Cram

Podcast 2 — In conversation

Speaker A: I spent the whole night cramming for the technical interview — algorithms, data structures, system design patterns. By the morning I was fairly confident I'd forgotten half of it already.

Speaker B: Cramming for a technical interview — the universal experience. The self-aware addition that you'd forgotten half of it already is precisely what the research predicts: cramming produces fragile, quickly decaying recall. Spaced practice across weeks produces far better retention, but most of us reach for the cram session when the deadline is tomorrow.

Speaker A: The schedule is absolutely crammed — back-to-back meetings from eight in the morning until six. There's not even a break for lunch built in.

Speaker B: Crammed schedule as an overloaded system — the calendar equivalent of the overstuffed suitcase. The no lunch break detail is the tell: when a schedule is crammed, the margins disappear first. And it is precisely those margins — the five minutes between meetings, the lunch break for thinking — that allow a mind to process and integrate what it has been doing.

Speaker A: We managed to cram twelve people into a room designed for six. It was a productive meeting, somehow, despite the elbows.

Speaker B: Cram as physical overcrowding — twelve into a space for six is the classic cram scenario. The somehow productive despite the elbows is very human: extreme physical proximity can paradoxically increase the intensity and focus of a meeting, removing all the ambient distractions of a comfortable space.

Speaker A: People sometimes confuse cramming and revision.

Speaker B: Revision is the general term for reviewing and consolidating learning ahead of an exam — it can be planned, spaced, and systematic. Cramming is a specific, intensive, last-minute form of revision that relies on proximity to the exam rather than quality of learning. All cramming is revision in the broad sense; careful, spaced revision is the opposite of cramming. Cram implies both the intensity and the lateness.

Speaker A: And cram is different from pack. To pack is to fill efficiently — putting things in so they fit well. To cram is to fill excessively — forcing things in beyond comfortable capacity. Packing is systematic; cramming is desperate. You pack for a holiday; you cram at the last minute when you realise you've packed too much.

Podcast 3 — Prompt engineering

Speaker B: Scenario one — a schedule capacity prompt.

Design a meeting schedule analyser that detects crammed calendars. The analyser reviews a user's calendar and flags: days where more than 80 percent of working hours are booked (crammed days), sequences of back-to-back meetings with no buffer time between them, meetings scheduled during lunch hour, and weeks where every day is crammed. For each crammed pattern found, the analyser suggests: where buffer time could be inserted by shortening meetings to 45 or 25 minutes instead of 60 or 30, which meetings could be declined or delegated, and what the cost of a crammed schedule is in terms of decision fatigue and reduced thinking time.

Speaker B: 25 and 45-minute meeting suggestions instead of 30 and 60 minutes is the Parkinson's Law intervention — meetings expand to fill the time allotted, and trimming 5 minutes creates buffer without reducing meeting quality.

Speaker A: Cost of a crammed schedule in terms of decision fatigue is the qualitative harm that most calendar tools don't surface — knowing that all your decisions after 3pm on a crammed day are lower quality is actionable information.

Speaker B: Next — a database storage capacity prompt.

Build a database storage capacity monitoring system that detects crammed tables and schemas. The system monitors PostgreSQL databases and alerts when: a table's row count is within 10 percent of the integer primary key limit, a varchar column's average length is within 20 percent of the column definition maximum (suggesting the column definition will need widening), a database instance's storage is above 80 percent capacity, and an index is crammed with duplicate values suggesting poor selectivity. For each crammed resource, the alert includes the current usage percentage, the projected time to limit at current growth rate, and the recommended remediation.

Speaker A: Varchar column average length within 20 percent of maximum is the silent cram that breaks applications without warning — when an application tries to insert a value longer than the column allows, it fails at runtime, not at design time.

Speaker B: Right. Now an exam preparation app prompt.

Design a smart exam preparation app that helps students move from cramming to spaced repetition. The app: detects cramming behaviour (all study sessions concentrated in the 48 hours before an exam), shows the user a comparison of predicted retention after cramming versus spaced practice using a forgetting curve visualisation, creates a spaced repetition study schedule from any exam date working backwards over the available weeks, and tracks whether the user is following the spaced schedule or falling back into cram patterns. An encouragement message appears when the user completes a scheduled session rather than cramming.

Speaker B: Forgetting curve visualisation comparing cramming versus spaced practice is the evidence-based motivation tool — seeing the predicted retention difference visually makes the abstract research immediately personal and persuasive.

Speaker B: Fourth — a content management prompt.

Write a content density analyser for a web page design tool that detects crammed layouts. A crammed layout is one where: the information density per screen area exceeds a readable threshold, there is insufficient white space between content elements, paragraphs contain more than 100 words without visual breaks, navigation menus contain more than 7 items, and hero sections attempt to communicate more than one primary message. For each crammed element identified, the analyser suggests: the content to remove or move to a secondary page, the white space to introduce, and the visual hierarchy change that would allow the most important content to breathe.

Speaker A: Navigation menus with more than 7 items as a cram signal is Miller's Law applied to UX — cognitive chunking limits human working memory to roughly seven items, and a menu that exceeds this is crammed beyond the user's comfortable processing capacity.

Speaker B: Fifth — a train capacity management prompt.

Design a real-time train capacity management system that prevents carriages from being crammed beyond safe occupancy. The system: receives live passenger count data from weight sensors on each carriage, displays current occupancy as a percentage of comfortable capacity and maximum safety capacity on a station platform display, sends alerts to train operators when any carriage exceeds 80 percent of comfortable capacity, suggests redistribution advice to platform staff (move passengers to less crammed carriages), and feeds data to a journey planning app so passengers can choose less crammed carriages before boarding.

Speaker B: Comfortable capacity versus maximum safety capacity as two distinct thresholds is the important distinction — a carriage crammed to maximum safety capacity is legally permissible but miserable; the system targets the comfortable threshold to improve experience before the safety threshold becomes relevant.

Speaker B: Sixth — a full small app.

Build a single-page web app called Cram Buster — a study session planner that replaces last-minute cramming with structured spaced practice. The user enters a subject to study and an exam date. The app creates a study schedule spread across the available days with: early sessions covering new material, middle sessions reviewing earlier material at spaced intervals, and final sessions doing light review rather than intensive cramming. Each session is logged when completed. A cram warning appears if the user hasn't studied in 5 or more days as the exam approaches. A completion tracker shows progress through the schedule. All data in localStorage.

Speaker A: Cram warning after 5 days of no study as the exam approaches is the gentle intervention — not preventing cramming, but making the pattern visible so the user can make a conscious choice.

Speaker B: That's the lesson — cramming fills the container but what goes in rarely stays. Whether it is a suitcase, a schedule, a train carriage, or a memory, cramming is the solution of insufficient time applied to a space that cannot quite hold what we are demanding of it. Better systems — spaced repetition, buffer time, measured capacity — work with the container's natural limits rather than against them.

Speaker B: Prompt example: