{"path":"research/societal-integration-outlook.md","content":"# Societal Integration Outlook: Expectations, Enemies, Allies\n\n**Date**: 2026-08-20 · **Type**: full-corpus synthesis, founder-prompted: *\"What are expectations we should have for the evolution and integration of Deliberus into society? What are the largest obstacles/frictions/enemies and what are the largest drivers/incentives/allies?\"* · **Grounding**: every claim below traces to a named corpus doc; nothing is asserted that a doc does not carry. Priors, not predictions — held in the same register as the wager itself: falsifiable, revisable, scored honestly.\n\n## §0 The one-paragraph answer\n\nExpect infrastructure-grade adoption, not product-grade: long quiet accumulation punctuated by niche-flips, with most of society never touching Deliberus directly — integration arrives through intermediaries (machine readers, facilitators, institutions in crisis) the way nobody \"uses\" DNS. The largest enemies are, in order of depth: power's structural need for ambiguity, the attention economy's opposite optimization, motivated human cognition, and the strategy-class adversaries that arrive with success. The largest allies are, in order of force: the AI epistemic crisis itself (the adversary IS the demand), machine readers as a consumer class, the falling cost curve of legibility, and the specific populations for whom clarity already pays. The deepest structural fact: the same AI wave powers both sides of every one of these pairs, so the race is endogenous — Deliberus rides the curve of its own adversary, and integration happens wherever three curves cross: falling cost of legibility × rising cost of illegibility × a population whose incentives already favor clarity.\n\n## §1 Expectations — what to actually expect, calibrated\n\n1. **Decadal timelines with AI-era compression uncertain.** Every precedent for substrate-grade infrastructure is slow: peer review took a century to universalize (it was near-absent at *Nature* until 1967 — [peer-review-and-the-reasoning-layer.md](peer-review-and-the-reasoning-layer.md)), arXiv took thirty years to become load-bearing, Wikipedia ~ten to become infrastructure. The one honest accelerant: the input friction that killed twenty predecessors is gone ([adoption-problem.md](adoption-problem.md)), and the crisis creating demand is moving on AI timescales, not institutional ones.\n2. **A steep engagement gradient forever.** The 90-9-1 shape is the baseline of every contribution system ever measured, confirmed on argument-mapping specifically by Kialo ([engagement-gradient-priors.md](engagement-gradient-priors.md)). A reader who never contributes is a success (the library test); the falsifier is within-tier decay, never tier populations. Do not expect — and never promise a funder — mass fine-grained participation.\n3. **Integration via intermediaries, not end-users.** Value scales with corpus coverage, not user count (Romer nonrivalry, [lowering-the-cost.md](lowering-the-cost.md) §5), which makes the largest plausible reader tier machine ([agents-as-a-consumer-class.md](agents-as-a-consumer-class.md)) and the human path run through facilitators, educators, reviewers, and institutions rather than a consumer app. Expect Deliberus to disappear into workflows the way checksums did.\n4. **Niches flip one at a time; society does not flip.** The eight legibility-pays conditions ([incentives-analysis.md](incentives-analysis.md) §2) name populations, not the public. Each stage must produce the asset the next needs (§6 there); skipping arrives empty-handed. Current best candidate for the first institutional flip: pre-submission review in drowning scientific publishing ([ai-slop-and-the-reasoning-layer.md](ai-slop-and-the-reasoning-layer.md) §2g).\n5. **Every rung re-breaks the instruments.** Taxonomies break per register (the three-runs law), mechanisms transfer free within an adversary class and need rebuilds across ([fractal-scales-and-temporal-frame.md](fractal-scales-and-temporal-frame.md)) — and no instrument here has yet met an interested party. Expect apparent success at each new scale to be the calm before that scale's adversary arrives.\n6. **The wager stays unresolved for years; the floor carries the value.** Legibility is the guaranteed deliverable each session; convergence is the bet the accumulated residue map tests ([convergence.md](../convergence.md)). Plan integration on the floor, not the wager.\n7. **Copying arrives with success.** The idea is now legible enough to imitate (the nanopub lineage already colonizes the scientific corner — [competitive-landscape.md](../competitive-landscape.md) § Nanoarguments; the well-kept wiki remains the honest rival — [structure-versus-scale.md](structure-versus-scale.md)). The durable moats are the accumulated corpus, the ontology depth, and the honesty discipline — the idea itself is not one.\n8. **Identity pressure recurs.** The corpus's own history is three successive legibility-shrinkages (product → personal tool → instrument) before the identity settled as public epistemic infrastructure. Expect every new audience to pull toward its own legible template, and expect the pull to be strongest exactly when funding or press attention arrives.\n\n## §2 The largest obstacles, frictions, enemies — ranked by depth\n\n1. **Power's need for ambiguity (the deepest enemy).** Institutions run on constructive ambiguity; some agreements survive only unspoken, and the ontology has no concept of load-bearing illegibility ([legibility-under-power-red-team.md](legibility-under-power-red-team.md) Hole 2). The actors with most to lose from legibility are precisely those whose position depends on vagueness — and they are not fringe: diplomacy, coalition politics, and organizational peace all run on the fudge. This enemy does not attack; it declines, obstructs, and refuses to adopt. Vagueness is armor (star three), and the powerful wear the most of it.\n2. **The attention economy (the largest enemy by mass).** The rival infrastructure for collective attention is engagement-optimized feeds with trillions in capex, optimizing for exactly what Deliberus refuses: speed, outrage, tribal reward. Stage-A's binding constraint is attention, not extraction ([incentives-analysis.md](incentives-analysis.md) §6b), and the competitor for every reader-minute is a machine tuned to win it.\n3. **Motivated human cognition (the most human enemy).** Comprehension capacity *worsens* polarization on contested questions; only curiosity reverses it ([curiosity-as-growth-fuel.md](curiosity-as-growth-fuel.md)). A gap displayed without a move produces avoidance, not descent (PACE). The tool's premise — that people want to understand why others disagree — is true only of a curious minority; the design answer (recruit the curious as ambassadors, never show a gap without a move) mitigates but does not dissolve this.\n4. **Strategy-class adversaries, arriving with success.** Gaming, sybils, selective legibility, weaponized decomposition, astroturf graphs, sacredness-as-shield — the full inventory in the two red teams ([convergence-wager-red-team.md](convergence-wager-red-team.md), [legibility-under-power-red-team.md](legibility-under-power-red-team.md) Holes 3+5), now with machine-speed variants empirically live in the wild (agent collusion, prompt injection — [the-scrutiny-gap.md](the-scrutiny-gap.md), [ai-slop-and-the-reasoning-layer.md](ai-slop-and-the-reasoning-layer.md) §2b). No coordination defense exists at any layer yet, and publishing detectors arms the adversary. The standing answer is payoff-removal before detection (fair meiosis), and it is a design discipline, not a solved problem.\n5. **The steering hazard from within.** LLM mediation can manufacture the very convergence the wager hopes for (the DeepMind result; five known flattening mechanisms, each blind to the others). If uncaught, the tool fabricates its own success signal and — when caught by others — discredits the whole class. The counter-instruments exist and must retain the power to falsify.\n6. **Operator fragility and the reflexivity gap.** A solo, currently unfunded operator; one unpaid bill already degrades the pipeline; no epistemic constitution, no governance story beyond the operator, fork rights without data rights ([legibility-under-power-red-team.md](legibility-under-power-red-team.md) Hole 7). The rung where the project would first be caught defecting from its own thesis is its own governance.\n7. **Extraction-without-reciprocity.** Stack Overflow's fate: AI consumed the value, contribution collapsed. The agent-readable surface maximizes harvestability; what an AI reader owes the graph is unanswered ([incentives-analysis.md](incentives-analysis.md) §7), and the graph's content still comes from humans.\n8. **Law and legitimacy at scale.** The permanent attributed record (Hole 6: supersession is not de-attribution; no deletion policy, no right-to-recant), ratification debt on mass-mapped named persons, and — at European scale — data-protection exposure on mapped opinions of identifiable people. Unremarkable at 25 sources; a legitimacy problem at 10,000 ([incentives-analysis.md](incentives-analysis.md) §6b).\n9. **Entropy (the quiet enemy).** Nobody's incentive is maintenance; strength is age-blind; a premise cannot go stale and a terminus cannot expire. The graph rots by default, and rot is the one adversary that never has to try.\n\n## §3 The largest drivers, incentives, allies — ranked by force\n\n1. **The AI epistemic crisis itself — the adversary IS the demand.** Machine persuasion out-runs human checking ([the-scrutiny-gap.md](the-scrutiny-gap.md)); machine slop drowns the scientific literature, closes preprint servers, and has publishing heads calling for radical change ([ai-slop-and-the-reasoning-layer.md](ai-slop-and-the-reasoning-layer.md)). Every worsening strengthens the case for checking infrastructure: Deliberus is counter-cyclical to epistemic decay, and the missing-layer absence now has a body count in the register funders read.\n2. **Machine readers as a consumer class.** Agents need dispute-structure that nothing else holds; value scales coverage × queries; structured reasoning is *distribution* in an AI-mediated information environment — whoever's reasoning is structured gets represented; whoever's is not gets summarized by someone else ([agents-as-a-consumer-class.md](agents-as-a-consumer-class.md), condition L8). This driver did not exist three years ago and grows with every model release.\n3. **The falling cost curve of legibility.** The friction that killed every predecessor is dissolved and keeps falling (Wright's law — [lowering-the-cost.md](lowering-the-cost.md)); the idea waited fifteen years for its parts, and the parts keep improving ([complexity-transitions-and-why-now.md](complexity-transitions-and-why-now.md) — the why-now argument is the strongest of the three).\n4. **Populations where clarity already pays.** The misrepresented, the audited, grant applicants, analysts, disagreeing counterparties with a decision at stake, anyone arguing with themselves (the eight conditions) — plus the supply-side exposure inversion: for already-published authors, mass mapping converts the contribution decision from courage to correction ([incentives-analysis.md](incentives-analysis.md) §6b). And Community Notes proves the cross-tribal incentive design works in production at hundreds-of-millions scale.\n5. **Institutions in crisis seeking legitimacy.** Drowning peer review (the pre-submission entry needs no institutional buy-in); deliberative-democracy practice (citizens' assemblies as facilitation-shaped allies); the EU open-internet funding ecosystem (NLnet already funds argumentation infrastructure — direct precedent); the AI-safety funding ladder (SFF, the Transformative AI Fund, Open Philanthropy) where the honesty-substrate need is independently argued from the machine side (Bengio/LawZero — [bengio-safety-from-honesty-and-deliberus.md](bengio-safety-from-honesty-and-deliberus.md)).\n6. **The curious, wherever they are.** Curiosity is the one disposition that reverses motivated reasoning, and it cuts across communities and tribes — the recruitment filter and the ambassador corps ([curiosity-as-growth-fuel.md](curiosity-as-growth-fuel.md)). The founder's own network already contains a documented 2011–2023 tier who engaged with the idea itself.\n7. **Adjacent builders as interop peers, not rivals.** The nanopub/Knowledge Pixels lineage (fact-layer interchange), discourse-graph and open-review communities, argumentation academia — each stalled on the friction Deliberus dissolves, each a potential organ in the same body ([science-reasoning-infrastructure.md](science-reasoning-infrastructure.md)).\n8. **The trust-collapse moment.** A culture losing default trust is a market for trust *infrastructure* — verify-then-trust mechanisms (Community Notes mainstreaming, open-source verification culture) are normalizing exactly the posture Deliberus institutionalizes ([the-missing-layer.md](../the-missing-layer.md)).\n\n## §4 The endogenous race, and where integration actually happens\n\nThe structural fact underneath both lists: **the same AI wave powers each enemy and its corresponding ally.** LLMs create the slop and dissolve the input friction; machine persuasion creates the checking demand and the checking tools; agents are the largest reader and the sybil threat; the attention economy's engines and the extraction pipeline are the same class of model. So the race is endogenous — there is no waiting it out, and no stable lead. Integration happens where three curves cross: the **falling cost of legibility** (technology, ours to ride), the **rising cost of illegibility** (the crisis, not ours to control, currently steepening), and a **population whose incentives already favor clarity** (ours to find — the eight conditions are the map). The strategy the corpus already converged on — floor-first, niche-by-niche, instruments-before-scale, payoff-removal-before-detection, confession-always — is exactly what riding that crossing looks like from inside.\n\n## §5 Funding-writing cross-reference (per founder, 2026-08-20)\n\nThe pieces of this synthesis and its feeders that belong in applications: the **why-now body count** (§3.1 + [ai-slop-and-the-reasoning-layer.md](ai-slop-and-the-reasoning-layer.md) §4's closing positioning paragraph) for every application's motivation section; the **six class-removals** (slop doc §4) as the mechanism story for NLnet (substrate/standards register) and the Transformative AI Fund (epistemic-security register); the **NLnet precedent** (Nanoarguments) + nanopub-interop offer for the NLnet application specifically; the **honest expectations** (§1 here — steep gradient, decadal integration, floor-not-wager) as the anti-overclaim ballast that keeps applications inside the corpus's honesty discipline; and the **enemies list** (§2) as the risks section no reviewer will have to write for us — a proposal that names power's need for ambiguity, the steering hazard, and its own governance gap as risks reads as the project keeping itself honest, which is the headline register. Worksheets live in `.private/funding/` (born-private); this doc is the public-side source they quarry.\n\n## Cross-references\n\n[incentives-analysis.md](incentives-analysis.md) (conditions, stages, supply-side) · [engagement-gradient-priors.md](engagement-gradient-priors.md) · [adoption-problem.md](adoption-problem.md) · [legibility-under-power-red-team.md](legibility-under-power-red-team.md) + [convergence-wager-red-team.md](convergence-wager-red-team.md) (the enemies at full strength) · [the-scrutiny-gap.md](the-scrutiny-gap.md) + [ai-slop-and-the-reasoning-layer.md](ai-slop-and-the-reasoning-layer.md) (the crisis-as-demand) · [agents-as-a-consumer-class.md](agents-as-a-consumer-class.md) · [lowering-the-cost.md](lowering-the-cost.md) + [complexity-transitions-and-why-now.md](complexity-transitions-and-why-now.md) (the curves) · [curiosity-as-growth-fuel.md](curiosity-as-growth-fuel.md) · [fractal-scales-and-temporal-frame.md](fractal-scales-and-temporal-frame.md) (adversary classes, rung discipline) · [governance-and-financing.md](governance-and-financing.md) · [the-missing-layer.md](../the-missing-layer.md) · [civilizational-vision.md](../civilizational-vision.md)\n"}