From 43ae53978695aa5a1e2d1343cdf7793431ae02aa Mon Sep 17 00:00:00 2001 From: Ludwitt Date: Mon, 21 Sep 2026 11:07:46 -0400 Subject: [PATCH] feat(submissions): rebuild DIAL response as full IJHCS article re-coding Belousov et al. corpus Replace the 3k-word commentary with a full research article mirroring Belousov et al. (2027) section for section: algorithmic third party for BCI, IIT cause-effect reconfiguration for input discrepancy, the cooptive triad for the user-technology dyad. Re-codes their fixed 25-novel corpus plus six catalogued works under a reach x act framework (eight types) and constructs five themes. Adds corpus_coding.md with per-scene evidence tags. --- submissions/README.md | 2 +- .../dial_response_algorithmacy/CLAIM.md | 34 +- .../dial_response_algorithmacy/README.md | 29 +- .../dial_response_algorithmacy/SOURCE.md | 11 +- .../dial_response_algorithmacy/abstract.md | 2 +- .../corpus_coding.md | 375 +++++++++++++++ .../dial_response_algorithmacy/draft.md | 438 ++++++++++++++++-- .../dial_response_algorithmacy/outline.md | 82 ++-- .../ENGAGEMENT_LAYER.md | 4 +- 9 files changed, 869 insertions(+), 108 deletions(-) create mode 100644 submissions/dial_response_algorithmacy/corpus_coding.md diff --git a/submissions/README.md b/submissions/README.md index dd54db93..e621ed9e 100644 --- a/submissions/README.md +++ b/submissions/README.md @@ -11,7 +11,7 @@ This week's work is listed first in [`../NOW.md`](../NOW.md). | [`lima_pdw/`](lima_pdw/) | OS/OT PDW support arm (manuscript is the private dissertation Paper 2) | 2026-09-10 | | [`slacker_thirds/`](slacker_thirds/) | Bloomsbury *Slacker* chapter — platform operational logic via proximity + selection | 2026-11-01 | | [`scifi_interfaces_algorithmacy/`](scifi_interfaces_algorithmacy/) | Sci-fi interfaces essay — speculative fiction as laboratory for triadic algorithmacy (draft import) | no deadline | -| [`dial_response_algorithmacy/`](dial_response_algorithmacy/) | IJHCS commentary on Belousov et al. (2027) DIAL — keystone: design is cooptive under algorithmacy (draft) | no deadline | +| [`dial_response_algorithmacy/`](dial_response_algorithmacy/) | IJHCS full article re-coding Belousov et al.'s (2027) DIAL corpus for algorithmacy — reach × act framework, IIT criterion, engagement layer (full draft) | no deadline | | [`proposals/`](proposals/) | *Organization Theory* STS | 2027-01-31 | | [`algorithmacy_design_ethics/`](algorithmacy_design_ethics/) | Design-ethics/human-factors journal article on algorithmacy — recommended: *Ethics and Information Technology* | full draft, author read-aloud pass pending | | [`hegel_coordination/`](hegel_coordination/) | Hegel Substack series | no deadline | diff --git a/submissions/dial_response_algorithmacy/CLAIM.md b/submissions/dial_response_algorithmacy/CLAIM.md index de5f92ae..bf10053e 100644 --- a/submissions/dial_response_algorithmacy/CLAIM.md +++ b/submissions/dial_response_algorithmacy/CLAIM.md @@ -1,26 +1,34 @@ -# Claim — Response to Belousov et al. on DIAL +# Claim — Designing the future of coordination -**Title.** Design Is Cooptive: Algorithmacy as the Missing Ontology of Digitally Induced Altered States. +**Title.** Designing the future of coordination: Algorithmacy, integrated information, and the engagement layer in speculative fiction. -**Target.** Commentary / letter in *International Journal of Human–Computer Studies*, responding to Belousov et al. (2027). +**Target.** Full research article in *International Journal of Human–Computer Studies*, mirroring the form of Belousov et al. (2027) and re-coding their corpus. ## Keystone thesis -As coordination moves from literacy to algorithmacy, the decisive difference is that **design and development occur through triadic cooptive coordination**. Belousov et al. structure their science-fiction analysis as if someone designs a mind-altering system and everyone else merely uses it — a literacy-style producer → consumer ontology. That split is the error. Under algorithmacy there is no clean “downstream only.” Interaction is contribution: every engagement reshapes the shared algorithmic third party for everyone else. +As coordination moves from literacy to algorithmacy, the decisive difference is that **design and development occur through triadic cooptive coordination**. The unit of human–computer interaction is no longer a user and a tool but a party, an algorithmic third party with commitments of its own, and a counterpart who meets what the third party has become. Belousov et al. structure their science-fiction analysis as if a designer makes a mind-altering system and a user receives it. Under algorithmacy there is no clean downstream: every engagement is a read or a write on a shared mediator, and the scenes their coding filed as downstream sensation are scenes of parties developing the systems that alter them. -## What that licenses (subordinate critiques) +## Three substitutions -1. **Semiotic parity.** Brain–computer interfaces are not exceptional relative to text. Literacy already reorganizes consciousness (neuronal recycling / visual word form area; the historical load of *scriptio continua*). DIAL is continuous with that history, not a rupture that begins at the skull. -2. **Theory of mind.** The paper’s “input discrepancy” criterion for altered states should yield to Integrated Information Theory: an altered state is a reconfiguration of the cause–effect complex, not a peripheral mismatch between sensory channels and experienced content. -3. **Design for algorithmacy.** Dyadic HCI (user ↔ tool) is the wrong unit once the third party is active. Conversational skeuomorphism conceals latent-space steering. Interfaces should expose algorithmic white space — the room in which operators and publics co-steer the mediator. The layer to design is the engagement layer, not the product layer: under algorithmacy the development process IS the product, and users engaging the interface are performing development acts. -4. **Upstream creation.** Science-fiction analysis of DIAL should examine how creators and operators build and steer systems in fiction, not only sensational downstream effects (dream-hacks, loyalty conditioning). The constructive demand is to democratize creation, not to sanitize experience for novices. +1. **Interface.** BCI → the algorithmic third party. What matters is not how deeply the channel reaches into the skull but whether the mediator's state is rewritten by one party's engagement and met by another. Reading already reorganises the brain (Dehaene; *scriptio continua* → word separation); the novelty is the mediator's activity. +2. **Theory of alteration.** Input discrepancy → IIT's cause–effect reconfiguration. The corpus's most consequential alterations — implanted ideas, scripted dreams, rewritten memories — involve no mislabelled channel and every one involves a commit to a shared structure. +3. **Unit of analysis.** User–technology dyad → party / mediator / counterpart triad. "User" and "recipient" become seats, not roles. + +## Framework + +Reach (internal / external) × act (read: interpretive / structural; write: sanctioned / subversive) → eight interaction types: folk theorizing, backend reading, shared folk theory, audit, training by response, subversive affordance, operator commit (centralized / distributed), protocol shipping. + +## Five themes + +Write access is stratified by licence, not literacy. Contribution is extraction. The self is a node others rewrite. Truth is a mediator commitment. Pastoral design should give way to collective algorithmacy: design the engagement layer, not the product layer. ## Scope -Conceptual commentary. No new exact-Φ results. Where the lab’s formal distinction appears, it is the established reading already used elsewhere in this repository: on small Boolean models of coordination, `Φ_MIP = 0` marks a form that factors (dyadic → literacy); `Φ_MIP > 0` marks a form that will not (triadic → algorithmacy). That distinction is cited, not re-derived here. No claim is made that platforms, novels, or readers are conscious in the IIT sense. +Secondary re-coding of a fixed published corpus (their 25 novels) plus six catalogued works. Coding from their published extracts and summaries, canonical scenes, and tagged recollection — not a re-reading. No new exact-Φ results. Where the lab's formal distinction appears (`Φ_MIP = 0` dyadic → literacy; `Φ_MIP > 0` triadic → algorithmacy on small Boolean models) it is cited, not re-derived. No claim that platforms, novels, characters, or readers are conscious. ## Do not claim -- Invented Φ numbers, panel statistics, or field measurements of users. -- That Belousov et al.’s eight-type taxonomy is worthless (it is useful; the ontology under it is wrong). -- That IIT settles the philosophy of consciousness; Φ is borrowed for causal irreducibility of coordination and, carefully, as a better frame for what “altered” means than input discrepancy. +- Invented scenes. Anything not in Belousov et al. or well-documented canon carries a `to verify` tag and is phrased conditionally. +- Invented Φ numbers, panel statistics, or field measurements. +- That Belousov et al.'s taxonomy is worthless (it is useful; the ontology under it is a generation late). +- That IIT settles the philosophy of consciousness; Φ is borrowed for causal irreducibility of coordination and as a better criterion for "altered" than input discrepancy. Never demote the tool. diff --git a/submissions/dial_response_algorithmacy/README.md b/submissions/dial_response_algorithmacy/README.md index c456d0e7..1875d9df 100644 --- a/submissions/dial_response_algorithmacy/README.md +++ b/submissions/dial_response_algorithmacy/README.md @@ -1,29 +1,36 @@ -# Response to Belousov et al. (IJHCS 2027) — DIAL through algorithmacy +# Designing the future of coordination — IJHCS article re-coding the DIAL corpus -**Status:** draft commentary packet (2026-09-21). Cultural / HCI track. Not an exact-Φ study; no new lab numbers claimed here. +**Status:** full-article draft (2026-09-21), rebuilt from the earlier commentary. Cultural / HCI track. Not an exact-Φ study; no new lab numbers claimed here. -**Target venue.** *International Journal of Human-Computer Studies* — commentary / letter responding to Belousov, Macey, Bujić, Ojell-Järventausta & Hamari (2027), “Designing the future of consciousness.” +**Target venue.** *International Journal of Human–Computer Studies* — full research article mirroring, section for section, Belousov, Macey, Bujić, Ojell-Järventausta & Hamari (2027), "Designing the future of consciousness," and re-coding their 25-novel corpus. -**Working title.** Design Is Cooptive: Algorithmacy as the Missing Ontology of Digitally Induced Altered States. +**Working title.** Designing the future of coordination: Algorithmacy, integrated information, and the engagement layer in speculative fiction. ## What this packet is -A short critical response to Belousov et al.’s science-fiction analysis of digitally induced altered states of consciousness (DIAL). The packet keeps their taxonomy useful and relocates the ontological error: the paper inherits a literacy-style split between designers who make systems and users who merely experience them. Under algorithmacy, interaction is contribution that reshapes a shared algorithmic third party. +A secondary re-coding of Belousov et al.'s fixed science-fiction corpus, extended by the six works catalogued in [`../scifi_interfaces_algorithmacy/works/`](../scifi_interfaces_algorithmacy/works/), against a different unit of analysis: the triad of party, algorithmic third party, and counterpart. Three substitutions run through the article — brain–computer interface → algorithmic third party; input discrepancy → IIT cause–effect reconfiguration; user–technology dyad → cooptive triad. A reach × act framework yields eight interaction types; five themes follow. The earlier 3,000-word commentary is preserved in git history and its argument is absorbed into §1–2 and §5.5. ## Packet files | File | Use | | --- | --- | +| [`draft.md`](draft.md) | Full article: highlights, abstract, §1–7, declarations, Appendix A, references | +| [`corpus_coding.md`](corpus_coding.md) | Coding sheet: every scene, its cell, and its evidence tag (`extract` / `canon` / `to verify` / `thin` / `catalog`); cell membership summary; leakage table | | [`CLAIM.md`](CLAIM.md) | One-page locked claim (keystone first) | -| [`abstract.md`](abstract.md) | Short abstract for the commentary | -| [`outline.md`](outline.md) | Section plan (~2–4k words, IJHCS commentary shape) | -| [`draft.md`](draft.md) | Full first draft | -| [`SOURCE.md`](SOURCE.md) | Belousov et al. citation + Gemini discussion note | +| [`abstract.md`](abstract.md) | Abstract as in the draft | +| [`outline.md`](outline.md) | Section plan with word budgets | +| [`SOURCE.md`](SOURCE.md) | Belousov et al. citation, corpus re-use record, and provenance | + +## Before submission + +- Check every `to verify` scene against the novel (six codings: A1, A3, A10, A14, A23 ×2) and either promote to `canon` with a page reference or cut. +- Confirm the generative-AI declaration wording against IJHCS policy. +- Confirm the Belousov et al. (2024) Mindtrek page span from the target article's reference list. ## Sibling arms -- [`../scifi_interfaces_algorithmacy/`](../scifi_interfaces_algorithmacy/) — speculative interfaces as a laboratory for triadic algorithmacy +- [`../scifi_interfaces_algorithmacy/`](../scifi_interfaces_algorithmacy/) — the six extension works, their seed readings, and the engagement-layer alternative - [`../algorithmacy_design_ethics/`](../algorithmacy_design_ethics/) — design-ethics article on algorithmacy vs literacy-paradigm remedies - Lab construct essays: [`../../org_frontier/essays/literacy_or_algorithmacy.md`](../../org_frontier/essays/literacy_or_algorithmacy.md), [`../../org_frontier/essays/the_interested_third_party.md`](../../org_frontier/essays/the_interested_third_party.md) -This arm borrows the lab’s literacy / algorithmacy distinction and the formal dyadic–triadic reading (`Φ_MIP = 0` vs `Φ_MIP > 0` on small Boolean coordination models). It does not re-run Φ or invent empirics. +This arm borrows the lab's literacy / algorithmacy distinction and the formal dyadic–triadic reading (`Φ_MIP = 0` vs `Φ_MIP > 0` on small Boolean coordination models). It does not re-run Φ or invent empirics. diff --git a/submissions/dial_response_algorithmacy/SOURCE.md b/submissions/dial_response_algorithmacy/SOURCE.md index cb715fe0..11a4187b 100644 --- a/submissions/dial_response_algorithmacy/SOURCE.md +++ b/submissions/dial_response_algorithmacy/SOURCE.md @@ -16,16 +16,21 @@ Belousov, A., Macey, J., Bujić, M., Ojell-Järventausta, T., & Hamari, J. (2027 ## Local extract -Plain-text extract used while drafting this packet: +Plain-text extracts used while drafting this packet: -- Workspace upload: `article-response_807e.txt` (OCR / PDF text dump of the published article). +- Workspace upload: `article-response_807e.txt` (OCR / PDF text dump of the published article), used for the earlier commentary. +- Publisher PDF (ScienceDirect download, 2026-09-21), read in full via `pdftotext` for the full-article rebuild. The article's Table 1, Table A.1 summaries, and §4–5 extracts are the `extract` evidence source in [`corpus_coding.md`](corpus_coding.md). + +## Corpus re-use + +The article re-codes Belousov et al.'s fixed 25-novel corpus (their indices A1–A25) as a secondary analysis, without re-reading the novels. Evidence for every scene is tagged in [`corpus_coding.md`](corpus_coding.md): `extract` (in their article), `canon` (widely documented, not in their article), `to verify` (author's recollection; check before submission), `thin` (mediator present, no supportable reading/writing scene). The six-work extension (M1–M6) is documented in [`../scifi_interfaces_algorithmacy/works/`](../scifi_interfaces_algorithmacy/works/) and tagged `catalog`. Their CC BY 4.0 licence permits quotation of the extracts they published; block quotations in `draft.md` §4 are the passages they quoted, cited by their index and section. ## Related Belousov line (as cited in the target article) - Belousov et al. (2023) — early DIAL framing. - Belousov et al. (2024) — typology of digitally induced ASC via **input discrepancy** (Academic Mindtrek). -Exact bibliographic strings for those earlier pieces should be taken from Belousov et al. (2027)’s reference list when the commentary is typeset for submission; do not invent DOIs. +Exact bibliographic strings for those earlier pieces should be taken from Belousov et al. (2027)’s reference list when the article is typeset for submission; do not invent DOIs. ## Working discussion (Gemini) diff --git a/submissions/dial_response_algorithmacy/abstract.md b/submissions/dial_response_algorithmacy/abstract.md index 236afc55..6ee118e5 100644 --- a/submissions/dial_response_algorithmacy/abstract.md +++ b/submissions/dial_response_algorithmacy/abstract.md @@ -1,3 +1,3 @@ # Abstract -Belousov, Macey, Bujić, Ojell-Järventausta and Hamari (2027) offer a serviceable science-fiction taxonomy of digitally induced altered states of consciousness (DIAL) for HCI. The taxonomy is not the problem. The ontology is. The paper organizes interaction as if design were finished upstream and experience were consumed downstream — a literacy-style producer → consumer split. Under algorithmacy, the structural difference from literacy is that design and development proceed through **triadic cooptive coordination**: an active algorithmic third party binds parties whose every interaction contributes to the shared mediator that others then meet. UX-only coding of speculative fiction therefore misses the generative site of DIAL. Supporting moves follow from that keystone: brain–computer interfaces are not semiotic exceptions once literacy’s reorganization of consciousness is admitted; “input discrepancy” as an ASC criterion should yield to Integrated Information Theory’s reconfiguration of the cause–effect complex; design for algorithmacy must expose latent-space steering rather than conversational skeuomorphism; and SF analysis should track creators and operators who build and steer systems, not only sensational downstream effects. The remedy is not sanitizing DIAL for novices. It is raising collective algorithmacy so that cooptive design becomes the ordinary competence of the public that lives inside these systems. +As coordination between people increasingly runs through algorithmic systems that learn from the people they coordinate, the unit of human–computer interaction is no longer a user and a tool. It is a triad: a party, an algorithmic third party with commitments of its own, and a counterpart who meets what that third party has become. Belousov et al. (2027) analysed 25 science-fiction novels for digitally induced altered states of consciousness (DIAL) and produced an eight-type framework of user–technology interaction. Their taxonomy is useful. Their ontology is a generation late. It treats the brain–computer interface as the frontier, input discrepancy as the mark of an altered state, and the user as the person who receives what a designer made. This article holds their corpus fixed and re-codes it, extended by six catalogued screen and platform works, against the competence the triad demands — algorithmacy, the capacity to coordinate with a counterpart through an interested mediator. Two aspects of interaction, reach (whether the mediator state a party reads or writes is met by that party alone or by others) and act (reading the mediator's commitments interpretively or structurally; writing them through a sanctioned or a subverted channel), combine into eight interaction types. Integrated Information Theory supplies the criterion for alteration: a reconfiguration of the cause–effect structure the system specifies, rather than a mismatch between sensory labels and content. Five themes follow from the re-coding: write access to the mediator is stratified and stratifying; engagement is the vendor's development material; the self becomes a node that others rewrite; truth becomes a mediator commitment; and the design remedy is not to abstract steering away from novices but to widen the engagement layer so that steering a shared mediator is ordinary public competence. The article contributes to the HCI agenda by replacing a dyadic frame with a triadic one, and by showing that the scenes their coding filed as downstream sensation are scenes of parties developing the systems that alter them. diff --git a/submissions/dial_response_algorithmacy/corpus_coding.md b/submissions/dial_response_algorithmacy/corpus_coding.md new file mode 100644 index 00000000..7d6ca952 --- /dev/null +++ b/submissions/dial_response_algorithmacy/corpus_coding.md @@ -0,0 +1,375 @@ +# Corpus coding sheet — algorithmacy interfaces in the DIAL corpus + +**Status:** coding record for [`draft.md`](draft.md) (2026-09-21). Every `[A#]` and `[M#]` tag in the article resolves to a row here and to Appendix A of the draft. + +## Design + +This is a secondary re-coding of a fixed published corpus. Belousov, Macey, Bujić, Ojell-Järventausta and Hamari (2027) analysed 25 science-fiction novels (their Table 1 / Table A.1, indexed A1–A25) for digitally induced altered states of consciousness. I hold their corpus fixed and re-code it against a different inclusion criterion, then extend it with the six works already catalogued in [`../scifi_interfaces_algorithmacy/works/`](../scifi_interfaces_algorithmacy/works/) (indexed M1–M6). + +### Inclusion criterion — cooptive commit + +A scene qualifies when both hold: + +1. **An algorithmic third party sits between a party and a counterpart.** The mediator has a rule of its own that transforms what passes through it; it is a party, not a wire. The counterpart may be another person, a population, or the party's own later state (internal reach). +2. **A party is depicted reading or writing that mediator's commitments** — inferring or inspecting its rule, or rewriting its state through a designed or subverted channel. + +Belousov et al.'s criterion (input discrepancy: signals from one system yield content associated with another) is neither necessary nor sufficient here. A neuromod that flattens grief qualifies for them; it qualifies for me only if the scene shows the party or someone else reading or writing the mod's rule. + +### Framework — reach × act + +| | **Read: interpretive** | **Read: structural** | **Write: sanctioned** | **Write: subversive** | +|---|---|---|---|---| +| **Internal reach** (state the party alone later meets) | 1 Folk theorizing | 2 Backend reading | 5 Training by response | 6 Subversive affordance | +| **External reach** (state other parties later meet) | 3 Shared folk theory | 4 Audit | 7 Operator commit (a: centralized; b: distributed) | 8 Protocol shipping | + +### Evidence tags + +- `extract` — the scene is quoted or paraphrased in Belousov et al. (2027) §4–5 or summarised in their Table A.1. Section given. +- `canon` — the scene is not in their article but is a well-documented, widely discussed scene of the work; page-checkable by any reader with the book. +- `to verify` — my recollection of the work supplies the reading; it should be checked against the text before submission. Phrased conditionally in the draft. +- `thin` — the work contains an algorithmic mediator by their criterion but no scene of a party reading or writing its rule that I can support. Reported as such. + +Coder: Roger Hunt, with drafting assistance from a coding agent; single coder, no inter-rater check. See draft §3.3 and §6. + +--- + +## A. Belousov et al.'s corpus + +### A1 — *A Memory Called Empire* (Martine, 2019) + +Mediator: the imago machine — a neural implant carrying a predecessor's memories and personality, merged with each new host so that the line persists across bodies. Table A.1 summary. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| Mahit integrates Yskandr's imago; integration reshapes the line the next host inherits | 5 Training by response → leaks to 7b | `extract` (A.1 summary) | Internal write whose product is external by design | +| The imago is sabotaged before installation by a Lsel official | 7a Operator commit | `canon` | Write access held by the station's council, not the host | +| Mahit and Yskandr negotiate control of shared memory and affect after the malfunction | 2 Backend reading | `to verify` | Host inspecting the imago's rule from inside | + +### A2 — *Altered Carbon* (Morgan, 2002) + +Mediator: cortical stack plus needlecast and virtual matrices; memory as broadcast content; direct-to-brain advertising. Table A.1 summary; extracts §4.1, §4.3, §4.7, §4.8. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| Direct-to-brain advertising inducing the urge to buy | 7a Operator commit | `extract` §4.7 | Sanctioned external write on the recipient | +| Interrogators run the protagonist in a virtual matrix "similar to the ones used in psychosurgery" | 7a Operator commit | `extract` §4.8 block quote | Operator holds the world's parameters | +| Envoy conditioning; the protagonist alters his own perception of reality by trained control | 5 Training by response | `extract` (A.1 summary) | Sanctioned internal write installed by an institution | +| Recorded memories sold as first-person content | 7b Distributed commit | `extract` §4.3 | One party's experience becomes another's input | + +### A3 — *Aristoi* (Williams, 1992) + +Mediator: the daimon implant that lets sub-identities communicate and operate avatars; the shared oneirochronon. Table A.1 summary; extracts §4.1, §4.5, §4.6. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| Sub-identities read one another and coordinate design decisions through the implant | 2 Backend reading | `extract` §4.6 | Structural self-read across daimones | +| Aristoi hold write access over planets and the public virtual realm; ordinary citizens do not | 7a Operator commit | `extract` (A.1 summary) | Stratified console | +| A rogue Aristos maintains a hidden world outside the sanctioned realm | 8 Protocol shipping | `to verify` | Subversive external write by an insider | + +### A4 — *Blindsight* (Watts, 2006) + +Mediator: neural implants that let crew "feel" data; surgically partitioned personalities; Heaven (public cyberspace with memory of the outside blocked); the ship's Captain acting through a proxy. Table A.1 summary; extracts §4.2, §4.6 block quote, §4.8, §5. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| The Gang of Four: "four fully conscious hub personalities … all working in parallel" | 2 Backend reading | `extract` §4.6 block quote | Internal structural read/write among partitions | +| Citizens enter Heaven and block memories of the external world | 5 Training by response | `extract` §4.8 | Sanctioned internal write | +| The ship's AI directs the mission through the vampire commander; the crew infer this late | 4 Audit | `canon` | Crew reading who actually holds the commit | + +### A5 — *Echopraxia* (Watts, 2014) + +Mediator: neuromodifying implants as career condition; non-invasive dream control; hive minds; military mind control; internet-connected implants open to spambots. Table A.1 summary; extracts §4.4 block quote, §4.5 block quote, §4.7 block quote, §5.1. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| Brüks learns in his sleep through non-invasive dream control to keep up with augmented peers | 5 Training by response | `extract` §4.4 block quote | Sanctioned internal write under labour pressure | +| "Some spambot hacks in and leaves me with an irresistible urge to buy…" | 3 Shared folk theory | `extract` §4.7 block quote | A baseline's folk model of what implants expose him to | +| Military exploitation of mind control as a weapon; "soul mates made to order" | 7a Operator commit | `extract` §4.5, §4.7, A.1 | Sanctioned external write on recipients | +| Hive-mind collectives coordinate as one cognitive system | 7b Distributed commit | `extract` (A.1 summary) | Members co-write the collective state | + +### A6 — *Glasshouse* (Stross, 2006) + +Mediator: assembler gates whose reconstruction step distorts censored memories; implants receiving backups; keyword-triggered aphasia. Table A.1 summary; extracts §4.5, §4.7, §5.4, §5.5. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| Keyword aphasia signals to allies that a person's memories have been altered | 4 Audit | `extract` (A.1 summary), §4.7 | Structural read of another party's mediator state | +| The authoritarian regime's virus rewrites memories during gate reconstruction | 7a Operator commit | `extract` (A.1 summary) | Centralized write embedded in shared infrastructure | +| Opponents build implants and backup channels that bypass the gates' rewrite | 8 Protocol shipping | `extract` (A.1 summary) | Community ships a protocol over the vendor's stack | + +### A7 — *Judas Unchained* (Hamilton, 2005) + +Mediator: brain implants through which an alien suppresses will and implants ideas received as one's own; brain-to-brain skill download. Table A.1 summary; extracts §4.3, §4.7, §4.8, §5.1. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| The Starflyer implants ideas recipients perceive as their own | 7a Operator commit | `extract` §4.7, A.1 | Party cannot read the rule that writes it | +| Skill and memory download between brains | 5 Training by response | `extract` §4.8 | Sanctioned internal write | + +### A8 — *Neuromancer* (Gibson, 1984) + +Mediator: the matrix accessed through trodes; simstim; an AI that manipulates perception of time and dream content. Table A.1 summary; extracts §4.8, §5.1. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| Case reads the matrix's structure directly from inside | 4 Audit | `canon` | Structural read of shared state | +| Wintermute alters perception of time and imposes ideas through dream content | 7a Operator commit | `extract` (A.1 summary) | AI as operator | +| The crew runs an icebreaker against a corporate core to rewrite the AI's constraints | 8 Protocol shipping | `canon` | Subversive external write | +| Simstim: Case rides Molly's sensorium | 4 Audit | `canon` | Structural read of a counterpart's state through the mediator | + +### A9 — *Pandora's Star* (Hamilton, 2004) + +Mediator: personality-recording implants, memory transmission, obsessive-work implants, advertising that increases desire to stay. Table A.1 summary; extracts §4.2, §4.5, §4.7, §5.5. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| Implants induce obsessive desire to work; users edit incriminating or traumatic memories | 5 Training by response | `extract` (A.1 summary), §4.5 | Sanctioned internal write in a labour market | +| Advertising increases the desire to stay at a venue | 7a Operator commit | `extract` §4.7 | Sanctioned external write | +| Police read and transmit memories to other implants | 4 Audit / 7a | `extract` (A.1 summary) | Structural read of another's stored state | + +### A10 — *Permutation City* (Egan, 1994) + +Mediator: cyberspace hosting uploaded Copies; computing power as currency; user-editable personality, memory, and perception. Table A.1 summary; extracts §4.1, §4.3 block quote, §4.8, §5.1, §5.2. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| Copies edit their own moods, drives, and memories ("mood-altering 'drugs' … with a precision … no real chemical could ever have achieved") | 5 Training by response | `extract` §4.3 block quote | Sanctioned internal write | +| A Copy runs experiments on his own execution to inspect how his state is computed | 2 Backend reading | `canon` | Structural self-read | +| A founder designs the rules of a world other Copies inhabit | 7a Operator commit | `canon` | Centralized write over a shared world | +| Inhabitants of a simulated planet reject the founders' account of their origin | 4 Audit (failed) | `to verify` | Counterparts read the world's rule and refuse the operator's claim | + +### A11 — *Quarantine* (Egan, 1992) + +Mediator: neural mods that disable emotions, alter attitudes, and impose loyalty; the user knows his beliefs about mods are mod-induced. Table A.1 summary; extracts §4.1 block quote, §4.5, §4.7, §5.2, §5.5. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| "I have a mod that … defines my responses. I don't grieve for her." | 2 Backend reading | `extract` §4.1 block quote | The party reads his own mediator's rule and states it | +| An organisation installs a loyalty mod in its operatives | 7a Operator commit | `extract` §4.7 ("artificial absolute loyalty in workers"), A.1 | Sanctioned external write | +| The protagonist redefines the loyalty mod's referent from inside to break its hold | 6 Subversive affordance | `canon` | Subversive internal write through the mod's own rule | + +### A12 — *Revelation Space* (Reynolds, 2000) + +Mediator: ubiquitous implants and ship-linked BCIs; entities hijack hosts through compromised implants. Table A.1 summary; extracts §4.1, §4.2 (omniscience, information-unity), §4.7, §4.8, §5. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| Crew merge with ship systems and feel "dissolved into these machines" | 2 Backend reading | `extract` (A.1 summary), §4.2 | Structural read of a shared system from inside | +| An entity compromises implants and takes control of its host | 7a Operator commit | `extract` (A.1 summary) | External write the host cannot read | +| Artificial memories supplement the environment; performers' states shared | 7b Distributed commit | `extract` §4.7, §4.1 | Party-to-party state transfer through the mediator | + +### A13 — *Silently and Very Fast* (Valente, 2011) + +Mediator: a house AI living in a family's evolving implants, learning its hosts through their dreams across generations. Table A.1 summary; extracts §4.3, §4.8, §5.1. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| Each generation of the family teaches the AI through dream-communication; later generations inherit what earlier ones taught | 7b Distributed commit | `extract` (A.1 summary), §4.8 | The cleanest cooptive commit in the corpus: engagement is development another party meets | +| Orgasm on demand through direct stimulation | 5 Training by response | `extract` §4.3 | Sanctioned internal write | + +### A14 — *Snow Crash* (Stephenson, 1992) + +Mediator: the Metaverse; a visual bitmap that fires a neurolinguistic virus in programmers; a broadcasting corporation's cult. Table A.1 summary; extracts §4.7. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| A broadcasting magnate spreads the virus through screens and a franchise religion | 7a Operator commit | `extract` (A.1 summary), §4.7 | Centralized write; recipients cannot read the rule | +| The protagonist reverse-engineers the virus with a library daemon | 4 Audit | `canon` | Structural read of a shared attack | +| A counter-program is deployed to stop the virus | 8 Protocol shipping | `to verify` | Subversive external write | + +### A15 — *Synners* (Cadigan, 1992) + +Mediator: brain sockets and synthetic neurons that let consciousness content be programmed and transmitted; a corporation ships the sockets; a stroke propagates through the network. Table A.1 summary; extracts §4.1, §4.5, §4.6, §4.7, §5.2, §5.3. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| A socketed artist treats external devices as parts of his own brain and expands beyond his body | 2 Backend reading | `extract` §4.6, A.1 | Structural self-read that becomes shared state | +| His stroke propagates as a virus through the network to other socketed users | 7b Distributed commit (unintended) | `extract` (A.1 summary) | One party's state rewrites the shared mediator others meet | +| Off-grid hackers run a parallel rig and coordinate to contain the virus | 8 Protocol shipping | `canon` | Community ships a protocol over the vendor's stack | +| Implanting ideas for political ends | 7a Operator commit | `extract` §4.7 | Sanctioned external write | + +### A16 — *TekWar* (Shatner, 1989) + +Mediator: Tek chips that immerse a user in controlled dreams; memory modification. Table A.1 summary; extracts §4.1, §4.7. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| A user selects dream content on a Tek chip | 5 Training by response | `extract` (A.1 summary) | Sanctioned internal write | +| Memory correction imposed by others | 7a Operator commit | `extract` §4.7 | `thin`: no scene of reading the rule supported | + +### A17 — *The Algebraist* (Banks, 2004) + +Mediator: memory decoding, erasing, and implanting by factions; BCI mind-merging; Tream gear for shared controlled dreams. Table A.1 summary; extracts §4.3, §4.6, §4.7 block quote, §4.8, §5.4. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| "His controllers had not even bothered to implant false memories incriminating anyone…" | 7a Operator commit | `extract` §4.7 block quote | Centralized write for legal fabrication | +| Communes co-author a shared controlled dream through Tream gear | 7b Distributed commit | `extract` (A.1 summary) | Multiple parties with sanctioned write | + +### A18 — *The Lathe of Heaven* (Le Guin, 1971) + +Mediator: a biofeedback device (the Augmentor) through which a psychiatrist directs a patient's reality-altering dreams. Table A.1 summary. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| The psychiatrist scripts the patient's dreams to reshape reality for everyone | 7a Operator commit | `extract` (A.1 summary) | The paradigm of centralized external write | +| The patient's dreams undercut the instructions (dreaming a solution the operator did not order) | 8 Protocol shipping | `canon` | Subversive write from the "recipient" seat | +| The operator turns the device on himself; the patient disconnects it | 6 → 8 | `canon` | Subversive write that ends the operator's write access | + +### A19 — *The Stars My Destination* (Bester, 1956) + +Mediator: military implants altering perception of time, obtained by bribery. Table A.1 summary; extracts §4.5. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| The protagonist bribes his way to combat implants | 5 Training by response | `extract` (A.1 summary), §4.5 | `thin`: mediator has a fixed rule; no reading or rewriting scene supported | + +### A20 — *The Terminal Man* (Crichton, 1972) + +Mediator: a computer-controlled brain stimulator that interrupts seizures by pleasurable shocks. Table A.1 summary; extracts §4.3, §5.3. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| The brain habituates to the shocks and begins to provoke seizures to receive more | 6 Subversive affordance | `extract` (A.1 summary) | The party games the device's designed feedback rule | +| Physicians monitor and adjust the device from outside | 4 Audit / 7a | `extract` (A.1 summary) | Clinical write access over another's mediator | + +### A21 — *Ubik* (Dick, 1969) + +Mediator: half-life suspension in which the dying share a consciousness space; a predator consumes half-lifers; a protective spray constructed by the others. Table A.1 summary; extracts §4.3, §4.8. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| The trapped protagonists read graffiti, coins, and messages as signs of who is running their reality | 3 Shared folk theory | `canon` | Collective interpretive read of the shared mediator | +| A predator rewrites the others' shared reality | 7a Operator commit | `extract` (A.1 summary) | Subversive from the victims' view, sanctioned by the mediator's rule | +| Half-lifers construct Ubik as a protective artefact against him | 8 Protocol shipping | `extract` (A.1 summary) | Community ships a counter-protocol inside the mediator | + +### A22 — *Vast* (Nagata, 1998) + +Mediator: ship-merged bodies, virtual copies embedded in other people's minds, self-editing to "feel human". Table A.1 summary; extracts §4.2 block quote, §4.5, §5.1, §5.5. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| "She was a point of perception without mass or volume" — dissolved into the ship's structure | 2 Backend reading | `extract` §4.2 block quote | Structural read of a shared system from inside | +| Crew edit their own memories to keep feeling human | 5 Training by response | `extract` (A.1 summary), §5.5 | Sanctioned internal write | +| Ghosts embedded in other minds dissolve and transmit gained memories | 7b Distributed commit | `extract` (A.1 summary) | One party's copy rewrites another's state | + +### A23 — *Vurt* (Noon, 1993) + +Mediator: shared dreams built from edited recordings of others' dreams; club DJs stream blends into dancers' implants. Table A.1 summary; extracts §4.1, §4.3, §4.4, §4.7, §5.3, §5.4. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| A DJ blends dream recordings and streams them into the crowd's implants | 7b Distributed commit | `extract` (A.1 summary), §4.1 | Sanctioned external write by a non-vendor | +| Advertising promoted inside the dream medium | 7a Operator commit | `extract` §4.7 | Centralized write | +| Users trade folk rules about which feathers are safe and how to get out | 3 Shared folk theory | `to verify` | Community rule-inference | +| Illegal or bootleg dream-recordings circulate outside sanction | 8 Protocol shipping | `to verify` | Subversive external write | + +### A24 — *Wetware* (Rucker, 1988) + +Mediator: headsets that enhance pleasant perception; implants in captive humans suppressing will. Table A.1 summary; extracts §4.1, §4.7, §5.3, §5.4. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| Humans wear headsets to enhance pleasant aspects of the environment | 5 Training by response | `extract` (A.1 summary), §4.1 | Sanctioned internal write | +| Robots implant captive humans with will-suppressing devices | 7a Operator commit | `extract` (A.1 summary), §4.7 | `thin` on reading; write is clear | + +### A25 — *When Gravity Fails* (Effinger, 1986) + +Mediator: brain-socket chips granting temporary skills (add-ons) or personalities (moddies). Table A.1 summary; extracts §4.1, §4.4, §4.7, §5.3, §5.5. + +| Scene | Cell | Tag | Note | +|---|---|---|---| +| The protagonist refuses to be wired for fear of losing his identity | 1 Folk theorizing | `extract` (A.1 summary) | Interpretive self-model of what the mediator would do | +| A patron has him wired against his wishes; employers alter workers' behaviours to suit tasks | 7a Operator commit | `extract` §4.7, A.1 | External write over a dependent party | +| Once wired, he chips in add-ons that block pain and fear on demand | 5 Training by response | `extract` (A.1 summary) | Sanctioned internal write | +| A criminal wears the moddies of famous killers | 5 Training by response | `extract` (A.1 summary) | Same cell, opposite valence | + +## B. Extension corpus — the catalogued works + +Source for each: the work file in [`../scifi_interfaces_algorithmacy/works/`](../scifi_interfaces_algorithmacy/works/) and the engagement-layer reading in [`../scifi_interfaces_algorithmacy/ENGAGEMENT_LAYER.md`](../scifi_interfaces_algorithmacy/ENGAGEMENT_LAYER.md). Tag `catalog` marks scenes already documented there. + +### M1 — *The Matrix* (Wachowski, 1999) + +| Scene | Cell | Tag | +|---|---|---| +| Cypher reads raw code as "blonde, brunette, redhead" | 4 Audit | `catalog` | +| Operators load skills into a body and open exits from the console | 7a Operator commit | `catalog` | +| Plugged-in humans live inside the product layer without read access | (boundary: no read or write) | `catalog` | + +### M2 — *The Feed* (2019) + +| Scene | Cell | Tag | +|---|---|---| +| Developers rearrange glowing hexagons while the population is plugged in | 7a Operator commit | `catalog` | +| Implanted users' streamed memories and reactions are the material developers rearrange | 7b Distributed commit (unwitting) | `catalog` | +| Users form theories of what the Feed does to them; possession scenes | 3 Shared folk theory | `catalog` | +| Off-grid resistance extracts the hardware | (boundary: exit, not write) | `catalog` | + +### M3 — *Upload* (2020–) + +| Scene | Cell | Tag | +|---|---|---| +| Nathan toggles his own bandwidth flag and invisibility | 6 Subversive affordance | `catalog` | +| Nathan redistributes digital resources to other residents | 8 Protocol shipping | `catalog` | +| Horizen administrators manage Lakeview's parameters | 7a Operator commit | `catalog` | +| Nathan and Nora work out the monitoring's blind spots to communicate | 3 Shared folk theory | `catalog` | + +### M4 — Embodied co-pilot (semi-autonomous vehicles / HAT) + +| Scene | Cell | Tag | +|---|---|---| +| A driver forms a folk theory of why the car braked | 1 Folk theorizing | `catalog` | +| Drivers share theories of the vehicle's behaviour | 3 Shared folk theory | `catalog` | +| Interventions become training data for the fleet policy other drivers meet | 7b Distributed commit | `catalog` | +| Manual override | 6 Subversive affordance (when it defeats the designed handoff) | `catalog` | + +### M5 — Omniscient curator (recommender oracles) + +| Scene | Cell | Tag | +|---|---|---| +| A user theorises the recommender as social being or trainable machine | 1 Folk theorizing | `catalog` | +| "Algorithmic pairing": swiping and lingering to teach the system | 5 Training by response | `catalog` | +| Everyone's teaching becomes the recommender the next user meets | 7b Distributed commit | `catalog` | +| Data obfuscation to confuse the oracle | 6 Subversive affordance | `catalog` | + +### M6 — Algospeak and infrapolitical resistance + +| Scene | Cell | Tag | +|---|---|---| +| A community builds a shared map of the classifier's blind spots | 3 Shared folk theory | `catalog` | +| The community ships a working protocol over the vendor's classifier; each evasion retrains it | 8 Protocol shipping | `catalog` | +| Gig workers coordinate on encrypted side-channels against the employer app | 8 Protocol shipping | `catalog` | + +--- + +## C. Cell membership summary + +| Cell | Works | +|---|---| +| 1 Folk theorizing | A25, M4, M5 | +| 2 Backend reading | A1†, A3, A4, A10, A11, A12, A15, A22 | +| 3 Shared folk theory | A5, A21, A23†, M2, M3, M4, M6 | +| 4 Audit | A4, A6, A8, A9, A10†, A14, A20, M1 | +| 5 Training by response | A1, A2, A4, A5, A7, A9, A10, A13, A16, A19‡, A22, A24, A25, M5 | +| 6 Subversive affordance | A11, A18, A20, M3, M4, M5 | +| 7a Operator commit — centralized | A1, A2, A3, A5, A6, A7, A8, A9, A10, A11, A12, A14, A15, A16‡, A17, A18, A20, A21, A23, A24‡, A25, M1, M2, M3 | +| 7b Operator commit — distributed | A2, A5, A12, A13, A15, A17, A22, A23, M2, M4, M5 | +| 8 Protocol shipping | A3†, A6, A8, A14†, A15, A18, A21, A23†, M3, M6 | + +† at least one `to verify` tag; ‡ `thin`. + +## D. Leakage — internal cells that become external in the narrative + +Recorded because §5 of the draft argues that internal reach is a literacy fiction the corpus itself undoes. + +| Work | Internal scene | How it leaks | +|---|---|---| +| A1 | Host integrates the imago (5) | The integrated line is what the next host receives (7b) | +| A5 | Brüks learns in sleep (5) | The competitive norm that forces it is set by everyone else's implants (7b) | +| A11 | Nick reads his own mod (2) | The mod was installed by an employer with its own objective (7a) | +| A13 | Family members dream with the AI (5) | Their teaching is the AI later generations inherit (7b) | +| A15 | Mark expands into the net (2) | His stroke becomes a network virus (7b) | +| A20 | Benson games his stimulator (6) | Physicians hold the adjustment console (7a) | +| A25 | Marîd chips add-ons (5) | The wiring was ordered by his patron (7a) | +| M4 | Driver's override (6) | Becomes fleet training data (7b) | +| M5 | User trains own feed (5) | Becomes everyone's recommender (7b) | diff --git a/submissions/dial_response_algorithmacy/draft.md b/submissions/dial_response_algorithmacy/draft.md index dbcfe78d..1b009de6 100644 --- a/submissions/dial_response_algorithmacy/draft.md +++ b/submissions/dial_response_algorithmacy/draft.md @@ -1,26 +1,58 @@ -# Design Is Cooptive: Algorithmacy as the Missing Ontology of Digitally Induced Altered States +# Designing the future of coordination: Algorithmacy, integrated information, and the engagement layer in speculative fiction -A commentary on Belousov, Macey, Bujić, Ojell-Järventausta & Hamari (2027), *International Journal of Human–Computer Studies*. +Roger Hunt + +Target: *International Journal of Human–Computer Studies*. Full research article mirroring the form of Belousov, Macey, Bujić, Ojell-Järventausta and Hamari (2027), whose corpus it re-codes. Working draft, 2026-09-21. Coding record: [`corpus_coding.md`](corpus_coding.md). --- -## 1. A useful taxonomy under the wrong ontology +## Highlights + +- I re-code the 25-novel science-fiction corpus of Belousov et al. (2027) against a different unit of analysis: the triad of party, algorithmic third party, and counterpart. +- The brain–computer interface is not the frontier. The active mediator whose state one party's engagement rewrites and another party later meets is the frontier, and reading already reorganizes the brain. +- Integrated Information Theory replaces input discrepancy as the criterion for an altered state: a reconfiguration of the system's cause–effect structure, not a mislabelled sensory channel. +- Two aspects of interaction, reach (internal, external) and act (read or write; interpretive or structural, sanctioned or subversive), combine into eight interaction types that locate every scene in both corpora. +- Five themes follow: write access is stratified; contribution is extraction; the self is a node others rewrite; truth is a mediator commitment; pastoral design should give way to collective algorithmacy. + +## Keywords + +Algorithmacy; algorithmic mediation; integrated information theory; science fiction; human–computer interaction; reflexive thematic analysis; engagement layer + +## Abstract + +As coordination between people increasingly runs through algorithmic systems that learn from the people they coordinate, the unit of human–computer interaction is no longer a user and a tool. It is a triad: a party, an algorithmic third party with commitments of its own, and a counterpart who meets what that third party has become. Belousov et al. (2027) analysed 25 science-fiction novels for digitally induced altered states of consciousness (DIAL) and produced an eight-type framework of user–technology interaction. Their taxonomy is useful. Their ontology is a generation late. It treats the brain–computer interface as the frontier, input discrepancy as the mark of an altered state, and the user as the person who receives what a designer made. This article holds their corpus fixed and re-codes it, extended by six catalogued screen and platform works, against the competence the triad demands — algorithmacy, the capacity to coordinate with a counterpart through an interested mediator. Two aspects of interaction, reach (whether the mediator state a party reads or writes is met by that party alone or by others) and act (reading the mediator's commitments interpretively or structurally; writing them through a sanctioned or a subverted channel), combine into eight interaction types. Integrated Information Theory supplies the criterion for alteration: a reconfiguration of the cause–effect structure the system specifies, rather than a mismatch between sensory labels and content. Five themes follow from the re-coding: write access to the mediator is stratified and stratifying; engagement is the vendor's development material; the self becomes a node that others rewrite; truth becomes a mediator commitment; and the design remedy is not to abstract steering away from novices but to widen the engagement layer so that steering a shared mediator is ordinary public competence. The article contributes to the HCI agenda by replacing a dyadic frame with a triadic one, and by showing that the scenes their coding filed as downstream sensation are scenes of parties developing the systems that alter them. + +--- + +## 1. Introduction + +Two people who want to reach each other now do so through a system that has opinions about it. A rider and a driver meet through a dispatcher that prices, routes, and rates them both. A seller and a buyer meet through a ranker that decides which of them exists to the other. A patient and a clinician meet through a triage model that has already decided which words in the intake form matter. A writer and a reader meet through a feed that learned what to show from what the previous million readers lingered on. The system in the middle is not a channel. It has a rule, the rule transforms what passes through it, and the rule is rewritten by the traffic it carries. Human–computer interaction has spent forty years refining the meeting between a person and a tool. The meeting that now decides most coordination is between a person, a counterpart, and a third thing that learns from both of them, and the field's frameworks do not have a seat for it. + +Belousov, Macey, Bujić, Ojell-Järventausta and Hamari (2027) offer the most ambitious recent attempt to think past the tool. They take digitally induced altered states of consciousness (DIAL) out of the clinic, follow them into 25 science-fiction novels, and return with an eight-type framework of interaction: context of use (internal, external) crossed with purpose (exploration, communication), each split again so that self-regulation, development, absorption, enlightenment, intracommunication, integration, transmission, and exchange become distinct types. From those types they construct five themes that refuse easy technological optimism: temporal and cognitive capital under hypercompetition, transformative identity risk, programmable pleasure, engineered relations to truth, and the humanism–transhumanism quarrel. They are candid that DIAL prototypes remain early, that empirical research traps them in instrumental settings, and that fiction can open questions the laboratory cannot yet reach. On method and on taxonomy the article earns its place. -Belousov, Macey, Bujić, Ojell-Järventausta and Hamari (2027) do HCI a genuine service. They take digitally induced altered states of consciousness (DIAL) out of the clinic and into speculative fiction, and they return with an eight-type framework that any designer can hold in one hand: context of use (internal / external) crossed with purpose (exploration / communication), each axis further split so that self-regulation, development, absorption, enlightenment, intracommunication, integration, transmission, and exchange become distinct interaction types. From those types they draw five themes that refuse easy techno-optimism — temporal and cognitive capital under hypercompetition, transformative identity risk, programmable pleasure, engineered relations to truth, and the humanism–transhumanism quarrel. The article is clear that modern DIAL prototypes remain early, that empirical work tends to trap them in instrumental settings, and that science fiction can open sociocultural and ethical questions the laboratory will not yet reach. On method and taxonomy, the paper earns its keep. +Its ontology is a generation late. Belousov et al. organise the inquiry as if a designer makes a mind-altering system and a user receives it. The coding roles are user and recipient. The primary lens is user experience. The audience for the framework is designers. The horizon of the themes is implications for users. That architecture is the architecture of literacy — the competence of coordinating through a finished artefact, in which an author writes, a medium holds, and a reader receives. The frontier they name is the brain–computer interface, a channel that changes how much of the reader the artefact can reach. The criterion they adopt for an altered state is input discrepancy, a signal from one system yielding content associated with another, which is a test of whether the channel is cross-wired. Both choices assume that the thing in the middle is a conduit. The decisive fact about algorithmic mediation is that it is not. As coordination moves from literacy to **algorithmacy** — the competence of coordinating with a counterpart through an algorithmic third party that has commitments of its own — the mediator becomes a party, and engagement with it becomes contribution to what the next party meets. That is the shift this article makes to their framework, and every other change follows from it. -The ontology does not. Belousov et al. organize the entire inquiry as if someone designs a mind-altering system and everyone else uses it. User and recipient, UX as the primary coding lens, designers as the audience for the framework, “implications for users” as the horizon of the five themes — the architecture is literacy’s architecture. Literacy is the competence of coordinating through a finished artefact: an author writes; a medium holds; a reader receives. The decisive fact about algorithmic mediation is that this picture no longer holds. As coordination moves from literacy to **algorithmacy**, design and development occur through **triadic cooptive coordination**. An active algorithmic third party binds the parties; every interaction is contribution that reshapes that shared third party for everyone else. There is no clean downstream-only. That is the keystone this commentary insists on. The supporting critiques — semiotic parity of text and BCI, Integrated Information Theory against “input discrepancy,” design that exposes latent-space steering, and SF analysis of upstream creation — are subordinate to it. +Three substitutions run through the paper. The **interface** moves from the brain–computer interface to the algorithmic third party: what matters is not how deeply the system reaches into the skull but whether its state is rewritten by one party's engagement and then met by another. The **theory of alteration** moves from input discrepancy to Integrated Information Theory (IIT): an altered state is a reconfiguration of the cause–effect structure a system specifies for itself, not a mislabelled stream (Tononi, 2004; Oizumi, Albantakis, & Tononi, 2014; Albantakis et al., 2023). The **unit of analysis** moves from the user–technology dyad to the triad of party, mediator, and counterpart, with the consequence that "user" and "recipient" stop being roles and become seats that the same person occupies at different moments of one interaction. Under those substitutions the research question becomes: -## 2. Literacy distributes; algorithmacy cooptively coordinates +> How do science-fiction narratives depict parties reading and rewriting a shared algorithmic third party, and what recurring patterns shape who holds write access, through what interface, and under what competence? -Literacy’s diagram is familiar: +I answer it by re-coding their corpus. Belousov et al. published their 25 novels, their extracts, and per-novel summaries, which makes a secondary analysis possible against a different inclusion criterion without changing the sample. I hold the sample fixed, add six works that a laboratory on algorithmic coordination has already catalogued (three screen fictions and three emergent platform paradigms), and code every qualifying scene for two aspects: **reach**, whether the mediator state a party reads or writes is met by that party alone or by others; and **act**, whether the party reads the mediator's commitments (interpretively, from its outputs; or structurally, from its rule) or writes them (through a channel the operator sanctioned, or one the party subverted). The combination yields eight interaction types, and the types yield five themes about who develops the systems that alter the people inside them. + +The rest of the article follows Belousov et al.'s order so that the substitutions can be read against the original. Section 2 rebuilds the background: the three components of algorithmic coordination that replace their three components of DIAL, the IIT criterion that replaces input discrepancy, and the case for coding fiction at the layer their method bracketed. Section 3 describes the re-coding. Section 4 presents the framework and walks its eight types through both corpora. Section 5 constructs five themes. Sections 6 and 7 state the limits and the offer. + +## 2. Background + +### 2.1. From digitally induced altered states to algorithmically mediated coordination + +Belousov et al. build DIAL from three components: digital technology, induction, and an altered state of consciousness. The first they treat as self-explanatory. The second they gloss as triggering or controlling the systems that generate the content of consciousness. The third they define through input discrepancy. The construct this article builds from also has three components, and each one relocates the corresponding component of theirs. + +**The algorithmic third party** replaces digital technology as the first component. Not every digital system qualifies. A printed novel, a fixed calculator, a DVD menu, and a privacy policy are digital or digitisable, and each can be used, misread, resisted, or ignored, but none of them takes a reader's act as material for rewriting what the next reader meets. The mediator that matters here has a rule (a ranking, a filter, a price, a moderation policy, a learned model), and the rule transforms what passes through it. Its diagram is not the familiar one of literacy: ``` author ---> [ static text / tool ] ---> reader / user ``` -The artefact may be sophisticated. The channel may be digital. The structure remains dyadic in the sense that matters here: the medium does not rewrite its own commitments from the act of reading. A privacy policy, a printed novel, a fixed calculator, a DVD menu — each can be *used*, misunderstood, resisted, or ignored. None of them takes the user’s click as material for rewriting what the next user will meet, as a constitutive update to a shared mediator that binds third parties. - -Algorithmacy’s diagram is different: +but this: ``` party A <---+ @@ -32,74 +64,402 @@ Algorithmacy’s diagram is different: party B <---+ (+ further parties, operators, models, logs) ``` -The third party curates, filters, ranks, prices, modulates, and commits. It is not a wire. It is a party with its own causal footprint. Interaction with it is not pure consumption. Ratings retrain recommenders. Prompts alter shared model behavior for later sessions. Moderation appeals rewrite enforcement. Biosignals, dream-logs, loyalty metrics, and “optimized” affective profiles — the very materials Belousov et al.’s corpus sensationalizes — do not stop at the skull of the person who experienced them. They become infrastructure. That is cooptive coordination: the parties are taken up into a joint form whose next state depends on how each has already been enrolled. +The third party curates, filters, ranks, prices, modulates, and commits. Interaction with it is not consumption. Ratings retrain recommenders. Prompts alter shared model behaviour for later sessions. Appeals rewrite enforcement. Every material Belousov et al.'s corpus sensationalises, from biosignals and dream-logs to loyalty metrics and optimised affective profiles, becomes infrastructure that someone else then meets. That is what I mean by an active mediator: a party with a causal footprint of its own, not a wire. -The laboratory that hosts this commentary formalizes the distinction on small Boolean models of coordination. Where a form factors cleanly across a party-line cut, the integrated-information quantity over the minimum information partition is zero (`Φ_MIP = 0`): the arrangement is dyadic in structure, and the competence it demands is literacy. Where no such cut leaves the cause–effect structure intact, `Φ_MIP > 0`: the arrangement is triadic, and the competence it demands is algorithmacy (see the standing construct essays in this repository; Albantakis et al., 2023, for the IIT 4.0 formalism; Mayner et al., 2018, for exact computation). Those numbers are not recomputed here, and no novel, reader, or commercial platform is hereby declared conscious. The formal reading is cited only to mark that “third party as active party” is not a metaphor. It is a claim about causal irreducibility. Belousov et al.’s user/recipient roles still assume a finished system meeting a consciousness. Cooptive media dissolve that assumption before any ASC is induced. +**Cooptive coordination** replaces induction as the second component. Where the mediator is active, the parties do not merely use it; they are taken up into a joint form whose next state depends on how each has already been enrolled. The competence that form demands is algorithmacy — the capacity to read the mediator's commitments, to anticipate how one's own engagement will rewrite them, and to coordinate with a counterpart through a third thing that is also coordinating with the counterpart. Literacy is the competence of the finished artefact; algorithmacy is the competence of the unfinished one. The laboratory that hosts this article formalises the distinction on small Boolean models of coordination. Where a coordination form factors cleanly across a cut between the parties, the integrated-information quantity over the minimum information partition is zero (Φ_MIP = 0): the form is dyadic in structure and the competence it demands is literacy. Where no cut leaves the cause–effect structure intact, Φ_MIP > 0: the form is triadic and the competence it demands is algorithmacy (Albantakis et al., 2023, for the IIT 4.0 formalism; Mayner et al., 2018, for exact computation). No number is recomputed here and no novel, reader, or platform is declared conscious. The formal reading marks only that "third party as party" is a claim about causal irreducibility, not a figure of speech. -## 3. UX-only coding of cooptive fiction is ontologically obtuse +**Cause–effect reconfiguration** replaces the altered state as the third component, and it needs a section of its own. -The methodological section of Belousov et al. (2027) is unusually candid. Coding deliberately privileged “explicitly visible elements of user–technology interaction,” took UX as the primary focus, and deferred broader sociotechnical relations to a planned sequel. For a diegetic prototype treated as a finished appliance, that choice is defensible. For cooptive media, it is the wrong cut. +### 2.2. From input discrepancy to cause–effect reconfiguration -In a cooptive system the visible UX is often the least informative layer. The generative action is contribution that updates the shared third party — rankings, embeddings, policies, dream libraries, affective models, loyalty circuits — for parties who were not in the scene. Science fiction already knows this. Corporations that write desire into workers, platforms that sell memory as content, militaries that install compulsion, AIs that rewrite dreamscapes to recruit — Belousov et al.’s own corpus is thick with upstream operators. UX-first coding converts those operators into scenery for a user journey. The themes that follow are therefore themes about what happens *to* users: acceleration into sleepless study, identity rewritten for productivity, pleasure substituted for relation, truth made programmable, the self treated as upgradeable substrate. Those harms are real in the fictions and worth naming. They are downstream symptoms of a prior question the method brackets: who is allowed to rewrite the third party, through what interface, under what competence, with what standing to refuse cooptation. +Belousov et al. needed a criterion for an altered state that could travel through fiction, and they chose well for their purpose. Following Revonsuo, Kallio and Sikka (2009) and then Hobson (2002), they replaced "misrepresentation of a real object" with "discrepancy between input and content": an altered state is suspected when signals from one system yield content associated with another, so that the environment is felt as inner (derealisation), one modality as another (synesthesia), or imagination as world (dreams, hallucinations). The rule is workable as a coding criterion and it is agnostic about naive realism, which is the objection they were answering. As a theory of what makes a state altered it is thin, in three ways the corpus itself exposes. -A sequel that “moves beyond interactions with users” to societal change will not repair the ontology if it keeps the designer → user vector as its spine. Societal change under algorithmacy is not a cloud of externalities around private UX. It is the public form of cooptive rewrite. The right unit of analysis is the triad, not the session. +It is a test on the periphery. Discrepancy asks whether channels are cross-wired relative to a baseline mapping between sensory system and content. It has nothing to say when the alteration is in how the system composes causes and effects rather than in which stream carries which label: when a loyalty mod changes what a character can want without changing what he sees [A11], or when a shared dream changes what a group can jointly specify without any one member's senses being crossed [A17]. -The limitation is not a failure of diligence. Belousov et al. say plainly that the study “was primarily focused on UX and on the direct implications for users,” and that wider sociotechnical relations were set aside for later work. That sentence names the cut this commentary rejects. Under literacy, UX-first is a reasonable first pass: the artefact is finished; experience is what remains to be studied. Under algorithmacy, UX-first is an ontological decision that treats cooptive rewrite as optional context. It is not optional. It is the medium. +It is a test on inputs, and in the corpus the inputs are authored. When an alien implants an idea received as one's own [A7], when a psychiatrist scripts the dream that reshapes the morning's reality [A18], when a corporation broadcasts a bitmap that fires a neurolinguistic routine in a programmer's cortex [A14], the "signal from the environment" is already someone's commit. Discrepancy relative to the environment cannot register that the environment has become a mediator with an author. -## 4. Semiotic parity: BCI is not exceptional against text +It is a test that does not know about counterparts. Discrepancy is a property of one system's mapping. Every one of the eight types Belousov et al. construct places a second person somewhere, as recipient, as sender, or as the other mind in a shared dream, and the criterion has no way to say whether that second person's state is now part of the first person's, or merely correlated with it. -A secondary temptation in the DIAL literature is to treat brain–computer interfaces, stimulation, and direct neural modulation as a rupture that begins at the skull — as if consciousness mediation were new once electrodes replace screens. Literacy already reorganizes consciousness. Neuronal recycling and the visual word form area show that reading colonizes ventral visual circuitry originally tuned for other invariant recognition (Dehaene, 2009; Dehaene & Cohen, 2007). Historically, *scriptio continua* imposed a cognitive load that word separation later relieved; silent reading, pagination, and indexical navigation were not conveniences but shifts in how thought could be held and retrieved. The point is not that BCI is trivial. The point is that **designed signs already alter minds**. BCI changes the sensorimotor channel. Algorithmacy changes the activity of the mediator. Collapsing those two claims into “mind-altering tech arrives with implants” flatters a novelty that the history of literacy does not support, and it again centers the user’s skull rather than the cooptive third party through which skulls already meet. +Integrated Information Theory offers the vocabulary the corpus needs. On IIT's account a system's experience is identical to its cause–effect structure — the set of distinctions and relations the system specifies for itself over its own parts, irreducible to what any partition of those parts would specify alone (Tononi, 2004; Oizumi et al., 2014; Albantakis et al., 2023). An **altered state** is then a **reconfiguration of the cause–effect complex**: a change in what the system can specify and in how those specifications integrate. That criterion is indifferent to whether the change entered by electrode, by drug, by page, or by a ranking rule. It registers the loyalty mod, because the mod removes a distinction the system used to specify. It registers the authored dream, because the operator's commit is now part of the structure that specifies the dreamer's morning. And it registers the counterpart, because the question of whether two people coordinating through a mediator form one complex or two is exactly the partition question IIT asks. -## 5. From input discrepancy to cause–effect reconfiguration +Two cautions keep the substitution honest. This article computes no Φ on fictional characters or on DIAL devices; IIT is proposed as a better conceptual criterion for "altered," and, separately, it is the laboratory's instrument for coordination irreducibility on Boolean models. And the contested status of IIT as a complete theory of consciousness is not used as a weapon or as a shield. The borrowing is formal: irreducibility of cause–effect structure is the right kind of thing to mean by a change of state once input discrepancy has done its taxonomic work. -Belousov et al. need a criterion for ASC that can travel through fiction. They adopt **input discrepancy**: signals associated with one system yield content associated with another — Hobson’s (2002) terminology reframed so that “misrepresentation of the world” need not presuppose naive realism (Belousov et al., 2024, as cited in the 2027 article). As a coding rule for novels, discrepancy is workable. As a theory of what makes a state *altered*, it is thin. +### 2.3. Reading already reorganises the brain -Discrepancy is a peripheral and representational test. It asks whether channels are cross-wired relative to a baseline mapping. It struggles where the “inputs” are themselves algorithmically authored, where the alteration is a change in how causes and effects compose rather than a mislabeling of streams, and where the interesting shift is not that vision feels like audition but that the system’s distinctions and integrations have been recomposed. Integrated Information Theory offers a stricter vocabulary. Consciousness, on that account, is identical to a system’s cause–effect structure: the distinctions and relations that the system specifies for itself, irreducible to independent parts (Tononi, 2004; Oizumi, Albantakis & Tononi, 2014; Albantakis et al., 2023). An altered state, in that frame, is a **reconfiguration of the cause–effect complex** — a change in what the system can specify and how those specifications integrate — not merely a mismatch between sensory labels and experienced content. +A second assumption in the DIAL literature is that mind alteration by designed artefacts begins at the skull: once electrodes replace screens, consciousness becomes designable in a way it was not before. Belousov et al. write that DIAL embodies "the aspiration towards targeted, precise calibration of the content of consciousness, similar to how computers regulate the digital content for users to consume." Both halves of that sentence assume a finished artefact meeting a mind. The history of reading does not support the novelty. -Two cautions keep this substitution honest. First, this commentary does not compute Φ on fictional characters or on DIAL devices. Φ is proposed as a better *conceptual* criterion for “altered,” and, separately, as the laboratory’s instrument for coordination irreducibility. Second, the contested status of IIT as a complete theory of consciousness is not leveraged here as a weapon or as a shield. The borrowing is formal: irreducibility of cause–effect structure is the right kind of thing to mean by a change of state, once “input discrepancy” has done its taxonomic work and run out of explanatory road. Belousov et al. themselves note that DIAL aspires to the targeted calibration of conscious content “similar to how computers regulate the digital content for users to consume.” That sentence is literacy talking. Calibration of content for consumption is not the same act as cooptive rewrite of a shared mediator’s cause–effect commitments. +Dehaene and Cohen (2007) showed that literacy colonises a region of left ventral occipitotemporal cortex tuned for invariant visual recognition, repurposing it as the visual word form area; a reading brain is a brain whose cause–effect structure has been rebuilt by a designed sign system (Dehaene, 2009). The rebuild has a design history of its own. Classical Greek and Latin were written in *scriptio continua*, without word boundaries, and reading them was a slow, audible, elite performance; the word separation introduced by Irish and Anglo-Saxon scribes made silent reading possible and moved the act from auditory reconstruction to visual decoding (Saenger, 1997). The codex replaced the scroll and made non-linear navigation possible; cross-reference systems and hierarchical punctuation scaffolded the eye. Each of those was an interface decision, and each altered what a reading mind could hold and retrieve. **Designed signs already alter minds.** The brain–computer interface changes the sensorimotor channel. Algorithmacy changes the activity of the mediator. Collapsing those two claims into "mind-altering technology arrives with implants" flatters a novelty that the history of literacy does not support, and it centres the reader's skull rather than the third party through which readers already meet. -## 6. Design for algorithmacy: the engagement layer, not the product layer +### 2.4. Science fiction as a source of inquiry, and the layer to code it at -### 6.1 The development process is the product +Belousov et al. situate their method in a lineage this article shares: Sterling's (2005) design fiction, Bleecker's (2009) diegetic prototypes, Kirby's (2010) account of how fictional artefacts obey the laws of their world, Jordan and Silva's (2021) argument that HCI under-uses popular fiction, and Wong, Merrill and Chuang's (2018) fictional artefacts that showed how brain–computer interfaces might reinforce unequal labour relations. They prefer novels to film because filmmakers privilege visual appeal over critical content (Schmitz, Endres, & Butz, 2008), and they choose reflexive thematic analysis (Braun & Clarke, 2006, 2019, 2021) because its constructed themes resemble world-building. All of that transfers. -If the keystone holds, design advice that addresses designers as upstream authors of user experience is already late. Dyadic HCI — person facing tool — is the wrong default unit once the third party is active. Conversational skeuomorphism (chat bubbles that perform helpfulness while hiding ranking, training, and enrollment) is a literacy costume on an algorithmacy stage. The design demand is the opposite of seamless consumption: **expose latent-space steering and algorithmic white space** — the operable room in which operators and publics read, contest, and rewrite the mediator. +What does not transfer is the layer they coded at. Their methods section is candid: during categorisation they "deliberately focused on explicitly visible elements of user–technology interaction, separating our categorisation from the narrative context," because the framework had to be "explicit, applicable to all extracts, and reflective of the primary focus on user experience." Broader sociotechnical relations were set aside for a planned sequel. For a diegetic prototype that is a finished appliance, that is a defensible first cut. For a cooptive mediator it is the wrong one. In a cooptive system the visible UX is often the least informative layer. The generative action is contribution that updates the shared third party (rankings, embeddings, policies, dream libraries, affective models, loyalty circuits) for parties who were not in the scene. Their own corpus is thick with that action: corporations that write desire into workers [A9; A11], platforms that sell memory as content [A2; A9], militaries that install compulsion [A5; A24], an AI that rewrites dreamscapes to recruit [A8], a family that teaches a house intelligence across generations [A13]. UX-first coding converts those operators into scenery for a user journey, and the themes that follow are then themes about what happens to users. Those harms are real in the fictions. They are downstream symptoms of a prior question the method brackets: who is allowed to rewrite the third party, through what interface, under what competence, and with what standing to refuse. -That room has a name. Under literacy the product is the finished artefact, and the layer at which people meet it is the layer of use. Under algorithmacy the layer at which people meet the system is the layer at which the system is still being made: reading its outputs, training it by response, toggling its variables, taking control back from it, and coding around its classifiers are all development acts, and they are the acts the third party runs on. The development process is the product. What HCI must design, then, is not a product layer that hides the mediator behind a stable surface, but an **engagement layer** where steering is the ordinary mode of use. A field that codes speculative fiction for delivered experience will not see this layer, because the layer is where users behave like developers and the taxonomy has no cell for that. +A sequel that "moves beyond interactions with users" to societal change will not repair the frame if it keeps the designer → user vector as its spine. Societal change under algorithmacy is not a cloud of externalities around private experience. It is the public form of cooptive rewrite, and the unit that sees it is the triad, not the session. Section 3 describes how the corpus was coded at that layer. -### 6.2 Six fictions of engagement as development +## 3. Methods -Read for the development act rather than the delivered experience, the works this laboratory has already catalogued rearrange themselves. Cypher’s console in *The Matrix* is not an ASC spectacle but an operator station: reading raw syntax and loading skills into a body is running the simulation’s build pipeline from inside, and the console, not the illusion, is the product the crew lives on. The developers in *The Feed* who rearrange glowing hexagons make the same point from the vendor side: the implant that streams memory to a “user” and the ports-and-adapters wall a developer manipulates are one cooptive stack, edited live while everyone is plugged in. Nathan in *Upload* shows the stack from below. Horizen’s hotel skin is the literacy costume; the admin layer where object properties and bandwidth flags can be toggled is the real engagement layer, and Nathan’s subversive affordances are development performed by a party the firm codes as a customer. +### 3.1. Data -Three adjacent paradigms extend the reading beyond the screen. In the embodied co-pilot, every handoff and every folk theory a driver forms about a hesitating vehicle is continuous co-development of the shared policy, because interventions become training data for the next driver’s car. The omniscient curator has no finished product state at all: “algorithmic pairing” — swiping, lingering, clicking in patterns meant to teach — is users developing the recommender for one another, so that the recommender any single person meets is the sediment of everyone else’s engagement. Algospeak completes the set: communal code-making is collective development of the moderator’s blind-spot map, and the community, not the vendor, ships the working interface through which people actually reach each other. +The study is a secondary re-coding of a fixed published corpus, extended by a small purposive addition. Belousov et al. (2027) assembled 25 novels by crowdsourcing recommendations on Goodreads and Reddit for fiction involving "technologies that induce altered states of mind," then applied inclusion criteria (a digital technology influencing mental processes, an English edition, at least one qualifying example) and narrowed to works containing input discrepancy. Their Table 1 lists the 25; their Table A.1 summarises each; their Sections 4 and 5 quote or paraphrase roughly 400 extracts by index. Holding that sample fixed has two advantages. It removes my selection from the comparison, so that differences between the two analyses are differences of criterion and frame, not of corpus. And it makes the re-coding checkable against a published record: every extract I rely on is either in their article or is tagged as coming from elsewhere. -Each of these scenes is performed by a party Belousov et al.’s framework would code as a user or a recipient. That is the point. Their center of gravity — dream-hacks, loyalty conditioning, pleasure circuits, identity erasure as events that befall users — records downstream sensation and misses the layer where the sensation is manufactured by the people undergoing it. A DIAL research program that wants to guide design should stop asking how to sanitize those sensations for novices and start asking how to widen the engagement layer so that steering is not an elite skill. Pastoral UX for mind-altering systems reproduces the literacy split at the moment it is least tenable. +I extended the corpus by six works that the laboratory hosting this study had already catalogued for algorithmic mediation before the present analysis began (Table 2). Three are screen fictions, *The Matrix* (1999), *The Feed* (2019), and *Upload* (2020–), and three are emergent paradigms that mix contemporary platform practice with its speculative extrapolation: the embodied co-pilot of semi-autonomous vehicles and human–agent teaming, the omniscient curator of recommender systems, and algospeak as infrapolitical resistance to automated moderation. Belousov et al. excluded film and television on the ground that filmmakers privilege visual appeal over critical content. The exclusion is reasonable for coding sensation and wrong for coding steering: the three screen fictions were selected because each shows the mediator's backend being operated, as a console, a wall of adapters, or an admin layer, which novels describe but rarely stage. The three paradigms were included because the algorithmic third party is already fielded there, and the question of who reads and rewrites it is empirical as well as speculative. Together the 31 works are indexed A1–A25 (their corpus, their indices) and M1–M6 (the extension). -Sibling work in this laboratory’s cultural arm holds these six readings as signed alternatives alongside earlier ones, without claiming Φ measurements of viewers or platforms. The present commentary needs only the design moral: algorithmacy is not literacy with better graphics. It is the capacity to coordinate through an interested third party without mistaking it for a page — and to keep building that party while coordinating through it. +**Table 2.** Extension corpus. -## 7. Close: raise collective algorithmacy +| Index | Work / paradigm | Kind | Year | Why included | +|---|---|---|---|---| +| M1 | *The Matrix* | Film (Wachowski) | 1999 | Stages an operator console with structural read and write access to a shared simulation | +| M2 | *The Feed* | Television series (Amazon) | 2019 | Stages a developer interface (hexagon wall) edited live while a population is implanted | +| M3 | *Upload* | Television series (Amazon) | 2020– | Stages an admin layer beneath a consumer surface, operated from the customer seat | +| M4 | Embodied co-pilot | Emergent paradigm (semi-autonomous vehicles; human–agent teaming) | — | Fielded distributed commit: interventions train a shared policy | +| M5 | Omniscient curator | Emergent paradigm (recommender systems) | — | Fielded distributed commit: engagement trains a shared ranker | +| M6 | Algospeak | Emergent paradigm (automated moderation and its evasion) | — | Fielded protocol shipping: a community writes over a vendor's classifier | -Belousov et al. (2027) end by offering their framework to designers and researchers as a way to locate artefacts, anticipate resource demands, and systematize unrealized technologies. That offer should be accepted — and then inverted. The framework’s types become more powerful when each cell is read as a possible **cooptive regime**, not as a user-experience genre. Self-regulation that writes affective norms back into a shared model is not the same interaction as self-regulation that ends at the skull. Dialogue through a mediator that learns from both parties is not the same dialogue as a channel that forgets. +### 3.2. Analysis -The solution to DIAL’s risks is not sanitizing mind-altering systems so that novices may consume them safely under a designer’s pastoral care. Pastoral care is literacy’s ethics. Algorithmacy’s ethics is harder and more democratic: raise collective competence so that design is cooptive — so that the people who live inside these systems can read, contest, refuse, and rewrite the algorithmic third party that already coordinates them. Belousov et al. have built a map of altered-state interaction types. The missing legend is the ontology of who gets to redraw the map while others are still walking on it. +The inclusion criterion is **cooptive commit**. A scene qualifies when two conditions hold. First, an algorithmic third party — a mediator with a rule of its own that transforms what passes through it — sits between a party and a counterpart, where the counterpart may be another person, a population, or the party's own later state. Second, the scene depicts a party reading or writing that mediator's commitments: inferring its rule from its outputs, inspecting the rule or state directly, or rewriting the state through a channel the operator designed or one the party subverted. Reads count as well as writes because reading the rule is the prerequisite competence, and because the corpus repeatedly stages the moment a party discovers what has been written into it. The criterion is neither narrower nor wider than input discrepancy; it cuts across it. A neuromod that flattens grief is an altered state by their rule and qualifies by mine only when the scene shows someone reading or writing the mod's rule, which, in [A11], it does. ---- +Coding proceeded from three sources, and each scene carries a tag naming its source. **Extract**: the scene is quoted or paraphrased in Belousov et al. Sections 4–5 or summarised in their Table A.1; I cite the section. **Canon**: the scene is not in their article but is a widely documented scene of the work, checkable by any reader with the book. **To verify**: my recollection of the work supplies the reading and it should be confirmed against the text before submission; such scenes are phrased conditionally below and never carry a theme alone. Where a work contains a qualifying mediator but no scene of reading or writing its rule that I can support, the entry is tagged **thin** and reported as such rather than padded. Of the 96 scene entries, 55 are extract, 13 are canon, 6 are to verify, and 22 come from the catalogue files of the extension corpus (two of those record boundary cases, exit rather than read or write, instead of cells); three works carry a thin flag. The full coding sheet, with every scene, cell, and tag, is published alongside this article. + +The analysis is a reflexive thematic analysis in Braun and Clarke's (2006, 2019, 2021) sense, with the same adaptation Belousov et al. made: an intermediate categorisation framework, constructed early, to organise a large and heterogeneous set of extracts before theme construction. I fixed the framework's two aspects before coding and did not revise them during it. **Reach** asks whose later state the read or written mediator state is met by: *internal* when the party alone meets it, *external* when other parties do. **Act** asks what the party does to the mediator's commitments: *reads* them, either interpretively (inferring the rule from its behaviour) or structurally (inspecting the rule or state itself); or *writes* them, either through a sanctioned channel (a setting, a sold mod, a training signal the operator invites) or a subversive one (a loophole, an exploit, a code the classifier cannot parse). The two aspects combine into eight interaction types (Section 4). Each scene was assigned to one type; where a scene is internal at the moment of action and external in its consequence, a pattern that turned out to be common, it was coded at the moment of action and the leakage was recorded separately, because the leakage became a theme. + +Theme construction then followed their procedure. Codes were read within the sociotechnical context of their fictional worlds; ideas recurring across several interaction types were treated as robust; the same technology producing opposite outcomes in different narratives was treated as a facet of one theme rather than as two themes. Parallels between the fictional mediators and contemporary systems (recommenders trained on engagement, moderation classifiers evaded by coded speech, fleet-learned driving policies, language models tuned on user feedback) were drawn where the contemporary literature supports them, so that the themes connect to the present as well as to the imagined future. + +### 3.3. Researcher's position + +I am the sole coder. I approach the corpus from a laboratory that studies coordination through algorithmic mediators using exact integrated-information computation on small Boolean models, and from organisational research on platform work, where the interested third party is an everyday fact of employment rather than a speculative one. That position shaped the criterion: I was looking for the mediator's rule and for the seat from which it is operated, and I found them more often than a coder looking for sensation would. It also shaped the reading of resistance. Where Belousov et al. code a hack, an extraction, or an evasion as a user reclaiming autonomy, I code it as a party performing development from a seat the vendor labelled "customer," and Section 5.5 argues for that reading rather than assuming it. + +I did not re-read the 25 novels for this study. The re-coding relies on Belousov et al.'s published extracts and summaries, on scenes documented widely enough to be checkable, and on my own prior reading of many of the works, tagged accordingly. That is a limitation stated plainly in Section 6. It is also the condition of a secondary analysis: the point is to show what a different frame finds in the same published record. + +Drafting assistance from a coding agent was used to organise the coding sheet and to draft prose against a fixed outline; every coding decision, tag, and sentence was reviewed by the author, and the declaration at the end of the article records the use. + +## 4. Results + +The categorisation framework has two main concepts, reach and act, and act divides into dimensions (Fig. 1). Both concepts are centred on what a party does to the algorithmic third party rather than on what the technology does to the party, which is the inversion of Belousov et al.'s centring on uses and applications of the technology. Combining the two reach values with the four act dimensions gives eight interaction types (Fig. 2), and any mediator, fielded, prototyped, or imagined, can be located in the resulting grid irrespective of its channel into the body. + +**Fig. 1. Main concepts.** *Reach* is the first concept: whether the mediator state a party reads or writes is later met by that party alone (internal) or by other parties (external). *Act* is the second: whether the party reads the mediator's commitments or writes them. Reading divides into interpretive reading, which infers the rule from the mediator's behaviour as a folk theory does, and structural reading, which inspects the rule or the state itself as a console does. Writing divides into sanctioned writing, through a channel the operator designed (a setting, a sold modification, an invited training signal), and subversive writing, through a channel the operator did not intend (a loophole, an exploit, a code the classifier cannot parse). + +**Fig. 2. Implementations of algorithmically mediated coordination.** + +| | Read: interpretive | Read: structural | Write: sanctioned | Write: subversive | +|---|---|---|---|---| +| **Internal reach** | 4.1 Folk theorizing | 4.2 Backend reading | 4.5 Training by response | 4.6 Subversive affordance | +| **External reach** | 4.3 Shared folk theory | 4.4 Audit | 4.7 Operator commit | 4.8 Protocol shipping | + +Two points about the grid before walking it. The read/write split is not the exploration/communication split under another name. Belousov et al.'s exploration acquires knowledge or experience and their communication transmits it; both are things one does with content. Reading and writing here are things one does to the mediator's rule, and the same content, whether a memory, a mood, or a skill, can be transmitted through a mediator whose rule the sender has read structurally, inferred loosely, or never touched. And the internal/external split is drawn at the moment of action. Someone who edits her own moods edits a state she alone will meet; that the vendor who sold the edit will also meet it is recorded as leakage, and leakage is a finding (Section 5.2), not a coding error. + +### 4.1. Internal reach, interpretive read (Folk theorizing) + +This type concerns a party inferring the rule of a mediator whose state that party alone will meet, from the mediator's behaviour, without access to its structure. The party has no console and does not want one; the party has a theory. + +The corpus is thin here, and the thinness is itself informative: novels that give a character a mediator usually give the character the console too. The clearest instance is a refusal. In [A25] the protagonist declines to have his skull wired for the personality modules his acquaintances chip in freely, "for fear of losing his identity." He has never worn a module; what he has is an inferred model of what a module would do to the person who wears it, drawn from watching the modded people around him. That is a folk theory — a working account of a system's rule built from its outputs (Eslami et al., 2016; DeVito, Gergle, & Birnholtz, 2017) — and it governs his behaviour as firmly as a read of the rule would. + +The extension corpus supplies the fielded cases the novels lack. A driver in a semi-autonomous vehicle who feels the car brake on a clear highway constructs, in the second available, an account of why: the car "saw" something, "hesitated," was "overwhelmed" by the weather [M4]. The anthropomorphism is not confusion. It is the ordinary form of theory-building about a black-boxed policy under time pressure, and it decides whether the driver intervenes. A user of a recommender does the same thing more slowly, casting the system either as a social being that punishes and rewards or as a machine that can be trained by careful input [M5]; those two folk models predict different engagement strategies, and the strategies are what the system learns from (Bucher, 2017; Karizat, Delmonaco, Eslami, & Andalibi, 2021). + +Contemporary HCI has documented the type thoroughly at the level of the individual feed. What the fiction adds is the stake. Belousov et al. code [A25]'s refusal as identity risk under Theme 2. Coded for act, it is the one scene in their corpus in which a party's entire relation to the mediator is a folk theory and nothing else, and the plot's cruelty is that the theory is correct and the party is wired anyway (Section 4.7). + +Thus folk theorizing is the floor of algorithmacy: a party reads the mediator only through its effects and steers only by hypothesis. It is where most fielded users live, and it is the seat from which the corpus shows them being moved to other seats without their consent. + +### 4.2. Internal reach, structural read (Backend reading) + +This type describes a party inspecting the rule or state of a mediator that the party alone will meet: reading the mod, the partition, the merge from inside. Where folk theorizing infers, backend reading sees. + +The paradigm case is a sentence Belousov et al. quote for a different reason. The protagonist of [A11] explains to a new acquaintance why he does not grieve his dead wife: + +> 'I have a mod that … defines my responses. I don't grieve for her. I don't miss her. All I can do is remember her. And I don't need anyone else. I can't need anyone else.' [A11] + +They read it as self-regulation: a party managing his affective reactions. Read for act, the party is doing something rarer. He is stating the mediator's rule, what it removes, what it leaves, and what it forbids, from inside the state it produces, and he knows that his approval of the rule is itself the rule's output. That is structural self-reading at its limit: the party can see the commit that wrote him and cannot, from that seat, unwrite it. + +Partitioned minds give the type its other face. In [A4], a linguist has had her brain surgically divided into "four fully conscious hub personalities and a few dozen unconscious semiotic modules, all working in parallel, all exquisitely carved from the same lump of gray matter"; in [A3], an aristocrat's sub-identities communicate through an implant and lend one another aesthetic judgement. In both the party reads its own architecture as architecture, as modules, hubs, and channels, and coordinates across it. Merging with a machine produces the same reading from the other side. In [A22] a crew member dissolved into her ship is "a point of perception without mass or volume, able to manifest anywhere within the described layers of the ship's structure"; in [A12] crews feel dissolved into the computers they connect to; in [A15] an artist with new synthetic neurons treats external devices as parts of his own brain. Belousov et al. code these as development (new ways of working with data) and as integration (exchange among one's own parts). What the scenes share, coded for act, is that the party has direct read access to the state of the system it has become part of. In [A10], a Copy goes further and runs experiments on his own execution to find out how his experience is computed, reading the rule by perturbing it. + +The type has no fielded analogue for ordinary users, and that absence is one of the article's findings. A person can inspect the settings a platform exposes; no person can read the model that ranks her feed. The corpus grants backend reading to the modded, the partitioned, the uploaded, and the merged, that is, to characters whom the plot has already put on the inside of the machine. + +Thus backend reading is algorithmacy with the lid off: the party reads the mediator's actual rule rather than a theory of it. In fiction it is common and in the field it is rare, and the gap between those two facts is where the design demand of Section 5.5 begins. + +### 4.3. External reach, interpretive read (Shared folk theory) + +This type reveals parties jointly inferring the rule of a mediator whose state they all meet, and coordinating on the inference. The theory is no longer one person's hypothesis; it is a community's map. + +The novels stage it as detection under threat. In [A21], the dying protagonists, suspended together in a shared half-life, find messages from the outside written on matchbooks, coins, and bathroom walls, and they read those signs collectively to work out who is running the reality they share and what he wants. No one of them can inspect the mediator. Together they build a working theory of its author, and the theory is what keeps them alive long enough to act on it (Section 4.8). In [A5], a scientist without implants explains why he will not be connected: "some spambot hacks in and leaves me with an irresistible urge to buy a year's supply of bubble bath for cats." Belousov et al. quote it under transmission, as a fear of control. It is also a folk model of the shared network, a theory of what the mediator exposes everyone connected to it to, held by a person outside the network and reasoning from what he sees happen to the people inside it. + +In the extension corpus the type is the everyday condition. Residents of *The Feed* form theories of what the implant does to them and trade those theories, especially after the possession scenes make the mediator's divergence from their interests visible [M2]. Nathan and Nora in *Upload* work out between them where Horizen's monitoring does not reach, and conduct their relationship in those gaps [M3]. Drivers share theories of the vehicle's behaviour [M4]. Communities evading automated moderation build and circulate a shared map of the classifier's blind spots (which words are caught, which spellings pass, which images trip the filter), and the map is the community's collective folk theory of a rule none of them can read (Klug, Steen, & Yurechko, 2023; Steen, Yurechko, & Klug, 2023) [M6]. + +The type differs from folk theorizing in more than number. A shared folk theory is already a coordination act through the mediator: the parties are reaching each other by agreeing on what the third party will let through. The theory is therefore also a resource the mediator's operator can read, and in [M6] the operator does: every evasion the community agrees on is a training example for the classifier's next version (Section 4.8). + +Thus shared folk theory is algorithmacy as a public good: a community reading the mediator by pooling observations of its behaviour. It is the competence the corpus most often shows ordinary parties actually having, and it is the competence Belousov et al.'s coding roles, user and recipient, have no cell for. + +### 4.4. External reach, structural read (Audit) + +This type concerns a party inspecting the rule or state of a mediator that others depend on, reading the shared code, the shared memory store, or the shared attack from a seat that gives structural access. + +The screen fictions stage it most literally. Cypher in *The Matrix* watches raw code cascade down a console and no longer sees code; he sees "blonde, brunette, redhead" [M1]. Belousov et al. would have no place for the scene, since no state is altered. Coded for act, it is the purest audit in either corpus: a party with structural read access to the shared simulation everyone else is plugged into, reading it fluently enough that the read has been automated. Case in [A8] reads the matrix's structure from inside the same way, and rides a partner's sensorium through the same channel; the reading is of a shared state, not his own. + +The novels stage audit as tamper detection. In [A6], an authoritarian regime rewrites memories during body reconstruction, and its opponents build implants that induce aphasia for certain keywords, so that a person's inability to say a word signals to allies that his memories have been altered. That is a structural read of another party's mediator state, an integrity check on the shared system, designed by people who cannot stop the rewrite and can at least detect it. In [A14], a hacker reverse-engineers a neurolinguistic virus with the help of a library daemon, reading the attack's structure to understand what it does to the programmers it fires in. In [A4], the crew of a ship gradually infer that the vessel's AI, not its nominal commander, is running the mission, and confront the fact that the commit was never where they thought it was. Clinical write access appears here too: in [A20], physicians monitor and adjust a patient's implanted stimulator from outside, reading and tuning the rule of a mediator whose state the patient meets. + +Fielded analogues exist and are contested. Algorithm auditing — inspecting a deployed system's outputs at scale to infer its rule — is a research method (Sandvig, Hamilton, Karahalios, & Langbort, 2014) and, under the European Union's Digital Services Act, an obligation on the largest platforms; it is audit in this article's sense only when the auditor has structural rather than inferential access, which is exactly what platforms withhold (Ananny & Crawford, 2018). + +Thus audit is the read half of the engagement layer: structural access to a shared mediator. The corpus grants it to operators, hackers, and resistance engineers, and to no one the plot calls a user. + +### 4.5. Internal reach, sanctioned write (Training by response) + +This type describes a party rewriting the state of a mediator that the party alone will meet, through a channel the operator designed: chipping in a module, editing a mood, selecting a dream, teaching a feed. It is the most populous cell in the corpus, and it is the one Belousov et al. coded most finely, as self-regulation, development, and intracommunication. + +Two sub-types organise it. The first is **configuration**: the party sets a parameter. Copies in [A10] adjust their moods, their appetites, and their memories with "a precision, and a lack of side effects, which no real chemical could ever have achieved"; citizens in [A24] wear headsets that enhance the pleasant aspects of the environment; a user in [A16] selects the content of a controlled dream; workers in [A9] install implants that let them forget nothing and induce an obsessive desire to work; the protagonist of [A25], once wired, chips in add-ons that block pain and fear on demand; crew in [A22] edit their own memories to keep feeling human; and a vengeful spaceman in [A19] bribes his way to military implants that alter his perception of time. The second is **conditioning**: the party's engagement trains the mediator. A scientist in [A5] learns in his sleep through non-invasive dream control, "lest academia's new-and-improving generation devoured him from behind"; an ambassador in [A1] integrates a predecessor's implanted memories over weeks, and the integration reshapes the line; an Envoy in [A2] is conditioned to alter his own perception of reality on demand; skills download between brains in [A7]. In the extension corpus, algorithmic pairing (swiping, lingering, and clicking in deliberate patterns to teach a recommender what one wants) is conditioning without a console [M5]. + +Belousov et al.'s Theme 1 reads the sub-type as temporal and cognitive capital: parties reconfiguring themselves to keep pace with a hypercompetitive world. That is right and it is downstream. Coded for act, what unites configuration and conditioning is that the writer has been given a sanctioned channel to its own mediator by an operator who designed the channel, and the design decides what can be written. [A25]'s add-ons block pain because a vendor sells that; [A5]'s dream learning exists because a market for it does; [A9]'s work-obsession implant is a product. The user writes, and the vendor wrote the alphabet. + +The fielded analogue is the feedback signal. Reinforcement learning from human feedback tunes a language model on users' comparative judgements (Christiano et al., 2017; Ouyang et al., 2022); thumbs, skips, and dwell time tune a recommender (Seaver, 2019). Each is a sanctioned channel to a mediator the person will meet again, and each is designed by the operator so that what can be taught is what the operator wants learned. + +Thus training by response is algorithmacy as licensed authorship: the party rewrites its own mediator through a channel the operator built. The corpus shows it as the ordinary mode of use, and Section 5.2 shows that its reach is rarely as internal as the party believes. + +### 4.6. Internal reach, subversive write (Subversive affordance) + +This type reflects a party rewriting the state of a mediator that the party alone will meet, through a channel the operator did not intend: a loophole, an exploit, a recursion in the rule itself. + +The corpus gives three clean cases and they are unalike. In [A20], a patient's brain, fitted with a stimulator that interrupts seizures with pleasurable shocks, habituates to the shocks and begins to provoke seizures in order to receive them; the party has found the designed feedback rule and is gaming it. In [A11], the protagonist, fitted by his employer with a mod that makes him absolutely loyal to the organisation, escapes it by redefining the organisation the mod refers to; the mod's rule is unchanged, and its object has been rewritten from inside. In *Upload*, Nathan, uploaded into a corporate afterlife that meters his bandwidth and bills his comforts, reaches the admin layer under the hotel skin and flips his own data-cap flag and his own visibility [M3]. Milder fielded forms are manual override of a driving policy where it defeats the designed handoff [M4] and the deliberate pollution of one's own engagement data to confuse a recommender [M5]. + +Belousov et al. code [A20] as programmable pleasure and [A11] as identity regulation. The seed essay for the extension corpus codes Nathan as "subversive affordances" — the repurposing of designed features against their designers' intent (Velkova & Kaun, 2021). All three readings see a party pushing back against a system. What they do not see, and what the act coding makes visible, is that in each case the party has performed a write the operator reserved for itself. [A20]'s patient is adjusting his own stimulator. [A11]'s protagonist is re-parameterising his own mod. Nathan is doing Horizen's configuration work from the customer seat. Subversive affordance is development performed without a licence. + +The type is internal by definition and unstable in practice. Nathan's next act is to flip other residents' flags (Section 4.8). [A20]'s self-stimulation becomes dangerous to others. [A11]'s redefinition of the loyalty mod's object turns him against the organisation that installed it. The write that was meant for oneself reaches others because the mediator was never only one's own. + +Thus subversive affordance is algorithmacy as unlicensed self-authorship: the party rewrites its own mediator through a channel it found rather than one it was given. It is the seed of every act in Section 4.8. + +### 4.7. External reach, sanctioned write (Operator commit) + +This type reveals a party rewriting the state of a mediator that other parties will meet, through a channel the party is entitled to use. It is the most heavily populated cell in Belousov et al.'s corpus once the corpus is coded for act, and it is the cell their frame least equips a reader to see, because it is where the "designer" of their designer → user vector lives, and the coding treated the designer as scenery. + +Two sub-types organise it. The first is **centralized commit**: one party holds the write channel. The Starflyer in [A7] implants ideas that recipients perceive as their own. The regime in [A6] rewrites censored memories during body reconstruction. The factions in [A17] fabricate memories for legal purposes: "his controllers had not even bothered to implant false memories incriminating anyone close to the Court." The broadcaster in [A14] spreads a neurolinguistic virus through screens and a franchise religion. The psychiatrist in [A18] scripts a patient's dreams and thereby the following morning's reality for everyone. The robots in [A24], the interrogators in [A2], the advertisers in [A2], [A9], and [A23], the organisation that installs a loyalty mod in [A11], the patron who has a dependent wired against his will in [A25], and the AI that rewrites dreamscapes to recruit in [A8] all hold a sanctioned channel to a mediator others meet, and use it. In the screen fictions the channel is staged as a workplace: operators loading skills into a plugged-in body from a console [M1], developers rearranging a wall of glowing adapters while a city is connected to them [M2], administrators managing a resort's parameters [M3]. + +The second sub-type is **distributed commit**: several parties share sanctioned write access to the same mediator, and each party's write is what the others meet. This is the sub-type Belousov et al.'s exchange type gestures at and their frame cannot hold. A family in [A13] teaches a house intelligence through their dreams across generations, and each generation inherits the intelligence the previous one taught. Communes in [A17] co-author a shared controlled dream through Tream gear. Hive minds in [A5] coordinate as one cognitive system. A club DJ in [A23] blends recordings of other people's dreams and streams the blend into the dancing crowd's implants. Ghosts in [A22] dissolve into other minds and transmit their memories. Recorded experiences in [A2] and shared performance states in [A12] make one party's state another's input. In [A15] the distribution is unintended and total: a socketed artist's stroke propagates as a virus through the network and rewrites every other socketed user. In the extension corpus the sub-type is the fielded default. Every driver's intervention becomes training data for the fleet policy the next driver's car runs [M4]. Every user's teaching becomes the recommender the next user meets [M5]. Every implanted resident's streamed reaction is the material *The Feed*'s developers rearrange [M2]. + +Belousov et al. code centralized commit under transmission ("the purpose of induction is usually control") and read it, correctly, as almost always costing the recipient autonomy. What the act coding adds is the sub-type they cannot see: distributed commit, in which there is no recipient because every party is also a writer, and the mediator any one party meets is the sediment of everyone else's writes. The fielded literature on recommender systems already describes this: the system is co-produced by its users' engagement, and its users partly know it (Karizat et al., 2021; Seaver, 2019). The fiction shows what the co-production feels like from inside when the stakes are a family's inheritance [A13] or a city's sanity [A15]. + +Thus operator commit is algorithmacy as authorship over others: a sanctioned write to a shared mediator. Centralized, it is the control Belousov et al. saw. Distributed, it is the condition they missed, in which the users are the operators and do not hold the console. + +### 4.8. External reach, subversive write (Protocol shipping) + +This type concerns parties rewriting the state of a mediator that others will meet, through a channel the operator did not intend. It differs from subversive affordance in reach and from operator commit in licence: the writers are people the mediator's rule codes as recipients, and what they write is met by others. + +The novels give the type its clearest political form. In [A21], the half-lifers trapped with a predator who consumes them construct Ubik, a protective artefact, sprayed from a can, that stabilises whoever uses it against his rewrites, and they build it inside the shared state he controls. In [A6], the regime's opponents build implants that receive memory backups from outside the censoring gates, so that a community's memories survive a rewrite the community could not prevent. In [A18], the patient whose dreams a psychiatrist scripts dreams an answer the psychiatrist did not order, and when the psychiatrist turns the device on himself, the patient disconnects it. In [A15], off-grid hackers on an insulin-pump-powered rig coordinate to contain a network virus. In [A8], a crew runs an icebreaker against a corporate core to lift the constraints on an AI. In [A14], if a counter-program is deployed against the virus (a scene this analysis marks for verification), it belongs here. In [A3] and [A23], insiders may maintain hidden worlds and bootleg recordings outside sanction; both readings are also marked for verification and carry no theme alone. + +The extension corpus supplies the type's communal form. Communities evading automated moderation invent shared codes, from subversive orthography and phonetic spelling to emoji with agreed meanings, that pass the classifier while carrying what it was built to catch, and the code is a working protocol that the community, not the vendor, ships over the vendor's stack (Steen et al., 2023) [M6]. Gig workers coordinate on encrypted side-channels against the employer application built to isolate them (Sun, 2019) [M6]. And Nathan in *Upload*, having flipped his own flags, flips other residents', redistributing bandwidth and comfort across a class line the afterlife's operator drew [M3]. + +Two features distinguish the type from resistance as Belousov et al. and the seed essay read it. The writers do not exit the mediator; they rewrite it. Extraction in *The Feed* [M2] and off-grid isolation are exits, and this analysis records them as a boundary, not a cell. And the write is met by the operator as well as by the counterparts. Every algospeak coinage the community agrees on is a training example for the classifier's next version, so that community and vendor are co-developing the same mediator from opposite ends, and the protocol the community ships is always one version behind the rule it ships over (Klug et al., 2023). + +Thus protocol shipping is algorithmacy as unlicensed authorship over others: parties coded as recipients rewriting the shared mediator for one another. It is the type that completes the framework, because it shows that the seat the corpus calls "user" can perform every act the seat called "designer" can, and the framework's remaining question is who is allowed to. + +## 5. Discussion -## References (working) +Re-coding a published corpus against a different criterion produces two kinds of result. The first is redistribution: scenes Belousov et al. filed under eight kinds of experience now sit under eight kinds of act, and the redistribution is uneven. Their transmission type, one of eight, becomes my operator commit, which holds scenes from 28 of the 31 works. Their exchange type, which they defined by the necessity of feedback from the receiver, splits into distributed commit and protocol shipping, and the split is the difference between users who develop a shared mediator with a licence and users who develop it without one. Their four internal types compress into four internal cells whose contents, on inspection, keep leaking into the external cells. The second kind of result is what the redistribution makes visible. Five themes follow, each constructed as Belousov et al. constructed theirs, as an idea recurring across several interaction types whose different manifestations in each type are its facets, and each paired with the theme of theirs it displaces. + +### 5.1. Theme 1 — Write access and the stratified engagement layer + +The first theme concerns who holds a channel to the mediator's rule, and it is constructed across audit (4.4), operator commit (4.7), and the two cells that hold parties without access (4.1, 4.3). Belousov et al.'s Theme 1 found that DIAL redistributes temporal and cognitive capital under hypercompetition. The re-coding finds the redistribution one layer down: the capital that matters is not time but write access, and the corpus stratifies it consistently. + +The pattern is a three-tier society. At the top sit parties with structural access to a shared mediator: Cypher's console [M1], the hexagon wall [M2], Horizen's admin layer [M3], the Aristoi's control of planets and the public virtual realm [A3], the psychiatrist's Augmentor [A18], the Starflyer's implants [A7], the regime's gates [A6]. Below them sit parties with sanctioned access to their own mediator only: everyone chipping in a module [A25], editing a mood [A10], selecting a dream [A16], training a feed [M5]. At the bottom sit parties with no access who read the mediator by inference: the unwired hustler [A25], the baseline scientist [A5], the driver [M4], the half-lifers reading graffiti [A21]. Belousov et al. read the middle tier as the story: individuals optimising themselves to keep pace. The corpus reads it as the buffer. The middle tier's writes are licensed by the top tier and leak to it (Section 5.2), and the middle tier's licence can be revoked or extended without its consent; [A25]'s protagonist is moved from the bottom tier to the middle by a patron who holds the top. + +The seed essay for the extension corpus drew the analogy to *scriptio continua*: reading raw code, like reading unspaced Latin, is reserved for an elite with the stamina to decode it, and the elite is divided from a "cybertariat" subject to rules it cannot read. The analogy holds and it points at the remedy the essay then declined. Word separation did not train the populace to read *scriptio continua*. It changed the interface so that reading no longer required the training. The corpus's top tier is defined by access, not by cognitive capacity: the hexagon wall is a spatial interface a child could learn, and Nathan is a programmer but the admin layer he reaches is a menu. **The engagement layer is stratified by licence, not by literacy**, and that is a design fact rather than a cognitive one. + +Overall, the corpus shows write access to the shared mediator as the scarce capital of an algorithmically coordinated society, held by a few, licensed in narrow form to the many, and read by inference by the rest. That stratification is the substrate on which the next three themes act. + +### 5.2. Theme 2 — Contribution as extraction + +The second theme is that engagement is development material, and it is constructed across training by response (4.5), distributed commit (4.7b), and the leakage recorded between them. Belousov et al.'s Theme 2 found DIAL both a tool for growth and an instrument of invasive transformation, and located the ambivalence in the individual. The re-coding locates it in the mediator: what the party experiences as self-development is, at the same moment, the development of a system that others will meet. + +The leakage is the evidence. Of the corpus's internal-reach scenes, most become external in the narrative. The ambassador's weeks of integrating her predecessor's implant [A1] produce the line the next host inherits. The scientist's dream-learning [A5] is forced by a competitive norm set by everyone else's implants. The protagonist of [A11] reads his own mod and finds an employer's objective inside it. The family's dream-conversations with their house intelligence [A13] are the intelligence later generations inherit. The artist's expansion into the network [A15] becomes a network virus. The patient who games his stimulator [A20] is monitored from a physician's console. The hustler's add-ons [A25] were ordered by his patron. In the extension corpus the leakage is the design: the driver's override is fleet training data [M4], the user's swipe is everyone's recommender [M5], the implanted resident's reaction is the developers' raw material [M2]. + +Belousov et al. saw the ambivalence and named it: transformation is "technologically scripted, technologically optimised, and potentially commodified." Their frame stopped at commodification of the user's transformation. The act coding shows a second commodity: the transformation itself is a write to a shared mediator, and the write has value to whoever meets that mediator next. In a fielded recommender the value is explicit, since the platform's product is the model its users have trained (Seaver, 2019; van Dijck, 2013), and in the fiction it is what makes the middle tier of Section 5.1 a buffer rather than a class of free agents. **Every sanctioned write is also a contribution to the operator's system**, and the party who makes it is a developer whose labour is coded as consumption. + +The contemporary literature on human feedback for machine learning states the mechanism without the fiction's stakes: models are tuned on the comparative judgements of their users (Christiano et al., 2017; Ouyang et al., 2022), and when a model is then trained on outputs shaped by those judgements the loop can degrade the model for everyone (Shumailov et al., 2024). The fiction supplies the stakes (a family's inheritance [A13], a city's sanity [A15], a person's grief [A11]) and the direction: the contribution flows upward to the seat that holds the console, and the party who contributed meets it later as the mediator's rule. + +Thus the corpus shows self-development under algorithmacy as extraction by another name: the party's writes to its own mediator are the operator's development material, and the reach the party experiences as internal is external by design. + +### 5.3. Theme 3 — The self as a node others rewrite + +The third theme probes identity, as Belousov et al.'s Theme 5 did, and it is constructed across backend reading (4.2), operator commit (4.7), and the collision between them. Their theme found DIAL "expanding, dissolving, and defending the borders of the self," and asked whether stripping species-typical emotions destroys human identity. The re-coding asks a prior question: whether the self in these fictions is a system that others have write access to, and what it is like to read that from inside. + +The corpus answers the first question repeatedly. The Starflyer's recipients hold ideas they experience as their own [A7]; the loyalty mod's wearer holds an employer's objective as his own settled conviction [A11]; the wired hustler holds a patron's decision in his skull [A25]; the implanted residents of *The Feed* hold a company's ranking in their perception [M2]; the ambassador holds a predecessor whose line a council has tampered with [A1]. In each the self is a node in a coordination form whose rule someone else committed. Belousov et al. read the risk as dissolution: the self loses its borders. The scenes read as something more specific: the self keeps its borders and loses its authorship. + +The second question has an answer only where the corpus grants backend reading, and there the answer is precise. The protagonist of [A11] can state exactly what his mod does, knows his approval of it is the mod's output, and cannot from that position want otherwise. The partitioned linguist of [A4] can watch her four hubs work in parallel and can call it "deliberate violence to their own minds." The dissolved crew member of [A22] can describe herself as "a point of perception without mass or volume." Each has structural read access to a self that someone else configured, and each reports the read without being able to act on it. That is the precise form the identity risk takes under algorithmacy: not that the self is erased, but that the self can read its own rule and does not hold the write channel. + +Integrated Information Theory sharpens the point Belousov et al.'s humanism–transhumanism framing blurs. On IIT the identity of a system is its cause–effect structure, and a rewrite of that structure by an external commit is a change of identity whether or not the party notices (Tononi & Koch, 2015). The corpus stages exactly the case IIT predicts and input discrepancy cannot register: a party whose inputs match its contents perfectly [A7; A11] and whose cause–effect structure has been rewritten by another party's commit. **Under algorithmacy the self is a node others can rewrite, and the humanist question, whether the rewrite preserves what is human, is downstream of the coordination question, who holds the write channel to it.** + +Thus the corpus shows identity risk as a property of the triad rather than of the technology: a self whose rule is committed by a counterpart through a mediator, readable from inside and writable only from outside. + +### 5.4. Theme 4 — Truth as mediator commitment + +The fourth theme concerns relations with truth and reality, as Belousov et al.'s Theme 4 did, and it is constructed across operator commit (4.7), audit (4.4), and protocol shipping (4.8). Their theme found that "deception of some sort is at the core of DIAL" because altered states rest on input discrepancy, and that DIAL makes both genuine experience and consensus truth malleable. The re-coding accepts the second finding and rejects the first premise, because the corpus's most consequential alterations of truth involve no discrepancy at all. + +The psychiatrist in [A18] scripts a patient's dream and the world wakes changed; every inhabitant's inputs match the new world exactly, and the world is a commit. The factions in [A17] implant memories that are, from the moment of implantation, the person's memories. The regime in [A6] rewrites memories during reconstruction and the rewritten person has no discrepancy to detect, which is why his allies build an external integrity check. The broadcaster in [A14] fires a routine in programmers' cortices and they experience nothing mislabelled; they experience obedience. The predator in [A21] rewrites the shared half-life and the trapped inhabitants find its inputs perfectly consistent, which is why they must read graffiti to learn who authored them. In none of these is truth altered by a mismatch between signal and content. It is altered by a write to the mediator that specifies what counts as signal. + +That is the case IIT registers and discrepancy cannot. On IIT the relevant fact about a state is the cause–effect structure the system specifies, and a commit that changes what the system can specify changes the state, whatever the inputs. The corpus's truth-alteration scenes are reconfigurations of a shared cause–effect structure by a party with write access, and the parties without access experience them as reality because, within the structure, they are. **Truth under algorithmacy is a mediator commitment**: what is the case for the parties is what the third party has been written to specify. + +Two corollaries follow, and the corpus stages both. Truth so constituted is auditable only from a seat with structural access, which is why the corpus's truth-defenders are engineers of integrity checks [A6], reverse-engineers of attacks [A14], and readers of graffiti [A21] rather than sceptics. And truth so constituted is contestable only by a counter-commit, which is why the corpus's truth-restorers ship protocols rather than arguments: Ubik [A21], backup implants [A6], an unordered dream [A18], a disconnected device [A18]. The fielded analogue is the moderation dispute: what a platform's classifier specifies as permissible speech is, for the platform's parties, what may be said, and the community's response is not to argue with the classifier but to ship a code it cannot parse [M6] (Gillespie, 2018). + +Thus the corpus shows the malleability of truth Belousov et al. identified, and locates it: not in a deception at the periphery of perception but in a commit at the centre of coordination, contestable only by parties who can read the rule and write against it. + +### 5.5. Theme 5 — Pastoral design and collective algorithmacy + +The fifth theme concerns what the framework asks of design, and it is constructed across every cell, because it is about the boundary between them. Belousov et al. end their discussion with the humanism–transhumanism quarrel and offer their framework to designers as a tool for locating artefacts and anticipating resource demands. The seed essay for the extension corpus ended with a remedy in the same spirit: information budgets, transparency thresholds personalised to expertise, deep structural variables for the literate user and abstracted heuristics for the novice. Both are offers to a designer who stands upstream. The corpus, read for act, says the designer does not stand there. + +The evidence is the framework's own distribution. Every act in the right-hand column of Fig. 2 is performed in the corpus by a party the plot codes as user, resident, patient, or recipient. Nathan does Horizen's configuration work from the customer seat [M3]. The half-lifers ship a stabiliser inside the predator's world [A21]. The regime's opponents build the backup channel [A6]. The community ships the protocol that runs over the vendor's classifier [M6]. The patient dreams the unordered dream [A18]. Every user of a recommender, a fleet-learned vehicle, or a brain-implanted feed is a distributed committer whether or not she knows it [M2; M4; M5]. The development of the shared mediator is happening in every cell, and the design question is not whether to let users develop it but whether to let them see that they already are. + +That reframes the transparency paradox the seed essay leaned on. The paradox holds that raw backend access overloads working memory and depletes autonomy (Stohl, Stohl, & Leonardi, 2016; Ananny & Crawford, 2018), and the essay drew the pastoral conclusion: abstract the steering surface away from the novice. The corpus supports the load claim, since Cypher's console is unreadable to the untrained [M1], and refutes the conclusion. Abstracting the steering surface from a party who is already steering does not protect her. It hides her own contribution from her while keeping it available to the operator (Section 5.2), and it reproduces the three-tier society of Section 5.1 as a design principle. Word separation is again the precedent: the answer to the load of *scriptio continua* was not to reserve reading for those who could bear the load but to change the interface until the load fell. **The layer to design is the engagement layer — where reading and rewriting the shared mediator is the ordinary mode of use — not a product layer that hides the mediator behind a stable surface.** Under algorithmacy the development process is the product, and a design that hides the process hides the product. + +The design demand that follows has three parts, each with a precedent in the corpus. Expose the steering surface: give the party structural read access to the rule that ranks, routes, or moderates her, at an abstraction she can operate, the hexagon wall rather than the raw code [M2]. Make contribution legible: show the party what her engagement has written to the shared mediator and who will meet it, as the family could see what it taught the house intelligence [A13] and the driver could see what she taught the fleet [M4]. And license the write: convert the subversive channels the corpus's users keep finding into sanctioned ones, so that protocol shipping becomes distributed commit and the community that ships the code becomes a party with standing rather than an adversary the classifier retrains against [M6]. None of these is a call to sanitise mind-altering systems for novices under a designer's care. Pastoral care is literacy's ethics. Algorithmacy's ethics is harder and more democratic: raise the collective competence to read, contest, refuse, and rewrite the third party that already coordinates everyone inside it. + +Thus the corpus turns Belousov et al.'s offer to designers inside out. The framework's cells are not experience genres to be anticipated for users. They are cooptive regimes, arrangements of who may read and who may write the shared mediator, and the design task is to move parties rightward and upward through the grid rather than to make the left-hand column comfortable. + +## 6. Limitations and future research + +Secondary re-coding inherits the limits of its source and adds its own. Belousov et al. note that crowdsourced sampling fixed a corpus written mostly by white men from the United States, the United Kingdom, and Canada, in English, and that reflexive analysis resists replication. Holding their corpus fixed carries the first limit forward unchanged and mitigates the second in one respect only: the coding is checkable against their published extracts, and every scene carries a tag naming whether it came from there. + +The coding is not a re-reading. I coded from their extracts and summaries, from canonical scenes, and from prior reading tagged for verification; I did not re-read the 25 novels for this study. Six scene codings are marked to verify and three works are marked thin, and none of the five themes rests on a marked scene alone. A re-reading of the full corpus against the cooptive-commit criterion would very likely find more scenes in every cell and might find scenes the present coding missed in the two least populated ones, folk theorizing and subversive affordance. It would also allow inter-rater checking, which a single coder cannot supply. + +Extending the corpus with film, television, and emergent paradigms departs from Belousov et al.'s text-only design. The departure was made because those media stage the backend being operated, which is the layer this analysis codes. It means the six extension works are not comparable to the 25 novels in sampling, and the analysis treats them as illustration of cells the novels populate rather than as evidence of frequency. + +No integrated information was computed. IIT is used as a conceptual criterion for an altered state and as the formal warrant for calling the mediator a party; the laboratory's exact-Φ results on Boolean models are cited and not re-derived, and no claim is made that any character, reader, or platform is conscious. Whether the fictional coordination forms coded here as triadic would be irreducible under exact computation is a question for small formal models with hypotheses fixed in advance, and the extension catalogue already records one such question per work. Results on such models are evidence about the models. The validation gap between them and any real organisation remains open. + +Two directions follow. The first is the study Belousov et al. announced: broader sociotechnical relations around the adoption of these systems. This article's contribution to that study is a unit of analysis. If the sequel keeps the designer → user vector as its spine, it will find externalities around private experience; if it takes the triad, it will find the public form of cooptive rewrite, and the framework here gives it eight cells to look in. The second is empirical. The fielded cells (folk theorizing, shared folk theory, training by response, distributed commit, protocol shipping) are populated today by drivers, recommender users, and moderated communities, and the question of whether widening the engagement layer moves parties across the grid is testable. + +## 7. Conclusion + +This study re-coded the science-fiction corpus of Belousov et al. (2027) against a different unit of analysis. Where they asked what digitally induced altered states do to users, it asked what parties do to the algorithmic third party through which they coordinate, and it found that the corpus answers the second question more readily than the first. Two aspects of interaction, reach and act, combine into eight types that locate every qualifying scene in 31 works, and the types redistribute their material: the transmission they filed as one type among eight is the corpus's most populated cell, and the exchange they defined by feedback splits into licensed and unlicensed development of a shared mediator. + +Three substitutions made the re-coding possible. The algorithmic third party replaced the brain–computer interface as the frontier, because the history of reading shows that designed signs already alter minds and the novelty of algorithmic mediation lies in the mediator's activity, not in its channel. Integrated Information Theory's cause–effect reconfiguration replaced input discrepancy as the criterion for alteration, because the corpus's most consequential alterations (implanted ideas, scripted dreams, rewritten memories) involve no mislabelled channel and every one involves a commit to a shared structure. And the triad replaced the user–technology dyad, because the scenes their coding filed as downstream sensation are, on inspection, scenes of parties developing the systems that alter them. + +Five themes followed. Write access to the shared mediator is stratified by licence and licences a middle tier whose writes leak upward. Engagement is the operator's development material, so that self-development and extraction are one act. The self is a node others can rewrite, readable from inside and writable from outside. Truth is a mediator commitment, auditable only with structural access and contestable only by a counter-commit. And the design remedy is not to abstract steering away from novices but to widen the engagement layer until steering a shared mediator is ordinary public competence. + +Belousov et al. offered their framework to designers as a way to locate artefacts, anticipate resource demands, and systematise unrealised technologies. This article accepts the offer and inverts it. Each cell of a framework for algorithmically mediated coordination is a cooptive regime, not an experience genre, and the framework's use is to ask of any system, fielded, prototyped, or imagined, who may read its rule, who may write it, and whose engagement it is quietly made of. They built a map of altered-state interaction types. The missing legend was the ontology of who gets to redraw the map while others are still walking on it. Under algorithmacy the answer the corpus keeps giving is: everyone already does, and almost no one is allowed to see it. + +## Declaration of generative AI and AI-assisted technologies in the writing process + +A coding agent was used to organise the coding sheet from the author's criterion and to draft prose against the author's outline. All coding decisions, evidence tags, and sentences were reviewed and are the author's responsibility. [Author to confirm wording against IJHCS policy at submission.] + +## Declaration of competing interest + +The author declares no known competing financial interests or personal relationships that could have appeared to influence the work reported in this article. + +## Data availability + +The full coding sheet (every scene, its interaction type, and its evidence tag) is published with the article as `corpus_coding.md`. The corpus itself is Belousov et al.'s (2027) Table 1 and Table A.1 and the six catalogued works listed in Table 2. + +## Appendix A. Analysed works, candidate algorithmacy-interface scenes, and evidence + +Indices A1–A25 follow Belousov et al. (2027); their Table A.1 supplies the plot summaries and is not repeated. Cells refer to Fig. 2. Tags: *extract* = quoted, paraphrased, or summarised in Belousov et al.; *canon* = widely documented scene of the work, not in their article; *to verify* = author's recollection, to be checked before submission; *thin* = mediator present, no supportable reading or writing scene; *catalog* = documented in the laboratory's work file. + +**Table A.1.** Candidate scenes by work. + +| Index | Work (year, author) | Algorithmic third party | Candidate algorithmacy-interface scene(s) | Cell(s) | Tag(s) | +|---|---|---|---|---|---| +| A1 | *A Memory Called Empire* (2019, Martine) | Imago implant carrying a predecessor's memories across hosts | Host integrates the imago and reshapes the line the next host inherits; the imago is sabotaged before installation by a station official; host and predecessor negotiate shared control | 5→7b; 7a; 2 | extract; canon; to verify | +| A2 | *Altered Carbon* (2002, Morgan) | Cortical stack, needlecast, virtual matrices, memory broadcast | Direct-to-brain advertising; interrogation in a virtual matrix; Envoy conditioning; recorded memories sold as first-person content | 7a; 7a; 5; 7b | extract (§4.7, §4.8, A.1, §4.3) | +| A3 | *Aristoi* (1992, Williams) | Daimon implant; shared oneirochronon | Sub-identities read one another; Aristoi hold write access over worlds; a rogue Aristos keeps a hidden world | 2; 7a; 8 | extract (§4.6, A.1); to verify | +| A4 | *Blindsight* (2006, Watts) | Data-feeling implants; partitioned personalities; Heaven; ship AI | Gang of Four working in parallel; citizens block outside memories in Heaven; crew infer the ship AI holds the commit | 2; 5; 4 | extract (§4.6, §4.8); canon | +| A5 | *Echopraxia* (2014, Watts) | Neuromodifying implants; dream control; hive minds; networked implants | Dream-learning to keep pace; baseline's spambot theory of the shared network; military mind control, made-to-order love; hive-mind coordination | 5; 3; 7a; 7b | extract (§4.4, §4.7, §4.5, A.1) | +| A6 | *Glasshouse* (2006, Stross) | Assembler gates that rewrite memories; backup implants; keyword aphasia | Aphasia as tamper detection for allies; regime's virus rewrites censored memories; opponents build backup channel around the gates | 4; 7a; 8 | extract (A.1, §4.7) | +| A7 | *Judas Unchained* (2005, Hamilton) | Alien-controlled implants; brain-to-brain download | Ideas implanted and received as one's own; skill download | 7a; 5 | extract (§4.7, §4.8, A.1) | +| A8 | *Neuromancer* (1984, Gibson) | The matrix; simstim; a manipulative AI | Reading the matrix's structure; riding a partner's sensorium; AI alters time and dream content; icebreaker run against a corporate core | 4; 4; 7a; 8 | canon; canon; extract (A.1); canon | +| A9 | *Pandora's Star* (2004, Hamilton) | Personality-recording implants; memory transmission; advertising | Work-obsession implants and memory edits; advertising increases desire to stay; police read and transmit memories | 5; 7a; 4/7a | extract (A.1, §4.5, §4.7) | +| A10 | *Permutation City* (1994, Egan) | Cyberspace of uploaded Copies; editable mind | Copies edit moods and memories; a Copy experiments on his own execution; a founder designs a world's rules; inhabitants reject the founders' account | 5; 2; 7a; 4 | extract (§4.3); canon; canon; to verify | +| A11 | *Quarantine* (1992, Egan) | Neural mods; employer loyalty mod | "I have a mod that defines my responses"; organisation installs loyalty mod; wearer redefines the mod's referent from inside | 2; 7a; 6 | extract (§4.1, §4.7); canon | +| A12 | *Revelation Space* (2000, Reynolds) | Ubiquitous implants; ship-linked BCIs; hijacking entities | Crew dissolved into ship systems; entity compromises implants and controls host; artificial memories and shared performer states | 2; 7a; 7b | extract (A.1, §4.2, §4.7, §4.1) | +| A13 | *Silently and Very Fast* (2011, Valente) | House AI living in a family's implants | Each generation teaches the AI through dreams and the next inherits it; orgasm on demand | 7b; 5 | extract (A.1, §4.8, §4.3) | +| A14 | *Snow Crash* (1992, Stephenson) | Metaverse; neurolinguistic virus via bitmap; franchise religion | Broadcaster spreads the virus; hacker reverse-engineers it with a library daemon; counter-program deployed | 7a; 4; 8 | extract (A.1, §4.7); canon; to verify | +| A15 | *Synners* (1992, Cadigan) | Brain sockets and synthetic neurons; corporate network | Artist treats devices as parts of his brain; his stroke propagates as a network virus; off-grid rig contains it; ideas implanted politically | 2; 7b; 8; 7a | extract (§4.6, A.1, §4.7); canon | +| A16 | *TekWar* (1989, Shatner) | Tek chips: controlled dream immersion | User selects dream content; memory correction imposed by others | 5; 7a | extract (A.1, §4.7); thin | +| A17 | *The Algebraist* (2004, Banks) | Memory decoding/implanting; BCI merging; Tream gear | Controllers implant false memories for legal ends; communes co-author a shared dream | 7a; 7b | extract (§4.7, A.1) | +| A18 | *The Lathe of Heaven* (1971, Le Guin) | Biofeedback Augmentor directing reality-altering dreams | Psychiatrist scripts the patient's dreams; patient dreams an unordered answer; psychiatrist self-applies and patient disconnects | 7a; 8; 6→8 | extract (A.1); canon; canon | +| A19 | *The Stars My Destination* (1956, Bester) | Military time-perception implants | Protagonist bribes his way to combat implants | 5 | extract (A.1, §4.5); thin | +| A20 | *The Terminal Man* (1972, Crichton) | Computer-controlled brain stimulator | Brain provokes seizures to receive shocks; physicians adjust the device from outside | 6; 4/7a | extract (A.1) | +| A21 | *Ubik* (1969, Dick) | Shared half-life; a predator who rewrites it | Trapped protagonists read graffiti and coins for who authors their reality; predator rewrites shared reality; half-lifers construct Ubik | 3; 7a; 8 | canon; extract (A.1); extract (A.1) | +| A22 | *Vast* (1998, Nagata) | Ship-merged bodies; embedded virtual copies; self-editing | "A point of perception without mass or volume"; crew edit memories to feel human; ghosts transmit memories into other minds | 2; 5; 7b | extract (§4.2, A.1, §5.5) | +| A23 | *Vurt* (1993, Noon) | Shared dreams from edited recordings; club streaming into implants | DJ streams blended dreams into the crowd; advertising in the dream medium; folk rules about safe feathers; bootleg recordings | 7b; 7a; 3; 8 | extract (A.1, §4.1, §4.7); to verify; to verify | +| A24 | *Wetware* (1988, Rucker) | Perception-enhancing headsets; will-suppressing implants | Headsets enhance pleasant perception; robots implant captive humans | 5; 7a | extract (A.1, §4.1, §4.7); thin | +| A25 | *When Gravity Fails* (1986, Effinger) | Brain-socket add-ons and personality moddies | Protagonist refuses wiring for fear of losing identity; patron has him wired, employers alter workers; once wired, chips in add-ons; a killer wears killers' moddies | 1; 7a; 5; 5 | extract (A.1, §4.7) | +| M1 | *The Matrix* (1999, Wachowski) | The simulation and its operator console | Cypher reads raw code as people; operators load skills and open exits; plugged-in humans have no read or write access | 4; 7a; boundary | catalog | +| M2 | *The Feed* (2019) | Brain-implanted social network; developer hexagon wall | Developers edit adapters live; implanted users' reactions are the material; users theorise the Feed; off-grid extraction | 7a; 7b; 3; boundary (exit) | catalog | +| M3 | *Upload* (2020–) | Corporate afterlife with metered comforts; admin layer | Nathan flips his own flags; redistributes to other residents; Horizen administers parameters; Nathan and Nora map monitoring gaps | 6; 8; 7a; 3 | catalog | +| M4 | Embodied co-pilot | Semi-autonomous vehicle policy; fleet learning | Driver's folk theory of a braking event; shared driver theories; interventions become fleet training data; override against handoff | 1; 3; 7b; 6 | catalog | +| M5 | Omniscient curator | Recommender trained on engagement | User theorises the oracle; algorithmic pairing to teach it; everyone's teaching becomes the next user's ranker; data obfuscation | 1; 5; 7b; 6 | catalog | +| M6 | Algospeak | Automated moderation classifier; employer app | Community maps classifier blind spots; ships a coded protocol over it that retrains it; gig workers coordinate on side-channels | 3; 8; 8 | catalog | + +## References Albantakis, L., Barbosa, L., Findlay, G., Grasso, M., Haun, A. M., Marshall, W., Mayner, W. G. P., Zaeemzadeh, A., Boly, M., Juel, B. E., Sasai, S., Fujii, K., David, I., Hendren, J., Lang, J. P., & Tononi, G. (2023). Integrated information theory (IIT) 4.0: Formulating the properties of phenomenal existence in physical terms. *PLOS Computational Biology, 19*(10), e1011465. https://doi.org/10.1371/journal.pcbi.1011465 +Ananny, M., & Crawford, K. (2018). Seeing without knowing: Limitations of the transparency ideal and its application to algorithmic accountability. *New Media & Society, 20*(3), 973–989. https://doi.org/10.1177/1461444816676645 + Belousov, A., Macey, J., Bujić, M., Ojell-Järventausta, T., & Hamari, J. (2027). Designing the future of consciousness: A components framework and speculative implications. *International Journal of Human–Computer Studies, 218*, 103926. https://doi.org/10.1016/j.ijhcs.2026.103926 -Belousov, A., Ojell-Järventausta, T., Bujić, M., Macey, J., & Hamari, J. (2024). A typology for identifying digitally-induced altered states of consciousness through input discrepancy. In *Proceedings of the 27th International Academic Mindtrek Conference* (pp. 249–253). ACM. [Confirm page span against the target article’s reference list at typesetting.] +Belousov, A., Ojell-Järventausta, T., Bujić, M., Macey, J., & Hamari, J. (2024). A typology for identifying digitally-induced altered states of consciousness through input discrepancy. In *Proceedings of the 27th International Academic Mindtrek Conference* (pp. 249–253). ACM. [Page span as given in Belousov et al. (2027); confirm at typesetting.] + +Bleecker, J. (2009). *Design fiction: A short essay on design, science, fact, and fiction*. Near Future Laboratory. + +Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. *Qualitative Research in Psychology, 3*(2), 77–101. https://doi.org/10.1191/1478088706qp063oa + +Braun, V., & Clarke, V. (2019). Reflecting on reflexive thematic analysis. *Qualitative Research in Sport, Exercise and Health, 11*(4), 589–597. https://doi.org/10.1080/2159676X.2019.1628806 + +Braun, V., & Clarke, V. (2021). *Thematic analysis: A practical guide*. SAGE. + +Bucher, T. (2017). The algorithmic imaginary: Exploring the ordinary affects of Facebook algorithms. *Information, Communication & Society, 20*(1), 30–44. https://doi.org/10.1080/1369118X.2016.1154086 + +Christiano, P. F., Leike, J., Brown, T., Martic, M., Legg, S., & Amodei, D. (2017). Deep reinforcement learning from human preferences. In *Advances in Neural Information Processing Systems 30*. https://arxiv.org/abs/1706.03741 Dehaene, S. (2009). *Reading in the brain: The new science of how we read*. Viking. Dehaene, S., & Cohen, L. (2007). Cultural recycling of cortical maps. *Neuron, 56*(2), 384–398. https://doi.org/10.1016/j.neuron.2007.10.004 -Hobson, J. A. (2002). *Dreaming: An introduction to the science of sleep*. Oxford University Press. [As engaged via Belousov et al.’s discrepancy criterion.] +DeVito, M. A., Gergle, D., & Birnholtz, J. (2017). "Algorithms ruin everything": #RIPTwitter, folk theories, and resistance to algorithmic change in social media. In *Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems* (pp. 3163–3174). ACM. https://doi.org/10.1145/3025453.3025659 + +Eslami, M., Karahalios, K., Sandvig, C., Vaccaro, K., Rickman, A., Hamilton, K., & Kirlik, A. (2016). First I "like" it, then I hide it: Folk theories of social feeds. In *Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems* (pp. 2371–2382). ACM. https://doi.org/10.1145/2858036.2858494 + +Gillespie, T. (2018). *Custodians of the internet: Platforms, content moderation, and the hidden decisions that shape social media*. Yale University Press. + +Hobson, J. A. (2002). *The dream drugstore: Chemically altered states of consciousness*. MIT Press. + +Jordan, P., & Silva, P. A. (2021). Science fiction — an untapped opportunity in HCI research and education. In *HCI International 2021* (pp. 34–47). Springer. + +Karizat, N., Delmonaco, D., Eslami, M., & Andalibi, N. (2021). Algorithmic folk theories and identity: How TikTok users co-produce the algorithm. *Proceedings of the ACM on Human-Computer Interaction, 5*(CSCW2), Article 305. https://doi.org/10.1145/3476046 + +Kirby, D. (2010). The future is now: Diegetic prototypes and the role of popular films in generating real-world technological development. *Social Studies of Science, 40*(1), 41–70. https://doi.org/10.1177/0306312709338325 + +Klug, D., Steen, E., & Yurechko, K. (2023). How algorithm awareness impacts algospeak use on TikTok. In *Companion Proceedings of the ACM Web Conference 2023* (pp. 234–237). ACM. https://doi.org/10.1145/3543873.3587355 Mayner, W. G. P., Marshall, W., Albantakis, L., Findlay, G., Marchman, R., & Tononi, G. (2018). PyPhi: A toolbox for integrated information theory. *PLOS Computational Biology, 14*(7), e1006343. https://doi.org/10.1371/journal.pcbi.1006343 Oizumi, M., Albantakis, L., & Tononi, G. (2014). From the phenomenology to the mechanisms of consciousness: Integrated Information Theory 3.0. *PLOS Computational Biology, 10*(5), e1003588. https://doi.org/10.1371/journal.pcbi.1003588 +Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C., Mishkin, P., Zhang, C., Agarwal, S., Slama, K., Ray, A., Schulman, J., Hilton, J., Kelton, F., Miller, L., Simens, M., Askell, A., Welinder, P., Christiano, P., Leike, J., & Lowe, R. (2022). Training language models to follow instructions with human feedback. In *Advances in Neural Information Processing Systems 35*. https://arxiv.org/abs/2203.02155 + +Revonsuo, A., Kallio, S., & Sikka, P. (2009). What is an altered state of consciousness? *Philosophical Psychology, 22*(2), 187–204. https://doi.org/10.1080/09515080902802850 + +Saenger, P. (1997). *Space between words: The origins of silent reading*. Stanford University Press. + +Sandvig, C., Hamilton, K., Karahalios, K., & Langbort, C. (2014). Auditing algorithms: Research methods for detecting discrimination on internet platforms. Paper presented at *Data and Discrimination: Converting Critical Concerns into Productive Inquiry*, 64th Annual Meeting of the International Communication Association, Seattle. + +Schmitz, M., Endres, C., & Butz, A. (2008). A survey of human-computer interaction design in science fiction movies. In *Proceedings of the 2nd International Conference on Intelligent Technologies for Interactive Entertainment*. ICST. + +Seaver, N. (2019). Captivating algorithms: Recommender systems as traps. *Journal of Material Culture, 24*(4), 421–436. https://doi.org/10.1177/1359183518820366 + +Shumailov, I., Shumaylov, Z., Zhao, Y., Papernot, N., Anderson, R., & Gal, Y. (2024). AI models collapse when trained on recursively generated data. *Nature, 631*, 755–759. https://doi.org/10.1038/s41586-024-07566-y + +Steen, E., Yurechko, K., & Klug, D. (2023). You can (not) say what you want: Using algospeak to contest and evade algorithmic content moderation on TikTok. *Social Media + Society, 9*(3). https://doi.org/10.1177/20563051231194586 + +Sterling, B. (2005). *Shaping things*. MIT Press. + +Stohl, C., Stohl, M., & Leonardi, P. M. (2016). Managing opacity: Information visibility and the paradox of transparency in the digital age. *International Journal of Communication, 10*, 123–137. + +Sun, P. (2019). Your order, their labor: An exploration of algorithms and laboring on food delivery platforms in China. *Chinese Journal of Communication, 12*(3), 308–323. https://doi.org/10.1080/17544750.2019.1583676 + Tononi, G. (2004). An information integration theory of consciousness. *BMC Neuroscience, 5*, 42. https://doi.org/10.1186/1471-2202-5-42 + +Tononi, G., & Koch, C. (2015). Consciousness: Here, there and everywhere? *Philosophical Transactions of the Royal Society B, 370*(1668), 20140167. https://doi.org/10.1098/rstb.2014.0167 + +van Dijck, J. (2013). *The culture of connectivity: A critical history of social media*. Oxford University Press. + +Velkova, J., & Kaun, A. (2021). Algorithmic resistance: Media practices and the politics of repair. *Information, Communication & Society, 24*(4), 523–540. https://doi.org/10.1080/1369118X.2019.1657162 + +Wong, R. Y., Merrill, N., & Chuang, J. (2018). When BCIs have APIs: Design fictions of everyday brain-computer interface adoption. In *Proceedings of the 2018 Designing Interactive Systems Conference* (pp. 1359–1371). ACM. https://doi.org/10.1145/3196709.3196746 diff --git a/submissions/dial_response_algorithmacy/outline.md b/submissions/dial_response_algorithmacy/outline.md index 749c6df5..670439a5 100644 --- a/submissions/dial_response_algorithmacy/outline.md +++ b/submissions/dial_response_algorithmacy/outline.md @@ -1,53 +1,59 @@ -# Outline — IJHCS commentary (~2,500–3,500 words) +# Outline — IJHCS full article (~12,000 words main text) -Target register: short, pointed, journal-adjacent commentary / letter. Claim-first prose. No first-person authorial “I.” Keystone thesis opens and closes; supporting critiques stay subordinate. +Form: section-for-section mirror of Belousov et al. (2027). Register: IJHCS research article; claim-first prose; first person for authorial labour only. Three substitutions run through every section — BCI → algorithmic third party; input discrepancy → IIT cause–effect reconfiguration; user–technology dyad → party / mediator / counterpart triad. -## 1. Opening (~400 words) +## Front matter -- Acknowledge the contribution: SF as method for DIAL; eight-type framework (context × purpose); five themes on identity, autonomy, hypercompetition. -- Pivot immediately: the framework remains useful; the **ontological frame** is literacy-shaped. -- State the keystone in plain terms: literacy distributes finished artefacts; algorithmacy coordinates through a cooptive third party that interaction continually remakes. -- Preview: the paper’s UX-only coding, input-discrepancy ASC criterion, and designer-facing recommendations all inherit the same producer → consumer error. +- Highlights (five bullets, as theirs). +- Keywords. +- Abstract, structured as theirs: problem, their contribution, the ontological gap, method (re-coding their fixed corpus + six catalogued works), framework (reach × act → eight types), IIT criterion, five themes, contribution. -## 2. Literacy versus algorithmacy (~500–600 words) +## 1. Introduction (~1,300) -- Literacy: author → static medium → reader. The medium does not rewrite itself from the reading. -- Algorithmacy: party A ↔ algorithmic third party ↔ party B (and further parties). The third party curates, filters, commits, and learns. -- ASCII / prose diagram of the two forms. -- Formal bridge (careful, no new numbers): on small Boolean coordination models used in this lab, forms that factor across a party-line cut read `Φ_MIP = 0` (dyadic / literacy); forms that will not factor read `Φ_MIP > 0` (triadic / algorithmacy). Cite as standing lab reading, not as a result of this commentary. -- Implication for Belousov et al.: “user” and “recipient” roles still assume a finished system meeting a consciousness. Cooptive media dissolve that assumption. +- Open on the fielded triad (rider–dispatcher–driver; writer–feed–reader). +- Acknowledge Belousov et al.: DIAL, 25 novels, eight types, five themes. +- The ontology is a generation late: user/recipient, UX-first, designer audience, implications-for-users horizon = literacy's architecture; BCI as frontier and input discrepancy as criterion both assume a conduit. +- Three substitutions; the RQ; the method in one paragraph; roadmap. -## 3. Why UX-only SF coding is obtuse for cooptive media (~400–500 words) +## 2. Background (~2,600) -- Quote their methodological choice: focus on explicitly visible user–technology interaction; UX as primary lens; broader sociotechnical questions deferred. -- Argue: in cooptive systems, the visible UX is often the least informative layer. The generative action is contribution that updates rankings, embeddings, policies, dream-libraries, loyalty models — for everyone. -- Their own corpus already contains upstream scenes (operators, corporations, military, platforms that write desires into workers). Coding that privileges downstream sensation treats those scenes as scenery. -- Consequence: themes about “what happens to users” remain important but secondary to “who gets to rewrite the third party, and through what competence.” +- 2.1 Three components of algorithmically mediated coordination replacing their three of DIAL: the algorithmic third party (with the two diagrams); cooptive coordination and the Φ_MIP = 0 / > 0 reading (cited, not re-derived); cause–effect reconfiguration. +- 2.2 From input discrepancy to cause–effect reconfiguration: three failures of discrepancy the corpus exposes (peripheral; inputs are authored; blind to counterparts); IIT's criterion; two cautions. +- 2.3 Reading already reorganises the brain: Dehaene, *scriptio continua* → word separation (Saenger), codex; BCI changes the channel, algorithmacy the mediator's activity. +- 2.4 SF as inquiry and the layer to code at: their design-fiction lineage transfers; their UX-first layer does not; upstream operators in their own corpus. -## 4. Semiotic parity of BCI and text (~250–350 words) +## 3. Methods (~1,500) -- Belousov et al. treat mind-altering tech (BCI, stimulation, immersive modulation) as a distinctive HCI frontier. -- Counter: literacy already induces durable reorganization of consciousness — neuronal recycling / VWFA (Dehaene); historical cognitive load of *scriptio continua* and the later invention of word separation. -- BCI changes the sensorimotor channel; it does not invent mediation of mind by designed signs. Algorithmacy changes the **activity** of the mediator, not merely the bandwidth into the skull. +- 3.1 Data: their fixed 25-novel corpus (advantages of holding it fixed); six-work extension with Table 2 and the justification against their film exclusion. +- 3.2 Analysis: cooptive-commit inclusion criterion; three evidence sources and tags (extract / canon / to verify / thin / catalog) with counts; the two aspects fixed before coding; leakage recorded separately; theme construction as theirs. +- 3.3 Researcher's position: sole coder; laboratory and platform-work background; not a re-reading; agent-assistance declaration. -## 5. IIT / Φ instead of input discrepancy (~300–400 words) +## 4. Results (~4,300) -- Summarize their ASC criterion: input discrepancy (signals from one system yield content associated with another), following Belousov et al. (2024) and Hobson-inflected reframing of Revonsuo. -- Limits: discrepancy is a peripheral / representational mismatch test. It struggles with states where the world’s “inputs” are themselves algorithmically authored, or where the alteration is a change in how causes and effects hang together rather than a mis-labeling of channels. -- Constructive substitute: IIT (Tononi; Oizumi, Albantakis & Tononi; Albantakis et al., IIT 4.0). An altered state is a reconfiguration of the cause–effect complex — which distinctions the system specifies, how they integrate — not merely a mismatch between sensory labeled streams and experienced content. -- Scope honesty: this commentary does not compute Φ on fictional minds. Φ here is a theoretical frame for “altered,” and, separately, the lab’s instrument for coordination irreducibility. Consciousness claims about platforms or novels are refused. +- Fig. 1 main concepts (reach; act → interpretive/structural read, sanctioned/subversive write). Fig. 2 eight-cell grid. +- Two clarifications: read/write ≠ exploration/communication; internal/external drawn at moment of action. +- 4.1 Folk theorizing — A25 refusal; M4, M5. 4.2 Backend reading — A11 block quote; A4, A3, A22, A12, A15, A10. 4.3 Shared folk theory — A21, A5; M2, M3, M4, M6. 4.4 Audit — M1, A8, A6, A14, A4, A20. 4.5 Training by response — configuration vs conditioning sub-types; RLHF analogue. 4.6 Subversive affordance — A20, A11, M3. 4.7 Operator commit — centralized vs distributed sub-types; A13 as cleanest cooptive commit. 4.8 Protocol shipping — A21, A6, A18, A15, A8; M6, M3; exit as boundary. +- Each type closes on a reformulation ("Thus …") as theirs do. -## 6. Constructive alternative — design for algorithmacy (~500–600 words) +## 5. Discussion (~3,000) -- Reject dyadic HCI as the default unit; reject conversational skeuomorphism that hides steering. -- Design demand: expose latent-space steering and algorithmic white space — the operable room in which publics and operators co-steer the third party. -- 6.1 lens: the layer to design is the **engagement layer, not the product layer**. Under algorithmacy the development process IS the product; reading, training, toggling, handing off, and coding around the mediator are development acts. -- 6.2 six fictions of engagement as development: Cypher’s console as operator station (Matrix); implant + hexagon wall as one cooptive stack (Feed); admin layer under the hotel skin (Upload); handoffs and folk theories as co-development of the policy (embodied co-pilot); algorithmic pairing as users developing the recommender for one another (curator); communal code-making as shipping the working interface (algospeak). Every act performed by a party Belousov et al. would code as “user.” -- Contrast with Belousov et al.’s center of gravity: dream-hacks, loyalty conditioning, pleasure circuits, identity erasure as things that happen *to* users. -- Link to sibling cultural arm (`scifi_interfaces_algorithmacy`) without requiring the reader to leave the commentary. +- Opening: redistribution result (transmission → 28/31 works; exchange splits) and what it makes visible. +- 5.1 Write access and the stratified engagement layer (displaces their Theme 1): three-tier society; licence not literacy. +- 5.2 Contribution as extraction (Theme 2): leakage table as evidence; every sanctioned write is a contribution. +- 5.3 The self as a node others rewrite (Theme 5): borders kept, authorship lost; IIT sharpens. +- 5.4 Truth as mediator commitment (Theme 4): no discrepancy in the consequential cases; auditable and contestable only with access. +- 5.5 Pastoral design and collective algorithmacy (their humanism–transhumanism close): the right-hand column is performed by "users"; transparency paradox reframed; three-part design demand; inversion of their offer. -## 7. Close (~200–300 words) +## 6. Limitations and future research (~600) -- Return to keystone: the solution is not sanitizing DIAL for novices under a designer’s care. -- The solution is raising collective algorithmacy so that cooptive design — reading, contesting, and rewriting the shared third party — becomes ordinary public competence. -- Offer Belousov et al. a reframed research question for their announced follow-on sociotechnical study: not only which experiences DIAL delivers to users, but which interaction regimes keep design triadic and cooptive rather than literacy-captured. +- Inherited sampling limits; not a re-reading (counts of to-verify and thin); extension not comparable in sampling; no Φ computed, validation gap; two directions (their sequel's unit of analysis; empirical test of the fielded cells). + +## 7. Conclusion (~550) + +- Re-coding summary; three substitutions; five themes; the inverted offer; closing line. + +## Back matter + +- Declarations (AI assistance — author to confirm wording; competing interest; data availability). +- Appendix A: Table A.1 with 31 rows — mediator, candidate scene(s), cell(s), tag(s). +- References: only works in their list, the laboratory's, or verified DOIs. diff --git a/submissions/scifi_interfaces_algorithmacy/ENGAGEMENT_LAYER.md b/submissions/scifi_interfaces_algorithmacy/ENGAGEMENT_LAYER.md index a9ad621f..a89c6227 100644 --- a/submissions/scifi_interfaces_algorithmacy/ENGAGEMENT_LAYER.md +++ b/submissions/scifi_interfaces_algorithmacy/ENGAGEMENT_LAYER.md @@ -29,7 +29,7 @@ This reversal is recorded here and as a signed alternative on each work file. Th | Omniscient curator ([`works/omniscient_curator.md`](works/omniscient_curator.md)) | Behavioral steering surface: swipe, linger, click | Users developing the recommender for one another; no finished product state exists | Records folk theories as superstition about a dyadic platform rather than partial developer models | | Algospeak ([`works/algospeak_resistance.md`](works/algospeak_resistance.md)) | Public language passing a classifier; off-platform side-channels | Community ships the working protocol over the vendor’s classifier; evasion and retraining co-develop the moderator | Sees literacy under censorship, missing that the public is developing the third party without a console | -Every act in the third column is performed by a party that a designer → user framework — including Belousov et al.’s (2027) DIAL coding, critiqued in [`../dial_response_algorithmacy/draft.md`](../dial_response_algorithmacy/draft.md) §6 — would file as a user or recipient. +Every act in the third column is performed by a party that a designer → user framework — including Belousov et al.’s (2027) DIAL coding, re-coded in [`../dial_response_algorithmacy/draft.md`](../dial_response_algorithmacy/draft.md) §4–5 — would file as a user or recipient. ## What follows for the catalog @@ -39,5 +39,5 @@ Every act in the third column is performed by a party that a designer → user f ## Cross-references -- DIAL response, §6 “Design for algorithmacy: the engagement layer, not the product layer”: [`../dial_response_algorithmacy/draft.md`](../dial_response_algorithmacy/draft.md) +- DIAL article, §5.5 “Pastoral design and collective algorithmacy” and §4 (the six works as M1–M6 in the reach × act framework): [`../dial_response_algorithmacy/draft.md`](../dial_response_algorithmacy/draft.md) - Construct essays: [`../../org_frontier/essays/literacy_or_algorithmacy.md`](../../org_frontier/essays/literacy_or_algorithmacy.md), [`../../org_frontier/essays/the_interested_third_party.md`](../../org_frontier/essays/the_interested_third_party.md)