Every model, one typed interface

In one sentence, explain what a neural network is.

A neural network is a computational model inspired by the human brain, consisting of interconnected layers of nodes (neurons) that process and learn patterns from data to make predictions or decisions.

Explain recursion with one analogy.

Recursion is when something solves a problem by doing a smaller version of the same thing. **Analogy:** Imagine standing between two mirrors facing each other. You see an image of yourself, and inside that image is another image of yourself, and inside that another, repeating smaller and smaller. That’s like recursion: a function “calls itself” on a smaller version of the problem, until it reaches a stopping point. - **Repeated pattern:** each mirror image contains another image - **Smaller version each time:** the images get smaller - **Stopping point:** eventually the images fade so you don’t keep noticing more In programming, recursion works the same way: 1. Do a small part of the work 2. Call the same function on a smaller problem 3. Stop at a **base case** so it doesn’t continue forever If you want, I can also show a tiny code example.

Explain why the sky is blue, in two sentences.

As sunlight enters Earth's atmosphere, gases and particles scatter the light in all directions. Blue light is scattered much more than other colors because it travels as shorter, smaller waves, causing the sky to appear blue to our eyes.

Explain caching using a kitchen analogy.

**Caching is like keeping your most-used ingredients on the counter instead of fetching them from the pantry every time.** In a kitchen, the pantry (or fridge) holds lots of food but is farther away and slower to access. The counter, spice rack, or a small “mise en place” area next to the stove holds a limited set of frequently needed items so you can grab them instantly while cooking. ### How it maps to computers - **Pantry / fridge** → Slow, large storage (disk, database, remote server). Plenty of capacity, but retrieving something takes time (I/O or network latency). - **Counter / spice rack** → Cache (CPU cache, RAM cache, CDN, browser cache, etc.). Small, fast, close to the “chef” (CPU or application). - **Chef cooking a dish** → Program or request that needs data. **Typical flow:** 1. You need salt (a piece of data). 2. Check the counter first (**cache lookup**). - **Hit**: It’s there → use it immediately. Fast. - **Miss**: Not there → walk to the pantry, get it, use it, and usually put a copy on the counter for next time (**populate the cache**). 3. Repeat. Over time the counter fills with what you actually use a lot. ### Extra kitchen details that match real caching - **Limited space**: The counter isn’t infinite. When it’s full you have to decide what to keep and what to put away (**eviction policy**). Common strategies: throw out the least recently used item (LRU), the oldest, or the least frequently used. - **Freshness**: Ingredients can go stale or get updated in the pantry. You might need to check whether the counter version is still good (**cache invalidation / TTL / consistency**). Sometimes you write changes to both places at once (**write-through**); sometimes you update the counter first and sync later (**write-back**). - **Warm-up**: A brand-new kitchen starts with an empty counter; the first few dishes are slower until frequently used items accumulate. - **Multiple levels**: Tiny ultra-fast area right by the stove (L1 cache), a slightly larger nearby rack (L2), then the pantry, then the grocery store (main memory → disk → network). The goal is the same in both worlds: avoid expensive trips for things you keep needing. Caching trades a bit of extra complexity and limited space for much lower latency on the common case. That’s the core idea—keep hot data close so the “cooking” (computation or serving requests) stays snappy.

Name three uses for a paper clip.

# Three Uses for a Paper Clip 1. **Fastening papers together** — The most common use; it holds multiple sheets in place without damaging them. 2. **Emergency straightened wire** — When straightened out, it can serve as a makeshift tool for cleaning small crevices, fixing electronics, or as a makeshift lock pick. 3. **Bookmark** — It can mark your place in a book or document without leaving any permanent marks.

Describe what an index does to a database query.

An index changes how a database query accesses data: instead of scanning an entire table, the database can use the index to find matching rows much faster. More specifically, an index is a separate data structure—commonly a B-tree, hash, or bitmap—that stores a sorted copy of one or more columns along with pointers back to the actual table rows. When you run a query, the database optimizer may choose to use an index for operations such as: - `WHERE` filters - `JOIN` conditions - `ORDER BY` sorting - `GROUP BY` aggregation - `DISTINCT` or uniqueness checks For example: ```sql SELECT * FROM employees WHERE last_name = 'Smith'; ``` Without an index on `last_name`, the database may perform a full table scan, reading every row to find matches. With an index on `last_name`, it can traverse the index structure to jump directly to the matching entries, reducing I/O and improving response time. In short, an index speeds up read queries by replacing full scans with efficient lookups, at the cost of slower writes and extra storage, because the index must be maintained whenever data is inserted, updated, or deleted.

Summarize photosynthesis in one line.

Photosynthesis is the process by which plants use sunlight to convert water and carbon dioxide into chemical energy (glucose) and oxygen.

Describe the water cycle in one sentence.

The water cycle is the continuous process in which water evaporates from the Earth's surface, condenses into clouds in the atmosphere, and falls back to the ground as precipitation.

Give a concise argument for testing public behavior.

Test public behavior because it’s the contract users depend on. If tests target internals, they become brittle: harmless refactors break tests even when the software still works. By testing observable inputs/outputs and side effects, you verify real correctness while preserving freedom to change the implementation.

How do I keep my sourdough starter alive over winter?

Give one vivid opening line for a story set in a lighthouse.

For three nights the lighthouse lamp burned blue, and on the fourth morning a drowned sailor knocked at the door.