle in mathematics and physics. group theory — the mathematics of symmetry — reveals why symmetric solutions tend to be elegant, why conservation laws exist, and why a rubik's cube has exactly 43,252,003,274,489,856,000 possible states.
+
+## what is a group?
+
+a group is a set with an operation that satisfies four properties:
+1. **closure**: combining two elements gives another element in the set
+2. **associativity**: (a·b)·c = a·(b·c)
+3. **identity**: there's an element that does nothing (like 0 for addition, 1 for multiplication)
+4. **inverses**: every element has an inverse that undoes it
+
+that's it. four properties. and from these four properties, an enormous amount of structure follows.
+
+examples:
+- the integers under addition: the identity is 0, the inverse of 5 is -5
+- rotations of a square: 0°, 90°, 180°, 270°. the identity is 0°. the inverse of 90° is 270°.
+- permutations: all possible rearrangements of n objects, with composition as the operation
+
+## symmetry = invariance under transformation
+
+a square has 8 symmetries: 4 rotations and 4 reflections. each symmetry is a transformation that leaves the square looking the same. the *group* of these 8 symmetries (called D₄) captures everything about the square's symmetry in an algebraic structure.
+
+this definition of symmetry — invariance under a group of transformations — is enormously general:
+- a circle has infinitely many rotational symmetries (the rotation group SO(2))
+- the laws of physics are symmetric under translation (the same experiment gives the same result in new york and tokyo) and rotation (it doesn't matter which direction you face)
+- a musical chord is symmetric under octave transposition (a C major chord sounds "the same" an octave higher)
+
+## noether's theorem: symmetry → conservation
+
+emmy noether proved one of the most beautiful theorems in [[physics|physics]]: every continuous symmetry of a physical system corresponds to a conservation law.
+
+- **translational symmetry** (physics doesn't depend on where you are) → conservation of momentum
+- **rotational symmetry** (physics doesn't depend on your orientation) → conservation of angular momentum
+- **time translation symmetry** (physics doesn't depend on *when* you do the experiment) → conservation of energy
+
+this is extraordinary. it means conservation laws aren't empirical accidents — they're mathematical consequences of symmetry. energy is conserved *because* the laws of physics don't change over time. this is physics at its most elegant: a deep truth derived from pure math.
+
+## the rubik's cube
+
+the rubik's cube is a group theory problem in disguise. each move is a permutation of the cube's pieces. the set of all possible states, with cube moves as the operation, forms a group.
+
+the group has 43 quintillion elements, but it's generated by just 6 basic moves (one for each face). every possible state can be reached by some sequence of these 6 moves, and every state can be solved (returned to the identity) in at most 20 moves. that number — 20 — is called "God's number," and proving it required both group theory and massive computation.
+
+the rubik's cube teaches a key group theory lesson: a complex system with an astronomical number of states can be understood through its *generators* (the basic moves) and *relations* (how moves interact). you don't need to enumerate all 43 quintillion states — you need to understand the structure.
+
+## music theory
+
+music has deep group-theoretic structure:
+- the 12 notes of the chromatic scale form a cyclic group Z₁₂ (after 12 half-steps, you're back where you started)
+- transposition (shifting a melody up or down) is a group action on this cyclic group
+- inversion (flipping a melody upside-down) and retrograde (playing it backwards) are group operations that composers use
+- chords can be classified by their symmetry group: a diminished seventh chord has more symmetry (it's invariant under transposition by 3 half-steps) than a major triad
+
+this isn't just a mathematical curiosity — it explains *why* certain musical structures sound the way they do. highly symmetric chords (diminished, augmented) sound ambiguous and unstable. asymmetric chords (major, minor) have a clear tonal center. symmetry determines character.
+
+## symmetry in problem-solving
+
+symmetric problems tend to have symmetric solutions. this is both a mathematical theorem (in many contexts) and a practical heuristic:
+
+- if a problem is symmetric in x and y, try solutions where x = y
+- if a function is even (f(-x) = f(x)), its fourier series has only cosine terms
+- if a physical system has spherical symmetry, look for spherically symmetric solutions first
+
+exploiting symmetry reduces complexity. a problem with n-fold symmetry effectively becomes n times smaller. this is why physicists love symmetry: it's not just aesthetically pleasing — it's computationally essential.
+
+## the deep point
+
+group theory shows that symmetry isn't just a visual property — it's an algebraic structure with deep consequences. symmetry constrains possibilities (noether's theorem), simplifies analysis (symmetry reduction), and reveals hidden connections (the same group appearing in different contexts means the same symmetry is at work).
+
+the [[the-organizational-lens|organizational lens]] here is: look for symmetries. what transformations leave the system unchanged? what's conserved? what simplifications does symmetry allow? these questions, powered by group theory, are among the most productive questions you can ask about any structured system. and the connection to [[topology-as-thinking|topology]] runs deep — both fields study invariants under transformation, and [[linear-algebra-as-thinking|representation theory]] bridges them by expressing group symmetries as linear transformations.
\ No newline at end of file
new file mode 100644
index 0000000..2d3f20a
@@ -0,0 +1,38 @@
+---
+visibility: public-edit
+---
+
+# the organizational lens
+
+> "for me, advanced math has great utility. the biggest use is the ability to organize and structure things."
+
+i wrote that in a chinese class essay about 无用之用 — the usefulness of the useless. the prompt was about why math matters, and my answer surprised me: it's not about computing things. it's about seeing things.
+
+a lot of advanced math — [[calculus-as-thinking|calculus]], [[linear-algebra-as-thinking|linear algebra]], [[set-theory-as-thinking|set theory]] — is really just an organizational lens applied to very normal things. velocity is calculus applied to position. a recommendation engine is linear algebra applied to preferences. a venn diagram is set theory applied to categories. the math doesn't create the structure — it reveals structure that was already there.
+
+## the vector space example
+
+the clearest example i know: semantic space. take a word — any word — and represent its meaning as a vector. now you can do math on meanings.
+
+dot product tells you how similar two meanings are. vector addition creates new meanings: "woman" + "king" - "man" = "queen." that's not a trick — it's [[linear-algebra-as-thinking|linear algebra]] applied to language, and it works because language has geometric structure that was invisible until someone thought to look for it.
+
+this is what word embeddings do. this is what makes modern AI work. and it's fundamentally an organizational insight: meanings have directions, and those directions live in a space you can navigate mathematically.
+
+## what the lens does in practice
+
+when i'm doing math modeling for [[engineering-and-modeling|HiMCM or MCM/ICM]], the hardest part is never solving the equations. it's figuring out which equations to write. that's the organizational lens — looking at a messy real-world problem (fire evacuation, drone routing, bus equity) and asking: what kind of structure does this have?
+
+- is it changing over time? → [[calculus-as-thinking|calculus]]
+- does it have multiple interacting dimensions? → [[multivariable-calculus-as-thinking|multivariable calculus]]
+- is it about categories and membership? → [[set-theory-as-thinking|set theory]]
+- does it have directions, transformations, stability? → [[linear-algebra-as-thinking|linear algebra]]
+- does it have symmetry? → [[symmetry-and-groups|group theory]]
+- does it have shape that matters more than measurement? → [[topology-as-thinking|topology]]
+
+answering "what kind of problem is this?" correctly is more than half the work. the rest is technique. [[patterns-and-estimation|pattern recognition]] gets you the first foothold — the rough shape before you formalize. the organizational lens is what turns that rough shape into a mathematical frame.
+
+## beyond math
+
+the organizational lens isn't limited to math problems. every time i look at a social situation and think "this is a coordination problem, not a motivation problem" — that's the lens. every time i look at a bug and think "this is a state problem, not a logic problem" — that's the lens. every time i frame a decision as "reversible vs irreversible" instead of "risky vs safe" — that's the lens.
+
+math teaches you to name the structure of problems. once you can name it, you can solve it. the naming is the hard part.
\ No newline at end of file
new file mode 100644
index 0000000..f22ecad
@@ -0,0 +1,73 @@
+---
+visibility: public-edit
+---
+
+# topology as thinking
+
+topology asks: what stays the same when you stretch, bend, and deform — but don't cut or glue?
+
+a coffee cup and a donut are topologically the same: they each have one hole. you can continuously deform one into the other without cutting or gluing. a sphere and a cube are topologically the same: zero holes, no matter how you squish or stretch.
+
+this sounds like a joke ("a topologist can't tell the difference between a coffee cup and a donut"). but the underlying idea — caring about structural properties that survive deformation — is one of the most powerful concepts in mathematics.
+
+## invariants: what doesn't change?
+
+the central question of topology is: given a transformation, what properties are *invariant*?
+
+the number of holes is a topological invariant. so is connectedness (is the shape one piece or multiple pieces?). so is orientability (does the surface have an "inside" and "outside," or is it like a möbius strip?).
+
+the thinking pattern generalizes far beyond geometry:
+
+- in business: what stays the same about your company when you change markets, products, or team members? the invariants are your actual competitive advantages.
+- in identity: what stays the same about *you* when your circumstances change? those invariants are your core values or personality traits.
+- in systems design: what properties of your system must be preserved under scaling, modification, or failure? those are your architectural invariants.
+
+identifying invariants is one of the most useful skills in abstract reasoning — it's the same instinct that drives [[symmetry-and-groups|group theory]], which asks what transformations preserve structure. it tells you what's essential vs what's accidental.
+
+## connectedness and paths
+
+a topological space is *connected* if you can get from any point to any other without leaving the space. the integers are not connected (there are gaps). the real numbers are connected (no gaps).
+
+*path-connectedness* is stronger: not just "no gaps" but "there exists a continuous path between any two points."
+
+this is directly relevant to:
+- **network analysis**: is your social network connected? can information flow from any person to any other? if not, what are the connected components?
+- **data clustering**: connected components in a similarity graph naturally define clusters
+- **reachability**: in a state machine or graph, can you get from state A to state B? this is a topological question
+
+## fundamental groups and higher structure
+
+in my algebraic topology course, we study the fundamental group — which classifies the "essentially different" loops you can draw in a space.
+
+on a plane, every loop can be shrunk to a point: the fundamental group is trivial. on a torus (donut surface), loops that go "around the hole" or "through the hole" can't be shrunk: the fundamental group is Z × Z (two independent types of loops).
+
+the fundamental group detects holes that you can't see by just looking at the space locally. this is the power of algebraic topology: converting geometric questions ("does this space have holes?") into algebraic questions ("what's the group structure?") that can be computed.
+
+higher homotopy groups detect higher-dimensional holes. homology gives you a computable version. cohomology gives you even more structure. it's a tower of increasingly sophisticated invariants.
+
+## topological data analysis
+
+topological data analysis (TDA) applies these ideas to real data. given a point cloud (a set of data points), TDA asks: what's the "shape" of this data?
+
+persistent homology — the main tool of TDA — tracks how topological features (connected components, loops, voids) appear and disappear as you vary a scale parameter. features that persist across many scales are "real" structure; features that appear briefly are noise.
+
+this has been applied to:
+- analyzing the shape of protein structures
+- detecting patterns in neural activity
+- understanding the structure of high-dimensional datasets that other methods miss
+
+## continuity and deformation
+
+the topological mindset is: don't get distracted by exact measurements. focus on structure.
+
+two problems are "topologically equivalent" if you can continuously deform one into the other. this means the specific numbers don't matter — what matters is the qualitative shape.
+
+this is liberating. instead of asking "what is the exact answer?" you ask "what *kind* of answer is it?" instead of "how big is the hole?" you ask "is there a hole at all?" instead of measuring precisely, you classify structurally.
+
+this connects to [[set-theory-as-thinking|set theory]] (classification by type) and contrasts with [[calculus-as-thinking|calculus]] (exact measurement of change). topology says: before you compute, understand the shape of the problem.
+
+## the deep point
+
+topology teaches you to ask: what's essential? what survives deformation? what are the invariants? these questions — applied to problems, organizations, systems, or ideas — cut through surface complexity to reveal underlying structure.
+
+the coffee cup = donut insight isn't trivial. it's a lesson in [[abstraction-as-power|abstraction]]: sometimes two things that look completely different are structurally identical, and sometimes two things that look similar are structurally different. topology gives you the tools to tell which is which.
\ No newline at end of file