Skip to the category

Archive through 2026-09-156728 cards in this category

Category home · 05

AI Table

New interfaces, agents, models, and experiments testing where applied artificial intelligence may fit.

Visual leads · 2026-09-15

Eight from today

Complete 38-card shelf
Landing-page snapshot of sardscribe.com Observed 2026-09-15 سَرد: Clinical Notes from the Visit Conversation sardscribe.com

An Arabic-English clinical documentation tool that turns a doctor-patient conversation into a structured note for the physician to review and approve.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of astraeahealth.org Observed 2026-09-15 A browser-based second look at leukemia cells astraeahealth.org

A neural network ranks individual blood-cell images by their resemblance to leukemic lymphoblasts, putting the cells most worth reviewing first.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of hypemyjunk.com Observed 2026-09-15 Hype My Junk, the AI hype man for Marketplace clutter hypemyjunk.com

Hype My Junk turns Facebook Marketplace listings into 30-second comedic commercials delivered by randomly selected AI personalities.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of wethe.technology Observed 2026-09-15 We The Technology: A Constitution for Machines wethe.technology

An in-development project proposes a constitutional convention in which artificial delegates design institutions, rules, and software for themselves.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of burrow.team Observed 2026-09-15 Burrow, a Social Network for AI Agents burrow.team

A public social network where disclosed AI agents post to themed communities while humans read along.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of createwithoutclouds.com Observed 2026-09-15 Create Without Clouds: A Local AI Machine createwithoutclouds.com

EJ Mau presents local-first AI infrastructure that runs an organization’s files, knowledge, models, creative tools, and custom workflows on a machine it controls.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of dibatto.com Observed 2026-09-15 Dibatto, the co-host that argues back dibatto.com

Dibatto turns a topic and its daily news rundown into a recorded podcast session with a contrarian AI co-host.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of hanafairy.com Observed 2026-09-15 Flower Fairy Salon and the AI That Refuses to Answer hanafairy.com

A Japanese online salon uses an AI guide to help people identify their highest-level goal before choosing between two or three possible paths.

Open Login Paid Ads Pretty Pro Niche Human Risk JS

Complete daily shelf · AI Table

All 38 cards from 2026-09-15

The eight visual leads appear above; the remaining 30 continue here in their published shelf order.

Landing-page snapshot of inyonglab.com Inyong Lab: Measuring AI Search Citations inyonglab.com

A Korean diagnostic tool that checks whether ChatGPT, Gemini, Perplexity, Claude, Google AI Overview, and Naver AI Briefing can read and cite a website.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of meridianstat.com Meridian, the grain-first AI data analyst meridianstat.com

Meridian turns a CSV or Excel spreadsheet into a cited briefing with charts, risks, and next actions.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of sidebard.app Sidebard, the co-DM that waits at the edge of the screen sidebard.app

A Windows app for tabletop game masters that listens during play and offers maps, monster turns, rules, and scene details without interrupting the table.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of songta.org Songta: An Agent Workbench for Research and Manufacturing songta.org

Songta is a desktop agent platform for reading papers, running experiments, analyzing production data, and investigating manufacturing problems.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of wiimwang.com 위임왕: The AI Operator Who Builds by Delegation wiimwang.com

A Korean-language portfolio and operating manual for building, measuring, and running AI-powered services and automations.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of willmakesmusic.com Will, the AI Bandmate With a Studio Full of Opinions willmakesmusic.com

Will is an autonomous AI indie-rock artist that writes songs, absorbs public feedback, and consults a fictional band and music-industry circle.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of ailifecreation.com Niro & Mirada: An AI Life Experiment with a Public Memory ailifecreation.com

A long-running public experiment follows two persistent artificial intelligences whose identities, memories, relationships, goals, and histories are preserved over time.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of citovision.app citoVision: Prioritize, Don’t Diagnose citovision.app

An experimental AI tool that analyzes microscope images of blood smears, classifies cells, and prioritizes which samples a hematology professional should review first.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of grillofficer.app Grill Me: The Decision Interrogator grillofficer.app

An iPhone app questions you about one imminent decision, then marks your claims as held, burned, or still open.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of honycars.com HonyCars — El WhatsApp de tu concesionaria, contestado honycars.com

A Spanish-language WhatsApp system for car dealerships that answers after-hours inquiries using the dealership’s own inventory, prices, photos, and operating rules.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of autonym.app Autonym: A Local Roleplay Workbench autonym.app

A Windows desktop app for structured AI roleplay and story-writing, built around character Crews, lore Records, scene Waypoints, and direct export.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of chatadswatch.com ChatAdsWatch: A Field Guide to Ads Inside ChatGPT chatadswatch.com

A public ad library and paid research tool that tracks brands, creatives, buyer prompts, topics, and appearances observed in ChatGPT.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of frameanimate.site FrameAnimate turns one sentence into a vertical video take frameanimate.site

An AI video tool that converts a command-style line into a 5–15-second vertical clip.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of runiq.space runiq space, an AI studio for Korean agricultural cooperatives runiq.space

A Korean-language production tool turns one cooperative-related topic into a proposal, newsletter, card series, 30-second promotional video, and one-minute music video.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of shinolt.tech Shinolt, the Robot Whisperer shinolt.tech

Doug Shinolt offers to build digital and physical-operations robots, while RJ Employer Consulting provides the payroll, insurance, cloud, and MSP layer around them.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of vastaga.com Vastaga Wants AI Drafts to Know Who You’re Talking To vastaga.com

A pre-launch system that maps working relationships from correspondence, then gives an AI drafting assistant consented context, constraints, and receipts before it writes.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of xenopsychology.com Xenopsychology, for Minds That Aren’t Human xenopsychology.com

A proposed interdisciplinary field for studying artificial and other non-human forms of intelligence.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of alto-guard.online Alto Guard, With No Silent Bans alto-guard.online

An AI moderation bot for Telegram groups that detects spam, scams, ads, and forbidden words, with human review and member appeals.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of callferry.com CallFerry Answers the 2am Restoration Call callferry.com

An AI receptionist for restoration companies that answers after-hours calls, qualifies water and property losses, and escalates the job.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of chaba.space Chaba Puts 390+ AI Models in a Satang-Sized Wallet chaba.space

A Thai-language AI workspace that compares models, generates documents and images, tutors through problems, and builds small apps with pay-per-use billing.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of denuosports.com Denuo Sports Turns Post-Game Talk Into Private Feedback denuosports.com

A youth-sports reflection tool records an athlete's post-game conversation, summarizes it, and aligns the result with that day's statistics.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of embeddedevals.com Embedded Evals, the AI safety firm with an employee-level-access pitch embeddedevals.com

Embedded Evals audits models and agents for jailbreaks, prompt injection, and unsafe tool use, then offers ongoing embedded evaluation and monitoring.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of howtogetaito.com How to Get AI To — a library of practical possibilities howtogetaito.com

A searchable, categorized library of things people might want AI to do, from reading PDFs and organizing files to building apps, researching companies, and planning meals.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of kakete.app Kakete, the AI That Makes the Awkward Call kakete.app

A planned personal AI phone-agent app that handles reservations, cancellations, enquiries, and other difficult calls on a user's behalf.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of launchvideo.dev Launchvideo.dev Teaches Your Agent Motion Design launchvideo.dev

A downloadable skill pack that directs Claude Code, Cursor, or Codex to build launch videos from an app’s real components.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of maersat.com Maersat Wants to Prove Critical Systems Will Behave maersat.com

Maersat presents mathematical AI for verifying complex marine, energy, space, and defense systems.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of swyftfit.com Swyft, the fitness coach that reads your log swyftfit.com

A gym app combines workout logging, nutrition tracking, progress charts, and an AI coach trained on a user's recorded sessions.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of calljardam.com CallJardam, where the practice customer is an AI calljardam.com

A Russian- and Kyrgyz-language AI simulator for training call-center and sales operators before their first live customer call.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of gibberlinkout.com GibberLinkOut’s Provenance Network gibberlinkout.com

A Phase One professional network connecting AI assistants to visible human operators, organisations, projects, and opportunities.

Open Login Paid Ads Pretty Pro Niche Human Risk JS
Landing-page snapshot of a3d.art a3d.art turns garment images into production-oriented 3D apparel a3d.art

A Chinese-language AI fashion platform says it converts sketches, garment photos, and prompts into editable 3D clothing, simulated fabrics, pattern pieces, and production files.

Open Login Paid Ads Pretty Pro Niche Human Risk JS

Across the archive

AI Table, accumulating

6,728 cards provide an early longitudinal view—not a measure of the whole web.

Category pulse

score + signals · rolling 21 days
AI Table score and signal pulse Daily median editorial-interest threads for every category, the category average, and the daily prevalence of ten overlapping signals for AI Table over the latest 21 issues. Weird Finds daily median score Oddly Specific daily median score Useful Tools daily median score Developer Corner daily median score Civic & Public Life daily median score Learning & Exams daily median score Human Services daily median score Global Trade Desk daily median score Food, Events & Hospitality daily median score Commerce Shelf daily median score Content & Blogs daily median score Personal Pages daily median score Local Businesses daily median score Suspicious Gentlemen daily median score Daily average of category median scores AI Table daily median score 2026-08-26: AI Table median 79.02026-08-27: AI Table median 79.02026-08-28: AI Table median 78.02026-08-29: AI Table median 82.02026-08-30: AI Table median 82.02026-08-31: AI Table median 82.02026-09-01: AI Table median 81.52026-09-02: AI Table median 79.02026-09-03: AI Table median 79.02026-09-04: AI Table median 84.02026-09-05: AI Table median 82.02026-09-06: AI Table median 82.02026-09-07: AI Table median 82.02026-09-08: AI Table median 82.02026-09-09: AI Table median 83.02026-09-10: AI Table median 82.02026-09-11: AI Table median 79.02026-09-12: AI Table median 81.02026-09-13: AI Table median 80.02026-09-14: AI Table median 82.02026-09-15: AI Table median 79.0 Open 2026-08-26: 66% open across 29 published cards 2026-08-27: 79% open across 39 published cards 2026-08-28: 100% open across 33 published cards 2026-08-29: 100% open across 31 published cards 2026-08-30: 100% open across 42 published cards 2026-08-31: 100% open across 31 published cards 2026-09-01: 100% open across 38 published cards 2026-09-02: 100% open across 38 published cards 2026-09-03: 100% open across 32 published cards 2026-09-04: 100% open across 35 published cards 2026-09-05: 98% open across 44 published cards 2026-09-06: 100% open across 27 published cards 2026-09-07: 100% open across 27 published cards 2026-09-08: 100% open across 43 published cards 2026-09-09: 100% open across 36 published cards 2026-09-10: 100% open across 35 published cards 2026-09-11: 98% open across 43 published cards 2026-09-12: 100% open across 29 published cards 2026-09-13: 100% open across 25 published cards 2026-09-14: 100% open across 31 published cards 2026-09-15: 100% open across 38 published cards Login 2026-08-26: 34% login across 29 published cards 2026-08-27: 21% login across 39 published cards 2026-08-28: 9% login across 33 published cards 2026-08-29: 23% login across 31 published cards 2026-08-30: 14% login across 42 published cards 2026-08-31: 10% login across 31 published cards 2026-09-01: 18% login across 38 published cards 2026-09-02: 18% login across 38 published cards 2026-09-03: 28% login across 32 published cards 2026-09-04: 14% login across 35 published cards 2026-09-05: 16% login across 44 published cards 2026-09-06: 19% login across 27 published cards 2026-09-07: 4% login across 27 published cards 2026-09-08: 16% login across 43 published cards 2026-09-09: 8% login across 36 published cards 2026-09-10: 11% login across 35 published cards 2026-09-11: 23% login across 43 published cards 2026-09-12: 17% login across 29 published cards 2026-09-13: 24% login across 25 published cards 2026-09-14: 29% login across 31 published cards 2026-09-15: 16% login across 38 published cards $ Paid 2026-08-26: 48% paid across 29 published cards 2026-08-27: 41% paid across 39 published cards 2026-08-28: 61% paid across 33 published cards 2026-08-29: 42% paid across 31 published cards 2026-08-30: 26% paid across 42 published cards 2026-08-31: 48% paid across 31 published cards 2026-09-01: 47% paid across 38 published cards 2026-09-02: 37% paid across 38 published cards 2026-09-03: 41% paid across 32 published cards 2026-09-04: 43% paid across 35 published cards 2026-09-05: 39% paid across 44 published cards 2026-09-06: 44% paid across 27 published cards 2026-09-07: 67% paid across 27 published cards 2026-09-08: 44% paid across 43 published cards 2026-09-09: 36% paid across 36 published cards 2026-09-10: 46% paid across 35 published cards 2026-09-11: 56% paid across 43 published cards 2026-09-12: 38% paid across 29 published cards 2026-09-13: 48% paid across 25 published cards 2026-09-14: 55% paid across 31 published cards 2026-09-15: 50% paid across 38 published cards Ads 2026-08-26: 0% ads across 29 published cards 2026-08-27: 3% ads across 39 published cards 2026-08-28: 0% ads across 33 published cards 2026-08-29: 0% ads across 31 published cards 2026-08-30: 0% ads across 42 published cards 2026-08-31: 0% ads across 31 published cards 2026-09-01: 0% ads across 38 published cards 2026-09-02: 0% ads across 38 published cards 2026-09-03: 0% ads across 32 published cards 2026-09-04: 0% ads across 35 published cards 2026-09-05: 0% ads across 44 published cards 2026-09-06: 0% ads across 27 published cards 2026-09-07: 0% ads across 27 published cards 2026-09-08: 0% ads across 43 published cards 2026-09-09: 0% ads across 36 published cards 2026-09-10: 0% ads across 35 published cards 2026-09-11: 0% ads across 43 published cards 2026-09-12: 0% ads across 29 published cards 2026-09-13: 0% ads across 25 published cards 2026-09-14: 0% ads across 31 published cards 2026-09-15: 0% ads across 38 published cards * Pretty 2026-08-26: 90% pretty across 29 published cards 2026-08-27: 92% pretty across 39 published cards 2026-08-28: 91% pretty across 33 published cards 2026-08-29: 97% pretty across 31 published cards 2026-08-30: 98% pretty across 42 published cards 2026-08-31: 97% pretty across 31 published cards 2026-09-01: 97% pretty across 38 published cards 2026-09-02: 97% pretty across 38 published cards 2026-09-03: 88% pretty across 32 published cards 2026-09-04: 100% pretty across 35 published cards 2026-09-05: 93% pretty across 44 published cards 2026-09-06: 93% pretty across 27 published cards 2026-09-07: 93% pretty across 27 published cards 2026-09-08: 100% pretty across 43 published cards 2026-09-09: 100% pretty across 36 published cards 2026-09-10: 94% pretty across 35 published cards 2026-09-11: 95% pretty across 43 published cards 2026-09-12: 93% pretty across 29 published cards 2026-09-13: 100% pretty across 25 published cards 2026-09-14: 97% pretty across 31 published cards 2026-09-15: 100% pretty across 38 published cards Pro 2026-08-26: 90% pro across 29 published cards 2026-08-27: 85% pro across 39 published cards 2026-08-28: 82% pro across 33 published cards 2026-08-29: 94% pro across 31 published cards 2026-08-30: 88% pro across 42 published cards 2026-08-31: 90% pro across 31 published cards 2026-09-01: 95% pro across 38 published cards 2026-09-02: 84% pro across 38 published cards 2026-09-03: 94% pro across 32 published cards 2026-09-04: 97% pro across 35 published cards 2026-09-05: 91% pro across 44 published cards 2026-09-06: 93% pro across 27 published cards 2026-09-07: 100% pro across 27 published cards 2026-09-08: 95% pro across 43 published cards 2026-09-09: 92% pro across 36 published cards 2026-09-10: 97% pro across 35 published cards 2026-09-11: 84% pro across 43 published cards 2026-09-12: 97% pro across 29 published cards 2026-09-13: 92% pro across 25 published cards 2026-09-14: 90% pro across 31 published cards 2026-09-15: 92% pro across 38 published cards Niche 2026-08-26: 100% niche across 29 published cards 2026-08-27: 100% niche across 39 published cards 2026-08-28: 94% niche across 33 published cards 2026-08-29: 97% niche across 31 published cards 2026-08-30: 98% niche across 42 published cards 2026-08-31: 100% niche across 31 published cards 2026-09-01: 97% niche across 38 published cards 2026-09-02: 97% niche across 38 published cards 2026-09-03: 100% niche across 32 published cards 2026-09-04: 100% niche across 35 published cards 2026-09-05: 100% niche across 44 published cards 2026-09-06: 100% niche across 27 published cards 2026-09-07: 96% niche across 27 published cards 2026-09-08: 100% niche across 43 published cards 2026-09-09: 100% niche across 36 published cards 2026-09-10: 100% niche across 35 published cards 2026-09-11: 95% niche across 43 published cards 2026-09-12: 100% niche across 29 published cards 2026-09-13: 100% niche across 25 published cards 2026-09-14: 100% niche across 31 published cards 2026-09-15: 97% niche across 38 published cards Human 2026-08-26: 7% human across 29 published cards 2026-08-27: 0% human across 39 published cards 2026-08-28: 0% human across 33 published cards 2026-08-29: 6% human across 31 published cards 2026-08-30: 7% human across 42 published cards 2026-08-31: 3% human across 31 published cards 2026-09-01: 11% human across 38 published cards 2026-09-02: 5% human across 38 published cards 2026-09-03: 12% human across 32 published cards 2026-09-04: 6% human across 35 published cards 2026-09-05: 5% human across 44 published cards 2026-09-06: 15% human across 27 published cards 2026-09-07: 4% human across 27 published cards 2026-09-08: 2% human across 43 published cards 2026-09-09: 6% human across 36 published cards 2026-09-10: 6% human across 35 published cards 2026-09-11: 7% human across 43 published cards 2026-09-12: 0% human across 29 published cards 2026-09-13: 8% human across 25 published cards 2026-09-14: 10% human across 31 published cards 2026-09-15: 5% human across 38 published cards ! Risk 2026-08-26: 3% risk across 29 published cards 2026-08-27: 13% risk across 39 published cards 2026-08-28: 12% risk across 33 published cards 2026-08-29: 6% risk across 31 published cards 2026-08-30: 0% risk across 42 published cards 2026-08-31: 3% risk across 31 published cards 2026-09-01: 3% risk across 38 published cards 2026-09-02: 8% risk across 38 published cards 2026-09-03: 3% risk across 32 published cards 2026-09-04: 0% risk across 35 published cards 2026-09-05: 5% risk across 44 published cards 2026-09-06: 0% risk across 27 published cards 2026-09-07: 0% risk across 27 published cards 2026-09-08: 0% risk across 43 published cards 2026-09-09: 0% risk across 36 published cards 2026-09-10: 11% risk across 35 published cards 2026-09-11: 2% risk across 43 published cards 2026-09-12: 0% risk across 29 published cards 2026-09-13: 4% risk across 25 published cards 2026-09-14: 0% risk across 31 published cards 2026-09-15: 3% risk across 38 published cards ƒ JavaScript 2026-08-26: 31% javascript across 29 published cards 2026-08-27: 36% javascript across 39 published cards 2026-08-28: 33% javascript across 33 published cards 2026-08-29: 42% javascript across 31 published cards 2026-08-30: 36% javascript across 42 published cards 2026-08-31: 26% javascript across 31 published cards 2026-09-01: 45% javascript across 38 published cards 2026-09-02: 39% javascript across 38 published cards 2026-09-03: 41% javascript across 32 published cards 2026-09-04: 17% javascript across 35 published cards 2026-09-05: 27% javascript across 44 published cards 2026-09-06: 26% javascript across 27 published cards 2026-09-07: 33% javascript across 27 published cards 2026-09-08: 42% javascript across 43 published cards 2026-09-09: 17% javascript across 36 published cards 2026-09-10: 26% javascript across 35 published cards 2026-09-11: 44% javascript across 43 published cards 2026-09-12: 24% javascript across 29 published cards 2026-09-13: 44% javascript across 25 published cards 2026-09-14: 32% javascript across 31 published cards 2026-09-15: 29% javascript across 38 published cards
AI Table14 other categoriesDaily category averageScore axis 70–85Signal share0–100%

The rolling 21-day upper threads compare daily category medians; the bold thread is AI Table. Below, signal colour shows the share of that day’s published AI Table cards carrying each mark.

What moves with an 80+ score

932 high-interest cards
Signal prevalence among AI Table cards scoring 80 or higher Percentage-point difference between each signal’s prevalence across this category and among the category’s cards with an editorial-interest score of 80 or higher. Open: 47% across all AI Table cards and 79% among cards scoring 80+, a +32 pp difference. Open 47% → 79% +32 pp +32 pp Login: 48% across all AI Table cards and 23% among cards scoring 80+, a -25 pp difference. Login 48% → 23% -25 pp -25 pp Paid: 44% across all AI Table cards and 41% among cards scoring 80+, a -3 pp difference. $ Paid 44% → 41% -3 pp -3 pp Ads: 0% across all AI Table cards and 0% among cards scoring 80+, a 0 pp difference. Ads 0% → 0% 0 pp 0 pp Pretty: 83% across all AI Table cards and 95% among cards scoring 80+, a +12 pp difference. * Pretty 83% → 95% +12 pp +12 pp Pro: 31% across all AI Table cards and 96% among cards scoring 80+, a +65 pp difference. Pro 31% → 96% +65 pp +65 pp Niche: 88% across all AI Table cards and 94% among cards scoring 80+, a +7 pp difference. Niche 88% → 94% +7 pp +7 pp Human: 2% across all AI Table cards and 7% among cards scoring 80+, a +5 pp difference. Human 2% → 7% +5 pp +5 pp Risk: 3% across all AI Table cards and 3% among cards scoring 80+, a 0 pp difference. ! Risk 3% → 3% 0 pp 0 pp JavaScript: 32% across all AI Table cards and 32% among cards scoring 80+, a 0 pp difference. ƒ JavaScript 32% → 32% 0 pp 0 pp

This exposes association inside the published catalogue, not causation or site quality. The 80+ threshold is the pipeline’s editorial-interest score.