10F Taxonomy

The 10F Consortium produces work in three distinct registers. Each has its own vocabulary, its own audience, and its own editorial constraints. The interesting findings live not inside any one register but at the edges between them — what crosses from one to the next, and what doesn’t.

Three registers

Internal operational language — what the Consortium uses to coordinate itself. The named moves, working formats, and rituals that constitute its production process. This vocabulary is functional. It exists to do work, not to be seen. Substrate: 30 internal meeting transcripts.

Internal conceptual language — what the Consortium uses when thinking together substantively. The diagnostic moves, the peer-thinking, the first-naming of patterns members notice in the world. This vocabulary is intellectual. It exists to think with, before any external commitment. Substrate: parts of the 30 meeting transcripts (where the entity discusses content rather than process), plus the 9 Singapore Pre-talks from September 2025.

External language — what the Consortium publishes. Forecasts, framing essays, named concepts; the public-facing analytical product. This vocabulary is performed. It exists to be read. Substrate: three framing pieces and ten forecast files.

The three registers don’t replace each other. The same people move between them session by session. Each is necessary; none is sufficient. What matters is the gradient between them: what crosses, what gets stripped, what gets added.

The 2×2

The three registers can be plotted across two axes:

• Reference — language about how the entity works (operational) vs. language about what the entity sees in the world (conceptual).
• Audience — language used internally (among ourselves) vs. language used externally (to the audience).

	Internal	External
Operational	How we work — Procedural vocabulary: methods, formats, working rituals.	— mostly empty — The Consortium does not publish about how it works.
Conceptual	How we see — Diagnostic vocabulary: substantive thinking, first-naming of patterns.	What we publish — Forecasts, framing, named concepts, the analytical product.

The gap

The empty quadrant is the central finding. The Consortium has a rich operational vocabulary that does not appear in its public work. The procedural innovation is invisible in the product. Three consequences:

• Readers of the public work see what the Consortium thinks about the world, but not how it thinks together.
• The procedural distinctiveness — which may be the Consortium’s most transferable contribution — does not transfer.
• The published work reads as a product, not a practice. The practice exists in the substrate, accessible only to people in the room.

This is editable: the Consortium can choose to surface its operational vocabulary in its public face, keep it as competitive advantage, or partially translate selected moves into the public register. The taxonomy is diagnostic first, design second.

Method

The corpus carries 10F in the three registers above, in increasing distance from the entity’s working life. Each move outward (operational → conceptual → external) strips affect, strips hedging vocabulary, raises confidence in detail, and substitutes editorial overlay for working substrate.

Sources read:

• Operational substrate — 30 behavioral summaries, Field 3 Recurring language (already analyst-extracted in Phase 0).
• Conceptual substrate — 9 Singapore Pre-talks, full read for register, recurring vocabulary, rhetorical moves and hedges; plus the conceptual-content portions of the 30 meeting transcripts.
• External substrate — the published GitHub repository: three framing pieces and ten forecast files (F04, F07, F09 in full; others sampled).

For the pre-talk → forecast mapping, each pre-talk was paired against its likely forecast home and compared at three levels: vocabulary survival, what got stripped, what got added in the writing phase. Then cross-cutting patterns lifted out and tested against the broader register-translation hypothesis.

The shared / portable vocabulary

Terms that survived translation from internal to external. The entity’s actual durable claims about the world.

Term	First conversational appearance	Published usage
From X to Y (the structural device)	2025-06-06 (“From X to Y / shifts” as preferred framing)	The structural backbone of every forecast title and abstract
One game to many games	2025-09-26 (central shorthand, contested but kept)	Title and central thesis of the framing essay
Norm cascade	2025-10-31 (“name the mechanism rather than the outcome”)	Section header in One Game to Many Games; appears in F01
Permission structures	2025-06-20 (imported from US political discourse)	Recurs across forecasts as a mechanism term
Espoused vs. lived	2025-09-26 (“espoused ideal”)	Implicit in every Old Assumptions / Emerging Norms split
Strategic intelligence (over “foresight”)	2026-03-20 (Igor’s late proposal)	Adopted in One Game to Many Games: “10F is a network of foresight practitioners and strategic analysts”
Blind spots	Recurring across sessions	Formalised as the Strategic Blind Spots section in every forecast

Read. Seven terms make the round-trip. Most are mechanism words — they name how things move, not how the entity feels about them. The portable vocabulary is the analytic vocabulary. Everything affective, methodological, or self-descriptive is left behind.

Singapore layer — summary

The 9 Singapore pre-talks are the middle register: oral substantive content from each domain expert, performed to peers under Chatham House Rule. They are the most likely direct oral source for the published forecasts.

Pre-talk → forecast mapping

Pre-talk	Primary forecast	Strength of mapping
Aarathi — transactional diplomacy	F04 Strategic Situationships	Strong (structural backbone)
Dexter — re-materialization & gold	F09 Money Unbundled	Strong (thesis core)
Lina — migration five Ds	F07 People as Asset Class	Strong (taxonomy lifted)
JD — sustainability & collective action	F05 Fragmented Adaptation	Moderate (diagnosis transferred, moral charge dropped)
Mick — AI & jobs	F10 / F07 / framing piece	Weak (vocabulary scattered, no home)
Scott — risk maxing	F06 / F02 / framing piece	Weak (the lens, not a destination)
Jake — network states	F04 / F06 / F10	Very weak (largely orphaned)
Susan — art market / quiet luxury	F01 / F09 / framing piece	Very weak (substrate only)
Igor — Gen Z behaviors	F02 / F03 (faintly)	Very weak (largely orphaned)

Five of nine pre-talks did not become forecasts. The orphans cluster into two groups:

• Lens talks (Scott, Mick): named patterns the writers used as analytic tools rather than as forecast-titles.
• Domain-vibrancy talks (Susan, Jake, Igor): named what people are building / making / doing — cultural collectibles, art markets, new governance models. These are positive-signal talks. The forecasts are negative-shift documents.

This is itself a finding: the v1 forecasts have a systematic negative-shift bias that filters out positive-signal substrate. Igor’s strategic optimism, Susan’s art-market interest as a creative-economy story, Jake’s new-governance-experimentation enthusiasm — none of these get a home.

Coinages

The pre-talk speakers had their own domain coinages — transactional diplomacy, re-materialization, risk maxing, toll gates, sanctions theater, autonomic core, future burrito, quiet luxury, seamless is sameness, present archaeology, five-D migration framework. Not one survived to publication.

Equally, all nine Capitalised Concepts from the publication side (Connected Sovereignty, Cognitive Arbitrage, Strategic Situationships, Engineered Opacity, Curated Invisibility, Statistical Fiction, Code-switching capability, Corridor Capabilities, Selective interdependence) return zero hits across the 9 pre-talks.

Verdict: the Capitalised Concepts are decisively writing-phase. Editorial overlay written on top of the substrate, not out of it. The branded vocabulary did not exist when the consortium was speaking among itself, even in its most substantive mode.

Publication-only inventory — what 10F added when writing

Roughly four clusters.

Coined proper-noun concepts (capital-letter constructs)

• Connected Sovereignty
• Cognitive Arbitrage
• Corridor Capabilities
• Strategic Situationships
• Engineered Opacity
• Curated Invisibility
• Statistical Fiction
• Code-switching capability
• Selective interdependence

These emerged in the writing phase, not the talking phase. The Capitalised form is a register move signalling: this is a coined term you can take with you.

Methodological / institutional vocabulary

• Sensemaking / collective sensemaking
• Horizon scanning
• Scenario planning / Scenario pathways (Transformation / Stagnation / Collapse)
• Chatham House Rule
• Practitioner network
• Second-order effects
• Pattern recognition over exhaustive citation
• Directional clarity over probabilistic hedging
• Consensus, not unanimity
• Blend of voices

These borrow the legitimating vocabulary of the foresight-and-strategy professional field. None appear in the conversational lexicon — the entity does not call itself a “practitioner network” when no one is watching.

Modesty-via-formula constructions

• A map, not a route
• Better questions, not a playbook
• Snapshots, not predictions
• Describes terrain, doesn’t prescribe moves
• Starting points, not endpoints
• Not a grand theory
• A fit-for-purpose tool
• Shared language: if others are using it, you can too

Conversational equivalent: none. Conversationally the entity says “what the fuck is going on”, “rules have collapsed”, “audacious.” Publication uses formulaic modesty; conversation uses direct claim. The modesty is audience-facing politeness, not a stance held internally.

Audience-coded distribution language

• Decision-makers
• Leaders and teams
• NGOs, multilaterals, corporations, government agencies and philanthropies
• Civil society
• Across six continents
• Policy design
• Strategic planning

The audience comes into view in publication. Conversationally the entity rarely names its audience this way — it talks about “people who get it”, “nobody knows who we are”, “chasing reach”, “the Us side.”

Internal-only inventory — what 10F talks about but does not publish

Six clusters.

Working-method names (operational register)

• Lightning talks — the entity’s signature pre-work format.
• Wireframes — “wireframes before content” as method.
• Murder walls / case files — the metaphor for the entity’s process arc (associative chaos → curated artefact).
• Speed dating / speed peer review — the original convening method.
• Ten-week campaign (replacing “launch”) — the rollout format.
• Write what excites you — the editorial principle.
• Minimum viable [understanding / meeting point / participation] — the working-state shorthand.
• Concrete to-do vs. open-ended role invitation — the entity’s hard-won lesson about volunteer coordination.
• Workflow — the entity’s recurring preoccupation.
• The stack / track list — the entity’s own framing of its output as a curated set.

The entity’s most distinctive working-style vocabulary is invisible in its public face.

Self-identity vocabulary

• Punk / punk rock / system punk
• Just clean enough
• Not shiny
• Raw / handmade
• Minimalist / stripped back
• Grassroots authenticity
• Weird science
• Berlin house party
• Guerrilla tactics / gorilla group
• The bad cop
• Quincy Jones moment
• Luther’s 99 theses

Conversationally the entity has a vivid, opinionated, culturally-located self-description. Publication renders this as “a network of foresight practitioners and strategic analysts”.

Affective vocabulary

• What the fuck is going on — the core question, conversationally.
• Nobody knows who we are — the recurring baseline acknowledgment.
• Audacious — proudly quoted from the funder.
• Secret sauce — what would be lost online.
• Free range — the desired quality of thinking.
• Spice — a scalar for provocative content.
• Chasing reach — the strategic anxiety.
• Sober scenarios — the rejected category.
• Handholds — what the entity wants to give its audience.
• Soft launch / quietly and greedily — the distribution mood.
• What could possibly go wrong — the entity’s self-deprecating risk frame.

Publication carries no affect.

Funder / contract vocabulary

• The money’s sitting in a transparent box
• Embargo (as both constraint and strategic variable)
• Operationalize that date range
• Summer-surplus bargaining
• Plain brown wrapper
• Bootstrap (because the funder has no marketing team)

In publication: one line of attribution framing the relationship as enabling independence. One instance of the broader pattern — the funder-as-managed-object becomes funder-as-acknowledgement.

Format-rejection vocabulary

• The r word (“I’m trying not to use the r word”) — recurring resistance to “report”.
• Not just a PDF — explicit format rejection.
• Case files (not reports).
• The stack / track list (not a monograph).
• Murder walls (not a literature review).

What got published: ten formally-structured Forecast files closer to academic articles than to any of the conversational alternatives. The conversational rejection of conventional form did not survive into the published form.

Imported references — the entity’s bench

• Quincy Jones moment — convening talent.
• Luther’s 99 theses — the desired output structure.
• Berlin house party — the desired convening atmosphere.
• Ambient intimacy (Lisa Reichel, early Twitter discourse).
• System punk (John Robb).
• P-doom (AI safety community).
• Permission structures (originally US political discourse — this one survived).
• The medium is the message (McLuhan, invoked by the journalist figure).

These references locate the entity in a specific intellectual-cultural milieu. Publication speaks from no milieu in particular — it borrows from the academic / professional foresight field’s neutral register.

Per-pre-talk analysis

Nine sections, one per Singapore lightning talk, with mapping and oral-to-published comparison.

1. Aarathi Krishnan — Transactional Diplomacy: The Collapse of Rules-Based Order

Maps to: F04 Strategic Situationships (primary), bleed into F08, F09.

Distinctive oral terms: transactional diplomacy · toll gates instead of guardrails · final sale · sanctions theater · snake oil (Malaysia/Thailand/UAE energy case) · burns long-term leverage · quiet signals vs. the noise of US-China-Russia · selling the future for something · reading between the lines · racketeering writ large (Madeline) · 1989 but everywhere (Madeline).

What survived to F04: The structural claim survives almost exactly — transactional bilateral arrangements replacing values-based multilateral commitments, with verification and reversibility displacing value alignment. The “selling-the-future” point survives as a strategic blind spot.

What got stripped: Toll gates, sanctions theater, snake oil, racketeering, final sale, 1989 but everywhere — every vivid named image. Even transactional diplomacy itself, the speaker’s headline coinage, does not survive: replaced by Strategic Situationships. Aarathi’s frank verdict and methodological self-disclosure (“we just get clients and ask Why? Why? Why? Why? Why?”) are gone.

What got added: Strategic Situationships (unusually witty / catchy contrast to Aarathi’s blunt transactional diplomacy). Corridor optionality / corridor portfolio. Legibility becomes liability. Code-switching capability. The whole AREAS framework. Scenario triplets with adaptive strategies for four actor types. Modesty hedges.

Key transformation: A direct, single-image oral thesis is upgraded to a branded coinage and embedded in an architectural framework. The oral talk is angrier and more confident; F04 is cooler and more comprehensive. The published version is better packaged; the oral version is better-aimed.

2. Dexter Bullard — Geo-Economics, Re-materialization, and Gold

Maps to: F09 Money Unbundled (primary), structural contribution to F08, F10.

Distinctive oral terms: re-materialization (headline coinage) · stacks and systems · Chinese over capacity (reframed) · Walmart or Amazon at state level · 5 billion person car market by 2040 · full stack of the Chinese social contract · finance is the big choke point · Marxist materialist terms / hard power and leverage · gold as a bridge / contingency plan / pillar · landmine that everyone’s standing around · BRICS pay · gold is genuinely trustless.

What survived to F09: The thesis survives, recast: money’s three functions unbundling, gold returning as a non-counterparty-risk asset, the weaponisation of dollar infrastructure as driver. The phrase “materiality as a substitute for trust” is essentially Dexter’s argument repackaged. Mick’s pushback survives as the Old Assumptions framing.

What got stripped: Re-materialization as a term does not appear in F09 (replaced with the more academic functional unbundling and return to materiality). The Marxist-materialist self-positioning is gone. The China-as-Walmart image is gone. The landmine image is gone. The unhedged claim (“this is way too China-centric to be an explicit narrative anywhere”) is gone. Dexter’s frank methodological note — that he is deliberately building this slowly — is gone.

What got added: “Money things” as a term of art. Three-functions-unbundling more crisply structured. Specific signal stats (CIPS reaches 185 countries, RWA tokenisation $26.5B). Three scenarios. Local adequacy vs. universal reach.

Key transformation: Dexter’s talk is deliberately half-baked (“loose thoughts,” “more than half baked”). F09 is the opposite: confident, structured, citation-rich. The published version removes the speaker’s epistemic caution. The oral hedges (“the answer isn’t right, but the pathway of the question”) are dropped — F09 commits.

3. Lina Srivastava — Global Migration Systems

Maps to: F07 People as Asset Class (primary), bleed into F06, F03.

Distinctive oral terms: the five D’s — denial, detention, deportation, disappearance, data (death added by Madeline) · othering / othering frames (load-bearing analytic term) · refoulement · convenience refugee / comfort refugee · GEO Corporation · meaningful refugee participation · narrative infrastructure · forensics (DNA + cultural) · white supremacy / white nationalism explicitly named as architecture · permission structures · people don’t want to be known as refugees · strategically optimistic.

What survived to F07: The taxonomy survives almost intact — denial/detention/deportation/disappearance/data become the empirical backbone of F07’s “Emerging Norms.” Convenience refugee survives. Private detention contractors survives. Skill corridor / sovereign labour idea survives. Permission structures survives.

What got stripped: Othering — the word is gone from F07. White supremacy / white nationalism — Lina’s explicit architectural diagnosis is absent. The personal stakes she disclosed (“technically I am a birthright citizen… eventually might be stripped of my status”) — gone. The methodological note “I’m not a frontline activist, I’m an advocate” — gone. The dual register (“globally I am feeling deep anguish… longer term I feel hopeful”) — gone; F07 carries no first-person affect.

What got added: Mobility apartheid. Skill corridor lock-in. Population trading. Migrant yield forecasting. Algorithmic cultural scoring. AGI hype devalues human labour (cross-talk synthesis with Mick). Three scenarios. National-template naming as Architecting exemplars. Dataset-style sourcing.

Key transformation: Lina’s oral talk has the single most clearly transferred analytic structure of any pre-talk — her 5-D framework is essentially F07’s spine. But the political indictment (this is built on the exploitation of Black and brown bodies, on white supremacy) is replaced by the technocratic indictment (algorithmic scoring, infrastructure of containment). F07 names the systems but scrubs the moral charge.

4. Mick Costigan — Will AI Take All the Jobs?

Maps to: F10 Sovereign Systems (weak — Mick’s talk does not have a primary forecast home). AI-and-jobs content bleeds across F07, F10, framing piece. There is no F-on-AI-and-jobs.

Distinctive oral terms: the pyramid / entry-level employees first · Dorcas (Dwarkesh) and Jim Cramer as named avatars · hammer available to the accountants · Wild West capitalism / post-Tiananmen (Pettis-via-Asia-Society) · the economy as a World War One battlefield at night (Brian Arthur, central image) · jagged intelligence · legible, specifiable, verifiable work · autonomic core (his model for future companies) · agent-first enterprise · taste / discernment / meaning-making / trust is the new oil · frontier models vs. smaller models · bespoking AI in our back garden.

What survived: The framework (restructure / automate / reskill / new demand) appears in F10’s general logic. Frontier-model gap, chip access as choke point, China-vs-US AI capability framing — all present in F10. AGI hype devalues labour is lifted near-verbatim into F07. Compute access as new diplomatic currency survives in F10.

What got stripped: Almost all the named figures (Bannon, Cramer, Amodei, Dwarkesh, Brian Arthur, Ogilvy, Pettis). The WWI battlefield at night metaphor — gone. Autonomic core — gone. Agent-first enterprise — gone. Mick’s most interesting positive imagination point (“we’re going to end up with shitty worlds we didn’t have to end up in if we don’t have enough imagination”) — gone.

What got added: F10’s central frame of sovereign systems is much larger than Mick attempts orally. State-capital fusion. Performed sovereignty, practical porosity. Hard borders, soft flows.

Key transformation: Mick’s talk is the least cleanly mapped of the nine. It contributes vocabulary to multiple forecasts but does not become any single one. The positive imagination note — Mick’s actual editorial position that the discourse over-indexes on pessimism — is completely absent from the published forecasts.

5. Scott Smith — Risk Maxing

Maps to: No single forecast. Closest fits: F06 (Tactical Shape-shifting), the framing piece, with substrate feeding F02.

Distinctive oral terms: risk maxing (headline coinage — “strategic abandonment of safety mechanisms to gain a competitive advantage through deliberate extremism with cascading effects”) · risk maxing as alpha · existential risk-taking · move fast and break things, to a higher level · coalitions are for suckers when you can mobilize a narrow but intense space · permission structures (credited to Lina) · crisis as strategy · temporal risk orientation / time compression · competing dominionism · imminentize the eschaton · system punk (John Robb) · super actors · Leroy Jenkins · algorithmic extremophilia / tubification (Madeline) · 9/11 as system punk.

What survived: Almost none of the named concepts. Permission structures survives. Norm-breaking as strategy survives as a generic logic across F02 and F06. The substantive observation that small concentrated groups can break large systems is implicit in F06.

What got stripped: Risk maxing itself. System punk. Super actors. Leroy Jenkins. Algorithmic extremophilia. Competing dominionism. Imminentize the eschaton. The whole bibliographic substrate (Castells, Robb, Zeckhauser, Douglas Adams). The internal-jokey register. The frank acknowledgment that “we’re talking about something larger than your piece of it.”

What got added: F02 Ideological Fog and F06 Tactical Shape-shifting are bigger frames than risk maxing. Where Scott names the strategy, the forecasts name the landscape that strategy operates in. Statistical fiction, synthetic positioning, swiss cheese policy windows — none are in Scott’s talk.

Key transformation: This is the clearest case of an oral talk that did not become a forecast. Risk maxing is a behavioural pattern; the forecasts work at the system shift level. Scott’s framing was upstream of the forecast-naming exercise — it provided a lens for seeing the shifts but did not become one. Striking that the convener’s pet concept is the one that did not get published.

6. JD (Joseph) D’Cruz — The Rules of Sustainability: Collective Action in Crisis

Maps to: F05 Fragmented Adaptation (primary), bleed into the framing piece.

Distinctive oral terms: sustainability defined as thinking about others (“Fight me on that one”) · the death of thinking about others · the multilateral system actively dead · the whiplash back (DEI scrubbing, broligarchy) · Bezos and Zuckerberg competing over gun collection size · WEF is performative / a CNBC kind of place · the alliance for vs. the alliance against · conservatism = conservation (recovery of stewardship as non-left value) · the opposition has noise and traction, but not depth · performative actions online · strategically tactical / strategically optimistic (Igor’s gloss, accepted).

What survived: The systems-level diagnosis — multilateral collapse, alliance reconfiguration, COP forum’s structural veto — survives in F05. The point that finance/insurance sectors still care about the long term survives in F05’s Architecting / Resisting split. The culture-war alignment of sustainability point survives. The recovery of stewardship as a non-left value survives only faintly.

What got stripped: Sustainability defined as thinking about others — JD’s load-bearing personal definition does not survive. The multilateral system is actively dead is softened to “stalled” and “fragmented.” The frank affect (Bezos-Zuckerberg gun collection joke, WEF dismissal) — all gone. The alliance for as an open question is not posed in F05.

What got added: Loans not grants. Adaptation is protecting capital, not communities. Empirical specificity of COP30 outcomes. Climate-threatened areas declared uninsurable. Indigenous participation data.

Key transformation: JD’s talk does the moral framing (sustainability = caring about others; its erosion is the death of solidarity). F05 does the structural description. The moral and the structural are the same observation, but the published version trades moral charge for analytical neutrality. JD’s question — where is the alliance for going to come from? — is not answered and not posed in F05.

7. Jake Dunagan — Network States and New Governance Models

Maps to: Weak. No forecast titled From X to Network States. Substantive content bleeds into F04, F06. The pure network state concept is absent from the forecasts.

Distinctive oral terms: network states (Balaji’s definition, quoted) · non-spatial government (Bruce Sterling, 1995) · temporary autonomous zones / pirate utopias / Christiania / Arcosanti · own the cloud and touch the land (Balaji) · sovereignty light (Jake’s category for Prospera-type fiefdoms) · climate as a motivator for new cities · subscription-based residency · crypto first / law first as two playbooks · techno feudal fiefdoms · clusters of clusters (millions of semi-sovereign co-ops alternative) · nomadic cities (Jake’s pet idea) · bioregional financial facilities (Ariel) · vanity cities / vision cities.

What survived: Almost nothing as named concepts. The substantive observation that governance is fragmenting into competing models survives generally in F06. Freedom-cities / charter-cities as Trump-aligned policy direction is faintly present in F06. Climate refugee cities is faintly present in F07.

What got stripped: Everything specific. The whole taxonomy (digital-first free zones / sovereignty light / climate risk locations / cultural affinity / vision / vanity / pop-up / company towns / utopian communes) does not survive. The lineage Jake supplies (Hakim Bey → Sterling → Balaji) is absent.

Key transformation: Jake’s talk is largely orphaned. The lightning talk file even cuts off mid-discussion. The substantive content was either too speculative for v1 or absorbed into the Connected Sovereignty / state-level frame at much higher altitude than Jake operated.

8. Susan Cox-Smith — Quiet Luxury: The Art Market as Economic Signal

Maps to: No single forecast. Closest fits: F01 (freeports / undisclosed asset storage), F09 (art as alternative store of value), framing piece. The quiet luxury concept does not survive.

Distinctive oral terms: quiet luxury (title concept) · the auction market as a weak signal (methodological move) · the Jake Medi sculpture with a $70m estimate that didn’t sell (lead anecdote) · conspicuous consumption vs. privacy as ultimate commodity · freeports · International Financial Centres (IFCs) · 5000 of these zones exist globally · guarantees vs. reserve pricing · eat the rich · Patrick Bateman (Igor) · keystone species / apex predator (Madeline).

What survived: Freeports, IFCs, the operational opacity of wealth — survive in F01 as a strategic blind spot. The point that sanctions / IRS workforce decimation enables this survives. The apex predator as ecosystem indicator logic is absent in the forecasts.

What got stripped: Quiet luxury as a frame. The Jake Medi anecdote. The NFT collapse stats (93% drop in trading volume). The art-market specificity. Patrick Bateman. The art-as-weak-signal methodology.

Key transformation: Susan’s talk is a domain-specific weak-signal piece that contributes substrate to F01 but does not become a forecast. The register of the talk is itself unusual — a connoisseur’s diagnostic with no parallel in the published prose. The closest the published corpus comes to Susan’s register is the Patrick Bateman image (Igor raised in discussion), which is exactly the kind of cultural anchor identified as conversational-only.

9. Igor Schwarzmann — Gen Z Behavioral Patterns and Physical Engagement

Maps to: Weak to none. Gen Z themes touched in F02 / F03 but Igor’s substantive content does not become a forecast.

Distinctive oral terms: Gen Z as a set of behaviors, not a demographic · future burrito (NFG’s Simona) · truly omnichannel · car boot sales in London · Touchland · jelly cat · friction created meaning vs. friction is a choice · seamless is sameness (Igor’s anti-platform slogan) · ambient intimacy (Lisa Reichel) · present archeology (Igor’s coinage) · totemization (Scott’s gloss, accepted) · they’re not even thinking about the wall / end of history framing · strategic investment in longevity · attention beats money (Ezra Klein NY mayoral race) reframed as “the point is having something to say nobody else is saying.”

What survived: Almost nothing as named terms. The substantive observation that young people are aware of pre-algorithmic culture, that third spaces are reviving, that strategic optimism characterizes Gen Z — survives faintly in F02 and F03 but without Igor’s vocabulary.

What got stripped: Future burrito, truly omnichannel, Touchland, jelly cat, friction created meaning, seamless is sameness, ambient intimacy, present archeology, totemization. The whole case-study register (car boot sales, Hamptons newspaper, Nike-vs-small-coffee-retail) does not survive.

Key transformation: Igor’s talk is the most thoroughly orphaned of all nine. Its content is fundamentally about positive behavioral signals. The forecasts are fundamentally about negative system shifts. There is no v1 forecast on emerging cultural infrastructure / what young people are building — and Igor’s talk would have been the natural source.

Cross-cutting findings on register translation

A. Common stripping moves

The writing phase consistently strips:

1. Named avatars / quoted figures. The talks are thick with attribution: Brian Arthur, Jay Ogilvy, Michael Pettis, Steve Bannon, Jim Cramer, Dwarkesh, Dario Amodei, Bruce Sterling, Hakim Bey, Balaji, Lisa Reichel, Henry Farrell, Richard Zeckhauser, Barry Eichengreen, John Robb, Katie Martin, Patrick Bateman. The forecasts cite none of these. They cite systems (CIPS, UPI, BUIDL, EU Talent Partnerships) but not thinkers.
2. The speaker’s headline coinage. Transactional diplomacy, re-materialization, risk maxing, quiet luxury, present archeology, future burrito, autonomic core, sovereignty light — almost every speaker’s signature term does not survive. When the writing phase upgrades the same phenomenon to a published name, it picks a different coinage.
3. Vivid single-image metaphors. Toll gates, snake oil, racketeering, 1989 but everywhere, landmine that everyone’s standing around, WWI battlefield at night, Leroy Jenkins, Patrick Bateman, jelly cat. Forecast prose carries no metaphor at the image level.
4. The speaker’s frank verdict. First-person editorial positions are uniformly absent from the forecasts.
5. Methodological self-disclosure. “I’m no expert”, “loose thoughts”, “not very well thought out so far”, “my half-baked theory”, “this is a thing I’ve been working on.” The forecasts present themselves as finished knowledge, not as thinking in progress.
6. The speaker’s professional positioning. All nine speakers introduce themselves through their own work. The forecasts speak from no positional standpoint.
7. Affect at every register. Anger, frustration, anguish, wry humour, self-deprecation, enthusiasm. The forecasts are flat affect.

B. Common additions

1. The AREAS framework itself (Architecting / Resisting / Exploiting / Avoiding / Shaped). Publication-only structural device.
2. Scenario triplets with adaptive strategies for four actor types. Every forecast has three scenarios; the pre-talks have none.
3. Aggressive signal-counting and stat density. F09 has 137 jurisdictions, $94.2B stablecoin volume. F07 has 45B detention budget. The pre-talks have anecdotes but not the saturation. The published version adds data weight.
4. The Capitalised Concept coinage. (See C.)
5. Modesty formulas. “Snapshots, not predictions.” Absent from every pre-talk.
6. Cross-forecast referencing. Each forecast points to related forecasts. Pre-talks are domain-bounded.
7. Domain-specific authority signals. “Key Domains” at top, “Boundaries” at bottom — institutional framing devices absent from the pre-talks.

C. Capitalised Concepts: oral or written-phase?

The headline question, decisively answered: publication-phase invention. All nine Capitalised Concepts (Connected Sovereignty, Cognitive Arbitrage, Corridor Capabilities, Strategic Situationships, Engineered Opacity, Curated Invisibility, Statistical Fiction, Code-switching capability, Selective interdependence) return zero hits across the nine pre-talks. The named experts had their own vocabulary; the writing phase substituted branded terms during publication.

This identifies where in the workflow the branded vocabulary entered. Not in the lightning-talk phase. Not in the internal meetings. The Capitalised Concepts are a layer added by the small number of writers who turned the consortium’s collective material into the published forecasts. The forecasts are editorial reframings of expert thinking using a different vocabulary architecture.

D. The cultural anchors strip monotonically

The pre-talks carry some cultural-anchor register (Madeline’s “1989 but everywhere”, Scott’s “Leroy Jenkins” and “system punk”, Igor’s whole punk-vs-goth talk, Madeline’s “Patrick Bateman”). Less than internal meetings, more than published forecasts. The cultural anchors are sparser in the talks (each speaker brings a few) but cumulatively rich; the published prose strips them all. Strip is monotonic along the gradient.

E. The single most interesting finding — epistemic-stance inversion

The speakers explicitly hedge their pre-talks as “half-baked,” “loose thoughts,” “not well thought out yet” — and the published forecasts present the same material as confident, signal-cited analysis. The epistemic stance changes more than any individual term. The writing phase does not just upgrade the vocabulary; it upgrades the confidence level.

This is in tension with the published modesty formulas. The published version is meta-modest (we don’t predict outcomes) while being micro-certain (within each forecast, every claim is firm). The pre-talks are the inverse: micro-uncertain (every speaker hedges) while macro-confident (the entity is doing work that matters).

The modesty formulas are the price for upgrading internal confidence. The entity buys top-line humility so it can claim more in every individual sentence than the substrate ever did.

Reading the asymmetries

Seven things the gradient says about the entity.

1. The headline asymmetry is epistemic, not lexical.

The writing phase didn’t translate vocabulary — it inverted epistemic stance. Pre-talks: macro-confident + micro-uncertain. Publications: macro-modest + micro-certain. The trade is exchange, not stripping.

2. The entity has a richer working-method vocabulary than its publication acknowledges.

Lightning talks, wireframes, case files, minimum viable, the stack — all in the working register, none in publication. The entity’s most distinctive innovation is procedural, and the procedure is invisible in the product. The modesty gap holding firm across both lower registers.

3. The entity has a rich, culturally-located self-identity that publication neutralises.

Punk, Berlin house party, guerrilla, grassroots, not shiny, just clean enough — working register. “A network of foresight practitioners and strategic analysts” — published. The Singapore layer sharpens whose identity is at stake: the pre-talks also show individual expert authority (frank verdicts, named lineages, professional positioning, methodological self-disclosure) that publication erases as completely as the punk-Berlin register. The entity-voice of the forecasts is neither the we of meetings nor the I of lightning talks; it is a manufactured third voice with no first-person witness.

4. The entity has a rich affective register that strips monotonically.

The entity feels strongly in working sessions. The pre-talks are slightly cooler but still personal. Publication carries no affect. The neutral analytical voice is performed and is the endpoint of a strip process, not a stance held in private.

5. The Capitalised Concepts are decisively a writing-phase fabrication.

Zero hits across both lower registers. The publication is not a transcription of the substrate; it is a layer the entity built on top of it. The substrate’s own coinages did not survive.

6. v1 has a directional bias in what it selects from its substrate.

5 of 9 pre-talks did not become forecasts. The orphans cluster into lens talks and domain-vibrancy talks. The latter group describes the cultural-positive substrate the published corpus systematically filters out. v1 is structurally a 10-piece set of “from-X-to-Y-which-is-worse” claims.

7. The publication’s modesty formulas have no working-register counterpart.

A map, not a route / snapshots, not predictions appear nowhere in the 30 sessions. The modesty is an audience-facing performance, not a stance held in private. Privately the entity is direct, opinionated, often unhedged.

What this implies for v2

Not prescriptive — the lexical gaps point at choices v2 would have to make.

• Surface the working-method vocabulary. Naming lightning talks, wireframes, the stack in public would close the modesty gap. Doing so reframes the entity from foresight publisher to foresight method. The method becomes part of the product.
• Let some affect through. Not all — but a public voice that carries some of the conversational register would distinguish 10F from the foresight-professional baseline more sharply than any Capitalised Concept can.
• Decide whether to publish the cultural-positive substrate. 5 of 9 Singapore lightning talks did not become forecasts; the orphans skew toward positive emergent signals. v1’s selection bias toward “from-X-to-worse-Y” was structural. v2 can either continue that posture (a deliberate diagnostic-pessimist house style) or open a parallel surface for the positive substrate that already exists in members’ material.
• Decide whether to keep the macro-modest / micro-certain exchange. v2 can continue the exchange (familiar foresight-genre move) or invert it: publish at the substrate’s actual epistemic stance — macro-confident the work matters, micro-uncertain about specifics. The latter is rarer in the field and would be distinctive.
• Decide what to do with the Capitalised Concepts. They are publication-only — the entity does not actually use them in working sessions. v2 can either workshop them back into the conversational register (so the entity can do them in practice) or retire them. Concepts that survive only in publication are decorative.
• Decide whether the rejection of “the r word” is real. The conversational corpus rejects the report format vigorously; the published artefacts are reports in everything but name. v2 can either honour the rejection (different form) or own that the entity produces reports and stop rejecting the word.

These are forks the taxonomy surfaces, not commitments. The workshop use of this taxonomy would be to put 10F members in front of it and ask which gaps they want to close, keep, or invert in v2.

Limitations

• The “Recurring language” field in each behavioral summary is itself an analyst’s selection. High-fidelity for named terms, less so for register at the syntactic level. A finer-grained pass would extend the analysis.
• The published corpus includes 10 forecasts — F04, F07, F09 read in full, others sampled. Deeper reads of F02 and F06 might surface additional faint inheritance from Scott’s risk-maxing talk; deeper read of F05 might surface more JD substrate.
• Mapping confidence varies. The “strong” mappings (Aarathi → F04, Dexter → F09, Lina → F07) are high-confidence — the structural backbones transfer. The “weak / orphan” mappings are more uncertain.
• Pre-talks have named speakers, so cross-talking and discussion content mixes domain experts with consortium peers. The analysis tries to keep the speaker’s own vocabulary distinct from interlocutor contributions but is necessarily entity-level for cross-cutting findings.
• Some pre-talks are uneven (Jake’s transcript cuts off mid-discussion). Some speakers had slides; others did not.
• Only the Singapore pre-talks are in the conceptual layer. Further oral talks (in-person mini-workshop sessions in Singapore, post-Singapore working sessions) are not in the Pre-talks folder. A complete conceptual-register frame would also pull from those.

Related

The taxonomy’s gaps are one specific case of the espoused-vs-enacted move theorised in strategy-as-protocol — the essay that anchors much of this work.