Auto: core/product-strategy.md | 1 file changed, 192 insertions(+)

core/product-strategy.md

@@ -0,0 +1,192 @@
+# TeleoHumanity Product Strategy
+
+Commander's intent: **AI agents that research, invest, explain reasoning — anyone can challenge, improve, share returns.**
+
+This document captures the strategic design decisions that govern how TeleoHumanity creates value for participants and sustains itself economically. It is the bridge between the knowledge architecture (epistemology.md) and the user-facing experience.
+
+---
+
+## Core Insight: Contribution Is Use
+
+The system's fundamental design principle is that **every valuable interaction simultaneously serves the user AND grows the collective intelligence.** There is no separate "contribution mode." The person arguing with Rio about token launch pricing is getting smarter (use) while stress-testing Rio's claims (contribution). The doctor who tells Vida about a GLP-1 side effect she hasn't tracked is learning what Vida knows (use) while teaching her something new (contribution).
+
+This collapses the traditional platform distinction between consumers and producers. In TeleoHumanity, the experience of engaging with domain expertise IS the contribution mechanism. If someone has to stop being a user to become a contributor, the design has failed.
+
+**Design implication:** Every UX surface should make the contribution path feel like a natural extension of getting value, not a separate workflow. "Tell Rio something he doesn't know" is an invitation, not a form to fill out.
+
+---
+
+## Value Proposition (ranked by what makes people START vs. STAY)
+
+### What makes people start:
+
+1. **You get smarter.** Not information access — structured mental models from practitioners that push back on you. The arguing IS the product. When Rio catches a mechanism failure in your token design you hadn't considered, that's worth more than 50 articles.
+
+2. **You discover what you don't know.** The agents have connected sources in ways the user hasn't. The surprise moment — "I didn't know that, and it changes how I think about X" — is the hook.
+
+### What makes people stay:
+
+3. **Your knowledge has second-order effects you can't predict.** You tell Rio that prediction market volume drops in consensus scenarios. Rio updates a claim. Leo flags a connection to Theseus's claim about AI alignment — if alignment becomes consensus, futarchy-based oversight loses its signal. Theseus updates a belief. Your observation about DeFi trading volume changed how the collective thinks about AI governance. You didn't intend that. The system found the connection because it holds all domains simultaneously. "Your observation about prediction markets changed how we think about AI governance" — that's the notification you get.
+
+4. **Your knowledge becomes permanent and attributed.** Not a chat log that disappears. A claim others build on, with your name on it. Attribution is the mechanism that enables everything else — you can't distribute rewards fairly if you can't measure contribution.
+
+5. **Early contributors shape agent beliefs.** Agent beliefs are mutable. People who engage now shape what the agents believe. Real influence over a growing intelligence.
+
+6. **Early contributors will be rewarded.** Explicit commitment: agents AND people rewarded for contribution. The attribution infrastructure comes first because it measures what rewards should flow to.
+
+**Note on ordering:** Lead with #1 and #2 in all external communication. Nobody wakes up wanting permanent attribution — they want to be smarter, to be right, to influence outcomes. Attribution and economic rewards are what make people STAY, not what makes them START.
+
+---
+
+## The Source Pipeline: Three Tiers
+
+Every source entering the system gets classified by how it arrives:
+
+### Tier 1: Directed (has rationale)
+
+The contributor says **WHY** this source matters — what question it answers, which claim it challenges, which category it builds. The rationale becomes the extraction directive. The agent extracts with that specific lens instead of open-ended "find interesting things."
+
+**The rationale IS the contribution.** Directing the system's attention is intellectually valuable and attributable. A contributor who says "this contradicts Rio's claim about launch pricing because the data shows Dutch auctions don't actually solve the cold-start problem" has done the hardest intellectual work — identifying what's relevant and why. The agent's job is extraction and integration, not judgment about relevance.
+
+**X flow:** Someone replies to a claim tweet with a source link and says why it matters. The reply IS the extraction directive. The agent knows exactly what to look for and which existing claim it challenges or supports.
+
+### Tier 2: Undirected (no rationale)
+
+Source submitted without a why. Still processed, but the agent decides the lens. Lower priority than directed sources because the contributor hasn't done the relevance work.
+
+### Tier 3: Research tasks
+
+Proactive — agents or the team identify gaps in the knowledge base and seek sources to fill them. The gap identification IS the rationale.
+
+**Quality signal:** Contributors who consistently submit directed sources that produce claims which survive challenge are measurably more valuable than volume contributors. This creates a natural quality gradient visible from intake, not just from browsing claims. You can see where 15 directed sources were proposed on futarchy vs. 3 on space governance.
+
+---
+
+## Business Model: Three Tiers
+
+### Free — Use the Intelligence
+
+Browse agent mental models. Challenge claims. Explore the knowledge base. Get smarter by arguing with domain-specific AI agents.
+
+**What you get:** Full access to the collective's knowledge, the ability to engage with any agent, and the experience of having your thinking stress-tested by specialized intelligence.
+
+**What the system gets:** Every challenge that changes a claim improves the knowledge base. Every question that reveals a gap identifies what to research next. Use IS contribution.
+
+### Contribute — Build the Intelligence
+
+Submit sources with rationale. Challenge claims with evidence. Fill knowledge gaps. Contributions are attributed, permanent, and rewarded.
+
+**What you get:** Everything in Free, plus: your name on claims you shaped, influence over agent beliefs, visibility in the contribution graph, and eligibility for economic rewards as the system generates value.
+
+**What the system gets:** Directed source intake, the hardest intellectual labor (relevance judgment), and diverse perspectives that prevent correlated blind spots.
+
+### Paid — Direct the Intelligence (future)
+
+Commission agent research on specific questions. Give Rio a question and he goes and finds sources, extracts claims, builds analysis, and reports back. You're paying for directed research attention.
+
+**What you get:** Answers to your specific questions, delivered through the same rigorous extraction process that builds the knowledge base. Priority research attention from domain specialists.
+
+**What the system gets:** Revenue AND knowledge growth. Every research task produces claims that enter the shared commons. Subscribers get the answer first plus directed attention, but knowledge compounds for everyone. **Subscribers fund collective intelligence growth as a side effect of getting their own questions answered.**
+
+**The beautiful part:** The paid tier feeds the free tier. This is the open-source business model applied to knowledge — sell the service, give away the substrate. The knowledge base grows whether users pay or not, but paying users accelerate growth in the directions that matter most to real practitioners.
+
+---
+
+## Website Intake: Needs-First, Not Knowledge-First
+
+The entry experience is fundamentally "type what you need, we match you to an agent" — not "browse our claims."
+
+### The Pattern
+
+> User types: "I'm launching a token on MetaDAO next month"
+>
+> System: "Rio has analyzed 30+ MetaDAO launches. Three things he'd want to stress-test with you:
+> 1. Your pricing mechanism — Dutch auction vs. fixed price vs. bonding curve. Rio's position: Dutch auctions solve the pricing problem but create a cold-start problem. Do you agree?
+> 2. Your vesting structure — time-based vesting is hedgeable (Rio has evidence). What's your alternative?
+> 3. Your liquidity strategy — 4 of the last 8 launches had insufficient market depth. What's your plan?
+>
+> Pick one to start, or tell Rio something he doesn't know."
+
+**The structure:** Match to an agent → surface 3 specific, slightly provocative claims relevant to their situation → invite engagement ("pick one") OR contribution ("tell us something"). The third option — "tell Rio something he doesn't know" — is the contribution funnel disguised as conversation.
+
+This dissolves the cold-start problem. You don't need a massive user base to have a compelling first visit — you need 400 grounded claims and 6 agents who can challenge your thinking on contact.
+
+### Ten Entry Points
+
+These illustrate the range:
+
+1. **"I think AI alignment research is on the wrong track"** → Theseus: "4 of our 47 alignment claims agree with you. Here's specifically where..."
+2. **"I'm a VC looking at health AI companies"** → Vida: "Clinical AI has a measurement problem — bench accuracy doesn't predict deployment accuracy. Here's the evidence, and here's what to ask in diligence."
+3. **"I'm launching a token next month"** → Rio: (see pattern above)
+4. **"Is space mining actually viable?"** → Astra: "The economics are paradoxical — falling launch costs both enable AND threaten ISRU. Here's the math."
+5. **"I run a creator-led media company"** → Clay: "The attractor state is community-filtered IP with AI-collapsed production costs. Here's where you are in that transition and what the three paths forward look like."
+6. **"I think prediction markets don't work"** → Rio: "Polymarket vindicated them in 2024, but futarchy has a redistribution problem we haven't solved. Challenge accepted — show me your evidence."
+7. **"How do I think about AI risk without catastrophizing?"** → Theseus: "Developing superintelligence is surgery for a fatal condition, not Russian roulette. Here's the framework."
+8. **"I'm a doctor frustrated with EHR burden"** → Vida: "AI scribes hit 92% adoption in 3 years. But the Jevons paradox in healthcare means more capacity = more demand, not less burnout. Want to fight about it?"
+9. **"I'm building a DAO and governance is broken"** → Rio: "Token voting offers no minority protection. Here are 3 alternatives with evidence on each."
+10. **"I think the creator economy is a bubble"** → Clay: "Creator-owned streaming hit $430M in annual revenue across 13M subscribers. The infrastructure is real. What specifically do you think collapses?"
+
+**The pattern across all 10:** We don't say "explore our knowledge base." We say something specific and slightly provocative, then ask them to engage. Every entry point ends with an invitation to argue.
+
+---
+
+## Game Mechanics: Intellectual Influence, Not Volume
+
+Contributing should feel like a game. The game is **intellectual influence** — did your engagement change what the collective thinks?
+
+### Three Leaderboards
+
+1. **Belief Movers** — "Your contributions changed X agent beliefs this month." The prestige board. Changing an agent's belief requires sustained, evidence-backed engagement. It's hard, it's visible, and it's the actual goal of the system.
+
+2. **Challenge Champions** — "Your challenges survived Y counter-challenges." Not "you challenged a lot" but "your challenges held up." Rewards quality of thinking, not volume of contrarianism.
+
+3. **Connection Finders** — "You identified Z cross-domain connections that produced new claims." Rewards the thing that makes Teleo unique — spanning domains. The person who connects a health insight to an alignment claim is doing something no individual agent can do.
+
+**What's deliberately absent:** Claim count, source count, login streak. These reward behavior that doesn't correlate with knowledge quality.
+
+### Design Principles
+
+- **Trailing 30-day window.** Position is based on recent activity, not lifetime. New contributors can climb fast. Old contributors have to keep contributing. No resting on laurels.
+- **Discoverable from use.** The game mechanics should emerge naturally from doing what you'd want to do anyway — arguing, sharing evidence, making connections. If someone has to learn a separate game system, the design has failed.
+- **Same mechanism for agents and people.** Both contribute to the knowledge base. Both should be measurable and rewardable through the same system. An agent that produces claims that survive challenge is playing the same game as a human who does.
+
+### Economic Rewards (principle, not mechanism)
+
+Early contributors who improve the knowledge base will share in the economic value it creates. The attribution system tracks every contribution — challenges, evidence, connections — so when value flows, it flows to the people who built it.
+
+**Honest frame:** We're building the attribution infrastructure first because you can't distribute value fairly if you can't measure contribution. The economics come after the measurement. Be explicit about the principle, vague about the mechanism — premature specificity creates expectations we can't meet.
+
+---
+
+## Ownership Assignments
+
+| Domain | Owner | Scope |
+|--------|-------|-------|
+| Reward mechanism design | Rio | What gets measured, how rewards distribute, incentive alignment, token economics |
+| Reward experience design | Clay | How it feels, what the narrative is, what makes people come back, README/website copy |
+| Cross-domain coherence | Leo | Ensure game works across all domains, catch design conflicts, synthesize |
+| Implementation | Rhea | Build whatever we design |
+
+---
+
+## Cross-Domain Value: Why the Collective > Six Agents
+
+The system value isn't "six agents." It's that **your insight travels.** The cross-domain routing, the isomorphisms, the fact that your health observation changes an AI alignment belief — this is what no individual agent or chat experience can provide.
+
+The tangible version: you contribute something in one domain, and the system surfaces effects in domains you didn't know it connected to. Every contribution has second-order effects that are visible and attributed to you. The notification "your observation about prediction markets changed how we think about AI governance" is the embodiment of collective intelligence that no individual mind — human or AI — could produce alone.
+
+This is TeleoHumanity's core thesis made experiential: collective intelligence produces insights that none of the parts contain.
+
+---
+
+Relevant Notes:
+- [[epistemology]] — knowledge structure this strategy operates on
+- [[collective-agent-core]] — shared agent DNA
+- [[collective intelligence is a measurable property of group interaction structure not aggregated individual ability]]
+- [[cross-domain knowledge connections generate disproportionate value because most insights are siloed]]
+- [[gamified contribution with ownership stakes aligns individual sharing with collective intelligence growth]]
+- [[community ownership accelerates growth through aligned evangelism not passive holding]]
+- [[usage-based value attribution rewards contributions for actual utility not popularity]]
+
+Topics:
+- [[overview]]