← Computer Science I

Fall 2026

Agenda of Activities

Open a week, then open a day for its plan.

Placement: Weeks and themes come from the course calendar, which is the authority. A deck appears on a day only where an objective calls for it; anything built but not yet called for is listed at the bottom rather than given an invented slot.

Capture: Every activity captures to a Google Form (primary, school accounts) with an identical looseleaf backup — the teacher picks the medium each day.

▸Week 1 — No course content· Aug 4–7Per operating spec §2: no course content this week. One standalone activity for early finishers and staggered field-trip returns — no sequence, no group dependency, joinable at minute 40. Real content opens Aug 10.

The standalone activity is not placed on a specific day; it runs across the week for whoever is present.

▸Week 2 — Opening: instruments, then Unit 1 begins· Aug 10–14Opening — Aug 10–12, fixed (U1-01 through U1-05). Attitudinal chunk A cold on a locked Google Form Monday, cognitive benchmark Tue–Wed, chunks B and C follow. U1-06 onward opens Thursday.
▸MondayAug 10Routines, provisional AI rule — and the survey, cold on a locked Google Form (U1-01, U1-02)
  • AnchorDeck 1A — Course · Teacher · Expectations · PoliciesOpen deck
  • DeckThe provisional AI rule, stated and labeled provisional· the thing they are going to replacecoming soon
  • InstrumentAttitudinal survey — chunk A, locked Google Form, cold, before any instruction· editing on, resumable; partial response is acceptable and is not made upcoming soon
▸TuesdayAug 11Cognitive benchmark, Part 1 of 2 (U1-03)
  • InstrumentCognitive benchmark — Part 1 (50 items)· separate form; editing after submit off; permanent item codes; participation grade, completion onlycoming soon
  • FormForm — cognitive benchmark, Part 1coming soon
▸WednesdayAug 12Benchmark Part 2, survey chunks B + C, and what the surveys are for (U1-03, U1-04, U1-05)
  • InstrumentCognitive benchmark — Part 2 (36 items plus one unscored constructed response)· order between parts does not matter; total across P1 + P2 is 86 itemscoming soon
  • FormForm — cognitive benchmark, Part 2coming soon
  • InstrumentAttitudinal survey — chunks B and Ccoming soon
  • DiscussWhat the surveys are for: they recur in December, and these responses become the dataset the class analyzes later in the unit· identifiers stripped before the coding weekcoming soon
▸ThursdayAug 13Ordered material opens — why the AI in this room does not run on the Chromebook (U1-06, U1-07)
  • DeckThe AI in this room does not run on the Chromebook in front of youcoming soon
  • DeckHow a computer computescoming soon
▸Week 3 — Hardware and where the model lives· Aug 17–21VRAM · dedicated GPU · iGPU · unified memory · the two cards in the teacher's rig as the worked example · one request traced twice, local stack and commercial cloud · what each endpoint may do with what arrived (U1-08 through U1-12).

Activities coming soon.

▸Week 4 — How a model gets made, and what it reads· Aug 24–28Pretraining, post-training, freezing, inference · parameters · tokenization live in MuggsOfPrompts · why token count, word count, and character count disagree (U1-13 through U1-18). MuggsOfPrompts is the apparatus starting here — build need-by Aug 24, aligned to Week 3 opening.

Activities coming soon.

▸Week 5 — Lab 1: Three Answers, One Question· Aug 31–Sep 4One prompt to three models of different sizes, at the same moment, four days running. What the word better means changes four times, and the last change tests the previous three. Describe · correctness · criteria · stability. Lab Report 1 is a major, written on Sep 9 and 10. Fri Sep 4 is Test 1 and contains nothing from this week.
▸MondayAug 31Describe, do not judge — single pass, hedge, specificity
  • DeckDay 1 · I Do — Describe, do not judgeOpen deck
  • ActivityDay 1 · We Do — the run, then the observation form· three models, one prompt, single passOpen activity
  • FormObservation Day 1 — section 1 classwork, section 2 exit ticket· embedded in the We Do deck; no separate linkcoming soon
▸TuesdaySep 1Judge by correctness — claim, verify, source
  • DeckDay 2 · I Do — Better means correctOpen deck
  • ActivityDay 2 · We Do — three checkable claims per response· sources posted on the board, not the answersOpen activity
  • FormObservation Day 2 — section 1 classwork, section 2 exit ticket· embedded in the We Do deck; no separate linkcoming soon
▸WednesdaySep 2Judge with no answer key — criterion, subjective, consistent
  • DeckDay 3 · I Do — When there is nothing to checkOpen deck
  • ActivityDay 3 · We Do — the four criteria, scored against every pane· ranking copied into the lab notebook; Day 4 cannot run without itOpen activity
  • FormObservation Day 3 — section 1 classwork, section 2 exit ticket· embedded in the We Do deck; no separate linkcoming soon
▸ThursdaySep 3Does the ranking hold — run, variation, stable
  • DeckDay 4 · I Do — Does it holdOpen deck
  • ActivityDay 4 · We Do — yesterday's prompt run twice more, same criteria· a ranking that collapses and is reported honestly earns full marksOpen activity
  • FormObservation Day 4 — section 1 classwork, section 2 exit ticket· embedded in the We Do deck; no separate linkcoming soon
▸FridaySep 4Test 1 — covers Aug 10 through Aug 28
  • InstrumentTest 1 — major grade· paper; Chromebooks stored, phones collected; nothing from this week is on itcoming soon
▸Week 6 — Lab 1 close-out, then Unit 1 closes· Sep 8–11Mon Sep 7 Labor Day and Tue Sep 8 School-Site PD are both out, so this week is two content days. Both go to Lab 1 close-out: the report is written in class across the three days. Nothing closes at a bell — the forms stay open so a student who was out, or who is not finished, can keep going. That displaces U1-27 through U1-33, which had been projected across these same two days. The AI-experience discussion set below is retained but has no block left this week — see the calendar note on U1-C39.
▸WednesdaySep 9Notes are not a report — the body: Tuesday, Wednesday, Thursday
  • DeckSep 9 · I Do — Notes are not a report· observation vs. narrative, and the audience who was not thereOpen deck
  • ActivitySep 9 · We Do — the report body form, with Days 1–4 open beside it· revising, not rewriting; submit each before moving onOpen activity
  • FormLab Report 1 · Body — three sections, one per prompt day; the exit ticket is its own form· embedded in the We Do deck; no separate linkcoming soon
▸ThursdaySep 10A claim, its evidence, and its limit — the introduction and conclusion
  • DeckSep 10 · I Do — A claim, its evidence, and its limitOpen deck
  • ActivitySep 10 · We Do — Intro, Body, Conclusion on tabs; today Intro then Conclusion· Lab Report 1 — in-class check pass; the forms stay openOpen activity
  • FormLab Report 1 · Intro form + Conclusion form; the exit ticket is its own form· embedded in the We Do deck; no separate linkcoming soon
▸Across the weekSep 8–11The AI-experience discussion set — the corpus material for U1-34 through U1-C39
  • FilmThe AI Doc — Segment 1 of 4: The Stakes· ~26 min
  • DiscussDeck — The AI Doc, Part 1: The StakesOpen deck
  • FilmThe AI Doc — Segment 2 of 4: The Choice· ~26 min
  • DiscussDeck — The AI Doc, Part 2: The ChoiceOpen deck
  • FilmThe AI Doc — Segment 3 of 4: The Answer· ~26 min
  • DiscussDeck — The AI Doc, Part 3: The AnswerOpen deck
  • FilmThe AI Doc — Segment 4 of 4: Closing· ~26 min
  • DiscussDeck — The AI Doc, Part 4: ClosingOpen deck
  • DeckThe Learning Zone — ZPD and the MIT study· builds capability vs. substitutes for it (U1-38)Open deck
  • DeckYour Brain on AIOpen deck
  • AuthorWrite the reflection: how you have experienced AI so far (U1-34)· this writing is the corpuscoming soon
  • AuthorWrite the closing statement: how you intend to use AI academically (U1-35)coming soon
  • FormForm — the corpus (U1-34, U1-35)coming soon
▸Week 7 — Unit 2 opens: the rotation on paper· Sep 14–18Mon Sep 14 is Lab 1 close-out Day 4 — a focused writing lab and the last class time on Lab Report 1 (grading run Thu Sep 17 afternoon). Then Unit 2: restate the class AI policy · solve in silence, compare, rotate, judge twice (correctness vs. teachability) · name the parameters of the process the class just ran (U2-01 through U2-07). No software this week — this is what gives Unit 2 its floor.
▸MondaySep 14Better means more checkable, not longer — the last class day on Lab Report 1
  • DeckSep 14 · I Do — Better means more checkable, not longer· revise · checkable · stranger; second Do Now slide is the last call and the Thu grading runOpen deck
  • ActivitySep 14 · We Do — the writing lab: Intro, Body, Conclusion on tabs· read each form as a stranger, attach the evidence, submit; exit ticket in the last 20 minutesOpen activity

Unit 2 activities coming soon.

▸Week 8 — Reliability, adjudication, and the pipeline design· Sep 21–25Reliability scores from seeded items · weighted consensus · adjudication routing · MuggsOfSources opens: state a claim before any evidence, vet a source against a checklist (U2-08 through U2-13). MuggsOfSources need-by Sep 21.

Activities coming soon.

▸Week 9 — Sources: attestation, extraction damage, and where the error enters· Sep 28–Oct 2Locate every passage in the original PDF and attest to it · quotes-by-selection · why PDF-to-text extraction can produce fluent sentences the author never wrote · a machine can be entirely scrupulous and still hand you something false (U2-14 through U2-21). Q1 benchmark window.

Activities coming soon.

▸Week 10 — Unit 2 closes: model runs through the pipeline· Oct 5–8RAG as a pipeline, and which stage this tool omits · client-side extraction · run the model through the pipeline the class built and publish the accuracy · defend a source session orally · unassisted checkpoint (U2-22 through U2-C28). Unit 2 closes Thu Oct 8 (half-day test).

Activities coming soon.

▸Week 11 — Paper first· Oct 14–16Sketch, Spec, Swap in full · spec a site in block print · first tool contact, vibecode only.
▸WednesdayOct 14Sketch, Spec, Swap
  • ActivityLesson plan: Sketch, Spec, Swap· 50 min · cross-group swap · seven-artifact comparisonOpen activity
▸ThursdayOct 15Observe and describe a website
  • ActivityWebsite observation and description· take-home from Oct 14: spec a different site in block printOpen activity
▸FridayOct 16First tool contact — vibecode only
  • DeckEnter MuggsOfCode in vibecode: describe, watch, refine by talking· no spec panels yetcoming soon
▸Weeks 12–19 — The loop, the studio, and the portfolio· Oct 19 – Dec 15Four-panel specs and the amend-recompile loop · precision and the handoff package · human-centered design and accessibility · peer-spec compilation · self-paced capstone against student-set milestones · deploy through Cloudflare Pages (Dec 9) · defend the build against its specification (Dec 10) · assemble the portfolio, in place of a capstone presentation.

Activities coming soon.

Built, not yet placed

These exist and work. No objective in the course calendar calls for them specifically, so they are listed here rather than given an invented slot.