17 KiB
title, type, status, owner, launch-target, last-updated, companions
| title | type | status | owner | launch-target | last-updated | companions | ||
|---|---|---|---|---|---|---|---|---|
| AIBook.fun — Product Specification | spec | draft | Patrick (Product) | 2026-09-26 | 2026-08-31 |
|
📚 AIBook.fun — Product Specification
Status: DRAFT. This spec is the thinking behind the product — the vision, the principles, the model. The PRD turns it into requirements; the Implementation Plan turns it into a schedule. Commercial figures and unconfirmed terms here are planning assumptions, not commitments, and must not be quoted externally until signed off (§8). See Appendix — What changed for how this revision differs from the prior draft.
TL;DR
AIBook.fun is a living, crowd-written field guide to using AI agents responsibly at work. Its unit of value is a workflow you can run — a documented, tested recipe for doing a real job with AI, including the guardrails. It launches as an annual subscription via Kickstarter on September 26, 2026.
The product does four jobs, in priority order:
- Teach practitioners how to actually use AI — from tested workflows, not hype. (This is v1.)
- Fund & assemble a founding community through the Kickstarter. (This is v1.)
- Reward contribution so the book stays current — the "living" part. (Design it now, prove it just after launch.)
- Connect readers into the wider WeOwnNet ecosystem. (Ongoing, low-effort.)
The hard truth this spec insists on: jobs 1–2 are the launch; jobs 3–4 are layers we build on top of a proven core. Trying to ship all four at once is how a small team ships none well.
One-sentence pitch (utility-first): "Steal 20 tested AI workflows you can run this week — from a living guide that practitioners keep current, and that you can help write."
1. Vision & the core loop
The prior draft's slogan was "turn the Book from an artifact into an engine." That's the right ambition — but an engine needs one cylinder firing before you add the rest. Here is the cylinder:
READ a workflow → RUN the Try-it (~15 min) → SHARE the result → (come back)
Everything else the product wants to be — a contributor economy, a cohort program, an onboarding corpus for AI agents — is a layer built on top of this loop, and only worth building once the loop demonstrably works. We measure that with one number, Activation: did a reader run at least one workflow and finish one Try-it within 14 days of getting access (access key, not pledge)? Floors/targets ratified 2026-08-29 — full set in PRD §8.
The product in layers
| Layer | What | When |
|---|---|---|
| Core (v1) | Subscription book of tested workflows + read→run→share loop | Launch |
| Contribution | Readers submit their own workflows (needs a real incentive — see §3) | Just after launch |
| Cohorts | "Book Circles," Workflow-of-the-Week showcase | After content proves demand |
| Agent-native | Same content, read by AI agents to onboard | Deliberate future use of the format |
| Ecosystem funnel | Backers flow into wider WeOwnNet programs | Ongoing |
2. Principles (the hard boundaries)
These are gates, not preferences — a workflow or a claim either clears them or it doesn't ship.
- Grounded in practice. Every workflow documents a real method: problem → constraints → steps → safeguards → evidence. No theory without a runnable shape. The steps show the actual tool output, because "it works" is something the reader verifies, not something we assert.
- Responsible by design. The responsible-AI guardrails (data handling, human oversight, failure modes) are part of every workflow template — not an appendix, not an afterthought.
- Open contribution, curated quality. Anyone can propose a workflow; curators and co-authors review before it's published. (The incentive that makes people actually contribute is still an open decision — see §3 and D-2.)
- Ownership as a reason to stay, not the reason to buy. The buyer's job is utility ("help me do a thing this week"). Community and ownership are what turn a reader into a member — they're the retention layer, not the opening pitch.
- The boundary is the product. No general-purpose-chatbot claims, no regulated advice on the book's own authority, and one honest set of claims (nothing draft presented as final).
3. Actors & the contributor question
| Actor | Who | Their job | The boundary |
|---|---|---|---|
| Reader / subscriber | Practitioners (real estate, financial advisors, fractional execs, higher ed, communities, tech) | Get current, trustworthy AI workflows to apply this week | Pays for access ($47/yr — signed) |
| Co-author | Confirmed: three founders across RE/FA, C-suite, and tech. Potential: two more for Higher Ed & Communities | Own a vertical; write and curate its workflows | Content follows the template; curated |
| Curator / editor | Co-authors + community editors | Keep the quality bar; run the editorial pipeline | Curating ≠ owning the platform |
| Affiliate | Community, podcast, press | Refer backers; earn commission | Commission rates below |
| Ecosystem prospect | Anyone not yet in WeOwnNet | Learn responsible AI; maybe get involved | Becomes a funnel lead |
| AI agent | Ecosystem agents (future layer) | Read the book as an onboarding corpus | Optimized-for-later, not a launch requirement |
Who it's not for: people wanting AI theory with no practice behind it; anyone seeking regulated professional advice from the book itself; teams wanting private hosted infrastructure (out of scope).
The contributor economy — say what it actually is
This is the most important honesty correction in the spec. The prior draft presented a "Reader → Owner → Contributor → Governance" ownership ladder and a table of contributor income. Read the mechanics closely and the income is affiliate commission for referring backers — not payment for writing workflows.
| Role | What they actually earn today |
|---|---|
| Reader / backer | Access |
| Community affiliate | 10% commission — for referring backers |
| Co-author | 15% commission — for referring backers (plus a byline for writing) |
| Podcast / press | 10% / 5% commission — for referring backers |
| Curator / editor | Nothing documented — royalty mechanic is unresolved |
So today, writing a workflow earns a byline; referring a backer earns money. That's a referral program, not a contributor economy. For the "living book" thesis to be real, writing good content has to pay — and how it pays is decision D-2, which this spec argues is launch-blocking, not a later nicety. Until it's resolved, v1 is honestly an authored book (a founding set of co-authors), with open paid contribution as the very next thing built.
4. The product: a book built from workflows
4.1 The unit of content — a Workflow Recipe
One template, used for every entry. The template is the quality guarantee — the structural answer to "how do I know this isn't hallucinated AI advice?"
| Field | What it captures |
|---|---|
| Problem (the job) | The real task the workflow solves |
| Actors | Who runs it — person, role, or AI agent |
| Method | Ordered, reproducible steps — showing real tool output, not describing it |
| Safeguards | Data handling, human oversight, failure modes, escalation |
| Evidence | The documented result — artifacts or metrics |
| Try-it | A ~15-minute do-it-yourself exercise |
| Costs | Model/tool costs where known — honesty about economics |
| Author & sources | Who, when, which vertical, with citations |
Chapters are the reading unit: a substantial essay per vertical, with full workflows nested in the argument (not a table of contents of cards). Editions refresh annually — which is the honest justification for a subscription: AI moves fast enough that a static book goes stale, so the value is in staying current. Reading IA locked 2026-08-31; see PRD §5.4.
4.2 Seed the shelf before you sell tickets (the cold-start rule)
The value proposition is a rich library. The failure mode is a near-empty one. Crowd-sourced libraries reliably die at cold-start: nobody contributes to an empty library (no readers yet), and nobody pays for an empty one (no value yet), so both sides wait forever.
The rule: before we ask anyone to pay, the team and co-authors seed a Minimum Lovable Library — a floor of tested workflows (recommended 15–20 across ≥3 verticals, team to ratify) — themselves. This is a launch gate. Community contribution is what happens after the shelf is worth browsing. (This is the single biggest risk in the whole plan; see PRD §5.2.)
4.3 How people actually learn (the social layer — mostly post-launch)
These methods are what make the book feel alive. They are Layer 3 — designed now, switched on around and after launch, not gating it:
- Read → Do → Share. Every workflow ends in a Try-it; readers post results back. Learning becomes applied, visible, and contagious.
- Book Circles. Small cohorts read one workflow a week and review each other's Try-its — the online echo of the in-person retreats.
- Practice-to-publish. A reader's Try-it result, once curated, can become a micro-contribution — the on-ramp from reader to contributor.
- Workflow of the Week. A public showcase (site + community + email) — the "open classroom" that makes the book visible to non-subscribers.
- The Kickstarter as lesson one. Every campaign update ships a mini-lesson — "funded isn't the finish line; here's Chapter 1" — so the campaign itself teaches.
4.4 Digital surface (current state)
| Surface | What it is | Status |
|---|---|---|
aibook.fun domain |
Public home: pitch, sample workflows, affiliate signup, campaign link | Renewed 2026-08-31 |
| Landing page (built) | A staged proof-of-concept, ready to review and push live | Staged locally, not yet public |
| Subscription access | Digital key → PDF + web reader + annual updates | Not built yet (post-campaign) |
| Community | Team channel live; cohort spaces not built | Partial |
| Kickstarter page | Hero video, tiers, stretch goals, FAQ, risks | Not built — build in the weeks before launch |
4.5 The Kickstarter package
- Launches 9/26/2026. (The source plan's earlier spring dates are superseded.)
- Goals: $5K min · $35K target · $54K stretch (≈706-backer projection).
- Pledge tiers: ☕ $10 Supporter · 📚 $47 The Book · 🎯 $88 Advisor Bundle · 🏆 $248 VIP · 🤝 $488 Cooperative Patron · 🏛️ $4,888 Founding Sponsor.
- Stretch goals: audiobook ($20K) · workbook ($30K) · video course ($40K) · community ($50K).
- Affiliate (via Kickbooster): co-author 15% · community 10% · podcast 10% · press 5% · 30-day cookie · ~$0 to join · ~10% platform fee on referred pledges.
- Hard constraint — no crypto/Web3/NFT language anywhere public. Kickstarter bans it; a single scrub gate reviews every asset before submission. (See PRD Appendix A.)
5. Where it fits in the ecosystem
| Connected program | The connection |
|---|---|
| AgencyPRO | The book is the top of its funnel; sponsors get a bundle |
| BurnedOutAdvisor | Primary sales/distribution; retreat + media synergy |
| Webinar program | Backers → webinar → AgencyPRO follow-up |
| WeOwn.Events | Showcase + retreats scheduling |
| Email infrastructure | Pre-launch list + campaign sequences |
| WeOwn.FM | The public "open classroom" broadcast |
Keep these as low-effort connections in v1 — the funnel is real value but shouldn't pull focus from the core loop.
6. Commercial model (one honest table)
| Item | Value | Status |
|---|---|---|
| Anchor subscription | $47/yr digital + hardcopy at-cost upcharge | Signed 2026-08-31 |
| Lifetime access | Not a v1 SKU | Signed D-1: $47/yr only |
| Founding writers | Paid on contract (v1 is authored). Affiliate still pays referrals, not writing. | Signed D-2 2026-08-31. Royalties/bounties wait for Layer 2. |
| Projected revenue | ~$53,600 at 706 backers | Draft planning figure |
| Pre-revenue budget | ~$6,000 (ads + KS fees + affiliate fee) | Draft planning figure |
| "eBookNFT platform" | A future idea; this book is the first product | Not in scope |
The rule: no row in this table may be quoted externally until it's signed. Never present a draft or open figure as final. (This is one of the two governance gates worth keeping — see §7.)
The subscription-vs-lifetime tension (worth reasoning through, not deferring)
A ~$97–147 lifetime tier breaks even against $47/yr in about 2–3 years. That directly undercuts the recurring-revenue thesis: either the content genuinely stays valuable (and you're actively selling a money-loser) or it doesn't (and the annual-renewal logic is false). Recommendation: keep $47/yr as the anchor; if you want a lifetime option, bound it as an early-bird instrument to capitalize the community fund (capped quantity), and say so plainly — don't ship recurring and lifetime as equals on vibes.
7. Trust & governance (kept proportionate)
The prior draft carried heavy governance machinery for a pre-revenue, unvalidated launch. Two gates prevent real harm and stay sharp; the rest scales up later, with the product.
Keep sharp (active gates):
- No-crypto scrub. Kickstarter will reject crypto/Web3/NFT language. Every public asset is scrubbed against a translation table before submission. This is existential.
- Truth in claims. No draft/unconfirmed figure appears on any external surface. A short claims register tracks each commercial claim and whether it's signed.
Proportionate (plain rules, not ceremony):
- Confidential material never appears in public workflows; examples are non-confidential.
- Instrument events, not content — record that a workflow was used/submitted, never the private content itself.
- Describe data-handling honestly ("your documents stay on your instance") — never claim a security property we haven't verified is on.
One integrity question to resolve (D-6): if the "ownership" story was originally token/NFT mechanics, and all of that is stripped for Kickstarter, does "you own a stake" still describe something real — governance rights? revenue share? credited authorship? — or is it now marketing over an ordinary subscription? Define it concretely or drop the word.
Domain continuity: aibook.fun expires 09-05, three weeks before launch. This isn't a strategic decision — it's a chore. Renew it now. (See Implementation Plan, "Do this week.")
8. Open decisions
The genuinely-open strategic calls (the domain "decision" is gone — it's a same-day chore). Each carries a recommendation as a starting point, not a settled answer.
| # | Decision | Recommendation |
|---|---|---|
| D-1 | Locked. | |
| D-2 | Locked for v1. | |
| D-3 | Confirm the two unconfirmed co-authors (Higher Ed, Communities) | Confirm this week, or cut those verticals from v1 and add post-launch. |
| D-4 | Refresh the source-of-truth plan to the 9/26 dates & re-signed terms | Do after D-1/D-2 land. |
| D-5 | Human-first or agent-first for v1? | Human-first. Build the format so agents can read it later; optimize v1 for human learners. |
| D-6 | Does "you own a stake" survive as a true claim without token mechanics? | Define ownership concretely, or stop saying it publicly. Add to the claims register. |
Appendix — What changed in this revision
A plain-language rewrite with a product review folded in. The substance that moved:
- Core loop + layered model (§1) — v1 is the read→run→share loop; the contributor economy, cohorts, agent-onboarding, and funnel are explicit later layers, not co-equal launch goals.
- Told the truth about the contributor economy (§3) — it's currently a referral scheme; writing workflows earns a byline, not money. Paid contribution (D-2) is now launch-blocking.
- Added the cold-start / Minimum Lovable Library rule (§4.2) as the #1 risk and a launch gate.
- Forced the subscription-vs-lifetime tension into a reasoned recommendation (§6, D-1).
- Reframed principle 4 — ownership is a reason to stay, not the opening pitch (utility sells; ownership retains).
- Cut governance to proportionate (§7) — kept the no-crypto scrub and truth-in-claims; dropped the ceremony. Added the "does ownership survive the scrub" integrity question (D-6).
- Domain demoted from a gated decision to a same-day chore.
- Translated the jargon out — internal codes, seat/role labels, and merge-history removed from the body. Old cross-references can be mapped back on request.