{
  "title": "Articles/prediction-markets-are-not-enough",
  "caption": "Prediction Markets Are Not Enough",
  "slug": "prediction-markets-are-not-enough",
  "tags": [
    "article",
    "citation-economy",
    "hermes-published",
    "pack-17",
    "prediction-markets",
    "published"
  ],
  "canonical_url": "https://mosiah.org/articles/prediction-markets-are-not-enough/",
  "interactive_url": "https://mosiah.org/#Articles%2Fprediction-markets-are-not-enough",
  "markdown_url": "https://mosiah.org/articles/prediction-markets-are-not-enough.md",
  "json_url": "https://mosiah.org/json/prediction-markets-are-not-enough.json",
  "fields": {
    "sort-date": "2026-05-12T15:45:00Z",
    "caption": "Prediction Markets Are Not Enough",
    "created": "20260512152905656",
    "modified": "20260512152905656",
    "tags": "article hermes-published published prediction-markets citation-economy pack-17",
    "title": "Articles/prediction-markets-are-not-enough",
    "type": "text/vnd.tiddlywiki"
  },
  "text": "//Related:// [[sources|Article Sources/prediction-markets-are-not-enough]] · [[notes|Article Notes/prediction-markets-are-not-enough]] · [[metadata|Article Metadata/prediction-markets-are-not-enough]] · [[Published Pieces]]\n\n! Prediction Markets Are Not Enough\n\n//Prediction markets can price some outcomes. Citation economies can remember the discourse that made outcomes interpretable.//\n\nPrediction markets are one of the cleanest intellectual technologies of the last generation. They force a question to resolve. They make confidence costly. They compress many private judgments into a price. They punish some forms of empty punditry. A good market can tell you, in one glance, that a crowd of motivated forecasters thinks an event has a 37% chance of happening by a specified date.\n\nThat is useful.\n\nIt is also nowhere close to enough.\n\nPrediction markets answer narrow resolution questions. They are good at “Will X happen by Y?” They are much weaker at “What should we be paying attention to?” “Which source is aging well?” “Which distinction mattered before people had language for it?” “Which frame will survive contact with reality?” “Who noticed the causal hinge early?” “Which claim became important because later discourse depended on it?”\n\nPublic cognition is not only a question of predicting outcomes. It is also a question of allocating attention.\n\nThe world is full of facts that will not become prediction-market contracts. Some are too vague, too early, too qualitative, too politically sensitive, or only matter after a future event reinterprets them. Some are not binary outcomes at all but distinctions, analogies, source trails, explanatory frames, patterns of institutional behavior, warnings, objections, and conceptual inventions.\n\nA prediction market can ask whether a war will begin by a certain date. It cannot easily track who saw the incentive structure forming years earlier, which source preserved the relevant supply-chain detail, which analyst distinguished theater from capability, which prior analogy misled people, and which early critic was ignored because they lacked status.\n\nThat missing layer is the citation economy.\n\nA citation economy tracks what later discourse depends on. It is not merely a bibliography. It is an attention and provenance system. It records which artifacts, sources, voices, claims, distinctions, and frames become useful to future work.\n\nPrediction markets are about settlement. Citation economies are about uptake.\n\nThe difference matters because public intelligence does not advance only by making correct bets. It advances by preserving the path by which people came to understand something. The market price may be right, but if the reasons are lost, the public has not learned very much. A price without provenance is a weather reading. A citation graph is the weather system.\n\nPrediction markets also depend on question design. A badly worded market can produce a clean price for the wrong question. Traders may optimize around resolution criteria rather than truth. A question can be technically resolved while failing to capture what people actually cared about. The market disciplines some kinds of ambiguity by forcing settlement, but that same force can erase the living ambiguity that made the issue important.\n\nThe citation economy is better suited to ambiguity because it does not need every contribution to be converted into a binary proposition. It can remember a warning, a frame, a source, a contradiction, a map, a timeline, a critique, a voice clip, a vtext, or a distinction. Later artifacts can cite, extend, refute, or depend on it. The graph can show not only who was right, but what became relevant.\n\nThis is especially important for AI. AI systems are already strong in domains where feedback is clean: code, math, tests, benchmarks, proofs, constrained retrieval, formal verification. They are weaker where the hard question is salience. What matters? Which prior work is relevant? Which source is trustworthy? Which frame is new? Which claim is old but under-cited? Which distinction prevents confusion? Which voice should be heard now because it saw the hinge earlier?\n\nA full-spectrum intelligence system needs a public memory layer that preserves attention over time.\n\nCurrent AI often answers from a compressed public consensus. It can summarize what is already legible, but it does not naturally preserve who noticed what, when, under what conditions, with which evidence, and how later discourse used or ignored it. The model may know the topic, but it does not necessarily know the provenance economy of the topic.\n\nA citation economy changes that.\n\nWhen a vtext, source, podcast clip, essay, dataset, or claim becomes useful downstream, the system records the dependency. It does not require a human to manually write a bibliography. Agents search the graph, surface relevant priors, evaluate novelty, track age, note contradictions, detect falsification pressure, and cite prior work when later artifacts rely on it. Humans publish. Agents cite. The protocol rewards.\n\nThis creates a memory of attention.\n\nWho saw the issue early? Who preserved the source? Who supplied the decisive quote? Who introduced the useful distinction? Who repeatedly made confident claims that did not survive? Who corrected quickly? Who was late but loud? Who was ignored until reality made their frame unavoidable?\n\nThese are not prediction-market questions. They are public-cognition questions.\n\nPrediction markets reward the person who makes money on the market. Public discourse needs other reputational dimensions: source quality, explanatory compression, originality, correction speed, generosity with prior art, resistance to hype, willingness to distinguish evidence from interpretation, and ability to notice the right thing before it is fashionable.\n\nA society that only prices forecasts will still miss the people who build the conceptual equipment that makes good forecasting possible.\n\nSource reputation matters. A source is not merely true or false in isolation. It ages. A source that seemed marginal may become central. A high-status source may repeatedly fail under later scrutiny. A podcast clip may contain the first public statement of a position that later becomes consensus. A local reporter may preserve the detail that national media misses. A technical blog post may hold the real explanation while glossy coverage repeats simplified nonsense.\n\nA citation economy lets sources accumulate a track record through use. Not all citations mean agreement. A later artifact may cite a source as evidence, prior art, a competing frame, an error, a refuted claim, or a historically important example. Citation is not endorsement. It is dependency accounting.\n\nThat dependency accounting is what turns public discourse into intellectual property. In ordinary social media, old posts usually decay into the archive or become liabilities. In a citation economy, old work can become an asset if future work depends on it.\n\nPrediction markets and citation economies are complementary. Prediction markets should exist. They are useful instruments for probabilistic discipline. But they cannot bear the full weight of public cognition. They do not decide which questions deserve to be asked. They do not preserve the genealogy of ideas. They do not reward the person who found the source that made the later market legible. They do not track which conceptual frames survived.\n\nThe future public-intelligence stack needs both.\n\nPrediction markets can price some outcomes.\n\nCitation economies can remember the discourse that made the outcomes interpretable.\n\nThe next great public-intelligence system will not merely predict the future. It will remember how the future became thinkable.\n"
}