This is a plan for future ACP topic series. Think of this as nested strata, not a linear blog list. Some arcs run in parallel; some depend on others.

ARC 0 — Orientation / Meta-Frame (implicit, ongoing)

Purpose

  • Establish ACP / Commonplace as a serious, reflective, non-hype project
  • Signal epistemic humility, institutional seriousness, and long-horizon thinking
  • Make clear: this is not product marketing, thought leadership, or evangelism

Key themes

  • AI as an emergent system (cosmology metaphor)
  • Claims, authority, compression, governance
  • “We are early; structures are still forming”
  • Why this way of thinking is rare but necessary

This arc is woven throughout, not published as a single series.

ARC 1 — Governance (FOUNDATIONAL)

Status: Substantially developed
Trigger: Substack governance articles (dissatisfaction)

Core questions

  • How does authority actually emerge?
  • Why policy and ethics frameworks fail
  • Governance as representation, not rules
  • Claim Authorization Matrix
  • NFPA analogy, hazard labels, interface as governance surface

Key outputs

  • Claim Authorization Matrix
  • AI Hazard Diamond (textual + visual)
  • Governance embedded in design
  • Pre-authorization vs post-hoc accountability

Why it matters
This arc establishes ACP as structurally different from “AI ethics,” “AI safety,” and “AI policy.”

ARC 2 — Failure States (DIAGNOSTIC)

Status: Substantially developed
Trigger: Guardian AI failure reporting

Core questions

  • Why failures repeat across domains
  • Why “hallucination” is a misdiagnosis
  • Compression, omission, authority drift
  • Institutional—not technical—failure modes

Key exemplars

  • Health summaries → medical advice
  • Admin summaries → records
  • Facial recognition → false arrests
  • AI Overviews
  • MCAS, opioids, DDT (cross-domain analogies)

Outcome
Failures reframed as predictable governance breakdowns, not model flaws.

ARC 3 — Design (STRUCTURAL)

Status: Actively forming

Core questions

  • Interface design as governance surface
  • Constraints vs instructions
  • Slow vs fast systems
  • Disclosure standards (nutrition labels, preflight checklists)
  • Where design embeds authority implicitly

Key analogies

  • Riverbanks / floodplains
  • Amazon River morphology
  • Nutrition labels (compression disclosure)
  • Preflight checklists
  • NFPA signage
  • Rules of engagement

ARC 4 — How AI Actually Works (DEMYSTIFICATION)

Status: Planned

Purpose

  • Strip away hype
  • Explain LLMs, ML, training, inference, limits
  • Distinguish capability from authority

Tone

  • Plain, accurate, non-technical
  • No mysticism, no AGI fantasies

Why important
Prevents both:

  • irrational fear
  • irrational trust

ARC 5 — Incentives & Power (POLITICAL ECONOMY)

Status: Planned, sensitive

Core questions

  • Why everything goes sideways
  • Venture capital, domination, scale incentives
  • SF tech culture, longtermism, Davos
  • Why “alignment” often means profit alignment

Constraint

  • Must be evidence-backed
  • Avoid conspiratorial framing
  • Structural incentives, not villains

ARC 6 — Commercial Misuse & Social Harm (DARK PATTERNS)

Status: Planned

Topics

  • “AI friend / partner” apps
  • EdTech misuse
  • Slop economies
  • Infinite scroll & persuasion
  • Propaganda flooding
  • Impersonation, nudification
  • Autonomous weapons
  • Labor displacement & inequality

Frame

  • Not “AI is evil”
  • “These incentives + these affordances = predictable harm”

ARC 7 — Positive, Legitimate Uses (CONSTRASTING)

Status: Underdeveloped, intentional

Domains

  • AlphaFold
  • Tumor detection
  • Astronomy, physics, math
  • Weather & climate
  • Public health

Purpose

  • Show ACP is not anti-AI
  • Identify conditions under which AI succeeds

ARC 8 — Unexplored / ACP-Native Uses

Status: Actively emerging

Examples

  • Language learning
  • Institutional memory
  • Story aggregation (conflict narratives)
  • Legislative drafting (bounded)
  • Immersive education
  • Organizational diagnostics

This is where Commonplace and Agora-AI live.

ARC 9 — Enterprise & Institutional Failure

Status: Planned

Questions

  • Why companies can’t make AI work
  • Reliability, hallucinations, exposure
  • Time waste, rework
  • False confidence
  • No failure detection

This bridges governance + operations.

ARC 10 — Regulation & Why It Fails

Status: Planned

Coverage

  • EU AI Act
  • Other national regimes
  • Why regulation lags reality
  • Why category-based bans fail
  • Why ACP-style constraint may succeed

ARC 11 — Worldviews That Conflict with ACP

Status: Planned

Examples

  • Tech solutionism
  • Accelerationism
  • Pure market fundamentalism
  • Strong AGI inevitabilism
  • Surveillance-first governance

Useful for clarity and boundary-setting.

ARC 12 — Meta / Authorial Arc

Status: Optional, late-stage

Question

  • What does this body of work say about the author?
  • Minority institutional perspective
  • Diplomatic failure vs structural insight
  • Why outsiders see things insiders miss

ARC 13 — Metaphors for AI: Ways of Thinking, Limits, and Survivability

This arc examines how different metaphors for AI shape expectations, policy, design, and risk--and how some metaphors survive contact with reality better than others.

Core idea:
We don’t reason about AI directly; we reason through metaphors. The metaphor chosen quietly determines what we build, tolerate, fear, or ignore.

Representative lenses (examples)

  • AI as tool (hammer, calculator, spreadsheet)
    Pros: clarity, bounded responsibility
    Cons: underestimates emergent behavior and scale
  • AI as employee / intern / assistant
    Pros: enables governance, feedback, training, limits
    Cons: invites anthropomorphism and misplaced trust
  • AI as organism / ecology
    Pros: captures emergence, feedback loops, pathologies
    Cons: weak on accountability and intentional design
  • AI as infrastructure (roads, power grids, plumbing)
    Pros: highlights systemic risk and public impact
    Cons: obscures agency and misuse at the edge
  • AI as oracle / prophet
    Pros: explains social behavior and authority drift
    Cons: historically catastrophic when believed
  • AI as market actor
    Pros: aligns incentives and failure modes
    Cons: normalizes harm as “externalities”
  • AI as weapon
    Pros: forces seriousness and restraint
    Cons: collapses nuance into threat framing

Survivability question

Which metaphors:

  • degrade gracefully under failure?
  • preserve human judgment under pressure?
  • resist hype and overreach?
  • allow correction without collapse?

ACP position (implicit, not doctrinal):
No single metaphor is sufficient. Survivable systems require metaphor pluralism, explicit boundary-setting, and mechanisms to detect when a metaphor has outlived its usefulness.

II. Phase Roadmap

This is the practical roadmap.

Phase 1 — Consolidate Foundations (NOW)

  • Finalize Governance + Failure + Design arcs
  • Stabilize core vocabulary
  • Produce canonical reference posts

Phase 2 — Expand Laterally

  • Add “How AI Actually Works”
  • Add Enterprise Failure
  • Begin Commercial Misuse arc

Phase 3 — Power & Regulation

  • Incentives & Political Economy
  • Regulation failures
  • Democracy & AI

Phase 4 — Positive & Constructive Futures

  • Legitimate uses
  • ACP-native experiments
  • Institutional redesign

Phase 5 — Reflection

  • Worldview conflicts
  • Meta arc
  • AI as Metaphor