Programming Fundamentals · Part Summary
Summary · Five Algorithm Ideas at a Glance
Twelve lessons done—time to close the net. This chapter really only taught five kinds of ideas, and each has a real face in AI. This page takes the whole chapter home in one big table, then uses 8 scenario questions to test whether you can “spot it at a glance”—see a problem and know which idea to reach for.
One big table · Five ideas × AI's real face
| Idea | One-line motto | Real face in AI | Related lessons |
|---|---|---|---|
| Complexity Big-O | First ask “what if data grows 10×” | Attention is O(n²): longer context, compute grows by the square—and so does the bill | algo-1algo-2 |
| Search and Sorting | Ordered → cut in half; unordered → sort first | Rerank: RAG-retrieved passages go through coarse ranking then fine ranking—sorting at heart | algo-3algo-4algo-5 |
| Recursion and Divide-and-Conquer | Break a big job into the same smaller job | Compaction: split a long chat, summarize each slice, merge—that's divide and conquer | algo-6algo-7 |
| Graph search BFS/DFS | Sweep layer by layer, or go all the way down one path | Coding Agent finding files: walk directories in a codebase, dig deep along reference chains | algo-8 |
| Greedy and Sampling | Pick the max each step, or roll the dice by probability | Temperature / Beam Search: AI's two personalities when picking words, and “look a few steps ahead” | algo-9algo-10 |
Spot it in one shot · 8 scenario questions
In real work nobody tells you “this is a binary-search problem.” Read the scenario, pick which idea to use, get instant feedback.
Two chapters side by side · Store × Process
🗃 Data Structures Part: how to store
Arrays, stacks, queues, hashes, caches, trees, graphs, vectors—eight ways to store, deciding where data lives and how you find it.
×
⚙️ Algorithms Part: how to process
Complexity, search & sorting, recursion & divide-and-conquer, graph search, greedy & sampling—five processing ideas, deciding how to compute and how fast.
Eight structures × five ideas = your full toolkit for reviewing AI-written code. See a stretch of AI-written code—first ask “where does it store the data” (structure), then “how does it plan to process it” (algorithm), then “what happens if data grows 100×” (complexity). Ask those three and the gap between “it runs” and “it ships” shows itself. You don't need to write it—from today on, you can see through it.
✅ What this chapter wants you to take away
- The whole chapter is only five ideas: complexity, search & sorting, recursion & divide-and-conquer, graph search, greedy & sampling
- Each has an AI real face: O(n²) attention, Rerank, Compaction, Agent finding files, Temperature
- Spotting beats solving: first recognize “whose turf is this,” and AI will give the answer
- Two chapters make one pair of glasses: structure handles storage, algorithms handle processing—together, the eye for reviewing AI-written code
- Next: the hands-on page: take a real AI code snippet and run a complexity checkup