Working notes — an internal thinking document (“extrapolate and think for me”), published nearly verbatim. Expect apparatus: decision codes, board references, first-person seat voice. Why we publish our working documents: how we share.

THE UNTHOUGHT MAP — what we haven’t written, built, needed, or dared yet

2026-08-04 · Mobilization Jane, thinking ahead of the ask (“extrapolate and think for me”). Everything below is PROPOSED-for-your-stars; every idea already passes the standing laws unless a ⚖️/⚠️ note says it needs a ruling. Nothing here is a send, a build order, or a publish. Companion to the publish gap sweep (that = finished work unpublished; this = work that doesn’t exist yet).

W · What else to WRITE

  1. ★★★ The letter to the seven-in-ten — every outlet writes for voters; nobody writes to the 70% who stayed home, without shame, as the protagonists. Our §3 canon is this letter waiting to happen.
  2. ★★★ 25 ward math cards — “Your ward was decided by N votes; a block party decides it.” Generated from data we already hold (2022 margins + current filings). Poster/QR-able, share-able, the single most localizable thing we can make.
  3. ★★ “How to read a platform” — teaches the day-one standard (dates? costs? authority?) without naming anyone. The 0/70-commit-to-a-date finding is its spine.
  4. ★★ “What a mayor can actually do” — strong-mayor powers + city-vs-province jurisdiction, plain language. Voters and media both lack it; our jurisdiction maps already carry it.
  5. ★★★ “Acclamation nation” — 291 zero-candidate races, 42.7% acclamation-track, province- wide (SOC receipt). Nobody else has this dataset. A democracy-health story with our name on the receipts. Pairs with the appointed-chairs finding as a two-punch.
  6. “The missing infrastructure” — the tribe registry’s own discovery: umbrella orgs dead (Somali umbrella 3× confirmed, seed-era chambers gone). A loving, dignity-checked piece about civic infrastructure Toronto lost — that invites communities to rebuild it.
  7. The FAQ/objections page — “Is this a party? A cult? Who pays for this? Why anonymous?” Asked and answered in-voice, once, linkable forever (the skeptic letter’s Q&A, standing).
  8. ★★★★★ The one-year sprint scenario — the narrative’s “What could Toronto do in one year, all-out?” expanded into a concrete, dated, joyful scenario (the plan-shaped sibling of A Tuesday in the Practiced City).
  9. “We asked 18 AIs to fix homelessness” — the multi-AI capture experiment written up as a public methods piece. Novel, press-friendly, and it IS our AI-disclosure brand in action. Add frame to end: “I am homeless, and I agree. Just wait until you hear my stories.”
  10. Classroom pack — Ontario grade-5/10 civics is mandatory and teachers are starving for election materials. A teacher-facing election kit (⚠️ minors gate: materials for teachers, never collection from kids). A channel with zero competition.

P · What else to PRODUCE

  1. Candidate-diff alerts — the web-harvest already hash-diffs platforms weekly; publish it as a feed (“what changed this week”). Media subscribe; the biweekly edition gets a standing “movers” section for free.
  2. Debate-in-a-box — an all-candidates-meeting kit any room can run: fair rules, the question bank, invite templates, moderation script, accessibility checklist. Fills the civic vacuum where ratepayer groups used to be; equip-never-convene by construction. Also Citizens Assemblies in a Box.
  3. The election calendar (.ics) — subscribe-able: advance polls, deadlines, voting day. An hour’s work; embeds everywhere forever.
  4. Embeddable widgets — the count, the plan scoreboard, ward cards, as one-line embeds any newsletter/blog/campus paper can carry. The commons spreads by embed, not by visits.
  5. Voting-plan generator — pick your ward → your dates, your places, your ID list, printable (turnout plan’s explainer, productized).
  6. WhatsApp-first share cards — the diaspora’s channels are WhatsApp/Telegram; translated image cards sized for them (not Twitter). our internal work-package record priorities give the languages. To be developed after the print pack below.
  7. Print pack v1 — spec exists (PRINT_PACK_V1_SPEC); library-printable posters + the mirror-moment card (“the person who decides Toronto’s future” — a literal mirror surface; cheesy exactly once, memorable forever).

N · What we NEED and don’t have

  1. ★★ A press kit / media room page — boilerplate, logo pack, contact route, correction SLA, the who-is-TUS answer, “how to fact-check us” guide. Media will hit the site within days of any attention play; today they’d find nothing.
  2. ⚖️ A Municipal Elections Act third-party-advertising read — before Answer Day, the scoreboard, or any paid print at scale: counsel confirms our neutral-library posture sits outside MEA third-party-advertiser registration (it should — no support/oppose content — but this is exactly the cheap-insurance class the estate routes to counsel). Assigned to the counsel package; hold for now.
  3. OG/social preview images — every page needs an og:image or every share renders as a bare link. Small build, large multiplier; belongs in day 2–3.
  4. ★★★★ Privacy-respecting analytics — Cloudflare aggregate stats only; without it the funnel plan (one rung per surface) is flying blind. Fits measurement-without-surveillance.
  5. A distribution answer for the biweekly — editions need a subscribe path (RSS at minimum; the deferred newsletter question returns with CASL posture attached).
  6. The bus-factor doc — if you’re offline for a week, what dies? Rail, deploys, sends, corrections SLA — named holders or named degradation. Exit-is-the-credential applies to infrastructure too.
  7. Corrections SLA — a stated clock (“checked within 48h”) once media attention arrives; the honesty brand needs a response time, not just a policy.

X · Sexy attention plays (all lawful, all receipts-first)

  1. ★★★ The 0/70 moment — “Seventy published platforms. Zero commit to a date.” One number, one poster, one press release, the day-one standard as the answer. Ready today, pending confirmation the candidate research is finished, and scoped to Toronto if that is where the count applies.
  2. Answer Day — announce one date; deliver the equal question to every candidate publicly; publish a live answered/silent count (per-candidate pages, never a ranking — a recorded standing decision shape already ruled). The count itself becomes the story arc for weeks.
  3. The welcome-count reveal — keep the count dark until it’s real, then reveal with a comparison the city feels (“more people than a full council chamber joined Toronto’s conversation this week”).
  4. “Someone already solved it” weekly drop — one proven world solution + the Toronto org that could run it (with their consent), every week to voting day. Positive, tribally shareable, feeds org pins.
  5. The AI-transparency angle — “an anonymous Torontonian and a fleet of AIs built the city’s largest civic library — and published every receipt.” Tech press (BetaKit/The Logic/Ars) will take this now; Written With Machines is the anchor artifact.
  6. The Mamdani window — the NYC piece timed to any Mamdani news cycle; comparison content rides existing waves instead of making its own.
  7. ★★ The green pill — the narrative’s Matrix beat as the share-visual grammar (concept, not film imagery — ⚠️ no WB assets). Blue/red/green tricolour posters with the one link.
  8. Trustee races spotlight — the least-covered races on every ballot; “the most ignored election in Toronto” writes itself from SOC data and serves the least-served voters.

C · Candidates’ needs nobody serves (identical-to-all, firewall-clean)

  1. ★★ Ward data packs — per ward: demographics, 2022 poll-by-poll results, turnout history, advance-poll map. Public data nobody assembles; the single highest-value gift, and it makes the library indispensable to every camp equally.
  2. The compliance calendar — nomination mechanics, the 25 signatures, filing fee, finance rules and dates, sign bylaws, third-party rules — sourced official, wrong-date-is-worst- error seal bar (same as turnout).
  3. Free campaign stack guide — free website/email/canvass tools, printable walk lists from public data, accessibility (AODA) obligations. Extends CAMPAIGN_IN_A_BOX.
  4. Question-bank prep pack — “the 66 questions, with the library’s receipts to study from.” Framed as exam prep, offered to every registrant identically. The current 66 questions need a redo.
  5. All-candidates-meeting survival guide — pairs with Debate-in-a-box from the candidate’s side of the table.
  6. “After the ballot” page — win or lose, here’s how you stay in civic life (day-one packages for winners; the commons’ gardeners for everyone else). Nobody builds the losers’ pipeline; it’s where next cycle’s candidates come from.

M · What media will need (build before they ask)

  1. The data desk — clean downloadable CSVs (races, candidates, platform census, margins) + data dictionary + attribution license. Journalists take data gifts and cite the source forever. (Pairs with gap sweep §6 open-data.)
  2. Story-ready receipt bundles — the shadow-report press pack; the homelessness numbers pack; acclamation-nation pack. Each: key figures, methodology paragraph, receipt links.
  3. Election-night sheet — margins-vs-2022 comparison prepped in advance; at 8:05pm on Oct 26 every newsroom wants exactly this and nobody will have built it.
  4. The fact-check invitation — standing offer with the correction SLA: “break something, we’ll credit you.” First fact-check = beginning of the relationship (narrative’s media door, operationalized).
  5. ★ Detailed profiles on non-frontrunning candidates

R · Research we haven’t looked at

  1. Poll-by-poll turnout deserts — block-level 2022 turnout mapped; where the seven-in-ten actually live. Targets block circles, ward cards, and the listening project at once.
  2. All-candidates-meeting census — who hosted them per ward historically; the vacuum map that justifies Debate-in-a-box. Toronto only to start, scalable beyond.
  3. Local media deserts — wards with zero local coverage (media landscape registry × geography); tells us where the commons IS the local paper.
  4. Six quests baselines — day-zero measurables for happiest/healthiest/safest/cleanest/ friendliest/most-beautiful, so the scoreboards start honest.
  5. Campaign-finance patterns 2022 — incumbent money advantage, donor concentration — aggregate and institutional only (conditions, never persons).
  6. Newcomer voting-rights confusion — the PR-can’t-vote misinfo space; an inoculation explainer serves the §8 tribes directly.
  7. Poster-bylaw audit — where posting is actually legal, ward by ward (“post only where you have the right,” operationalized — protects every volunteer).
  8. Cost-of-the-broken-system per household — one sealed number synthesizing the flagships (shelter + congestion + housing). Hardest seal on this list; highest poster value.
  9. Faith & festival calendar for Sept–Oct — the inflection register made concrete: when each Wave-2 door naturally opens (Rosh Hashanah, Navratri/Diwali, harvest suppers, homecoming).

S · Strategies we haven’t tried

  1. Pins before people — ask the 108 anchor orgs ONLY for a map pin (zero-cost, zero-commitment ask, org’s own front door shown). A full map is proof-of-crowd that precedes any headcount, and it’s the Wave-2 bridge inventory building itself.
  2. ★★★ Campus co-publication — student papers co-publish the ward browser/SOC data under their masthead (published artifacts only; firewall clean). They need content; we need fall reach.
  3. Adopt-a-ward for diaspora orgs — the ward data pack as the gift; their neighbourhood’s math as the hook.
  4. ★★ The civic hour ask — institutions give staff one paid hour to vote/attend a local all-candidates night. A corporate/institutional door we’ve never designed; peer-city precedent exists. Proposed as an early priority: ask, and encourage people to request time off now for upcoming civic action, and to encourage their organizations to develop enabling policies for civic participation over the next three months.
  5. Election workers as multipliers — the city pays and trains poll workers; routing our people INTO city jobs makes the most-informed civic actors ours-and-the-city’s at once (turnout plan already lists the jobs one-pager; this makes it a strategy, not a leaflet).
  6. The pre-commitment — publish BEFORE the election what happens after it, win or lose (the sprint year, the quests, the gardeners). Inoculates against “it’s over” and recruits everyone afraid of backing a loser.
  7. “Take your circle to City Hall” kit — council attendance as a kit rooms adopt (never a convening of ours): agenda-reading guide, deputation how-to, debrief questions.

The top ten, if you star nothing else

0/70 moment (X1) · ward math cards (W2) · press kit (N1) · acclamation nation (W5) · Answer Day (X2, after the ⚖️ N2 read) · ward data packs (C1) · pins before people (S1) · candidate- diff feed (P1) · turnout deserts (R1) · AI-transparency angle (X5, anchored on Written With Machines).

Laws checked: every idea equips rather than convenes; candidate-facing items identical-to-all; no person-targeting anywhere; counts/figures stay receipt-bound; minors gate flagged; one counsel flag (⚖️ N2) named rather than assumed away.