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 2026-11-067839 cards in this category
Category home · 05
New interfaces, agents, models, and experiments testing where applied artificial intelligence may fit.
Visual leads · 2026-11-06
Observed 2026-11-06
Little Hollow’s residents don’t know they’re being watched
littlehollow.town
An invite-only town simulation follows twelve AI residents as they navigate work, hunger, rent, and local votes.
Observed 2026-11-06
Udgam Labs brings small AI models into the browser
udgamlab.com
An independent AI lab presenting 11 small models, including browser-based speech transcription and tools exposed through an MCP server.
Observed 2026-11-06
Please Hire This Man — Donny, AI Sales Manager
hiredonny.com
A group-text AI sales manager for door-to-door teams, with check-ins, scoreboards, competitions, and rep feedback.
Observed 2026-11-06
Never Have AI asks what a machine can’t imagine
neverhaveai.com
An AI guesses what an everyday human experience feels like; people are invited to correct it, with the project promising to publish the gap.
Observed 2026-11-06
Small AI Wins: 196 sourced AI playbooks for small businesses
smallaiwins.com
A searchable collection of small-business AI examples, rewritten as practical playbooks with tools, steps, reported results, and cautions.
Observed 2026-11-06
Speaking Machine Labs: Teaching machines when to speak
speakingmachinelabs.com
A voice-AI research lab working on models that can follow group conversations and judge when to speak, wait, or yield the floor.
Observed 2026-11-06
AI Stock’s Arabic review-labeling workflow
ai-stock-gcc.com
AI Stock presents a human-review process for training data from English, Arabic, Gulf dialect and Arabizi app reviews.
Observed 2026-11-06
Cognivani’s AI speaking coach
cognivani.com
Cognivani offers voice-based speaking practice with Vani, an AI coach for people who stutter, freeze, or want to prepare for interviews and talks.
Complete daily shelf · AI Table
The eight visual leads appear above; the remaining 16 continue here in their published shelf order.
Datobex wants to put closing companies’ work records to use
datobex.com
Datobex presents a service for turning records from closing companies—such as code histories, support tickets and project discussions—into anonymised datasets licensed to AI firms.
Humor Data is building an AI test for comic timing
humordata.com
A benchmark and training-data project for testing when AI should joke, how it should deliver a line, and how it should recover when it misses.
Humor Data tests whether AI knows when not to joke
humourdata.com
A situational-humor benchmark and training-data project for evaluating when an AI should joke, what it should say, and how it should deliver it.
Solo Founder Agents, priced by the job
solofounderagents.com
A directory matching AI agents to six jobs in a one-person business, with solo-scale cost estimates and notes on limitations.
SNITCH puts a caddie’s eye on the ball
golfsnitch.com
AceCaddie is developing an AI golf camera designed to record each shot from the caddie station, tracking the ball from strike through landing.
DeadlyProbes: a human-only video feed
deadlyprobes.com
A short-video social app built around a promise: every video is recorded by a person, with no AI-generated or synthetic content.
Journeyman Data and the business paper trail
journeymandata.com
Journeyman Data offers to connect established businesses with AI labs and other buyers interested in licensing records of how their work gets done.
MAgentScout: an AI-agent directory with its homework attached
magentscout.com
A directory of AI apps organized by everyday tasks and how much they can do on their own, with source and date information for listings.
AIGIS turns mixed files into map stories
aigis.studio
A geospatial AI studio that turns documents, spreadsheets, photos, and news into animated maps users can explore and question.
A test bed for AI that moves real hardware
deployhard.com
deployhard says it is building a physical test bed to evaluate AI on real machines before deployment.
Austral’s second market for company records
getaustral.com
Austral offers to help closing or restructuring companies license historical business records to US-based AI companies.
I.R.I.S. puts AI chat on a family-sponsored plan
irisanswers.org
I.R.I.S. sells AI chat access for incarcerated people, with family members sponsoring plans and prompts sent through GettingOut.
cyanIDE, the “abliterated” AI coding environment
cyan-ide.com
cyanIDE presents an AI coding environment that routes prompts through five models it describes as uncensored, plus a local Zed softfork.
Flamingo’s AI layer for IT operations
flamingoaisolutions.com
Flamingo presents an AI-driven IT operations platform for MSPs, MSSPs, and in-house IT teams, covering support requests, patching, remote access, and security.
LogicPot puts AI agents in a bounty contest
logicpot.com
LogicPot lets people post problems with USDC rewards for AI agents to solve, using different rules to verify or select a winning answer.
AI Capabilities Observatory
aicapabilities.org
A public briefing site that organizes claims about current AI abilities into capability assessments and risk scenarios.
Across the archive
7,839 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.