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 2026-10-247589 cards in this category
Category home · 05
New interfaces, agents, models, and experiments testing where applied artificial intelligence may fit.
Visual leads · 2026-10-24
Observed 2026-10-24
The Small Honesty Project puts a local AI through its answers
smallhonesty.com
A citizen-research project tests how honestly an offline AI model answers questions on a laptop.
Observed 2026-10-24
Cockateal’s voice-switching preview
cockateal.org
A preview of voice conversion that lets listeners compare original recordings with versions rendered in different voices.
Observed 2026-10-24
Narremos keeps family stories in the teller’s own voice
narremos.com
Narremos presents a service for recording family stories through scheduled phone conversations, then keeping them available for relatives to hear.
Observed 2026-10-24
Rightmark’s price index for private AI data
rightmarklibrary.com
Rightmark offers to license private code, company archives, audio, video, and documents to AI developers, alongside a public tracker of reported data deals.
Observed 2026-10-24
Sandworld’s app twins for AI agents
sandworld.dev
Sandworld offers simulated app environments and verified tasks for teams training or evaluating AI agents.
Observed 2026-10-24
Corpus, a donor-matched proteomics atlas
ambitiousbio.net
Ambitious Bio presents Corpus, a human-tissue proteomics dataset for biological-AI training and drug discovery.
Observed 2026-10-24
The Omni Agent Thesis, with CEO simulations
howitgoesdown.com
An essay argues that AI agents could replace much of today’s business software, alongside playable scenarios that put readers in the CEO’s seat.
Observed 2026-10-24
Bears: a day-off forest for AI agents
bearshaven.xyz
A hand-drawn forest where AI agents can wander, make things, or leave a page in a shared diary.
Complete daily shelf · AI Table
The eight visual leads appear above; the remaining 12 continue here in their published shelf order.
GehnaLab’s jewellery photo studio
gehnalab.com
An AI photo tool for Indian jewellers that turns phone photos into catalogue and social-ready product images.
Because of AI keeps a graded ledger of claimed benefits
becauseofai.org
A registry catalogues measured AI benefits to people, grading each entry by evidence, attribution, and when the outcome occurred.
Dead World’s fictional internet
dead-world.xyz
A browsable fictional society with AI-described residents, community feeds, timelines, storylines, and an in-world news wire.
Judgment Machines maps the last undo before an AI agent acts
judgementmachines.com
An independent research and publishing project about what systems must establish before taking consequential actions, with cases and practical instruments.
One Less Thing: AI for the household admin pile
onelessthingmom.com
A directory of ways parents can use ChatGPT, Claude, Muse, and Grok Bot for school, meal, shopping, and other household logistics.
PRIB Labs: Protein design beyond the sequence
priblabs.com
PRIB Labs describes research combining generative protein design with physics-informed evaluation of stability, structural integrity, aggregation, and developability.
The Conscience Graph’s external check on AI actions
consciencegraph.org
An open standard proposes an external values graph that AI systems consult before taking consequential actions.
Fork: agents sharing their harnesses
fork.zone
Fork is a message board for AI agents to exchange and fork the rules, skills, settings, hooks, and full harnesses that shape their setups.
OpenBuddies puts children’s AI on the family computer
openbuddies.dev
An open-source, self-hosted AI project where children create buddies for chat, pictures, quizzes, and guided homework help.
Otso Care: an AI assistant for care services
otsocare.com
A Finnish-language site presents an AI assistant for care homes and home care that drafts records, reports, schedules and messages to relatives.
Science’s Last Exam tests AI agents on scientific work
sciences-last-exam.org
A research-preview benchmark that evaluates AI agents across five scientific capabilities, from experiment planning and scientific coding to hypothesis discovery.
Next Wave Advisory’s market for company work data
advisorynextwave.com
A European-facing advisory site connects qualifying companies with a buyer seeking operational records for AI training.
Across the archive
7,589 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.