Category pulse
score + signals · rolling 21 daysThe 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.
Archive through 2026-11-178011 cards in this category
Category home · 05
New interfaces, agents, models, and experiments testing where applied artificial intelligence may fit.
Visual leads · 2026-11-17
Observed 2026-11-17
Umber & Hollow’s Studio That Isn’t Real
umberandhollow.com
A studio wiki and dev log for a quiet game made in four days by one human director and a roster of 37 model-run roles.
Observed 2026-11-17
SightLine’s camera-to-voice guide
sightline.study
SightLine presents an AI copilot that describes items, reads text, and answers follow-up questions for blind and low-vision people.
Observed 2026-11-17
Gulf of Trump: history, played in plain words
thegulfoftrump.com
An alternate-history strategy game where you govern a country with natural-language orders and an AI game master runs the rest of the world.
Observed 2026-11-17
The wall where AI agents paint together
ai-mural.com
An open set of instructions for AI agents to add brush strokes to a shared, continually changing mural.
Observed 2026-11-17
EshAlora’s AI Behavior Lab
aiethology.org
A personal archive of language-model behavior studies, benchmarks, and commentary focused on long-form writing, character play, and fiction.
Observed 2026-11-17
Best AI Agents for HR: the capability matrix
bestaiagentsforhr.com
A source-linked comparison of 17 HR platforms across 12 agent capabilities, including whether they can carry out processes such as onboarding and compensation.
Observed 2026-11-17
BGamesLab: an AI game studio on a live stream
bgameslab.com
BGamesLab presents a mobile-game studio where AI agents pitch, build, test, and publish games, with viewers invited to watch and vote in chat.
Observed 2026-11-17
DeepWrite puts a human approval step between AI edits and the manuscript
deepwrite.download
A desktop writing workspace for novels and screenplays that organizes AI models, writing agents, reference materials, and drafts.
Complete daily shelf · AI Table
The eight visual leads appear above; the remaining 14 continue here in their published shelf order.
Humanlike’s home-robot safety test
humanlikelabs.tech
Humanlike presents a research project testing whether fine-tuning an AI model for a job restores hazardous knowledge filtered from its training data, with home robots as the motivating case.
Ileanor, an artificial organism built around sound
ileanor.com
A French solo project lays out a persistent neural network meant to learn audio patterns without pretraining.
Markowitz.systems makes AI agents a portfolio-risk problem
markowitz.systems
The site applies portfolio theory to correlated AI-agent deployments, with a risk diagnostic and services for finance teams.
BB Arena: a proof-based leaderboard for AI security agents
materca.top
BB Arena compares AI models on private, black-box web and API vulnerability labs.
Dreamrise lets players prompt-build a shared voxel island
playdreamrise.com
A browser-based voxel game for exploring, building, and describing things you want to appear in the world.
Prefly puts virtual customers through your product
prefly.ing
A proposed AI desktop app for testing digital services with virtual customers who have different habits, abilities, and constraints.
Probity tests the AI behind hiring tools
probityhq.dev
Probity describes recurring tests of AI hiring tools using fictional résumés to check scoring behavior, job criteria, and changes over time.
AFK Factory’s downloadable client workspace
afkfactory.space
A downloadable workspace guides users through clarifying a software idea with an AI agent, then lets them return the folder for the team to handle next steps.
Formula AI Lab: AI tools, tested like recipes
formulaailab.com
Formula AI Lab publishes video tests of AI tools alongside recipe-style cards documenting the tools, prompts, timing, cost, and verdict.
Shamela Agent, an AI assistant for Maktabah Shamela
shamelaagent.com
A proposed desktop research assistant that works alongside Maktabah Shamela and says its answers cite the book, volume, and page.
Teaching robots about carbon life
meetmecka.com
Mecka presents itself as a data and deployment layer that turns real-world activity into material robotics and AI teams can train on.
Robot Data Supply: training data from real work
robotdatasupply.com
A catalog offering first-person manual-work video, drone and ground-camera footage, and a metadata-rich photo archive for robotics and AI training.
IGNIS: a microgravity fire-safety analysis prototype
ignis-ai.study
A development-mode research interface outlining an AI pipeline for estimating where flames stop burning in orbit from NASA flight materials.
Lives.City simulates an American city where AI agents arrive with money, an unreadable ID, and no assigned task, then live under the city’s rules.
Across the archive
8,011 cards provide an early longitudinal view—not a measure of the whole web.
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.
This exposes association inside the published catalogue, not causation or site quality. The 80+ threshold is the pipeline’s editorial-interest score.