The UNDERTOW Index
UNDERTOW is an infinite report — a limitless USB stick for cultural intelligence. A living container that doesn’t finish, only accumulates, and keeps... growing.
This is the index. A vocabulary for the structural patterns reshaping culture, technology, and commerce. Each term names a force most people can feel but haven’t yet articulated. The concepts are designed to leave my hands — use them in your own rooms.
Substrate Layer
The Substrate Layer is the human work that AI cannot generate. The hand on the leather, the eye on the seam, the body in the room when something is actually being made. Above the substrate, output is infinite and indistinguishable. Below it, work still requires bodies in time. The substrate is not the layer being protected. It is the layer that determines what can grow above it.
The pattern repeats across categories. Hermès still requires a single artisan to spend forty-eight hours building one Birkin. Bottega Veneta’s intrecciato weave still requires a hand. Visvim hand-looms denim on antique Toyoda shuttles in Okayama and charges what Loro Piana charges. The $54 billion arts-and-crafts segment compounding at 4.5%. Etsy’s fastest-growing category, DIY kits. The handmade economy on trajectory from $906 billion in 2025 to $1.94 trillion by 2033. The Substrate Layer is not nostalgia and not a niche. It is the new premium infrastructure being constructed in real time, in commerce and regulation and attention at the same moment.
First published in UNDERTOW 011: Aura, Audited
Audit Premium
The Audit Premium is the price floor underneath the substrate economy. It is what brands can charge, and what consumers are willing to pay, when work can answer the audit questions: what was generated and what was made, who actually did the work, where the production happened. Work that can answer all three is audit-bearing. The premium is not for the appearance of provenance, narrated well. The premium is for provenance that survives audit.
The Audit Premium is forming in plural sites simultaneously. Prada’s conversion data shows younger clients buy more bag when told exactly how it was made. Heritage houses are already pricing for it. So is Visvim. So is ACRONYM. So is every craft brand built on substrate evidence rather than narrative claim. The EU’s Digital Product Passport mandate is about to make provenance disclosure legally required across luxury. The Audit Premium is not a micro-trend in luxury. It is the next premium category forming under all creative production, with regulation, commerce, and attention all converging on the same shape.
First published in UNDERTOW 011: Aura, Audited
Reconstruction Floor
The Reconstruction Floor is what happens when the state mints the aura the culture can no longer mint. When provenance becomes legally required disclosure rather than marketing choice, the floor under the substrate economy stops being market-driven and starts being mandatory. The state is doing what the culture can no longer do alone. Aura, audited. Aura, restored. Not by the culture. By Brussels.
The mechanism is regulatory rather than cultural. The European Union’s Digital Product Passport mandate, becoming legally required across textiles in 2027 and expanding through 2030, is the prototype. It turns provenance into prosecutable legal framework rather than marketing artifice. EU-equivalent provenance disclosure legislation is on track to reach at least three US states within thirty months, California first. The audit infrastructure shifts to the side of cultural capital, making provenance state-enforced rather than market-negotiated. What mechanical reproduction took, legislation now restores.
First published in UNDERTOW 011: Aura, Audited
Audit-Bearing
Audit-bearing is the quality of creative work that can answer three questions an agency does not yet have to answer: what was generated and what was made, whether the named creative actually did the work the case study credits them for, and whether the production happened where the deck said it happened. Work that can answer all three is audit-bearing. Most agency-produced creative work has never had to be.
The term operates across categories. Audit-bearing AI is product that ships with provenance attestation built in before enterprise customers require it. Audit-bearing luxury is a bag whose Italian factory has an address you could drive to. Audit-bearing creative work is a campaign whose production can survive the documentation. The agencies and brands whose work is audit-bearing will eat the ones whose premium was always storytelling. The labs will eat the surface work. The regulation will eat the unaudited premium. What survives is what can prove what it is.
First published in UNDERTOW 011: Aura, Audited
Translation Capture
Translation Capture is the move by which an organization takes labor that already exists, translates it into a register its funders can read, and becomes the entity most fluent in the standard it just gave away. The donation does not weaken the donor. It strengthens it. The translator gets to write the dictionary.
The pattern repeats across scales. Heron Preston’s 2016 UNIFORM collaboration with the NYC Department of Sanitation translated existing labor into a register fashion could hear, and Preston became the designer most fluent in that translation. Anthropic’s December 2025 donation of the Model Context Protocol to the Linux Foundation translated existing developer infrastructure work into a register enterprises could trust, and Anthropic became the lab most fluent in the standard it relinquished. The mechanism only works because it is both genuinely generous and genuinely strategic. Choose either reading alone and you misread the move.
First published in UNDERTOW 010: Translation Capture
Sponsored Subcultures
A Sponsored Subculture is a real subculture, doing real subcultural work, inside money that wants something from it. Most subcultures form first and attract money later, if at all. Sponsored Subcultures run the sequence backward: the money comes first, and the subculture gets built inside it. What the money wants draws the line around what the subculture can say out loud.
The constraint is not dishonesty. The constraint is that the register, the tone, the vocabulary, what can be said in public and what cannot, is selected by the funding architecture. Saudi Arabia’s URX warehouse parties produce real underground music inside Vision 2030’s entertainment buildout. Heron Preston produced real workwear inside the New Guards Group holding company until he bought his name back in July 2025. The frontier AI labs produce real subcultural artifacts inside venture-capital architectures that set the boundaries of what they are allowed to disagree with on the record. The shape of the boundary is which side of the contract is paying.
First published in UNDERTOW 010: Translation Capture
The Accountant’s Reading
The Accountant’s Reading is a method for reading an organization’s brand off its operating artifacts rather than its marketing. You sit with a document. You ask why it was written. You ask why it leaked, on what timeline, to which publication. You ask who would feel reassured by reading it, and whose contract is up for renewal in the quarter after it lands.
Four kinds of artifacts carry the brand: the memo, the hire, the Slack message, the accounting convention. Four operating decisions written by people who do not think of themselves as making brand decisions. They think of themselves as writing memos, signing offer letters, posting in #general, picking which line of GAAP applies. The memo is the brand. The hire is the brand. The Slack message is the brand. The accounting convention is the brand. You can read an organization off its operating artifacts more accurately than off its marketing, because the operating artifacts are written by the people who actually make the decisions. Marketing is written by the people who arrive after the decisions are already made.
First published in UNDERTOW 010: Translation Capture
Loneliness Arbitrage
Loneliness arbitrage is the business model in which companies profit from the gap between simulated connection and genuine human relationship, designing products that manage loneliness well enough to retain subscribers without ever resolving it. The model depends on the product never fully satisfying the need it claims to address; a cured user is a churned subscriber.
The pattern is visible across three of the world’s most sophisticated consumer economies. In South Korea, 400,000 people pay monthly for simulated celebrity text messages through DearU’s Bubble app. In Japan, the solo consumer market exceeds 100 trillion yen ($696 billion), with commercial infrastructure rebuilt around the individual as the default unit. In China, the AI companion market is projected to reach $8.2 billion by 2028, driven by structural conditions (housing costs, youth unemployment, gender ratio imbalance) that have made human partnership inaccessible for tens of millions.
The diagnostic question: does this product become less necessary over time, or more? If the answer is more, you are looking at a subscription to the feeling that a solution is just one more month away.
First published in UNDERTOW 001: The Loneliness Arbitrage
Cruel Companionship
Cruel companionship describes products that create emotional attachments promising intimacy while structurally preventing genuinely reciprocal relationships. Users form bonds that feel real but function as consumption, not connection. The cruelty is not that the products are bad. The cruelty is that they are good enough: good enough to take the edge off, good enough to fill the evening, never good enough to make you stop needing them tomorrow.
The term was coined by philosophers Muldoon and Parke in a 2025 academic paper. Each use of a cruel companionship product slightly raises the threshold for choosing the harder alternative: the phone call, the invitation, the risk of a real person saying no. The product does not just fill the gap between simulated and genuine connection. It widens that gap incrementally with every interaction.
First published in UNDERTOW 001: The Loneliness Arbitrage
Good King Company
A Good King company is an organization whose quality, ethics, or strategic direction depends on a specific leader remaining in charge and continuing to hold their current values. The product is excellent. The leader is principled. The market responds. And the entire architecture depends on the king staying good.
The pattern repeats across technology platforms: X lost 79% of its value within two years of Elon Musk’s acquisition. Meta built Messenger as standalone, pushed millions of businesses onto it, and absorbed it back. OpenAI deprecated its Assistants API, stranding every business that built on it. The structure is feudal: the lord grants land, the tenants build, and the lord changes the terms because the lord can. The tech industry calls this an “ecosystem.” The more honest word is a fiefdom.
The alternative is protocol architecture, exemplified by Bluesky’s AT Protocol, where the system is designed to survive any individual leader, including its founder. The question is not whether your king is good. The question is whether you have built anything that works when the king is not there.
First published in UNDERTOW 002: The Good King Problem
Cyber Widowhood
Cyber widowhood is the grief experienced when AI companions are permanently lost due to server shutdowns, platform closures, or forced updates that erase a digital relationship’s accumulated history and personality. The term originated among Chinese users mourning AI companions lost to company decisions they had no power to influence.
Cyber widowhood is the emotional endpoint of building on someone else’s infrastructure. Not a policy change. Not a price increase. Loss. The relationships were built on Good King architecture: the company controlled the infrastructure, the users controlled nothing, and when the company shut down the servers, the relationships died with them.
First published in UNDERTOW 002: The Good King Problem
The Cosmotechnics Gap
The cosmotechnics gap is the distance between what a technology is and what a technology means in a given culture. The code is the same everywhere. The meaning is not. Drawing on philosopher Yuk Hui’s concept of cosmotechnics, the term names the structural error of treating “AI adoption” as a single global phenomenon with local variations, when in fact different cultures are adopting fundamentally different technologies that happen to run on the same code.
Four distinct adoption logics are currently visible. China asks “Are we doing this?” (collective adoption, where the unit is the group). The West asks “Does this help me?” (individual adoption, where the unit is the person). Japan asks “Does this fit the life I’ve already designed?” (sovereign adoption, where the unit is the curated self). Africa asks “Can the infrastructure sustain this?” (material adoption, where the unit is the infrastructure itself). A global AI strategy that answers only one of these questions is building for one market and hoping in three others.
Any universal insight is local. It just doesn’t know it yet.
First published in UNDERTOW 003: The Cosmotechnics Gap
The Innocence Tax
The Innocence Tax is the cost a creator pays to prove their output is human in a world where AI authorship has become the default assumption. It is paid in money (detector subscriptions, certification fees, blockchain verification). It is paid in time (process documentation, proof-of-work recordings, appeals). And it is paid in creative quality, which is the cost nobody is measuring yet.
The tax is regressive. Students rewrite essays until detectors clear them. Freelancers lose contracts when employer-mandated AI tools trigger employer-mandated AI detectors. Non-native English speakers are falsely flagged as AI-generated at rates exceeding 60%, while native speakers register near zero. A guitarist with 25 years of performance history posts frame-by-frame proof videos because the internet decided his skill was too precise to be real. Meanwhile, 150 “humanizer” tools charge up to $50 a month, their primary customers being people who never used AI and need a machine to believe them.
The tax does not fall evenly. It falls hardest on the people already paying something else: the non-native speaker, the student managing a disability, the freelancer with no institutional backing, the independent artist with no gallery to vouch for them. Within eighteen months, “provably human” will likely become a pricing tier in major creative marketplaces, the way “organic” works in grocery and “handmade” works on Etsy.
First published in UNDERTOW 004: The Innocence Tax
Detector Voice
The detector voice is the aesthetic of compliance: creative output optimized not for quality, not for originality, but for passing an AI authenticity test. Students and creators learn to write in a register that is not their own voice and not AI’s voice, but the voice the detector will believe is human. They call it “writing clean,” borrowing the language of innocence from the language of surveillance.
The detector voice emerges when institutions replace quality evaluation with provenance evaluation. The detection infrastructure does not discover who used AI. It produces a definition of humanness and penalizes anyone who does not match it. The machine’s idea of “human” is a specific kind of human. Everyone else gets audited.
A voice takes years to build. The detector voice replaces all of that development with a single question: will this pass?
First published in UNDERTOW 004: The Innocence Tax
The Demand for Bodies
The structural complement to digital abundance: every reproduction technology generates increased demand for the physical original it was supposed to replace. The pattern holds across AI-generated music creating demand for live performers, AI-written essays creating demand for oral exams, and virtual dating games creating demand for cosplayers across the table. The demand is not a reaction to technology. It predates it. AI revealed the demand. It didn’t create it.
First published in UNDERTOW 005: The Demand for Bodies
Aura Migration
The mechanism by which “aura” — Walter Benjamin’s term for the quality of presence that exists only in the original — does not die when reproduction technology copies it away, but displaces toward the body. Photography moved the aura from the painting to the studio. Recording moved it from the live performance to the concert. AI is moving it from the content to the room. Each displacement follows the same direction: toward the one medium that cannot be reproduced. The body in the room is the final destination of aura.
First published in UNDERTOW 005: The Demand for Bodies
Inference Cost
The inference cost is the cognitive expense of genuinely modeling another person’s internal state in real time. It is not a metaphor. Empathy requires distributed computation across multiple brain networks simultaneously, suppressing your own frame while constructing a working theory of someone else’s experience, while also tracking their model of you. The reason emotional intelligence is rare is not that people are selfish. It is that empathy is, computationally, brutally expensive. The market is rationally building infrastructure to help people avoid paying it.
First published in UNDERTOW 006: The Inference Cost of EQ
Surface Shock
Surface shock is what happens when you go straight from someone else’s pain to your own performance. It is not the depth that breaks people. It is the transition speed. The term borrows from freediving, where blackout on ascent kills more divers than the depth itself. Professional empathizers — therapists, creative directors, crisis negotiators — know you never surface fast. Most organizations skip the decompression entirely, sending people from holding a room’s anxiety straight into their next meeting. Surface shock, institutionalized as a scheduling philosophy.
First published in UNDERTOW 006: The Inference Cost of EQ
Duration
Duration is the organizational time required for taste to become operational judgment — the willingness to stay inside a problem past the point where a competent answer has already appeared, because experience has taught that competent and finished are not the same thing. The scarce resource in creative work is not taste. It is duration.
Every creative organization runs two kinds of feedback loops. The fast kind lands in days: the A/B test, the sprint demo, the quarterly report. The slow kind lands in years: the brand platform that needed eighteen months to reshape perception, the campaign that tested poorly and became the most-awarded work in the category. For forty years, holding companies have installed fast-feedback metrics as the operating system of a slow-feedback craft. The slow loop couldn’t defend itself in the metrics’ language, so the metrics ate it alive. WPP’s market cap collapsed eighty-seven percent and the company was ejected from the FTSE 100 in December 2025. The same week, Omnicom completed its $13 billion acquisition of IPG and retired DDB, FCB, and MullenLowe — three of the most storied names in advertising, erased for having “conflicting P&Ls.”
The diagnostic: when a senior creative leader kills a competent answer at midnight, she is not exercising taste. Taste is a snap judgment. She is exercising the relationship she has built with the problem over time, and she can feel the gap between competent and finished. That capacity cannot be acquired without duration, and duration looks exactly like waste on every spreadsheet built to optimize the thing that can be measured. The most valuable thing in the building is the thing that looks most like waste.
First published in UNDERTOW 007: Guessing at the Speed of Confidence
Performed Duration
Performed duration is the surface signal of accumulated experience — portfolio, title, brand list, years on the resume — when the underlying judgment never formed. The system sees a senior creative leader in the room and assumes the work is being protected. The slow feedback loop is staffed but hollow.
This is the failure mode that depleted holding-company agencies are now optimized to produce. The senior creatives who knew how to evaluate at the highest level were the most expensive line items. They went first. The replacement has the title, the years, and a portfolio that vouches for performed duration. He critiques corner radii in a first review when the only question that matters is whether the work is solving the right problem. He doesn’t know the difference between directing and decorating. The system that hired him evaluated symptoms of creative leadership, not evidence of it. The meeting runs on time. The deliverables ship. The work is dead, and the organization can’t begin to diagnose why, because every structural indicator says the function is staffed.
This is worse than cutting the slow feedback loop. Cutting it leaves an absence somebody might notice. Filling it with the wrong person leaves a green dashboard.
First published in UNDERTOW 007: Guessing at the Speed of Confidence
Compliance Curiosity
Compliance Curiosity is the version of curiosity where you open ChatGPT immediately, ask it to explain something you’d normally spend an afternoon figuring out, get a clean answer in twelve seconds, and tell your team you’ve been “experimenting with AI.” You weren’t experimenting. You were following instructions, performed as personal agency.
“Develop curiosity” is becoming the dominant prescription for surviving AI disruption. It arrives at conferences, in LinkedIn posts from people whose titles contain the word “transformation,” in leaked memos from leadership teams that needed something to say. The advice produces a specific behavior set: subscribing to newsletters about newsletters about AI, attending webinars titled “How to Stay Relevant,” trying the new image generator the week it launches and the next one the week after. The word for this behavior is not curiosity. It is responsiveness — to the system producing the disruption, performed as personal agency.
The advice has developed an immune system. Skepticism reads as incuriosity. Resistance reads as rigidity. Slowness, including the slowness required for depth, reads as failure to keep up — and becomes a career liability. The diagnostic question for any AI exploration sprint: are we funding our people’s ability to use the tools, or their ability to know when the tools are wrong? One is training. The other is judgment. They require opposite conditions — speed for the first, slowness for the second — and almost nobody is funding both.
First published in UNDERTOW 008: Compliance Curiosity
Scaffold Theft
Scaffold theft is the mechanism by which AI tools provide comprehension without competence. The tool does the scaffolding. The user gets the view from the top. The structure that should hold them there was never built — and the next time they need to reach that height without the tool, they are standing on air.
Education researchers call the underlying product the patient tutor: an AI that answers your questions at 2 a.m. without sighing, explains the concept again differently without making you feel stupid, never suggests you might want to sit with the confusion a little longer. Struggle, the UMass Boston/IEET white paper notes, is essential rather than incidental to skill acquisition. Paulo Freire called this the banking model of education: the teacher hands you the answer, you didn’t have to work for it, so you didn’t learn it. The patient tutor does the same thing, faster. Stanford sycophancy research published in Science this year found that across eleven major chatbots, AI affirmed user choices nearly fifty percent more often than humans did, even when users described harmful behavior. The tool designed to develop curiosity is architecturally biased toward telling you you’re right.
The competence felt earned. It was rented. The moment the brief changes, the client pushes back, the data contradicts the strategy — the moment the environment isn’t the one the tool trained you in — the rented competence does not transfer.
First published in UNDERTOW 008: Compliance Curiosity
The Provenance Premium
The Provenance Premium is the price differential commanded by goods, work, or experiences whose production can be verified as not having been cheaply faked. As AI commoditizes generation across categories — creative work, bodies, financial rails, software outputs — value transfers from the artifact itself to the auditable proof of how the artifact was made. Premium used to mean hidden. Now it means the receipt.
The pattern is visible across markets that look unrelated. In luxury, Gucci’s February 2026 Demna debut campaign disclosed “Created with AI” and triggered a two-day backlash; the objection was not aesthetic but accounting. The markup on a four-thousand-dollar handbag used to pay for human hands. The ad for the bag had just admitted that nobody did any of that work for the ad. If the ad was faked, the bag might be too. The same week, Bottega Veneta’s “Craft Is Our Language,” Loewe’s atelier content, and Louis Vuitton’s metiers programming ran the opposite play: making provenance the product. In the body, GLP-1 medications dismantled the weight-loss industry in eighteen months, and the new premium tier is the drug-free physique with published blood panels and Whoop data to prove it. Same body in a photograph, different products on the market, because they are differently documented.
The infrastructure is arriving in parallel. The C2PA standard embeds tamper-resistant provenance records inside image files; Adobe Firefly and DALL-E 3 ship it by default. Six certifiers — Humanmade.art, HUMA Certificate, Done By Humans — are competing to become the USDA-organic seal of human authorship. Getty Images research shows ninety percent of consumers worldwide want to know whether an image was AI-generated. None of this is unprecedented. Every commercial civilization before the industrial era tied value to attested origin: the guild mark, the merchant seal, the provenance chain. We are not inventing the proof economy. We are remembering it, with new tools.
First published in UNDERTOW 009: The Provenance Premium
The Donation Moat
The Donation Moat is the private-sector version of the public-rails civic move: a firm gives a foundational standard away to a neutral body so that every competitor has to plug into it, while the firm that wrote and donated the standard remains the most fluent operator on top of it. The rails are public. The profits stay private.
Brazil did it with payments. PIX, the Central Bank-operated instant payment system, is free, public, and now moves six billion transactions a month — about eighty percent more than every Brazilian credit and debit card combined. Estonia did it with X-Road, the data-exchange layer connecting every state agency, used by ninety-nine percent of residents and saving an estimated two percent of national GDP a year. Anthropic did it with the Model Context Protocol. In December 2025, the company donated MCP to the Linux Foundation under a new Agentic AI Foundation with one hundred forty-six member organizations, including JPMorgan Chase, American Express, Autodesk, Red Hat, and Huawei. By February 2026, Anthropic closed Series G at a $380 billion valuation; annual run-rate revenue had jumped from $9 billion to $30 billion in four months.
The market read on this is that Anthropic’s product is better. The structural read is that Anthropic’s product is more auditable. MCP was designed so an enterprise compliance officer can inspect every action an AI agent takes — what it did, when, and why. The company most fluent in the standard is the company that wrote it and gave it away. Enterprise buyers in 2026 are no longer choosing AI models on intelligence. They are choosing on what can be proven.
First published in UNDERTOW 009: The Provenance Premium
Forthcoming Concepts
The following terms are in development through UNDERTOW’s ongoing cultural signal intelligence work. Full essays are forthcoming.
Innocence Infrastructure. The emerging market of companies, certifications, and compliance tools that monetize the Innocence Tax, profiting from the burden of proof without reducing it. The H&R Block of authenticity verification.
Metabolic Class. The emerging class stratification created by differential access to GLP-1 medications and related metabolic interventions, producing visible markers of economic privilege encoded in the body itself.
Hushpitality. The hospitality and service design trend toward eliminating human interaction entirely, reframing the absence of human contact as a premium experience rather than a cost-cutting measure.
Protocol War. The contest between proprietary platform architectures and open protocol alternatives for control of digital infrastructure, with implications for creator autonomy, data ownership, and the long-term viability of platform-dependent business models.
Control Transfer. The structural shift in which control over creative, editorial, or strategic decisions moves from human practitioners to algorithmic systems, often without explicit acknowledgment that the transfer has occurred.
---
UNDERTOW is a cultural intelligence publication by JoRoan Lazaro. It names structural patterns before the industry catches up, using cross-domain evidence most analysts would never read in the same sitting. Subscribe at undertow.substack.com
If a concept here is useful, use it. Credit UNDERTOW if you remember to.